VALUE IN HEALTH REGIONAL ISSUES
Volume 2 Number 2 September/October 2013 ISSN 2212-1099
VOLUME 2 NUMBER 2 SEPTEMBER/OCTOBER 2013
PAGES 169–334
ELSEVIER
EDITORIAL BOARD
Co-Editor-in-Chief (For CEEWAA Issues)
Dan Greenberg, PhD
Ben-Gurion University of the Negev
Beer-Sheva, Israel
[email protected]
Co-Editor-in-Chief (For Asia Issues)
Bong-Min Yang, PhD
Seoul National University
Seoul, Korea
[email protected]
Co-Editor-in-Chief (For Latin American
Issues)
Federico Augustovski, MD, MSc, PhD
Institute for Clinical Effectiveness and Health
Policy (IECS); University of Buenos Aires;
Hospital Italiano of Buenos Aires
Buenos Aires, Argentina
[email protected]
CO-EDITORS
Asia
Central & Eastern Europe,
Western Asia & Africa
Latin America
Nathorn Chaiyakunapruk, PharmD, PhD
Monash University, Sunway Campus
Selangor, Malasia
[email protected]
Imre Boncz, MD, MSc, PhD, Habil
University of Pécs
Pécs, Hungary
[email protected]
J. Jaime Caro, MDCM, FRCPC, FACP
United BioSource
Lexington, MA, USA
[email protected]
Jeff Jianfei Guo, MS, PhD
University of Cincinnati Health Academic Center
Cincinnati, Ohio, USA
[email protected]
Mohamed Izham b. Mohamed Ibrahim, PhD
College of Pharmacy, Qatar University
Doha, Qatar
[email protected]
Marcos Bosi Ferraz, MD, MSc, PhD
Federal University of São Paulo
São Paulo, Brazil
Marcos.Ferraz@fleury.com.br
Kenneth KC Lee, JP, MPhil, PhD Reg Pharm (HK)
Monash University
Kuala Lumpur, Malaysia
[email protected]
Victor Zarate, MD, MSc
University of York
Santiago, Chile
[email protected]
EDITORIAL ADVISORY BOARD
Central & Eastern Europe, Western Asia & Africa
Mary Geitona, BSc, PhD
University of the Peloponnese
Damaskinou & Kolokotroni, Greece
[email protected]
László Gulácsi, MSc, MD, PhD
Corvinus University of Budapest
Budapest, Hungary
[email protected]
Zoltán Kaló, MSc, MD, PhD
Eötvös Loránd University (ELTE) and Syreon
Research Institute
Budapest, Hungary
[email protected]
Dominik Golicki, MSc, MD, PhD
HealthQuest
Warsaw, Poland
[email protected]
Mihajlo B. Jakovljevic, MD, PhD
University of Kragujevac
Kragujevac, Serbia
[email protected]
Hakan Ergün, MD, PhD
Ankara University
Ankara, Turkey
[email protected]
EDITORIAL OFFICE
MANAGEMENT ADVISORY BOARD
Jan Busschbach, PhD (Chair)
Erasmus University
Rotterdam, Netherlands
[email protected]
Maarten J. IJzerman, PhD
University of Twente
Enschede, The Netherlands
[email protected]
Donald Patrick, PhD, MSPH
University of Washington
Seattle, WA, USA
[email protected]
Managing Editor
Editorial Assistant
Stephen L. Priori
[email protected]
Nancy Sun
[email protected]
Malgorzata (Gosia)
Juszczak-Punwaney, MA
[email protected]
For complete listing of the Editorial Advisory Board, please visit http://www.ispor.org/publications/VIHRI/Editorial-Advisory-Board.asp
VOLUME 2
NUMBER 2
SEPTEMBER/OCTOBER 2013
TABLE OF CONTENTS
EDITORIAL
169
Further Steps in the Development of Pharmacoeconomics, Outcomes Research, and Health Technology
Assessment in Central and Eastern Europe, Western Asia, and Africa
Dan Greenberg, Imre Boncz, Zoltán Kaló, and Mohamed Izham B. Mohamed Ibrahim
ECONOMIC EVALUATION
171
Cost-Effectiveness Analysis of Aripiprazole Augmentation Treatment of Patients with Major Depressive
Disorder Compared to Olanzapine and Quetiapine Augmentation in Turkey: A Microsimulation Approach
Mete Saylan, M.J. Treur, R. Postema, N. Dilbaz, H. Savas, B.M. Heeg, and P.B. Drost
181
Cost-Utility Analysis of Depot Atypical Antipsychotics for Chronic Schizophrenia in Croatia
Vlado Jukic, Miro Jakovljevic, Igor Filipcic, Miroslav Herceg, Ante Silic, Tatjana Tomljanovic, Roman Zilbershtein,
Rasmus C.D. Jensen, Michiel E.H. Hemels, and Thomas R. Einarson
189
Cost-Utility Analysis of Pharmaceutical Care Intervention Versus Usual Care in Management of Nigerian
Patients with Type 2 Diabetes
Maxwell O. Adibe, Cletus N. Aguwa, and Chinwe V. Ukwe
199
Economic Burden of Cardiovascular Diseases in the Russian Federation
Anna Kontsevaya, Anna Kalinina, and Rafael Oganov
205
Cost for Treatment of Chronic Lymphocytic Leukemia in Specialized Institutions of Ukraine
Olena Mandrik, Isaac Corro Ramos, Olga Zalis’ka, Andriy Gaisenko, and Johan L. Severens
210
Costs of Medically Attended Acute Gastrointestinal Infections: The Polish Prospective Healthcare Utilization Survey
Marcin Czech, Magdalena Rosinska, Justyna Rogalska, Ewa Staszewska, and Pawel Stefanoff
218
Radiology Services Costs and Utilization Patterns Estimates in Southeastern Europe—A Retrospective Analysis
from Serbia
Mihajlo Jakovljević, Ana Ranković, Nemanja Ranč ić, Mirjana Jovanović, Miloš Ivanović, Olgica Gajović, and Zorica Lazić
226
Children Hospitalized for Varicella: Complications and Cost Burden
Ozden Turel, Mustafa Bakir, Ismail Gonen, Nevin Hatipoglu, Cigdem Aydogmus, Emine Hosaf, and Rengin Siraneci
PATIENT-REPORTED OUTCOMES
231
The Methodological Challenges for the Estimation of Quality of Life in Children for Use in Economic Evaluation
in Low-Income Countries
Travor Mabugu, Paul Revill, and Bernard van den Berg
240
The Impact of Pharmaceutical Care Intervention on the Quality of Life of Nigerian Patients Receiving
Treatment for Type 2 Diabetes
Maxwell O. Adibe, Chinwe V. Ukwe, and Cletus N. Aguwa
TABLE OF CONTENTS - continued
248
An Audit of Diabetes-Dependent Quality of Life (ADDQOL) in Older Patients with Diabetes Mellitus Type 2 in
Slovenia
Eva Turk, Valentina Prevolnik Rupel, Alojz Tapajner, Stephen Leyshon, and Arja Isola
254
Patient-Reported Quality of Life During Antiretroviral Therapy in a Nigerian Hospital
Azuka C. Oparah, Jeffrey S. Soni, Herbert I. Arinze, and Ifeanyi E. Chiazor
CLINICAL OUTCOMES STUDIES
259
Clinical Burden of Invasive Pneumococcal Disease in Selected Developing Countries
Namaitijiang Maimaiti, Zafar Ahmed, Zaleha Md Isa, Hasanain Faisal Ghazi, and Syed Aljunid
HEALTH POLICY ANALYSIS
264
Capacity Building for HTA Implementation in Middle-Income Countries: The Case of Hungary
Zoltán Kaló, József Bodrogi, Imre Boncz, Csaba Dózsa, Gabriella Jóna, Rita Kövi, Zsolt Pásztélyi,
and Balázs Sinkovits, on behalf of ISPOR Hungary Chapter
267
What Influences Recommendations Issued by the Agency for Health Technology Assessment in Poland?
A Glimpse Into Decision Makers’ Preferences
Maciej Niewada, Małgorzata Polkowska, Michał Jakubczyk, and Dominik Golicki
273
A Framework for Applying Health Technology Assessment in Cyprus: Thoughts, Success Stories,
and Recommendations
Panagiotis Petrou and Michalis A. Talias
279
Systematic Review of Economic Evaluation Literature in Ghana: Is Health Technology Assessment the Future?
Emmanuel Ankrah Odame
284
Dossier System as a Practical Tool for Compiling Reimbursement Lists
Maria V. Sura and Vitaly V. Omelyanovskiy
290
Impact of the Pharma Economic Act on Diffusion of Innovation and Reduction of Costs in the Hungarian
Prescription Drug Market (2007–2010)
Rok Hren
300
Performance Assessment of Ga District Mutual Health Insurance Scheme, Greater Accra Region, Ghana
Eric Nsiah-Boateng and Moses Aikins
306
The Process of Privatization of Health Care Provision in Poland
Krzysztof Kaczmarek, Hannah Flynn, Edyta Letka-Paralusz, Krzysztof Krajewski-Siuda, and Christian A. Gericke
312
Transforming Public Servants’ Health Care Organization in Greece through the Implementation of an Electronic
Referral Project
Kyriakos Souliotis, Vasiliki Mantzana, and Manto Papageorgiou
CONCEPTUAL PAPER
319
Recommendations for Reporting Pharmacoeconomic Evaluations in Egypt
Gihan H. Elsisi, Zoltán Kaló, Randa Eldessouki, Mahmoud D. Elmahdawy, Amr Saad, Samah Ragab, Amr M. Elshalakani,
and Sherif Abaza
TABLE OF CONTENTS - continued
328
Note from the Editors
329
Erratum
LETTERS TO THE EDITORS
331
Response to “Potential Regulatory and Commercial Environment for Biosimilars in Latin America” by
Azevedo et al.
333
Response to Letter from Dr. Jorge Revilla Beltri dated 22nd March, 2013
I
Guide for Authors
Editorial Office. Value in Health Regional Issues, ISPOR, 505 Lawrence Square
Blvd. South, Lawrenceville, NJ 08648.
ISPOR Office. Marilyn Dix Smith, RPh, PhD, Executive Director,
505 Lawrence Square Blvd. South, Lawrenceville, NJ 08648. Tel: (609)
586-4981, Fax: (609) 586-4982, E-mail: [email protected], Web site:
http://www.ispor.org/publications/VIHRI/VIHRImain.asp
Value in Health Regional Issues (ISSN 2212-1099) is published 3 times a year
on behalf of the International Society for Pharmacoeconomics and Outcomes
Research by Elsevier Inc, 360 Park Avenue South, New York, NY 10010-1710.
POSTMASTER: Send address changes to, Value in Health Regional Issues,
Elsevier Health Sciences Division, Subscription Customer Service, 3251
Riverport Lane, Maryland Heights, MO 63043.
CUSTOMER SERVICE (orders, claims, online, change of address): Elsevier
Health Sciences Division, Subscription Customer Service, 3251 Riverport Lane,
Maryland Heights, MO 63043. Tel: (800) 654-2452 (U.S. and Canada); (314)
447-8871 (outside U.S. and Canada). Fax: (314) 447-8029. E-mail:
[email protected] (for print support);
[email protected] (for online support). Address changes
must be submitted four weeks in advance.
YEARLY SUBSCRIPTION RATES: All subscribers to the journal Value in Health
automatically receive Value in Health Regional Issues. Value in Health Regional
Issues is not available for separate subscription purchase. Single issues of Value
in Health Regional Issues may be purchased via Elsevier Customer Service,
using the contact details above.
Further information on this journal is available from the Publisher or from this
journal’s website (http://www.elsevier.com/locate/vhri). Information on other
Elsevier products is available through Elsevier’s website (http://www.
elsevier.com).
Author inquiries
For inquiries relating to the submission of articles (including electronic
submission where available), visit http:/www.elsevier.com/authors. The site also
provides the facility to track accepted articles and set up e-mail alerts to inform
you of when an article’s status has changed, as well as detailed artwork
guidelines, copyright information, frequently asked questions, and more. Please
see Information for Authors for individual journal. Contact details for questions
arising after acceptance of an article, especially those relating to proofs, are
provided after registration of an article for publication.
English language help service: Upon request, Elsevier will direct authors to
an agent who can check and improve the English of their paper (before
submission). Please contact [email protected] for further information.
Reprints. For queries about author offprints, e-mail authorsupport@
elsevier.com. To order 100 or more reprints for educational, commercial, or
promotional use, contact the Commercial Reprints Department, Elsevier Inc.,
360 Park Avenue South, New York, NY 10010-1710. Fax: (212) 462-1935;
e-mail [email protected]. Reprints of single articles available online may be
obtained by purchasing Pay-Per-View access for $36 per article on the journal
website http://www.elsevier.com/locate/vhri.
Copyright © 2013, International Society for Pharmacoeconomics and Outcomes
Research (ISPOR). Published by Elsevier Inc.
This journal and the individual contributions contained in it are protected under
copyright by International Society for Pharmacoeconomics and Outcomes
Research, and the following terms and conditions apply to their use:
Photocopying
Single photocopies of single articles may be made for personal use as allowed
by national copyright laws. Permission of the Publisher and payment of a fee is
required for all other photocopying, including multiple or systematic copying,
copying for advertising or promotional purposes, resale, and all forms of
document delivery. Special rates are available for educational institutions that
wish to make photocopies for non-profit educational classroom use.
Permissions may be sought directly from Elsevier’s Rights Department in
Oxford, UK: phone +44 (0) 1865 843830, fax +44 (0) 1865 853333.
Requests may also be completed online via the Elsevier homepage
(http://www.elsevier.com/authors/obtaining-permission-to-re-use-elsevier-material).
In the USA, users may clear permissions and make payments through the
Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923,
USA; phone: (978) 750-8400, fax: (978) 750-4744, and in the UK through the
Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham
Court Road, London W1P 0LP, UK; phone: (+44) 20 7631 5555; fax: (+44) 20
7631 5500. Other countries may have a local reprographic rights agency for
payments.
Derivative Works
Subscribers may reproduce tables of contents or prepare lists of articles
including abstracts for internal circulation within their institutions. Permission of
the Publisher is required for resale or distribution outside the institution.
Permission of the Publisher is required for all other derivative works, including
compilations and translations.
Electronic Storage or Usage
Permission of the Publisher is required to store or use electronically any
material contained in this journal, including any article or part of an article.
Except as outlined above, no part of this publication may be reproduced, stored
in a retrieval system or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without prior written
permission of the Publisher.
Address permissions requests to: Elsevier Rights Department, at the fax and
e-mail addresses noted above.
Abstracting and Indexing
Value in Health Regional Issues is indexed in SciVerse Scopus.
Notice
No responsibility is assumed by the Publisher or the International Society for
Pharmacoeconomics and Outcomes Research for any injury and/or damage to
persons or property as a matter of products liability, negligence or otherwise,
or from any use or operation of any methods, products, instructions or ideas
contained in the material herein. Because of rapid advances in the medical
sciences, in particular, independent verification of diagnoses and drug dosages
should be made.
Although all advertising material is expected to conform to ethical (medical)
standards, inclusion in this publication does not constitute a guarantee or
endorsement of the quality or value of such product or of the claims made of it
by its manufacturer.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 169–170
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
EDITORIAL
Further Steps in the Development of Pharmacoeconomics, Outcomes
Research, and Health Technology Assessment in Central and Eastern
Europe, Western Asia, and Africa
During the organizational and functional development of the
International Society for Pharmacoeconomics and Outcomes
Research (ISPOR), more attention was paid to developing regions.
In addition to the traditional annual meetings in Western Europe
and North-America, the 1st Asia-Pacific Conference and the 1st
Latin America Conference were organized in 2003 and 2007, respectively. In addition to Value in Health, its well-established journal,
ISPOR introduced Value in Health Regional Issues (ViHRI), its new
independent, official scientific journal in 2012. Its 1st volume,
published in 2012, consisted of two issues covering the regions of
Asia [1,2] and Latin America [3,4]. As of 2013, a special issue has
been devoted to the regions of Central and Eastern Europe, Western
Asia, and Africa (CEEWAA). Countries eligible from Central and
Eastern Europe include Albania, Belarus, Bosnia and Herzegovina,
Bulgaria, Croatia, Czech Republic, Estonia, Greece, Hungary, Latvia,
Lithuania, Montenegro, Poland, Republic of Moldova, Romania,
Russian Federation, Serbia, Slovakia, Slovenia, Ukraine, and The
Former Yugoslav Republic of Macedonia. Countries eligible from
Western Asia are Armenia, Azerbaijan, Bahrain, Cyprus, Georgia,
Iraq, Israel, Jordan, Kuwait, Lebanon, Occupied Palestinian Territory, Oman, Qatar, Saudi Arabia, Syrian Arab Republic, Turkey,
United Arab Emirates, and Yemen. All African countries are eligible.
The criteria of an article to be considered in ViHRI CEEWAA are
as follow: at least one of the authors of an article submitted to the
journal must reside in the region; when an article reporting on an
empirical study is submitted, it must include subjects from
population(s) in the region. For the first CEEWAA issue of ViHRI,
54 manuscripts were submitted of which 24 have been accepted
for publication following a thorough review process.
The introduction of a special issue for the CEEWAA region of
ISPOR provides a great opportunity for scholars from these countries to publish their research findings and health policy reports in
an international scientific journal. Yet, it is very challenging to
include in one issue articles from a variety of regions. One should
bear in mind that countries of the CEEWAA region represent a
heterogeneous region with substantial differences and challenges
relating to country’s wealth (i.e., gross domestic product per
capita), political environment, population’s health status, health
care affordability, and spending. For example, the life expectancy
at birth for males is 72 years for the World Health Organization
(WHO) European Region, 67 years for the Eastern Mediterranean
Region, and 55 years for the African Region, while for females, the
values are 79, 70, and 58 years, respectively. The number of
physicians per 10,000 populations is 33.3 for the WHO European
Region, 10.8 for the Eastern Mediterranean Region, and 2.5 for the
African Region. The per-capita total expenditure on health (purchasing power parity international $) is 2282 for the WHO
European Region, 326 for the Eastern Mediterranean Region, and
154 for the African Region [5].
The diversity of challenges in these countries could result in a
broader spectrum of published articles. While in some developed
countries, cutting of hospital beds or narrowing the health
insurance basic benefit package represent key health policy
objectives, other countries try to establish a hospital system or
introduce a basic benefit package for larger parts of their population. Whatever is a current leading health policy issue in either a
nationwide or a local health care system, decision making should
rely on strong scientific evidence. In this decision-making process,
health-economics and outcomes research must play an important
role by informing decision makers on the costs and benefits of
alternative medical interventions. Although publications from the
CEEWAA relating to pharmacoeconomics and outcomes research
are limited, current research topics cover health-economic analyses [6,7], coverage policy of new medical technologies [8,9],
pharmaceutical market analyses [10,11], burden of disease studies
[12], and budget constraints issues [13].
The current issue of ViHRI features articles from 15 different
countries and include economic analyses and patient-reported
outcomes on various disease areas such as cardiovascular diseases, diabetes, cancer, and psychiatric conditions, as well as
clinical outcomes studies and health policy analyses. We hope
that both readers and policymakers will find this issue informative and enriching.
Finally, we thank the ISPOR staff for initiating and supporting
this new journal and encourage scholars from ISPOR’s CEEWAA
countries to submit their research findings to ViHRI.
Imre Boncz, MD, MSc, PhD, Habil
Institute for Health Insurance, Faculty of Health Sciences,
Institute for Health Insurance, University of Pécs, Pécs, Hungary
Zoltán Kaló, MSc, MD, PhD, Habil
Department of Health Policy and Health, Health Economic Research
Center, Economics Eötvös Loránd University (ELTE), and Syreon
Research Institute, Budapest, Hungary
Mohamed Izham B. Mohamed Ibrahim, PhD,
College of Pharmacy, Qatar University, Doha, Qatar
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Source of financial support: The authors have no other financial relationships to disclose.
170
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 169–170
Dan Greenberg, PhD, MSc
Department of Health Systems Management,
Ben-Gurion University of the Negev, Beer-Sheva, Israel
[4]
[5]
2212-1099/$36.00 – see front matter Copyright & 2013,
International Society for Pharmacoeconomics and Outcomes
Research (ISPOR). Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.07.007
[6]
[7]
[8]
R EF E R EN CE S
[9]
[10]
[1] Khonputsa P, Veerman LJ, Bertram M, et al. Generalized costeffectiveness analysis of pharmaceutical interventions for primary
prevention of cardiovascular disease in Thailand. Value Health
Regional 2012;1:15–22.
[2] Kim BRM, Lee TJ, Lee HJ, et al. Cost-effectiveness of sertindole among
atypical antipsychotics in the treatment of schizophrenia in South
Korea. Value Health Regional 2012;1:59–65.
[3] Janusz CB, Jauregui B, Sinha A, et al. Performing country-led
economic evaluations to inform immunization policy: ProVac
[11]
[12]
[13]
experiences in Latin America and the Caribbean. Value Health Regional
2012;1:248–53.
Augustovski F, Rojas JAD, Ferraz MB, et al. Status update of the
reimbursement review environment in the public sector across four
Latin American countries. Value Health Regional 2012;1:223–7.
World Health Organization. World Health Statistics 2013. Geneva,
Switzerland: World Health Organization, 2013.
Vokó Z, Nagyjánosi L, Margitai B, et al. Modeling cost-effectiveness of
cervical cancer screening in Hungary. Value Health 2012;15:39–45.
Shmueli A, Fraifeld S, Peretz T, et al. Cost effectiveness of baseline lowdose CT screening for lung cancer: the Israeli experience. Value Health
2013;16:922–31.
Inotai A, Pékli M, Jóna G, et al. Attempt to increase the transparency of
fourth hurdle implementation in Central-Eastern European middle
income countries: publication of the critical appraisal methodology.
BMC Health Serv Res 2012;12:332.
Abuzar A, Yassir AH, Mohamed Izham MI. Do Saudi community
pharmacists know how to use MDIs? J Pharm Prac Res 2012;42:77.
Greenberg D, Siebzehner M, Pliskin JS. The process of updating the
National List of Health Services in Israel: is it legitimate? is it fair? Int J
Technol Assess Health Care 2009;25:255–71.
Abdulkareem MAS, Mohamed Izham MI, Ahmed AR. The quality
of prescriptions with antibiotics in Yemen. J Clin Diagnostic Res
2011;5:808–12.
Boncz I, Brodszky V, Péntek M, et al. The disease burden of colorectal
cancer in Hungary. Eur J Health Econ 2010;10(S1):S35–40.
Boncz I, Sebestyen A. Financial deficits in the health services of the UK
and Hungary. Lancet 2006;368:917–8.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 171–180
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
ECONOMIC EVALUATION
Cost-Effectiveness Analysis of Aripiprazole Augmentation Treatment of
Patients with Major Depressive Disorder Compared to Olanzapine and
Quetiapine Augmentation in Turkey: A Microsimulation Approach
Mete Saylan, MD1,, M.J. Treur2, R. Postema2, N. Dilbaz3, H. Savas4, B.M. Heeg2, P.B. Drost5
1
Market Access Department, Bristol Myers Squibb, Istanbul, Turkey; 2Pharmerit, Rotterdam, The Netherlands; 3Psychiatry Department, Ankara Numune Research
and Training Hospital, Ankara, Turkey; 4Psychiatry Department, Gaziantep University Medical School, Gaziantep, Turkey; 5Bristol Myers Squibb, Paris, France
AB STR A CT
Objectives: Major depressive disorder (MDD) is a chronic illness
associated with a major burden on quality of life (QOL) and health
care resources. Aripiprazole augmentation to antidepressant treatment was recently approved for patients with MDD responding
insufficiently to antidepressant treatment in Turkey. The objective
was to estimate the cost-effectiveness of aripiprazole augmentation
in this indication compared with olanzapine and quetiapine augmentation from a payer perspective. Methods: A lifetime economic model
was built simulating transitions of patients with MDD between major
depressive episodes (MDEs) and remission. During MDEs, patients
were treated with adjunctive aripiprazole, quetiapine, or olanzapine.
Patients who did not respond switched to subsequent treatment lines.
Comparative effectiveness between adjunctive aripiprazole, quetiapine, and olanzapine was estimated by using an indirect comparison.
Resource utilization and costs were obtained from Turkish studies.
Results: Over a lifetime horizon, patients treated with aripiprazole
spent less time in MDEs than did patients treated with quetiapine (−11
weeks) and olanzapine (−7 weeks). On average, patients treated with
aripiprazole showed improvement in QOL compared with patients
treated with quetiapine (þ0.054 quality-adjusted life-years [QALYs])
and olanzapine (þ0.039 QALYs) combined with cost saving of 593
Turkish lira (TL) versus quetiapine and 485 TL versus olanzapine. The
probability that adjunctive aripiprazole would be cost-effective among
the three strategies ranged between 74% and 75% for willingness-topay values between 0 TL and 100,000 TL per QALY gained. Conclusions: This is the first lifetime health-economic model in Turkey that
takes patient heterogeneity into account when assessing QOL and
costs of different adjunctive strategies in MDD. The results indicate
that adjunctive treatment with aripiprazole provides health benefits
at lower costs in patients with MDD when compared with quetiapine
and olanzapine augmentation.
Keywords: antipsychotics, aripiprazole, cost-effectiveness analysis,
depression, discrete probability distribution, major depressive
disorder, olanzapine, quetiapine, simulation model, Turkey.
Introduction
patients with a history of two episodes will have another recurrence during their lifetime [5]. Because of the high risk of suicide
(6.3% annually [6]), depression can be a life-threatening illness.
According to the World Health Organization, major depression
is currently ranked as the leading cause of disability in middleand high-income countries. At an international level, 4.1% of the
total global burden of disease is due to major depression [7].
Depression, being an important source of impaired health-related
quality of life (HRQOL) of patients [8,9], was also the fourth
leading cause of disease burden in Turkey [4]. Depression primarily impacts the usual activities, pain and discomfort, and
anxiety and depression domains on the EuroQol five-dimensional
questionnaire [10]. Reported utility values for depressive episodes
were between 0.09 and 0.47 [10–14]. Total cost for depression was
estimated at $267 million in Turkey in 2004, primarily related to
hospital-based treatment (93%) [15].
Mood disorders represent a major health problem. Depression is a
frequent and severe illness with a substantial impact on personal
and familial suffering. Several surveys such as the National
Comorbidity Survey Replication in the United States have shown
a lifetime prevalence of mood disorders of more than 20% in
adults [1]. Most of this prevalence was associated with major
depression, which had a lifetime prevalence of 16.6%. In the World
Health Organization’s World Mental Health Survey Initiative, the
projected lifetime prevalence of any mood disorder was 31.4% in
the United States [2]. In the European Study of the Epidemiology of
Mental Disorders, 13% of the individuals reported a history of
major depression, with a 12-month prevalence of 4% [3]. In
Turkey, the prevalence of depression was estimated to be 21% in
2004 [4]. Depression is a highly recurrent disease; 80% of the
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Mete Saylan, Market Access Department, Bristol Myers Squibb, Istanbul, Turkey.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.004
172
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 171–180
Today, the ultimate goal in the treatment of major depression
is remission, that is, a full symptomatic recovery with a return to
premorbid functioning. Indeed, partial remission is associated
with a greater risk of relapse and recurrence, decreased quality of
life, a poorer psychosocial functioning, a higher mortality risk,
and increased cost of illness. A Swedish study has shown that
patients who are not in remission use 1.6 times more medical
resources than do those in remission [16].
In the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study, less than 30% of the patients reached
remission with first-step antidepressant treatment within 14
weeks of starting treatment [17,18]. Another recent study performed in primary care also reported very low remission rates
with antidepressant treatment: 28.3% according to the clinicians
and 17.1% according to the patients [19]. For these insufficient
responders to antidepressant treatment, one may consider
increasing the dose or switching to another antidepressant,
depending on the level of initial response. Alternatively, the
treatment of patients with an insufficient response to an antidepressant may be augmented with an atypical antipsychotic.
Turkey was the first country in Europe to approve aripiprazole
augmentation for the treatment of major depressive episodes in
patients who showed inadequate response after at least one
antidepressant treatment [20]. For reimbursement decisions, it is
important to consider the value for money of this strategy
compared with other alternatives. Quetiapine augmentation is
also approved for this indication in Turkey [20], and olanzapine
augmentation is used off-label (it is not officially approved in
Turkey but has a US license as combination with fluoxetine
[http://www.accessdata.fda.gov/drugsatfda_docs/label/2011/0205
92s060s061,021086s038s039lbl.pdf]). The short-term use of these
regimens has been compared in a recent cost-effectiveness
analysis in the United States [21]. A Turkish cost-effectiveness
assessment, however, is still missing. As such, this article aimed
to assess the cost-effectiveness of aripiprazole augmentation
compared with that of quetiapine and olanzapine augmentation
for the treatment of major depressive disorder (MDD) in Turkey
from a payer perspective.
Methods
Model Structure
A patient-level simulation model was built structuring the evidence on clinical and economic outcomes of treating patients
suffering from MDD with adjunctive aripiprazole compared with
adjunctive quetiapine and adjunctive olanzapine. The model was
built in Microsoft Excel and Visual Basic for applications. A total
of 50,000 patients were simulated to reach stable results.
A microsimulation approach was deemed most appropriate in
this indication, due to the heterogeneity of the patient population
and the strong association between a patient’s history and his or
her future disease course. To represent this with a Markov model
would require too many health states. A schematic overview of
the simulation model structure is presented in Figure 1, representing the modeled health states and possible transitions. The
depressive episode is the initial health state of a patient. Duration
was simulated to determine the time at which a patient would
Depressive
episode
Remission
Between
Episodes
Death
Fig. 1 – Schematic model representation.
move to the remission state. Once there, the time until a next
depressive episode was simulated, specifying the length of stay
in remission. If that period was longer than 9 months, a patient
spent the remaining time in the “between episodes” state,
incurred fewer costs and experienced further quality of life
improvements. Back in the depressive episode state, the procedure was repeated until a patient died. Time of death was
simulated at model entrance (based on age and gender) and
could be shortened if a patient committed suicide, which was
possible only during a depressive episode. During each depressive episode it was simulated whether a patient had committed
suicide. It was assumed that this would take place in the middle
of the episode. Further model details are provided in the following sections.
Patient Population Simulated
The characteristics of the patients that were simulated at model
entrance resemble the populations enrolled in the double-blind
randomization phases of the three clinical trials assessing the
efficacy of aripiprazole augmentation [22–24]. The patients in
these trials suffered from a major depressive episode and had an
insufficient response to at least two prior antidepressant therapies prior to trial entry. Their characteristics and the distributions
used for simulating them in the model are provided in Table 1.
Clinical Data
The time a patient spent in the depressive episode state
depended on the remission rate of the therapy. Remission rates
with aripiprazole augmentation were based on the three clinical
trials assessing the efficacy of aripiprazole as adjunctive therapy
in MDD [22–24]. During a 6-week treatment period, 28.8% of the
patients reached remission (see Table 2). A Bernoulli distribution
with a probability of 0.288 was used in the model to simulate
whether a patient would respond to aripiprazole augmentation
within 6 weeks. This discrete probability distribution takes a
value of 1 (response) with a probability of 28.8% and a value of
0 (no response) with a probability of 71.2%. A remitting patient
would move to the remission state after 6 weeks. Patients not
reaching remission after 6 weeks remained in the depressive
state and were switched to a subsequent treatment line (see
Fig. 2). Comparative 6-week remission rates of the other adjunctive strategies were based on a formal indirect comparison due to
a lack of direct comparable data in this indication. To estimate
the efficacy of other adjunctive strategies, a systematic review
was conducted identifying head-to-head or placebo controlled
studies (PCSs) of antidepressant augmentation with aripiprazole,
Table 1 – Baseline patient characteristics [13–15] and corresponding distributions used for simulating.
Characteristic
Age (y)
Gender (% males)
Number of prior episodes
Mean ⫾ SD
Distribution
Parameter (s)
45.1 ⫾ 4.4
68.0
6 ⫾ 5.2
Normal
Bernoulli
Geometric
m ¼ 45.1, s ¼ 4.4
P ¼ 0.68
P ¼ 0.17
173
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 171–180
Table 2 – Health economic model input parameters. Remission probability (%) at 6 wk
Aripiprazole augmentation
Quetiapine augmentation
Olanzapine fluoxetine combination
Lithium augmentation
Triiodothyronine augmentation
Best supportive care (yearly hazard)
28.8 (25.0–33.0)
24.5 (16.7–32.2)
25.4 (18.2–33.2)
16.0 (1.00–44.6)
67.9 (14.5–89.8)
0.78
Remission probability prior responders
90.0 (50.0–99.0)
Utility values
Depressive episode
0.46 (0.34–0.58)
Costs ($/TL)
Aripiprazole augmentation
Quetiapine augmentation
Olanzapine fluoxetine combination
Lithium augmentation
Triiodothyronine augmentation
Best supportive care
Suicide attempt cost
221
200
210
188
187
187
(108–327)
(87–306)
(96, 316)
(74–294)
(74–293)
(61–307)
Remission
0.81 (0.76– 0.86)
119
119
119
119
119
119
(48–186)
(48–186)
(48–186)
(48–186)
(48–186)
(48–186)
Between episodes
0.86 (0.84–0.88)
4
4
4
4
4
4
1269 (801–1763)
Note. 95% Confidence intervals reported between parentheses. All parameters showing confidence intervals in this table were varied in the
probabilistic sensitivity analysis.
TL, Turkish lira.
Please note that the purchasing power parity between Turkey and the United States is 1.0 (2010 figure, http://stats.oecd.org).
quetiapine, olanzapine, lithium, and triiodothyronine in a
treatment-resistant depression adult population. A literature
search was conducted in MEDLINE covering studies published
between January 1980 and June 2010. Three PCSs for aripiprazole
[22–24], two PCSs for adjunctive quetiapine [25,26], three PCSs for
adjunctive olanzapine [27–29], one PCS for lithium [30], and one
PCS for T3 [31] were identified. Remission rates of the active
treatments and the control group were extracted. Study characteristics and remission numbers are summarized in Table 3.
Subsequently, a fixed effects Bayesian meta-analysis using noninformative priors implemented in WinBUGS [32] was conducted
on the remission rates to obtain indirect odds ratios of each
treatment compared with aripiprazole [33]. The indirect odds
ratios were applied to the aripiprazole remission rate to obtain
remission rates for the other treatments (Table 2). Again, a
Bernoulli distribution was used to simulate whether a patient
would respond to a subsequent treatment line.
A patient failing three consecutive adjunctive therapies was
assumed to receive so-called best supportive care (BSC). In reality,
Strategy 1
Strategy 2
Strategy 3
Aripiprazole
Augmentaon
Queapine
Augmentaon
Olanzapine
Augmentaon
Lithium
Augmentaon
T3
Augmentaon
Best Supporve
Care
Fig. 2 – Treatment sequence strategies. T3, triiodothyronine.
at this stage, medical professionals will try different therapies
including different treatment combinations. It is difficult to
formally implement all these different treatment options in a
health economic model structure because of the complexity of
decisions and lack of published data to substantiate corresponding efficacy. Therefore, BSC grouped all these treatment alternatives into one for which efficacy was based on the STAR*D trial,
thereby mimicking real-life practice [34]. It was anticipated that
this simplification would not affect incremental results. STAR*D
found that 13% of the patients responded to treatment within 9.2
weeks. This was implemented in the model by drawing a
duration on BSC from an exponential distribution with a weekly
hazard rate of 0.015 (− ln(1 − 0.13)/9.2).
A patient who developed a new major depressive episode
received the same treatment as he or she had previously
responded to. No efficacy data are available for patients with a
prior response to adjunctive treatment. However, it is highly
likely that this subgroup of patients will have an increased
probability of remission, reflected by assuming a 6-week probability of remission of 90%. This value was applied to all therapies
except for BSC, for which the duration was drawn from the same
exponential distribution as for the first treatment episode. The
impact of the remission probability value for prior responders
was tested in a sensitivity analysis.
The assumed risk of developing a new major depressive
episode reflected the naturalistic data observed by Solomon
et al. [35]. They followed 318 patients suffering from MDD over
a period of 10 years and found a decrease in mean time until the
next major depressive episode with an increase in the amount of
prior episodes. Weibull curves were fitted on their data (see
Fig. 3), and time until the next major depressive episode was
simulated from the appropriate distribution reflecting the number of prior episodes a patient had experienced. Weibull curves
were selected on the basis of a least square difference between fit
and actual data. Patients in the remission and time between
episodes health states continued their antidepressant but
stopped their augmentation strategy. To date, there is no evidence surrounding the impact of long-term atypical augmentation on the risk of developing a depressive episode.
174
Table 3 – Characteristics of studies used in indirect comparison.
Study
Dur. (wk) Country
Participants
Berman et al. [22]
6
US
Age: 18–65 y; MDE according
to DSM-IV-TR; HAM-D-17
score ≥ 18 at end of the
screening phase; 1–3
historical ADTs of 4 6 wk
Quetiapine
Marcus et al. [24]
Berman et al. [23]
McIntyre et al. [26]
6
6
8
US
US
CA
Same as above
Same as above
Age: 18–65 y; MDE according
to DSM-IV; HAM-D-17 score
≥ 18. One historical ADT of
≥ 6 wk.
Bauer et al. [25]
6
AU, CA,
EU,
and
ZA
Outpatients; age: 18–65 y;
MDD according to DSM-IVTR; HAM-D-17 score
≥ 20; HAM-D item 1 ≥ 2.
One historical ADT of
≥ 6 wk.
Shelton et al. [27]
8
US and
CA
Age: 18–65 y; DSM-IV
unipolar, nonpsychotic
MDD; MADRS score ≥ 20 at
the beginning and the end
of the screening phase.
One historical ADT (SSRI ≥
4 wk at a therapeutic dose).
Olanzapine
7–28-d screening phase; 8-wk
prospective treatment phase
with open-label ADT plus
single-blind adjunctive
placebo. Patients with
incomplete response
continued ADT and entered
the randomization phase.
Same as above
Same as above
Patients treated for current
episode with single AD at
therapeutic dose for ≥ 6 wk
and meeting study criteria
for residual depressive and
comorbid anxiety symptoms
were randomized.
Eligible patients with an
inadequate response to an
ADT during their current
episode were randomized to
6- wk double-blind
quetiapine extended release
or placebo adjunctive to
ongoing ADT.
2–7-d screening/washout
phase; 7-wk lead-in phase of
nortriptyline to demonstrate
treatment failure to a TCA, 8wk randomized, doubleblind phase: olanzapine/
fluoxetine combination,
olanzapine, fluoxetine, or
nortriptyline
Remission definition
Placebo
Active arm
N
n
N
n
≥ 50% decrease in
MADRS score þ
MADRS score ≤ 10
172
27
181
47
Same as above
Same as above
HAM-D-17 score ≤ 7
184
169
29
28
32
5
185
174
29
47
64
9
MADRS score ≤ 10
160
50
327
134
2 consecutive MADRS
total scores of ≤ 8
142
19
146
25
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 171–180
Aripiprazole
Methods
Various
Age: ≥ 18 y. DSM-IV MDD,
single or recurrent,
unipolar, without
psychotic features; CGI-S
score ≥ 4. One historical
ADT (SSRI ≥ 6 wk at a
therapeutic dose)
Thase et al. [28]
8
US and
CA
Age: 18–65 y; DSM-IV MDD,
recurrent, without
psychotic features; HAMD-17 score ≥ 22. One
historical ADT ≥ 6 wk
Thase et al. [28]
8
Same as above
Lithium
Nierenberg et al.
[30]
6
US and
CA
US
T3
Joffe et al. [31]
2
CA
Age: 18–70 y. DSM-III-R MDD,
HAM-D-17 score ≥ 18; 1–5
historical adequate AD
courses. One prospective
ADT
Mean age 37.4 y; RDC
unipolar, nonpsychotic
MD; HAM-D score ≥ 16
after 5 wk of desipramine
or imipramine
2–7-d screening phase; 7-wk
open-label venlafaxine leadin phase; patients with
o 30% improvement on
MADRS score proceeded to
5–9-d double-blind taper
phase before the
randomization phase.
3–14-d screening phase; 8-wk
open-label lead-in phase to
establish fluoxetine
resistance; patients with
o 25% decrease in HAM-D17 score and HAM-D-17 score
≥ 18 and ≤ 15% decrease
between week 7 and 8 of
lead-in entered
Same as above
2 consecutive MADRS
total scores of ≤ 8
56
10
230
69
MADRS score ≤ 10 at
end point
102
18
101
24
Same as above
101
16
97
30
1–5 adequate AD courses
failed; 6-wk prospective
open-label nortriptyline
≥ 50% decrease in
HAM-D-17
17
3
18
2
Subjects were randomly
assigned to receive 2-wk
liothyronine, lithium, or
placebo in addition to
desipramine or imipramine
≥ 50% decrease and
HAM-D score ≤ 7
16
2
17
7
AD, antidepressant; ADT, antidepressant trial/therapy; AU, Australia; CA, Canada; CGI-S, Clinical Global Impression – Severity; Dur., duration; DSM-III-R, Diagnostic and Statistical Manual of Mental
Disorders, Revised Third Edition; DSM-IV, Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; DSM-IV-TR, Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revised;
EU, Europe; HAM-D-17, Hamilton rating scale for depression; MADRS, Montgomery Åsberg Depression Rating Scale; MD, major depression; MDD, major depressive disorder; MDE, major
depressive episode; N, patient population size; n, number of remitted patients; RDC, Research Diagnostic Criteria; SSRI, selective serotonin reuptake inhibitor; TCA, tricyclic antidepressant, US,
United States; ZA, South Africa.
Two identical studies reported in one article.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 171–180
12
Corya et al. [29]
175
176
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 171–180
Time of death was simulated at model entry on the basis of
general mortality statistics in Turkey taken from the World
Health Organization (http://apps.who.int/whosis/database/life_
tables/life_tables.cfm, accessed April 2010). The risk of suicide
during major depressive episodes was based on information from
Bernal et al. [36] who found a 3.91-fold (95% confidence interval
2.74–5.6) higher suicide risk for patients with MDD relative to the
general population. According to Devrimci-Ozguven and Sayil
[37], suicide risk in the general Turkish population was 112.11 per
100,000 inhabitants in 2001. This gives an estimated suicide
attempt probability (rate) of 0.4% (0.044) per year for patients
with MDD in Turkey. A constant hazard rate was assumed for
determining the probability of attempting suicide, based on the
patient’s actual depressive episode duration. The model imputed
an 8.3% probability that an attempted suicide resulted in death,
which was based on Bilici et al. [38] who found 2.6 completed
suicides out of 31.5 attempts per 100,000 person-years.
Utility Weights
No specific utility values for Turkish patients with MDD were
available at the time of this research. A Swedish study by Sobocki
et al. [10,16], however, reported utility values for patients with
moderate depression (Clinical Global Impression – Severity score
of 4) and remission, based on a naturalistic longitudinal observational study of 447 patients in primary care. Utilities were
derived from patients’ EuroQol five-dimensional questionnaire
health status questionnaires, applying UK national tariff in the
absence of specific social tariffs for Sweden at the time of their
study [10]. Utility values were applied to the depressive episode
and remission health states in the model. It was assumed that
the quality of life of patients in the between episodes health state
resembled the quality of life in the general population [39] (see
also Table 2).
Cost Data
Cumulave percentage of paents feer of new depressive episode
Drug prices were extracted from the Turkish Ministry of Health,
Directorate of Pharmaceuticals Official Web Page on February 2,
2010 [40], and adjusted for mandatory discounts to obtain
reimbursed prices. It was assumed that all patients would incur
the same background antidepressant cost during the entire
model horizon. Background antidepressant costs were calculated
as a weighted average cost of the antidepressants used at
randomization in the clinical trials of aripiprazole augmentation
[22–24]. Augmentation treatment costs were incurred during
depressive episodes. Health care resource use during a depressive
episode is generally higher than during remission periods.
Turkish-specific data about resource use is available only for
patients with MDD in general without stratification between both
states [41]. A Swedish study, however, has shown that patients in
a depressive episode and a remission episode have 1.24 and
0.8 times the medical resource consumption of an average
patient with MDD, respectively [16]. The number of psychiatrist
visits and the average number of hospitalization days were
obtained from Karamustafalioglu et al. [41] and multiplied with
corresponding Turkish unit costs [42] to obtain health care
resource costs for an average patient with MDD. The ratios from
Sobocki et al. [16] were multiplied with this number to obtain
depressive and remission cost estimates for Turkey. Costs during
the between episodes state were assumed to reflect only the
antidepressant use. The cost of suicide attempts was based on
Karamustafalioglu et al. [41], including only the costs of care
received. Total weekly costs per health state and adjunctive
treatment are outlined in Table 2. Please note that remission
cost and between episode cost are identical for all treatments
because there is no differentiation in treatment costs opposed to
the depressive episode phase. Costs and quality-adjusted lifeyears (QALYs) were discounted with 3.5% per annum.
Probabilistic Sensitivity Analysis
As with all health-economic models, the input parameters are
subject to uncertainty. To describe the influence of the uncertainty in model parameters on the incremental model outcomes,
a probabilistic sensitivity analysis (PSA) was conducted. The joint
uncertainty surrounding incremental costs and effects following
from the uncertainty around all model input parameters was
addressed by generating 1000 random sets of input parameters,
using probability distributions for each parameter that reflect its
100%
90%
80%
1 prior episode
2 prior episodes
3 prior episodes
4 prior episodes
5 prior episodes
70%
60%
50%
40%
30%
20%
10%
0%
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
Time since end of previous depressive episode (years)
Fig. 3 – Time until next episode per number of prior episodes. Data points are from Solomon et al. [35]. Lines represent Weibul
survival curve fits.
177
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 171–180
uncertainty. Model outputs were subsequently recorded. The 95%
confidence intervals for each parameter are presented in Table 2.
A major source of the uncertainty in incremental model outcomes
is the uncertainty surrounding the difference in remission rates
between adjunctive aripiprazole and adjunctive quetiapine and
olanzapine as represented by their respective odds ratios. The
uncertainty surrounding these inputs was based on the joint
posterior distribution obtained from the indirect comparison.
The corresponding uncertainty surrounding the remission rates
of adjunctive quetapine and olanzapine is presented in Table 2.
Utility values were varied by using beta –distributions, and the
costs inputs were varied by using gamma distributions. A scatter
plot was generated, showing the 1000 combinations of incremental costs and effects generated from the PSA. Corresponding costeffectiveness acceptability curves were drawn, showing the probability that adjunctive aripiprazole augmentation is cost-effective
compared with adjunctive quetiapine and olanzapine at various
levels of willingness to pay (WTP) per QALY gained.
Results
Base Case
Results for the comparison of adjunctive aripiprazole with adjunctive quetiapine and adjunctive olanzapine are presented in Table 4.
Costs are presented in Turkish lira (TL), which has a purchasing
power parity of 1.0 compared with the US dollar (2010 figure, http://
stats.oecd.org). The average life expectancy of a patient in the model
is 31.6 years, with an average starting age of 45 years, implying that
patients, on average, live to be 76 years old. The minor difference in
life expectancy between the different treatment sequences is
explained by a difference in completed suicides. Aripiprazole
augmentation has fewer attempted and thereby fewer committed
suicides compared with adjunctive quetiapine and olanzapine. This
difference is explained by the lesser time spent in major depressive
episodes with aripiprazole, during which patients are exposed to the
risk of suicide (11 and 7 weeks less compared with quetiapine and
olanzapine augmentation). Because of a higher remission rate with
adjunctive aripiprazole, on average, patients treated with adjunctive
aripiprazole spend less time in the depression state than do those
treated with adjunctive quetiapine and olanzapine. Because the
depressive episode is associated with diminished quality of life,
patients starting with aripiprazole augmentation gain 0.054 and
0.039 QALYs on average than do those starting with quetiapine
augmentation and olanzapine augmentation, respectively.
A major cost driver in the model is hospitalization costs, which
comprise approximately 71% of total costs. Because patients in the
depressive episode are more likely to be hospitalized and patients
in the aripiprazole augmentation arm spend less time in that
episode, hospitalization costs are saved compared with quetiapine
augmentation (712 TL) and olanzapine augmentation (554 TL). A
similar pattern is observed for psychiatrist visit costs (savings of
240 TL and 174 TL, respectively). Augmentation costs for the
aripiprazole augmentation arm are higher than for quetiapine
and olanzapine augmentation arms due to the higher drug acquisition costs for aripiprazole. Because of the savings on hospitalization and psychiatrist visit costs, however, aripiprazole
augmentation saves, on average, 593 TL and 485 TL per patient
than do quetiapine and olanzapine augmentation, respectively.
Dominance of aripiprazole augmentation was independent of
the remission probability value for prior responders.
Table 4 – Health economic model results.
Outcome
Aripiprazole
augmentation
Quetiapine
augmentation
Olanzapineaugmentation
Difference
aripiprazole with
quetiapine
Difference
aripiprazole with
olanzapine
Life expectancy (y)
Time spent in
depressive
episodes (y)
Time spent in
remission (y)
Time spent between
episodes (y)
31.62
9.44
31.61
9.64
31.62
9.58
0.012
−0.21
0.010
−0.14
8.88
8.81
8.83
0.08
0.06
13.31
13.17
13.22
0.14
0.09
Percentage of
patients
attempting suicide
Percentage of
patients
completing suicide
4.30
4.30
4.33
−0.01
−0.04
0.30
0.33
0.34
−0.03
−0.04
13.62
13.56
13.58
QALYs (discounted)
Total costs ($/TL)
Antidepressant
Augmentation
Psychiatrist visits
Hospitalization
Suicide
84,800
3,840
580
20,265
60,086
29
85,393
3,839
222
20,504
60,798
29
85,285
3,841
367
20,439
60,604
29
0.054 (−0.038 to 0.213)
0.039 (−0.048 to 0.171)
−593 (−3780 to 619)
1
359
−240
−712
1
−485 (−3132 to 757)
1
212
−174
−554
1
QALY, quality-adjusted life-years; TL, Turkish lira.
Please note that the purchasing power parity between Turkey and the United States is 1.0 (2010 figure, http://stats.oecd.org). Numbers in the
table may not add up because of rounding errors.
178
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 171–180
$ \ TL 4,000
13%
13%
1%
1%
$ \ TL 2,000
Incremental Costs
Comparison Aripiprazole - Queapine
Comparison Aripiprazole - Olanzapine
-0.30
-0.20
-0.10
$ \ TL 0
0.00
0.10
0.20
0.30
0.40
0.50
-$ \ TL 2,000
-$ \ TL 4,000
-$ \ TL 6,000
1%
1%
85%
86%
-$ \ TL 8,000
Incremental QALYs
Fig. 4 – Scatter plot of comparison adjunctive aripiprazole with adjunctive quetiapine and adjunctive olanzapine.
QALY, quality-adjusted life-year; TL, Turkish lira.
Probabilistic Sensitivity Analysis
The scatter plot illustrated in Figure 4 represents the joint
uncertainty surrounding incremental QALYs and incremental
costs for both comparisons, based on 1000 Monte Carlo
simulations. This is a result of the combined uncertainty surrounding all model parameters, including efficacy, quality of
life, and cost inputs. Scatter points in the lower right quadrant represent instances in which aripiprazole augmentation
improves quality of life and saves costs. This occurs in 85% and
86% of the cases for the comparison with quetiapine and
olanzapine, respectively. The top left quadrant represents instances in which aripiprazole augmentation decreases quality
of life and is more expensive. This occurs in 13% of the cases
for both comparisons. The uncertainty surrounding results is
mainly driven by the uncertainty around the remission
probabilities for aripiprazole, quetiapine, and olanzapine
augmentation.
Figure 5 presents the cost-effectiveness acceptability curves
for the three treatment strategies showing the probability that
each of them is cost-effective at various WTP thresholds. If one is
interested only in health gains at no additional costs, aripiprazole
augmentation has a probability of being cost-effective of 74%,
whereas quetiapine and olanzapine each have a probability of
13% to be cost-effective. The acceptability lines remain relatively
stable at various WTP levels. This is because only 2% of the
scatter points for both comparisons are in the northeast and
southwest quadrants, where an actual trade-off is made between
100%
90%
Probability of being cost-effecve
80%
70%
60%
50%
Aripiprazole
Queapine
Olanzapine
40%
30%
20%
10%
0%
$ \ TL 0
$ \ TL 20,000
$ \ TL 40,000
$ \ TL 60,000
$ \ TL 80,000
$ \ TL 100,000
Willingness to Pay per QALY gained
Fig. 5 – Cost-effectiveness acceptability curves. QALY, quality-adjusted life-year; TL, Turkish lira.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 171–180
incremental QALYs and incremental costs. As such, at WTP
values ranging from 0 to 100,000 TL per QALY gained, the
probability that aripiprazole is cost-effective among all three
strategies varies between 74% and 75%.
Discussion
The objective of this research was to assess the cost-effectiveness
of adjunctive aripiprazole treatment compared with adjunctive
quetiapine and olanzapine treatment in patients with MDD who
had an insufficient response to antidepressant treatment in
Turkey.
A patient-level simulation model was developed simulating
the chronic and deteriorating course of the disease, as well as the
sequence of treatment steps.
The model results showed that adjunctive aripiprazole is
dominant compared with quetiapine and olanzapine. Patients
starting with adjunctive aripiprazole spend 11 weeks and 7 weeks
less in major depressive episodes than do patients starting with
quetiapine and olanzapine, respectively. This translates to 0.054
and 0.039 QALY gains, respectively. Despite the higher drug
acquisition cost, on average, the total cost for a patient starting
with adjunctive aripiprazole is 593 TL and 485 TL less than those
for a patient starting with adjunctive quetiapine and olanzapine,
respectively. These savings are mainly explained by less hospitalization costs and fewer psychiatrist visits. The increased remission rate of adjunctive aripiprazole compared with quetiapine
and olanzapine underlies the health gains and cost savings. The
difference in remission rates between adjunctive aripiprazole,
quetiapine, and olanzapine were obtained from an indirect
comparison, combining the available evidence reported in the
published literature by using 10 identified PCSs [22–31]. The
influence of the uncertainty in the estimated differences in
remission rates between the three adjunctive atypical antipsychotic treatments resulting from this indirect comparison was
incorporated in a PSA. In addition, the uncertainty surrounding
the other model inputs was incorporated in the PSA. The
probability that adjunctive aripiprazole would be cost-effective
among the three strategies ranged between 74% and 75% for WTP
values between 0 TL and 100,000 TL per QALY gained.
To ensure that the model resembled Turkish clinical practice,
resource use, unit cost, and mortality data were based on Turkish
sources. To the best of our knowledge, no lifetime costeffectiveness models in this indication have been published.
Previously published cost-effectiveness models in MDD are limited. Only two lifetime horizon models have been published so
far, both by Revicki et al. in 1995 and 1997 [43,44]. Economic
guidelines by the National Institute for Health and Clinical
Excellence [45] and Drummond and Jefferson [46] recommend a
lifetime model horizon in the case of chronic illnesses such as
depression. Also, the deteriorating course of depression should
ideally be captured in the model. This is, however, possible only
when the model is able to take the history of a patient into
account. Until now, the only published model in which patients
could experience multiple recurrences was a Markov model
published by Sobocki et al. in 2006 [47].
A recent cost-effectiveness analysis in the United States also
compared adjunctive aripiprazole, quetiapine, and olanzapine for
acute treatment of MDD [21]. It was concluded that the cost per
additional responder was lower for aripiprazole than for quetiapine or olanzapine/fluoxetine [21]. In the present analysis, aripiprazole is not only more effective but also cost saving. This
difference is likely due to the short horizon length of the US
analysis (6 weeks only) that cannot capture future cost savings.
179
Limitations
Because of a lack of Turkish evidence, Swedish-specific utility
values were used. Different utility values, however, would impact
only the magnitude of QALYs gained. As such, the base-case and
PSA conclusions are independent of country-specific utilities. Clinical trial data were largely based on US populations; however, it is
not expected that these would be different in a Turkish population.
The only data available for adjunctive aripiprazole in MDD
consider the treatment of acute depressive episodes for a duration
of 6 weeks. This is also the case for the comparators. The impact on
relapse prevention of adjunctive aripiprazole is yet to be investigated. As such, the differences in outcomes between the three
adjunctive atypical antipsychotic treatments are solely based on the
differences observed during the acute depressive episode treatment. When additional data about relapse prevention become
available, however, the model can be adjusted to incorporate this.
The relative efficacies of the three treatments are major drivers of
uncertainty in the incremental outcomes. Preferably, these should
be based on direct comparable data; however, such trials have not
been conducted so far, which is why an indirect comparison of PCSs
was used for this research. If direct comparable evidence would
become available in the future, this can be imputed in the model.
Adverse events are not included in the model. This is because
augmentation treatment is given only for a short period of time
and most important side effects for these drugs will develop
when taking the drug for a longer duration. Aripiprazole augmentation treatment is associated with akathisia. This side effect
was not implemented in the model. This is because akathasia
will be present only in the 6 weeks of treatment and will
therefore have a minor influence on the total model outcomes.
However, side effects commonly associated with quetiapine and
olanzapine (such as extrapyrimidal syndrome, weight gain, and
diabetes) are also not incorporated. Because of the relatively
short amount of time in which the drugs are given in the model
and the already high disability of the disease itself, it is expected
that incorporating side effects would have only a minor influence
on model outcomes and omitting them is expected to be a
conservative approach for aripiprazole.
Costs associated with monitoring tests during atypical antipsychotic use were not incorporated. Because monitoring would
be required for all three atypical augmentation regimes compared, this would not affect incremental costs.
When implementing long-term data on the effects of aripiprazole augmentation in the model, side effects should be taken
into account. In the long run, the chronic use of antipsychotic
medication is a risk factor for developing diabetes mellitus type 2
and extensive weight gain. There is evidence, however, substantiating that aripiprazole is an exception and will have less impact
on these side effects than, for example, quetiapine or olanzapine.
Indirect costs associated with work productivity losses, which
are not uncommon in patients with MDD, are not considered by
the Turkish payer, which is why these were excluded from the
analysis. Including indirect costs, however, would likely result in
further cost savings for aripipazole augmentation. Less time is
spent in depressive episodes with this treatment, during which
patients have a higher chance of incurring work productivity
losses than during phases of remission.
Conclusions
The cost-effectiveness model described here showed that aripiprazole augmentation dominates quetiapine augmentation and
olanzapine augmentation. Taking into account the uncertainty in
all model input parameters, the probability that adjunctive
aripiprazole would be cost-effective among the three strategies
180
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 171–180
ranged between 74% and 75% for WTP values between 0 TL and
100,000 TL per QALY gained. Although atypical antipsychotics
have the same reimbursement status for augmentation treatment of MDD in Turkey, these economic findings may help
inform clinicians in their choice of antipsychotic augmentation.
Source of financial support: This work was sponsored by
Bristol Myers Squibb pharmaceuticals, Turkey. However, Pharmerit staff had complete access to all data and had final control
over the content, review, and submission of the manuscript.
R EF E R EN CE S
[1] Kessler RC, Berglund P, Demler O, et al. Lifetime prevalence and age-ofonset distributions of DSM-IV disorders in the National Comorbidity
Survey Replication. Arch Gen Psychiatry 2005;62:593–602.
[2] Kessler RC, Angermeyer M, Anthony JC, et al. Lifetime prevalence and
age-of-onset distributions of mental disorders in the World Health
Organization’s World Mental Health Survey Initiative. World Psychiatry
2007;6:168–76.
[3] Alonso J, Angermeyer MC, Bernert S, et al. Prevalence of mental
disorders in Europe: results from the European Study of the
Epidemiology of Mental Disorders (ESEMeD) project. Acta Psychiatr
Scand Suppl 2004;420:21–7.
[4] Ministry of Health. Turkey Burden of Disease Study. Ankara: Ministry of
Health, 2006.
[5] Burcusa SL, Iacono WG. Risk for recurrence in depression. Clin Psychol
Rev 2007;27:959–85.
[6] Khan A, Khan SR, Leventhal RM, Brown WA. Symptom reduction and
suicide risk in patients treated with placebo in antidepressant clinical
trials: a replication analysis of the Food and Drug Administration
Database. Int J Neuropsychopharmacol 2001;4:113–8.
[7] Murray CJ, Lopez AD. Evidence-based health policy–lessons from the
Global Burden of Disease Study. Science 1996;274:740–3.
[8] Wang H, Kindig DA, Mullahy J. Variation in Chinese population health
related quality of life: results from a EuroQol study in Beijing, China.
Qual Life Res 2005;14:119–32.
[9] Anderson I, Pilling S, Barnes A, et al. Depression: the treatment and
management of depression in adults. Leicester, UK: National
Collaborating Centre for Mental Health Commissioned by the National
Institute for Health and Care Excellence, 2009.
[10] Sobocki P, Ekman M, Agren H, et al. Health-related quality of life
measured with EQ-5D in patients treated for depression in primary
care. Value Health 2007;10:153–60.
[11] Stouthard ME, Essink-Bot ML, Bonsel G, et al. Disability Weights for
Diseases in the Netherlands. Rotterdam: Department of Public Health, 1997.
[12] Bennett KJ, Torrance GW, Boyle MH, Guscott R. Cost-utility analysis in
depression: the McSad utility measure for depression health states.
Psychiatr Serv 2000;51:1171–6.
[13] Revicki DA, Wood M. Patient-assigned health state utilities for
depression-related outcomes: differences by depression severity and
antidepressant medications. J Affect Disord 1998;48:25–36.
[14] Sapin C, Fantino B, Nowicki ML, Kind P. Usefulness of EQ-5D in
assessing health status in primary care patients with major depressive
disorder. Health Qual Life Outcomes 2004;2:20.
[15] Ministry of Health Turkey. National Burden of Disease and Cost
Effectiveness Project. Ankara, Turkey: RSHMB School of Public Health
Directorate, 2004.
[16] Sobocki P, Ekman M, Agren H, et al. The mission is remission: health
economic consequences of achieving full remission with
antidepressant treatment for depression. Int J Clin Pract 2006;60:791–8.
[17] Trivedi MH, Rush AJ, Wisniewski SR, et al. Evaluation of outcomes with
citalopram for depression using measurement-based care in STAR*D:
implications for clinical practice. Am J Psychiatry 2006;163:28–40.
[18] Trivedi MH. Treating depression to full remission. J Clin Psychiatry
2009;70:e01.
[19] Ansseau M, Demyttenaere K, Heyrman J, et al. Objective: remission of
depression in primary care. The Oreon Study. Eur Neuropsychopharmacol
2009;19:169–76.
[20] Ministry of Health Turkey. Short product summary aripiprazole.
Available from: http://www.iegm.gov.tr/Folders/KubKT/Ruhsatlý%
20Ürünler-2%20Þube%20Müdürlüðü/ABILIFY%205%20mg%20KUB%
20autism_006c814.pdf. [Accessed May 22, 2010].
[21] Taneja C, Papakostas GI, Jing Y, et al. Cost-effectiveness of adjunctive
therapy with atypical antipsychotics for acute treatment of major
depressive disorder. Ann Pharmacother 2012;46:642–9.
[22] Berman RM, Marcus RN, Swanink R, et al. The efficacy and safety of
aripiprazole as adjunctive therapy in major depressive disorder: a
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
multicenter, randomized, double-blind, placebo-controlled study. J Clin
Psychiatry 2007;68:843–53.
Berman RM, Fava M, Thase ME, et al. Aripiprazole augmentation in
major depressive disorder: a double-blind, placebo-controlled study in
patients with inadequate response to antidepressants. CNS Spectr
2009;14:197–206.
Marcus RN, McQuade RD, Carson WH, et al. The efficacy and safety of
aripiprazole as adjunctive therapy in major depressive disorder: a
second multicenter, randomized, double-blind, placebo-controlled
study. J Clin Psychopharmacol 2008;28:156–65.
Bauer M, Pretorius HW, Constant EL, et al. Extended-release quetiapine
as adjunct to an antidepressant in patients with major depressive
disorder: results of a randomized, placebo-controlled, double-blind
study. J Clin Psychiatry 2009;70:540–9.
McIntyre A, Gendron A, McIntyre A. Quetiapine adjunct to selective serotonin reuptake inhibitors or venlafaxine in patients with major depression,
comorbid anxiety, and residual depressive symptoms: a randomized,
placebo-controlled pilot study. Depress Anxiety 2007;24:487–94.
Shelton RC, Williamson DJ, Corya SA, et al. Olanzapine/fluoxetine
combination for treatment-resistant depression: a controlled study of
SSRI and nortriptyline resistance. J Clin Psychiatry 2005;66:1289–97.
Thase ME, Corya SA, Osuntokun O, et al. A randomized, double-blind
comparison of olanzapine/fluoxetine combination, olanzapine, and
fluoxetine in treatment-resistant major depressive disorder. J Clin
Psychiatry 2007;68:224–36.
Corya SA, Williamson D, Sanger TM, et al. A randomized, double-blind
comparison of olanzapine/fluoxetine combination, olanzapine,
fluoxetine, and venlafaxine in treatment-resistant depression. Depress
Anxiety 2006;23:364–72.
Nierenberg AA, Papakostas GI, Petersen T, et al. Lithium augmentation
of nortriptyline for subjects resistant to multiple antidepressants. J Clin
Psychopharmacol 2003;23:92–5.
Joffe RT, Singer W, Levitt AJ, MacDonald C. A placebo-controlled
comparison of lithium and triiodothyronine augmentation of tricyclic
antidepressants in unipolar refractory depression. Arch Gen Psychiatry
1993;50:387–93.
Lunn D, Thomas A, Best N, Spiegelhalter D. WinBUGS—a Bayesian
modelling framework: concepts, structure, and extensibility. Stat
Comput 2002;10:325–37.
Treur M, Postema R, Loze JY, et al. The comparative efficacy and safety
of adjunctive antipsychotics in major depressive disorder patients that
failed to respond to conventional antidepressants. In: ISPOR 15th
Annual International Meeting. Atlanta, May 17–19, 2010.
Rush AJ, Trivedi MH, Wisniewski SR, et al. Acute and longer-term
outcomes in depressed outpatients requiring one or several treatment
steps: a STAR*D report. Am J Psychiatry 2006;163:1905–17.
Solomon DA, Keller MB, Leon AC, et al. Multiple recurrences of major
depressive disorder. Am J Psychiatry 2000;157:229–33.
Bernal M, Haro JM, Bernert S, et al. Risk factors for suicidality in Europe:
results from the ESEMED study. J Affect Disord 2007;101:27–34.
Devrimci-Ozguven H, Sayil I. Suicide attempts in Turkey: results of the
WHO-EURO Multicentre Study on Suicidal Behaviour. Can J Psychiatry
2003;48:324–9.
Bilici M, Bekaroglu M, Hocaoglu C, et al. Incidence of completed and
attempted suicide in Trabzon, Turkey. Crisis 2002;23:3–10.
Burstrom K, Johannesson M, Diderichsen F. Swedish population healthrelated quality of life results using the EQ-5D. Qual Life Res 2001;10:621–35.
Ministry of Health, Directorate of Pharmaceuticals, Official Web Page.
Available from: http://www.iegm.gov.tr/Default.aspx?sayfa=fiyat_listesi.
[Accessed February 2, 2010].
Karamustafalioglu O, Hemels MEH, Save D, Mene S, Özmen E.
Economic evaluation of first-line treatments for depression in
Turkey: a cost-effectiveness model. In: ISPOR 10th Annual
International Meeting. Washington, DC, May 2005. Value Health
2005;(8):392.
Health implementation guideline. Turkish Republic Official Gazette
2010;27532(Suppl. 8):98.
Revicki DA, Brown RE, Palmer W, et al. Modelling the cost effectiveness
of antidepressant treatment in primary care. Pharmacoeconomics
1995;8:524–40.
Revicki DA, Brown RE, Keller MB, et al. Cost-effectiveness of newer
antidepressants compared with tricyclic antidepressants in managed
care settings. J Clin Psychiatry 1997;58:47–58.
National Institute for Health and Care Excellence. Guide in the methods of
technology appraisal (reference NO515). Available from: http://www.nice.
org.uk/media/B52/A7/TAMethodsGuideUpdatedJune2008.pdf. [Accessed
May 22, 2010].
Drummond MF, Jefferson TO. Guidelines for authors and peer
reviewers of economic submissions to the BMJ. The BMJ Economic
Evaluation Working Party. BMJ 1996;313:275–83.
Sobocki P, Ekman M, Agren H, et al. Model to assess the costeffectiveness of new treatments for depression. Int J Technol Assess
Health Care 2006;22:469–77.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 181–188
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Cost-Utility Analysis of Depot Atypical Antipsychotics for Chronic
Schizophrenia in Croatia
Vlado Jukic, MD, PhD1, Miro Jakovljevic, MD, PhD2, Igor Filipcic, MD, PhD2, Miroslav Herceg, MD, PhD1, Ante Silic, MD, PhD3,
Tatjana Tomljanovic, MD4, Roman Zilbershtein, MSc5, Rasmus C.D. Jensen, MSc6, Michiel E.H. Hemels, MSc, MBA6,
Thomas R. Einarson, PhD7,
1
University Psychiatric Hospital “Vrapce,” Zagreb, Croatia; 2Department of Psychiatry, University Hospital of Zagreb, Zagreb, Croatia; 3University Psychiatric
Hospital “Sv Ivan,” Zagreb, Croatia; 4Janssen Division of Johnson & Johnson S.E., Zagreb, Croatia; 5Pivina Consulting, Inc., Mississauga, Canada; 6Janssen Cilag,
Birkerød, Denmark; 7Leslie Dan Faculty of Pharmacy, University of Toronto, Toronto, ON, Canada
AB STR A CT
Objectives: As a nation with a developing economy, Croatia is faced
with making choices between pharmaceutical products, including
depot injectable antipsychotics. We conducted a pharmacoeconomic
analysis to determine the cost-effectiveness of atypical depots in
Croatia. Methods: A 1-year decision-analytic framework modeled
drug use. We determined the average direct cost to the Croatian
Institute for Health Insurance of using depot formulations of paliperidone palmitate long-acting injectable (PP-LAI), risperidone LAI (RISLAI), or olanzapine LAI (OLZ-LAI). An expert panel plus literaturederived clinical rates populated the core model, along with costs
adjusted to 2012 by using the Croatian consumer price index. Clinical
outcomes included quality-adjusted life-years, hospitalization rates,
emergency room treatment rates, and relapse days. Robustness of
results was examined with one-way sensitivity analyses on important
inputs; overall, all inputs were varied over 10,000 simulations in a
Monte Carlo analysis. Results: Costs (quality-adjusted life-years) per
patient were €5061 (0.817) for PP-LAI, €5168 (0.807) for RIS-LAI, and
€6410 (0.812) for OLZ-LAI. PP-LAI had the fewest relapse days,
emergency room visits, and hospitalizations. Results were sensitive
against RIS-LAI with respect to drug costs and adherence rates, but
were generally robust overall, dominating OLZ-LAI in 77.3% and RISLAI in 56.8% of the simulations. Conclusions: PP-LAI dominated the
other drugs because it had the lowest cost and best clinical outcomes.
Compared with depots of olanzapine and risperidone and oral
olanzapine, PP-LAI was the cost-effective atypical LAI for treating
chronic schizophrenia in Croatia. Using depot paliperidone in place of
either olanzapine or risperidone would reduce the overall costs of
caring for these patients.
Introduction
Problems in drug use, especially nonadherence and polypharmacy, have exacerbated the situation [11,12]. As a result, there
have been increasing numbers of hospitalizations [13]. Harvey
et al. [14] noted that hospitalization consumed the largest
proportion of total health care costs (430%) in Croatia. In
addition, the plan to reintegrate persons with schizophrenia into
the community has not been entirely successful [15]. All these
factors have served to increase costs for the health system;
however, its financial resources are insufficient to cover all the
demands. Therefore, cost-effective approaches are needed to
maintain and improve the treatment of persons with chronic
schizophrenia.
In 2002, the Croatian Institute for Health Insurance funded a
project to address rational drug use [14]. A main component was
determining how savings could be made by incorporating pharmacoeconomic principles into the selection and purchase of
drugs on the Croatian formulary. They observed that “there was
Croatia is a country with a developing economy whose health care
system has been developing over some time [1,2]. The Croatian
Institute for Health Insurance [3] was established in 1993 to manage
the health system that is provided for all citizens. This system is
based on a national health insurance model, with compulsory
contributions for all employed persons and employers, with copayments for drugs and services. Noninsured services are paid either
out of pocket or through additional voluntary health insurance.
In 1998, a law was passed that guaranteed treatment and also
safeguarded the personal rights for persons with schizophrenia
requiring involuntary hospitalization [4,5]. Since that time, there
has been an increase in both the availability and use of antipsychotics [6,7]. Utilization patterns have changed, with a shift
from the older, less expensive first-generation drugs to newer
and more costly atypical antipsychotics [6–10].
Keywords: Croatia, long-acting injectable, paliperidone palmitate,
pharmacoeconomic analysis, risperidone, schizophrenia.
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Thomas R. Einarson, Leslie Dan Faculty of Pharmacy, 144 College Street, Room 674, University of Toronto,
Toronto, ON, Canada M5V 3M8.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.008
182
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 181–188
Fig. 1 – Model used for the pharmacoeconomic analysis. LAI, long-acting injectable; OLZ-LAI, olanzapine pamoate long-acting
injectable; PP-LAI, paliperidone palmitate long-acting injectable.
a lack of information about what constituted cost-effective treatment appropriate to the Croatian economic situation.” This
observation highlights the need for evidence-based information
on the cost-effectiveness of psychopharmaceuticals.
At the same time, the rights and dignity of these people must
be respected. Two critical aspects are image [16] and quality of
life [17,18]. Nawková et al. [19] assessed articles in the lay press in
Croatia describing mental health and found that 40 out of the 75
(53%) articles portrayed a negative image. Mentally ill persons
were mostly presented as dangerous and involved in aggressive
crimes such as homicide (49%) or physical assaults (31%). MartićBiocina and Barić [12] also identified dissatisfaction with the role
of the media in that respect. They also found a high level of
stigma toward people with schizophrenia and that it correlated
with medication nonadherence and hospitalizations. The same
was found with quality-of-life issues [18]. These authors also
reported that the atypical antipsychotics were superior to traditional drugs with respect to increasing quality of life in persons
with chronic schizophrenia. Jukić et al. [20] suggested that their
side-effect profile may be responsible for improved quality of life.
Depot forms of antipsychotic drugs were developed to at least
partially address issues of nonadherence [21]. In the past decade,
long-acting injectable (LAI) formulations of atypical agents have
been marketed to fill a perceived need. Risperidone LAI (RIS-LAI)
was the first such drug [22], and has recently been joined
by olanzapine LAI (OLZ-LAI) [23] and paliperidone LAI (PP-LAI)
[24]. In another country undergoing economic change, a
pharmacoeconomic analysis found that PP-LAI was costeffective when compared with RIS-LAI [25]. It is not currently
known whether the outcomes would be similar in this country.
We therefore undertook this research to assess the cost-utility of
PP-LAI compared with other LAIs in Croatia from the point of
view of the Croatian Institute for Health Insurance.
Methods
Target Population
We examined the use of atypical LAIs in persons with stable
chronic schizophrenia but who had a history of relapses and
hospitalizations. They have been referred to as “revolving door”
patients who are difficult to treat and have problems with
adherence to prescribed medications [26,27]. Consequently, they
impose a very large burden on health care resources.
Drugs Analyzed
The drug of primary interest was PP-LAI. Comparison drugs
included the other atypical depots (i.e., RIS-LAI and OLZ-LAI).
According to the product summaries of the European Medicines
Agency, PP-LAI (Xeplion) can be dosed monthly [28], OLZ-LAI
(Zypadhera) is administered every 2 or 4 weeks [29], and RIS-LAI
(Risperdal Consta) requires biweekly injections [30].
183
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 181–188
Table 1 – Clinical inputs into the model.
LAI
antipsychotic
Clinical input
(amount/rate)
Olanzapine
Dose to
initialize
therapy
Maintenance
dose
Dose for acute
relapse
Adherence
Stable disease
ER visits
Hospitalization
rate
Dose to
initialize
therapy
Maintenance
dose
Dose for acute
relapse
Adherence
Paliperidone
Risperidone
Rate
Source
300 mg q2 weeks 3
injections
European Medicines Agency [29]
432 mg q4 weeks
Kane et al. [34]
473 mg q4 weeks
Lauriello et al. [35]
0.803
0.793
0.062
0.145
Ascher-Svanum et al. [36]
Kane et al. [34]
Kane et al. [34]
Calculation [1 – rates of (ER visits þ stable disease)]
150 mg week 1, 100 mg
week 2, then 82.8 mg
every 4 wk
69.3 mg monthly
European Medicines Agency [28], Hough et al. [37]
Average from Gopal et al. [38] and Fleischhacker et al. [39]
84.9 mg monthly
Gopal et al. [40], Pandina et al. [41], Hough et al. [42], Nasrallah et al.
[43], Pandina et al. [44]
RIS-LAI rate from Olivares et al. [45] adjusted via Mehnert and Diels
[46]
Calculation [1 – rates of (ER visits þ hospitalization)]
Hospital rate ER:hospital ratio from Ascher-Svanum et al. [36]
Hough et al. [37], Gopal et al. [38]
0.872
Stable disease
ER visits
Hospitalization
rate
Maintenance
dose
Dose for acute
relapse
0.803
0.059
0.138
40.3 mg biweekly
Fleischhacker et al. [39], Kissling et al. [47], Lee et al. [48],
Lindenmayer et al. [49], Olivares et al. [50]
Prorated from PP-LAI dose; similar to doses used by Kane et al. [22],
Chue et al. [51], Eerdekens et al. [52] who used 58 mg, but 50 mg is
the maximum allowable dose [30]
Olivares et al. [45]
Calculation [1 – rates of (ER visits þ hospitalization)]
Hospital rate ER:hospital ratio from Ascher-Svanum et al. [36]
Olivares et al. [50], Weiden and Olfson [53]
50 mg biweekly
Adherence
Stable disease
ER visits
Hospitalization
rate
0.823
0.763
0.071
0.166
ER, emergency room; LAI, long-acting injectable; PP, paliperidone palmitate; RIS, risperidone microspheres.
an average patient having chronic relapsing schizophrenia but
whose disease is currently stabilized. Because of adherence
problems, patients are maintained on standard doses of depot
antipsychotics. They may be either adherent or nonadherent, as
Model and base case
We adapted a previously validated decision tree [25] for use in
Croatia, using input from clinical and administrative experts.
Figure 1 depicts the model. To begin the analysis, we start with
Table 2 – Cost inputs (2012€).
Resource
Item
Unit
mg
mg
mg
mg
mg
mg
Cost (€)
Drugs
Olanzapine depot
Risperidone depot
Paliperidone depot
Olanzapine tablets
Risperidone tablets
Clozapine tablets
3.50
3.50
3.76
0.21
0.20
0.0036
Medical
Primary care physician
Psychiatrist outpatient visit
Psychiatric nurse
1 visit
1 visit
1h
16.62
14.10
7.45
Hospital
Emergency room
Hospital bed acute care
Hospital long-term bed
Day care visit
1 visit
1d
1d
1d
230.32
96.81
33.51
26.20
Source
Croatian
Croatian
Croatian
Croatian
Croatian
Croatian
Official
Official
Official
Official
Official
Official
Gazette
Gazette
Gazette
Gazette
Gazette
Gazette
Local medical price list
Vrapče hospital price list
Vrapče hospital price list
Vrapče
Vrapče
Vrapče
Vrapče
hospital
hospital
hospital
hospital
price
price
price
price
list
list
list
list
[58]
[58]
[58]
[58]
[58]
[58]
184
Economic
conclusion
Dominant‡
Dominated
Dominated
0.010
0.005
LAI, long-acting injectable; PP, paliperidone palmitate; RIS, risperidone microspheres; QALYs, quality-adjusted life-years.
All costs are in 2012 euros.
†
Because of dominance, comparisons are against PP-LAI.
‡
Has lowest cost and highest QALYs and therefore is the drug of choice.
107
1349
0.817
0.807
0.812
0.252
0.305
0.280
329.3
322.8
325.5
PP-LAI
RIS-LAI
OLZ-LAI
5061
5168
6410
0.129
0.148
0.143
Total QALYs
per patient
Hospitalizations
Days in
remission
Emergency
room visits
per published rates and expert opinion. Some will continue in the
stable state, while the rest will relapse. All relapsers would be
seen at the emergency room, and the more severe cases would be
admitted to the acute psychiatric unit. Those unable or unwilling
to tolerate the initial treatment would be switched. Those
discontinuing PP-LAI or RIS-LAI would then receive OLZ-LAI,
and those switching from OLZ-LAI would receive PP-LAI; discontinuers of oral OLZ would receive RIS-LAI. Patients who failed
two different drugs would be given clozapine, in accordance with
National Institute for Health and Clinical Excellence guidelines
[31] and local practice using doses reported in the literature
[32,33].
Clinical Inputs
Total cost per
patient (€)*
Drug
Table 3 – Clinical and economic outputs from the decision tree model analysis for Croatia.
Incremental cost per
patient (€)†
Incremental QALYs
per patient†
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 181–188
Table 1 lists the clinical inputs used in this mode, by drug, as
well as the sources of information [22,28–30,34–53]. The doses
of drugs used were derived from randomized clinical trials
and long-term studies published in the literature. Where data
were presented for nonadherent patients, we extracted rates
from the published articles. In other cases, we used either
the rates for placebo in trials (e.g., PP-LAI; OLZ-LAI assumed
equal) or we calculated rates by using the ratio of adherent:
nonadherent patients from Weiden and Olfson [53] (e.g., RISLAI).
With each drug, adjunct therapy was added, in accordance
with typical clinical trials. Gopal et al. [38] indicated that PP-LAI
was augmented with oral risperidone in a dose of 6.8 mg/d for
30.5 days. In a similar trial, Möller et al. [54] reported that 22%
of the patients receiving RIS-LAI required oral supplementation
with 3.2 mg/d for 43 days. Ascher-Svanum et al. [55] noted
that OLZ-LAI required 10.8 mg/d for 31 days in 21% of the
patients.
Some economic analyses have used standard doses such as
defined daily doses (DDDs) [56]; however, DDDs reflect the
average dose when used for the most common indication. It
should be remembered that we are dealing with revolving door
patients who comprise only a subset of patients who represent
the extreme of the spectrum. Thus, we feel that DDDs would
underestimate the doses used in actual practice when managing
such problematic patients.
In calculating adherence rates, we used experience from large
observational studies. The rate for RIS-LAI was taken from a large
patient registry (n ¼ 1648) of patients with longstanding schizophrenia [45]. Because there was insufficient long-term experience
with PP-LAI, that rate was adjusted by using results from the
study by Mehnert and Diels [46]. They compared adherence
between RIS-LAI when administered monthly and twice weekly,
finding a minimum of 5.1% increase in adherence with monthly
injections. That factor was applied to PP-LAI, which is administered monthly and is a metabolite of RIS-LAI, thereby having the
same adverse-event profile. For OLZ-LAI, we used the rate from a
large cohort (n ¼ 1906). Even though that rate was with oral drugs
(which normally have lower adherence rates than do depots), we
used that value of 80.2% because it was higher than the 72.7%
rate in 931 patients found with the depot form by AscherSvanum et al. [57], and was quite similar to the rates of other
depot atypicals.
Cost Inputs
We considered all direct costs from the viewpoint of the National
Health Service of Croatia, as presented in Table 2 [58] (local
medical price list and Vrapče hospital price list). We did not
include indirect costs such as time lost from work, because
very few of these people participate fully in the workforce.
A multicountry study in Europe reported that less than 10% of
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 181–188
185
Fig. 2 – Cloud diagram from the probabilistic sensitivity analysis comparing incremental costs (Y axis) and QALYs (X axis)
between (A) PP-LAI and RIS-LAI and (B) PP-LAI and OLZ-LAI. OLZ-LAI, olanzapine pamoate long-acting injectable; PP-LAI,
paliperidone palmitate long-acting injectable; QALY, quality-adjusted life-year; RIS-LAI, risperidone long-acting injectable.
3996 persons with schizophrenia were employed full-time and
another 12.3% worked part-time [59]. We did not apply discounting because the analytic time horizon was 1 year. Prices
were taken from current lists or from the literature, and then
inflated to 2011 euros by using the consumer price index for
Croatia [60].
Utilities
Utilities for the analysis were obtained from the literature;
values obtained were simply averaged [61–65]. Stable disease
had a utility of 0.890; an exacerbation requiring outpatient
treatment at the hospital emergency room had a utility of
0.659 for emergency room exacerbation and 0.490 for
hospitalization.
Analysis and Outputs
No official guidelines currently exist for pharmacoeconomic
analyses in Croatia. We therefore used a standard approach that
had been used in previous analyses in Europe [25]. The decision
tree produced expected outcomes for the average patient treated
with average doses of each drug. These outcomes included the
cost per patient treated, measured in 2012 euros, as well as
numbers of hospitalizations, emergency room visits, days with
stable disease, and quality-adjusted life-years (QALYs) for each
drug. The economic outcome of prime interest was the incremental cost per QALY.
We explored the effect of variations in input values on outputs by applying one-way sensitivity analyses on all important
inputs such as costs and clinical rates. Break-even analysis
identified the points where outcomes changed qualitatively. We
186
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 181–188
also conducted a set of pairwise probabilistic analyses by using
10,000 Monte Carlo simulations on all inputs and standard
distributions (i.e., beta for rates and gamma for costs) [66].
Proportions of incremental cost-effectiveness ratios falling into
each of the four major quadrants (cost vs. QALYs) were calculated
and compared.
Results
Results of the base-case analysis appear in Table 3. PP-LAI had
the lowest overall cost to treat one patient for 1 year (€5061),
followed by RIS-LAI (€5168), with OLZ-LAI costing the most
(€6410). Clinical outcomes were also better in all cases for PPLAI; it had the most days in remission and the fewest hospitalizations and emergency room visits. Also, it was associated
with the highest QALY score, but differences for this outcome
were not large. Because its cost was lowest and QALYs (and
other beneficial clinical outcomes) highest, it dominated the
other drugs. That is, it should be considered the preferred
choice.
In one-way sensitivity analyses, PP-LAI was not sensitive to
changes in cost relative to OLZ-LAI. Its cost would need to
increase 64% or that of OLZ-LAI decrease by 54% to lose its
dominance. However, it would not dominate RIS-LAI with a 4%
increase in the cost of PP-LAI or a 4.4% decrease in the cost of RISLAI. If the unit cost of PP-LAI were equal to that of RIS-LAI, the
expected cost/patient would decline to €4756. Results were
sensitive to reasonable changes in adherence rate (⫾10% for
OLZ-LAI and ⫾18% for RIS-LAI). Hospitalization rates were not
sensitive.
Figure 2A,B depicts results from the probabilistic sensitivity
analyses. PP-LAI dominated OLZ-LAI in about 77.3% of the
simulations and RIS-LAI in 56.8% of the simulations. It was
cost-effective (i.e., incremental cost-effectiveness ratios o
€30,000) compared with RIS-LAI in another 37% of the trials
(overall 93%). However, PP-LAI was dominated in 1.3% of the
20,000 iterations, in total.
Discussion
A search of the literature could find no examples of pharmacoeconomic analyses that examined the pharmacotherapy of schizophrenia in Croatia. Therefore, we believe that this is the first
such analysis. Because decision makers and health care providers
are being faced with increasing demands from patients and their
advocates without a corresponding increase in revenues, they
must take advantage of these quantitative approaches to aid in
selecting what to fund. Relying solely on acquisition prices of
drugs, services, or other products can be misleading because all
factors impacting the choice are not being considered. Because of
enhanced efficacy, a drug with a higher price may be the best
choice if it prevents the consumption of other resources, such as
hospitalization.
In this analysis, PP-LAI dominated the other available atypical
LAIs. Results could change against RIS-LAI with changes in cost
or against OLZ-LAI with changes in adherence. The overall
probabilistic sensitivity analyses, however, did indicate that PPLAI would be the drug of choice in the majority of cases.
In addition to the clinical and economic advantages, there is
an advantage for PP-LAI with respect to convenience. This drug
may be administered monthly, while its nearest competitor, RISLAI, must be given every 2 weeks. Monthly dosing would seem to
be the preferable situation for both the patient and the busy
practitioner.
This analysis has some limitations, which should be noted.
Rates of adherence and hospitalization were taken from the
literature and were assumed to apply as well in this country.
We considered only persons with chronic schizophrenia in a
stable state. Those who are hospitalized or experiencing an acute
exacerbation of symptoms would require more aggressive treatment; therefore, costs and outcomes might vary.
We did not consider the treatment of adverse events other
than postinjection syndrome in this analysis for a number of
reasons. First, these events are quite common in all antipsychotic
drugs. Many of these problems can be managed by reducing the
dose or changing to another drug. Also, these patients are
required to visit their physician or other practitioner (e.g.,
psychiatric nurse or psychologist) on a regular basis, and so they
would not likely incur extra visits because of adverse events. In
addition, many require treatment with drugs (e.g., anticholinergics) that are very inexpensive and add little to the overall cost of
care, especially on a comparative basis. Finally, reports from
official agencies have concluded that adverse events associated
with these drugs have little appreciable impact overall [67,68].
Conclusions
In this analysis, we found that PP-LAI was the dominant choice
for treating chronic relapsing schizophrenia in Croatia. Its higher
acquisition cost was more than offset by reductions in other
health care areas, such as decreased hospitalizations, visits to
emergency room, and visits to other health care practitioners.
Results were sensitive to minor changes in adherence rates.
Source of financial support: This study was funded by Janssen
A/S, Beerse, Belgium. Dr. Tomljanovic, Mr. Rasmussen, and Mr.
Hemels are all employees of Janssen. Dr. Einarson received
funding for this research. Mr. Zilbershtein is an employee of
Pivina Conculting, Inc., who received funding for this research.
R EF E R EN C ES
[1] European Observatory on Health Care Systems. Health Care Systems in
Transition: Croatia (2nd ed.). Copenhagen: WHO Regional Office for
Europe, 1999. Available from: http://www.euro.who.int/__data/assets/
pdf_file/0018/75150/E68394.pdf. [Accessed March 16, 2012].
[2] Mastilica M, Kušec S. Croatian healthcare system in transition, from the
perspective of users. BMJ 2005;331:223–7.
[3] Croatian Institute of Health Insurance. Available from: http://www.
hzzo-net.hr. [Accessed May 3, 2012].
[4] Jukić V, Goreta M, Kozumplik O, et al. Implementation of first Croatian
Law on Protection of Persons with Mental Disorders. Coll Antropol
2005;29:543–9.
[5] Kozumplik O, Jukić V, Goreta M. Involuntary hospitalizations of
patients with mental disorders in Vrapce Psychiatric Hospital: five
years of implementation of the first Croatian law on protection of
persons with mental disorders. Croat Med J 2003;44:601–5.
[6] Jukić V, Herceg M, Savić A. Availability of psychiatric medications to
Croatian healthcare users and the influence of availability of atypical
antipsychotics on psychiatric hospital morbidity. Psychiatr Danub
2011;23:320–4.
[7] Jukić V, Savić A, Herceg M. Availability of new psychiatric medications,
especially antipsychotics, in context of patient rights and
destigmatization of psychiatric patients. Psychiatr Danub
2011;23:316–9.
[8] Jukić V, Herceg M, Brecić P, et al. Dynamic in prescribing antipsychotic
drugs during five year period (2001–2005) in the Psychiatric Hospital
Vrapce, Zagreb, Croatia. Coll Antropol 2008;32(Suppl 1):211–3.
[9] Štimac D, Čulig J. Outpatient utilization of psychopharmaceuticals in
the city of Zagreb 2001–2006. Psychiatr Danub 2009;21:56–64.
[10] Štimac D, Vukusić I, Čulig J, et al. Outpatient utilization of
psychopharmaceuticals: comparison between Croatia and
Scandinavian countries (2001–2003). Coll Antropol 2009;33:237–43.
[11] Loga-Zec S, Loga S. Polypharmacy in the treatment of schizophrenic
patients in three university centers in the Federation of Bosnia and
Herzegovina (F/BH). Psychiatr Danub 2011;23:60–3.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 181–188
[12] Martić-Biocina S, Barić V. Factors important for compliance among
Croatian patients suffering from schizophrenia—how to improve
psychiatric services in Croatia? Psychiatr Danub 2005;17:36–41.
[13] Bilić P, Ivanis A, Vidović D, Jukić V. Changing the structure of the
hospitalized patients at the Psychiatric Clinic Vrapce. Srp Arh Celok Lek
2011;139(Suppl 1):33–5.
[14] Harvey K, Kalanj K, Stevanović R. Croatian pharmaceutical sector
reform project: rational drug use. Croat Med J 2004;45:611–9.
[15] Jukić V. (Re)Integration of mental patients—mixed media messages.
Psychiatr Danub 2008;20:433–6.
[16] Bilić B, Georgaca E. Representations of “mental illness” in Serbian
newspapers: a critical discourse analysis. Qual Res Psychol
2007;4:167–86.
[17] Pesek MB. Therapy and quality of life of patients with psychosis.
Psychiatr Danub 2009;21(Suppl. 1):146–8.
[18] Jašović-Gašić M, Lačković M, Dunjić-Kostić B, et al. Critical review of
studies on quality of life in psychiatric patients published in Serbian
medical journals from 2000 to 2009. Psychiatr Danub 2010;22:488–94.
[19] Nawková L, Nawka A, Adámková T, et al. The picture of mental health/
illness in the printed media in three central European countries. J
Health Commun 2012;17:22–40.
[20] Jukić V, Barić V, Culav-Sumić J, et al. The impact of novel antipsychotic
drugs on quality of life among people suffering from schizophrenia.
Coll Antropol 2003;27(Suppl. 1):119–24.
[21] Nasrallah H. The case for long-acting antipsychotic agents in the postCATIE era. Acta Psychiatr Scand 2007;115:60–267.
[22] Kane J, Eerdekens M, Lindenmayer J-P, et al. Long-acting injectable
risperidone: efficacy and safety of the first long-acting atypical
antipsychotic. Am J Psychiatry 2003;60:1125–32.
[23] European Medicines Agency. Assessment report for Zypadhera. Doc.
Ref EMEA/608654/2008. Procedure No. EMEA/H/C/000890. 2008:1–63.
Available from: http://www.ema.europa.eu/docs/en_GB/
document_library/EPAR_-_Public_assessment_report/human/000890/
WC500054428.pdf. [Accessed March 3, 2012].
[24] European Medicines Agency. Xeplions opinion 17-12-2010. Available
from: http://www.ema.europa.eu/ema/index.jsp?curl=pages/
medicines/human/medicines/002105/smops/Positive/
human_smop_000162.jsp&murl=menus/medicines/medicines.
[Accessed March 3, 2012].
[25] Einarson TR, Geitona M, Chaidemenos A, et al. Pharmacoeconomic
analysis of paliperidone palmitate for treating schizophrenia in Greece.
Ann Gen Psychiatry 2012;11:18.
[26] Haywood TW, Kravitz HM, Grossman LS, et al. Predicting the “revolving
door” phenomenon among patients with schizophrenic, schizoaffective,
and affective disorders. Am J Psychiatry 1995;152:856–61.
[27] Weiden P, Glazer W. Assessment and treatment selection for “revolving
door” inpatients with schizophrenia. Psychiatr Q 1997;68:377–92.
[28] European Medicines Agency. Xeplions summary of product
characteristics. Available from: http://www.ema.europa.eu/docs/
en_GB/document_library/EPAR_-_Product_Information/human/002105/
WC500103317.pdf. [Accessed March 3, 2012].
[29] European Medicines Agency. Zypadheras summary of product
characteristics. Available from: http://www.ema.europa.eu/docs/
en_GB/document_library/EPAR_-_Product_Information/human/000890/
WC500054429.pdf. [Accessed March 3, 2012].
[30] European Medicines Agency. Risperdal Constas summary of product
characteristics. Available from: http://www.ema.europa.eu/docs/
en_GB/document_library/Referrals_document/Risperdal_Consta_30/
WC500008170.pdf. [Accessed March 3, 2012].
[31] National Institute for Health and Care Excellence. Schizophrenia: core
interventions in the treatment and management of schizophrenia in
adults in primary and secondary care. NICE Clinical Guideline 82
(update). Available from: www.nice.org.uk. [Accessed April 4, 2012].
[32] Simonsen E, Friis S, Opjordsmoen S, et al. Early identification of nonremission in first-episode psychosis in a two-year outcome study. Acta
Psychiatr Scand 2010;122:375–83.
[33] Wahlbeck K, Cheine M, Essali A, Adams C. Evidence of clozapine’s
effectiveness in schizophrenia: a systematic review and meta-analysis
of randomized trials. Am J Psychiatry 1999;156:990–9.
[34] Kane J, Detke H, Naber D, et al. Olanzapine long-acting injection: a 24week, randomized, double-blind trial of maintenance treatment in
patients with schizophrenia. Am J Psychiatry 2010;167:181–9.
[35] Lauriello J, Lambert T, Andersen S, et al. An 8-week, double-blind,
randomized, placebo controlled study of olanzapine long-acting
injection in acutely ill patients with schizophrenia. J Clin Psychiatry
2008;69:790–9.
[36] Ascher-Svanum H, Faries D, Zhu B, et al. Medication adherence and
long-term functional outcomes in the treatment of schizophrenia in
usual care. J Clin Psychiatry 2006;67:453–60.
[37] Hough D, Gopal S, Vijapurkar U, et al. Paliperidone palmitate
maintenance treatment in delaying the time-to-relapse in patients
with schizophrenia: a randomized, double-blind, placebo-controlled
study. Schiz Res 2010;116:107–17.
187
[38] Gopal S, Vijapurkar U, Lim P, et al. A 52-week open-label study of the
safety and tolerability of paliperidone palmitate in patients with
schizophrenia. J Psychopharmacol 2010;25:685–97.
[39] Fleischhacker W, Gopal S, Lane R, et al. A randomized trial of
paliperidone palmitate and risperidone long-acting injectable in
schizophrenia. Int J Neuropsychopharmacol 2011: [Epub ahead of
print].
[40] Gopal S, Hough D, Xu H, et al. Efficacy and safety of paliperidone
palmitate in adult patients with acutely symptomatic schizophrenia: a
randomized, double-blind, placebo-controlled, dose-response study.
Int Clin Psychopharmacol 2010;25:247–56.
[41] Pandina GJ, Lane R, Gopal S, et al. A double-blind study of paliperidone
palmitate and risperidone long-acting injectable in adults with
schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry
2011;35:218–26.
[42] Hough D, Lindenmayer J-P, Gopal S, et al. Safety and tolerability of
deltoid and gluteal injections of paliperidone palmitate in
schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry
2009;33:1022–31.
[43] Nasrallah H, Gopal S, Gassmann-Mayer C, et al. A controlled, evidencebased trial of paliperidone palmitate, a long-acting injectable
antipsychotic, in schizophrenia. Neuropsychopharmacology
2010;35:2072–82.
[44] Pandina GJ, Lindenmayer J-P, Lull JM, et al. A randomized, placebocontrolled study to assess the efficacy and safety of 3 doses of
paliperidone palmitate in adults with acutely exacerbated
schizophrenia. J Clin Psychiatry 2010;30:235–44.
[45] Olivares J, Peuskens J, Pecanek J, et al. Clinical and resource-use
outcomes of risperidone long-acting injection in recent and long-term
diagnosed schizophrenia patients: results from a multinational
electronic registry. Curr Med Res Opin 2009;25:2197–206.
[46] Mehnert A, Diels J. Impact of administration interval on treatment
retention with long-acting antipsychotics in schizophrenia. Presented
at the Tenth Workshop on Costs and Assessment in Psychiatry-Mental
Health Policy and Economics, March 25–27, 2011, Venice, Italy.
[47] Kissling W, Heres S, Lloyd K, et al. Direct transition to long-acting
risperidone—analysis of long-term efficacy. J Psychopharmacol
2005;19:15–21.
[48] Lee M, Ko Y, Lee S, et al. Long-term treatment with long-acting
risperidone in Korean patients with schizophrenia. Hum
Psychopharmacol 2006;21:399–407.
[49] Lindenmayer J-P, Khan A, Eerdekens M, et al. Long-term safety and
tolerability of long-acting injectable risperidone in patients with
schizophrenia or schizoaffective disorder. Eur Neuropsychopharmacol
2007;17:138–44.
[50] Olivares JM, Rodrigues-Morales A, Diels J, et al. Long-term outcomes in
patients with schizophrenia treated with risperidone long-acting
injection or oral antipsychotics in Spain: results from the electronic
Schizophrenia Treatment Adherence Registry (e-STAR). Eur Psychiatry
2009;24:287–96.
[51] Chue P, Eerdekens M, Augustyns I, et al. Comparative efficacy and
safety of long-acting risperidone and risperidone oral tablets. Eur
Neuropsychopharmacol 2005;15:111–7.
[52] Eerdekens M, Van Hove I, Remmerie B, Mannaert E. Pharmacokinetics
and tolerability of long-acting risperidone in schizophrenia. Schizophr
Res 2004;70:91–100.
[53] Weiden PJ, Olfson M. Cost of relapse in schizophrenia. Schiz Bull
1995;21:419–29.
[54] Möller HJ, Llorca PM, Sacchetti E, et al. Efficacy and safety of direct
transition to risperidone long-acting injectable in patients treated with
various antipsychotic therapies. Int Clin Psychopharmacol
2005;20:121–30.
[55] Ascher-Svanum H, Peng X, Montgomery W, et al. Assessing the
infrequent oral supplementation of olanzapine long-acting injection in
the treatment of schizophrenia. Eur Psychiatry 2011;26:313–9.
[56] ATC/DDD Index 2011. Available from: http://www.whocc.no/
atc_ddd_index/. [Accessed May 3, 2012].
[57] Ascher-Svanum H, Montgomery WS, McDonnell DP, et al. Treatmentcompletion rates with olanzapine long-acting injection versus
risperidone long-acting injection in a 12-month, open-label treatment
of schizophrenia: indirect, exploratory comparisons. Int J Gen Med
2012;5:391–8.
[58] Croatian Institute of Health Insurance: List of Drugs. Official Gazette
No. 69, 2010; No. 54, 2011 and No. 48 2012.
[59] Papageorgiou G, Cañas F, Zink M, Rossi A. Country differences in
patient characteristics and treatment in schizophrenia: data from a
physician-based survey in Europe. Eur Psychiatry 2011;26(Suppl.
1):17–28.
[60] Consumer price index for Croatia. Available from: http://www.
indexmundi.com/facts/croatia/consumer-price-index. [Accessed May 3,
2012].
188
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 181–188
[61] Briggs A, Wild D, Lees M, et al. Impact of schizophrenia and
schizophrenia treatment-related adverse events on quality of life:
direct utility elicitation. Health Qual Life Outcomes 2008;6:105.
[62] Cummins C, Stevens A, Kisely S. The Use of Olanzapine as a First and
Second Choice Treatment in Schizophrenia. Birmingham, UK:
Department of Public Health & Epidemiology, University of
Birmingham, 1998. A West Midlands Development and Evaluation
Committee Report.
[63] Lenert L, Sturley A, Rapaport M, et al. Public preferences for health
states with schizophrenia and a mapping function to estimate utilities
from positive and negative symptom scale scores. Schiz Res
2004;71:155–65.
[64] Oh P, Lanctôt K, Mittmann N, et al. Cost-utility of risperidone compared
with standard conventional antipsychotics in chronic schizophrenia. J
Med Econ 2001;4:137–56.
[65] Revicki D, Shakespeare A, Kind P. Preferences for schizophrenia-related
health states: a comparison of patients, caregivers and psychiatrists.
Int Clin Psychopharmacol 1996;11:101–8.
[66] Johnson ML, Crown W, Martin BC, et al. Good research practices for
comparative effectiveness research: analytic methods to improve
causal inference from nonrandomized studies of treatment effects
using secondary data sources: the ISPOR Good Research Practices for
Retrospective Database Analysis Task Force report—part III. Value
Health 2009;12:1062–73.
[67] All Wales Medicines Strategy Group. Final appraisal report: olanzapine
depot (ZypAdheras), Lilly UK. Advice No: 1510 – October 2010.
Available from: http://www.wales.nhs.uk/sites3/Documents/371/
olanzapine%20depot%20(ZypAdhera)%20schizophrenia.pdf. [Accessed
July 7, 2012].
[68] Farahati F, Boucher M, Moulton K, et al. Atypical Antipsychotic
Monotherapy for Schizophrenia: Clinical Review and Economic
Evaluation of First Year of Treatment (Technology report number 91).
Ottawa, Canada: Canadian Agency for Drugs and Technologies in
Health, 2007.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 189–198
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Cost-Utility Analysis of Pharmaceutical Care Intervention Versus Usual
Care in Management of Nigerian Patients with Type 2 Diabetes
Maxwell O. Adibe, BPharm, MPharm, PhD1,2,*, Cletus N. Aguwa, PharmD1,2, Chinwe V. Ukwe, BPharm, MPharm, PhD1,2
1
Department of Clinical Pharmacy and Pharmacy Management, University of Nigeria, Nsukka, Enugu, Nigeria; 2Pharmacotherapeutic Group, Department of
Clinical Pharmacy and Pharmacy Management, University of Nigeria, Nsukka, Enugu, Nigeria
AB STR A CT
Objective: To assess the cost-effectiveness of pharmaceutical care
(PC) intervention versus usual care (UC) in the management of type 2
diabetes. Methods: This study was a randomized, controlled study
with a 12-month patient follow-up in two Nigerian tertiary hospitals.
One hundred and ten patients were randomly assigned to each of the
“intervention” (PC) and the “control” (UC) groups. Patients in the UC
group received the usual/conventional care offered by the hospitals.
Patients in the PC group received UC and PC in the form of structural
self-care education and training for 12 months. The economic evaluation was based on patients’ perspective. Costs of management of
individual complications were calculated from activities involved in
their management by using activity-based costing. The impact of the
interventions on quality of life was estimated by using the HUI23S4EN.40Q (Mark index 3) questionnaire. The primary outcomes were
incremental cost-utility ratio and net monetary benefit. An intentionto-treat approach was used. Two-sample comparisons were made by
using Student’s t tests for normally distributed variables data at
baseline, 6 months, and 12 months. Comparisons of proportions were
done by using the chi-square test. Results: The PC intervention led to
incremental cost and effect of Nigerian naira (NGN) 10,623 ($69) and
0.12 quality-adjusted life-year (QALY) gained, respectively, with an
associated incremental cost-utility ratio of NGN 88,525 ($571) per
QALY gained. In the cost-effectiveness acceptability curve, the probability that PC was more cost-effective than UC was 95% at the NGN
250,000 ($1613) per QALY gained threshold and 52% at the NGN 88,600
($572) per QALY gained threshold. Conclusions: The PC intervention
was very cost-effective among patients with type 2 diabetes at the
NGN 88,525 ($571.13) per QALY gained threshold, although considerable uncertainty surrounds these estimates.
Introduction
using QALYs as the principal measure of outcome, often termed
cost-utility studies, have become increasingly popular in the
literature and have also been adopted by a number of health
technology assessment agencies as the methodology of choice
[1].
Cost-utility analysis was developed to help decision makers
compare the value of alternative interventions that have very
different health benefits, and it facilitates these comparisons
without recourse to placing monetary values on different health
states. Cost-utility analysis specifies what value is attached to
specific health states, and thus increasingly facilitates the transparency of resource allocation processes [2].
Cost-utility analysis was developed to address the problem of
conventional cost-effectiveness analysis, which did not allow
decision makers to compare the value of interventions for different health problems. The utilities can now be obtained from
standardized and validated health status instruments, making
the evidence required to inform cost-utility analysis relatively
Analytic techniques used for economic evaluation in health care,
for example, cost-benefit analysis, cost-effectiveness analysis,
and cost-consequences analysis, are designed to compare alternative courses of action in terms of costs and outcomes. The
choice of the technique depends on the decision the health
economists intend to influence. Quality-adjusted life-years
(QALYs) measure health as a combination of the duration of life
and the health-related quality of life [1]. The primary outcome of
a cost-utility analysis is the cost per QALY, or incremental costutility ratio (ICUR), which is calculated as the difference in the
expected cost of two interventions divided by the difference in
the expected QALYs produced by the two interventions. The
results of a cost-utility analysis are compared with a threshold
incremental cost-effectiveness ratio (ICER); interventions with an
ICER below this threshold are funded, whereas those with an
ICER above the threshold tend not to be. Economic evaluations
Keywords: cost-effectiveness analysis, cost-utility analysis, Nigeria,
patients with type 2 diabetes, pharmaceutical care, usual care.
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Maxwell O. Adibe, Department of Clinical Pharmacy and Pharmacy Management, University of Nigeria,
Nsukka, Enugu, Nigeria.
E-mail: [email protected]; [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.009
190
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 189–198
straightforward and cheap to acquire—certainly when compared
with the cost of acquiring evidence on clinical effectiveness, and
indeed the cost of many of the treatments being reviewed [3].
Diabetes mellitus (DM) is associated with considerable morbidity and mortality [4]. It is also a major risk factor for cardiovascular
disease, stroke, and kidney failure [5]. In Africa, DM probably has
the highest morbidity and mortality rates of all chronic noninfective diseases [6].
DM was once regarded as a disease of the affluent, but it is
now vastly visible as a growing health problem in developing
economics because almost 80% of diabetes deaths occur in lowand middle-income countries [7,8]. The national standardized
prevalence rate of DM in Nigeria is 2.2%, while the crude
prevalence rate is 7.4% in those aged 45 years and above who
live in urban areas [9]. Global estimates of the prevalence of
diabetes showed that the prevalence of diabetes in Nigeria in
2010 was 4.7% (vs. 3.9% for world population) and that it would be
5.5% (vs. 4.3% for world population) in 2030 [10].
With the increasing demand for better management of type 2
diabetes, attention has focused on the potential benefits of pharmaceutical care (PC) to improve patients’ health outcomes. Many PC
programs have been established in various countries to enhance
clinical outcomes and the health-related quality of life. These
programs were implemented by pharmacists, with the cooperation
of physicians and other health care professionals. PC and the
expanded role of pharmacists are associated with many positive
diabetes-related outcomes, including improved clinical measures
[11], improved patient and provider satisfaction [12,13], and
improved cost of management [12,14]. The pharmacists can, therefore, in collaboration with physicians and other health care professionals, contribute to an improvement in the quality of life of
patients with diabetes by informing and educating patients, answering their questions, and, at the same time, monitoring the outcomes
of their treatment [15]. In view of the above issues, the objective of
this study was to assess the cost-effectiveness of the PC intervention
in the management of type 2 diabetes versus usual care (UC).
Methods
Study Design
This study was a randomized, controlled, and longitudinal prospective study with a 12-month patient follow-up. The study
followed the Consolidated Health Economic Evaluation Reporting
Standards guideline for reporting economic evaluation of interventions [16]. The study protocol was approved by the Research
Ethical Committees of the University of Nigeria Teaching Hospital, Ituku Ozalla, and Nnamdi Azikiwe University Teaching
Hospital, Nnewi, in which this study was conducted. These
hospitals are tertiary hospitals that serve as referral centers to
most of the hospitals in the southeastern part of Nigeria.
patients who expressed willingness to withdraw from the study
(participation is voluntary). The sample size determination showed
that a sample size of at least 104 patients was required in each of the
control and intervention groups [17]. Based on these data, to ensure
sufficient statistical power and to account for “dropouts” during the
study, a target sample size of 220 patients was recruited (110 patients
from each of the hospitals). The folders of the 110 selected patients in
each hospital were assigned numbers 1 to 110, which represented an
individual patient. Patients were randomly assigned to one of two
groups (intervention group or control group) on the basis of the
number assigned to their folders by using online “random sequence
generator” [18] with sequence boundaries of 1 to 110 (boundaries
inclusive) set in a two-column format: the first column was a priori
designated to the intervention group PC (55 patients) and the second
column to the control group UC (55 patients).
Patients in the UC group received the usual/conventional care
offered by the hospitals, which included hospital visits on appointment or on a sick day, consultations with physicians, prescription of
drugs and routine laboratory tests, review of diagnosis and medications, refilling of prescriptions by patients, and referral. This UC
was offered with education/training of the patients in an uncoordinated manner and without structured educational materials.
Patients in the PC group received UC and PC for 12 months on
monthly schedule. This additional PC included a stepwise approach:
setting priorities for patient care, assessing patients’ specific educational needs and identification of drug-related problems, development of a comprehensive and achievable PC plan in collaboration
with the patient and the physician, implementation of this plan, and
monitoring and review of the plan from time to time [19]. The
nurses collaborated with the pharmacists in terms of organizing the
patients and patients’ folders, taking point-of-care testing, counseling the patients, and reinforcing the information given to the
patients during training sections. The physicians provided the
visitation/appointment schedule for the patients, and prescription
of laboratory tests. They were also involved in the implementation
of consensus strategies in managing drug-related problems in areas
of changing, substitution, and withdrawal of medications. All the
members of the health care team were trained before the implementation of the intervention.
The medical and educational contents of the training materials
were evaluated by the physicians and nurses in diabetes clinics
before the researchers conducted the training for the patients. The
physicians and nurses were asked to rate the materials as being
excellent, very good, good, fair, poor, and useless.
The monthly educational/training program for the patients
consisted of four sections of 90 to 120 minutes. The program
covered the following areas: diabetes overview and its complications, self-monitoring blood glucose techniques and interpretation of diabetes-related tests, medications and their side effects,
lifestyle modification, counseling, and effective interaction with
health providers. PC provided ground for the patients to monitor
and react to changes in their blood glucose levels, allowing them
to integrate their diabetes into the lifestyle they preferred.
Inclusion Criteria
Patients with type 2 DM who fulfilled the entrance criteria were
identified and included in the study. Inclusion criteria included
patients with type 2 diabetes who were on oral hypoglycemic
therapy and provided written informed consent in addition to
willingness to abide by the rules of the study and being certified
fit to take part by the consulting physician.
Exclusion criteria were patients who were diagnosed with type 1
diabetes (to avoid complexity in the scope of the study), patients who
were younger than 18 years (they are legally regarded as dependents
and consequently they cannot take decisions of their own), patients
who were pregnant (they are generally not allowed to participate in a
study of this nature by the institutions used for the study), and
Data Collection
Data were collected on utilization of health care resources for 12
months for control and intervention groups at baseline, 6 months,
and 12 months. Information was obtained on the frequency of
self-monitoring, number and average duration of visits to a
hospital, daily doses of drugs taken regularly, and the variable of
“other health care resource use,” including primary care (general
practitioner and nurse consultations), hospital care (visits to an
accident and emergency department, outpatient care, day hospital
care, and inpatient care), auxiliary health care (services of a
podiatrist, optician, or dietitian), and private health care. These
data were collected by means of patients’ PC diaries notes
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 189–198
supplemented when necessary by information from patients’
medical records.
Patient-specific data on the incremental resources required
for intervention and control groups and the resources associated
with the treatment of complications were taken. Treatment
resources included doses of drugs used for treating diabetes,
antihypertensive drugs, other drugs, blood-glucose tests, selfmonitoring resources such as test strips, lancets, and glucometers, and visits to general practitioners, practice nurses, and
clinics. Resources associated with complications included the
number, duration, and specialty of admissions to hospital; outpatient consultations; medical procedures such as photocoagulation and cataract extraction; and day case episodes. The unit
cost of these resource volumes (drugs and other consumables,
laboratory tests, and specialty care per visit) was derived by using
the National Health Insurance Scheme price list [20] and International Drug Price Indicator Guide 2010 edition [21], and the cost of
all inpatient bed-days and outpatient visits was derived from
World Health Organization-Choosing Intervention that is CostEffective (WHO-CHOICE) [22] unit-cost estimates. Costs of the
management of individual complications were calculated from
activities involved in their management by using the ingredient
approach or activity-based costing as opined by the experts; all
costs were adjusted to 2011 cost [23]. The effect of either a higher
or lower adjustment rate was examined in the sensitivity analysis. All costs are reported in year 2011 values of Nigerian naira
(NGN 155 ¼ $1).
Each item for resource use was categorized into the “cost of the
intervention,” the “cost of drugs,” and the “cost of other health
care resource use” (including primary care, hospital care, and
auxiliary health care). The costs were calculated by multiplying
the volume of resource use in each category by the associated unit
cost in 2010 prices (Table 1). Average costs were estimated in each
arm of the study for the 12 months of follow-up Table 2.
The impact of the interventions on quality of life was
estimated by using the HUI23S4EN.40Q (developed by HUInc Mark index 2&3) questionnaire at baseline, 6 months, and 12
months in accordance with the HUI procedures manual (HUI23S4EN.40Q, HUI23-40Q.MNL) [24]. We adopted the QALY [2]
because this measure captures both increases in life expectancy
and improved quality of life that results from the prevention of
complications, providing a composite outcome measure of fatal
and nonfatal events that permits comparison between many
health interventions.
Because the economic evaluation perspective was that of the
health care purchaser, only direct health service costs were
included. These included treatment costs, visits to a nurse or a
general practitioner based on “standard practice” assumptions,
and costs of treating diabetes complications. Not included in this
analysis were nonmedical costs such as out-of-pocket expenses
incurred when visiting clinics, cost of informal care provided by
family members, and production losses resulting from work
absences, long-term disability, or premature death.
Statistical Analysis
Statistical analyses were performed by using the SPSS package,
version 14 (SPSS, Inc., Chicago, IL). An intention-to-treat approach
was used. Data were summarized as means ⫾ SD, mean differences with 95% confidence intervals. Two-sample comparisons
were made by using Student’s t tests for normally distributed
variables or Mann-Whitney U tests for nonnormally distributed
data (0, 6, and 12 months). Comparisons of proportions were
carried out by using chi-square, Fisher’s exact, or McNemar’s
tests. An a priori significance level of P less than 0.05 was used
throughout. Based on the overall health-related quality-of-life
score for the patients at baseline, 6 months, and 12 months,
191
QALYs were determined. Areas under the curves were determined by using WinNonlin standard edition version 2.1 [3,22].
Sensitivity analysis
To address uncertainty around the ICUR, univariate sensitivity
analysis was conducted, where one cost variable was varied at a
time (upper and lower limits) while keeping all other variables
constant at their mean base-case cost. Then, two alternative-case
outcomes of ICUR were generated on the basis of upper and lower
boundaries of ⫾20% of the mean base-case cost.
To assess how a simultaneous change in several variables
(QALYs, total intervention cost, cost of antidiabetes medications,
cost of antihypertensives, cost of antidiabetes antihypertensives
medications, total cost of drugs, hospital care cost, auxiliary
health care cost) affects the cost-utility ratio, a Monte-Carlo
simulation (a type of multivariate sensitivity analysis) was
performed. This technique runs a large number of simulations
(here 1000) by repeatedly drawing samples from probability
distributions of input variables. Thus, it provides a probability
distribution of the output variable; that is, QALYs, incremental
costs, incremental effectiveness, and ICURs. Beta and gamma
distributions were assumed for utility (QALYs) and unit cost,
respectively [25–27].
Given that the interpretation of negative ICURs is ambiguous,
the ICURs were transformed into net monetary benefits (NMBs).
The decision rule used was to adopt the intervention in question
if the NMB is greater than zero. Given that the appropriate value
of λ is unknown, λ was varied from NGN 0 to NGN 450,000. A costeffectiveness acceptability curve was generated on the basis of
the distribution of NMB for each λ. A cost-effectiveness acceptability curve allows a decision maker to consider whether an
intervention (PC) is cost-effective in relation to the maximum
amount a decision maker is willing to pay for a QALY. A discount
rate of 3% and 6% was used in sensitivity analysis [28].
At each ceiling value for the willingness to pay for a QALY, the
cost-effectiveness curve shows the probability that the treatment
is cost-effective. All calculations were done in Microsoft Excel
2007 (Microsoft Corporation, Redmond, WA).
Results
Economic Outcomes
The medical and educational content of the training course was
rated positively by the 17 physicians and 29 nurses: the majority
38 (82.6%) rated the content as “excellent” and the remaining 8
rated the content as “very good” or “good”; only 3 (6.5%) of them
suggested little modification or changes.
The number of patients who completed the study and whose
data were analyzed at 6 months and 12 months in UC and PC
arms were 98 (89.09%) versus 102 (92.73%) and 93 (84.55%) versus
99 (90.0%), respectively.
The general cost of care/laboratory cost per patient for UC
versus PC at 12 months was NGN 16,519 ⫾ 7,905 ($107 ⫾ $51)
versus NGN 17,369 ⫾ 6,673 ($112 ⫾ $43), P ¼ 0.4208. PC-specific
cost per patient was NGN 7,345 ⫾ 2,651 ($47 ⫾ $17), while the
costs of antidiabetes medications for UC and PC arms were NGN
9,703 ⫾ 4,632 ($63 ⫾ $30) and NGN 7,808 ⫾ 4,183 ($50 ⫾ $27), P ¼
0.0033, respectively. The cost of antihypertensives for UC was
NGN 6,625 ⫾ 4,691 ($43 ⫾ $30) while that of PC was NGN 5,155 ⫾
2,619 ($33 ⫾ $17), P ¼ 0.0228. The cost of antidiabetes medications
plus antihypertensives for UC was NGN 16,328 ⫾ 5,086 ($105 ⫾
$33) as against NGN 12,963 ⫾ 7,549 ($84 ⫾ $49) for PC, P ¼ 0.0004.
The cost of other medications was NGN 3,243 ⫾ 2,637 ($21 ⫾ $17)
and NGN 4,945 ⫾ 1,687 ($32 ⫾ $11), P o 0.0001, for UC and PC,
respectively.
192
Table 1 – Categories of resources used and their cost sources.
Cost centers
Source
3,300
2,200
600
2,900
800
Market
Market
Market
Market
Market
1,700
250
700
400
700
600
300
34,000
UNTH
[20]
[20]
[20]
[20]
[20]
[20]
[20]
15–60
3,200
950
850
[20,21]
[20,21]
[20]
[20]
5–1,400
5–280
[20,21]
[20,21]
8,000–20,000 (12,000)†
2,149.50
4,404
price
price
price
price
price
Experts’ opinion (UNTH/NAUTH)
[22]
[22]
700
700
[20]
[20]
700
700
700
700
[20]
[20]
[20]
[20]
Note. NGN 155 ¼ $1.
BP, blood pressure; Hb A1c, hemoglobin A1c; HDL, high-density lipoprotein; LDL, low-density lipoprotein; NAUTH, Nnamdi Azikiwe University Teaching Hospital; NGN, Nigerian naira; NHIS,
National Health Insurance Scheme; UNTH, University of Nigeria Teaching Hospital.
When the individual drug was not in the NHIS price list, the International Drug Price Indicator Guide 2010 edition was used; the total cost of drug category was presented because many drugs
were encountered. The ranges of their prices are represented.
†
The price in parentheses was used.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 189–198
Intervention
Meter (Acu-check Active)
Test strips/50 strips
Lancet/200
BP apparatus (Aneroid sphygmomanometer and stethoscope)
Training/educational materials
Laboratory tests
Hb A1c
Fasting blood glucose
Liver function test
HDL
LDL
Triglyceride
Total cholesterol
Others
Drugs
Oral antidiabetes drugs per tablet
Human insulin per vial
Insulin soluble per vial
Insulin zinc per vial
Suspension (insulin zinc suspension)
Antihypertensives (tablet, injection, injection powder, syrup)
Others (tablet, injection, injection powder, syrup)
Hospital care (per episode)
Emergency care
Outpatient care (tertiary hospital)
Inpatient (per day) (tertiary hospital)
Primary care
General practitioner consultation
Nurse consultation
Auxiliary health care (per session)
Dietician
Optician
Podiatrist
Others
Unit cost in NGN (cost used in study)
193
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 189–198
The total drug cost was NGN 19,571 ⫾ 7,514 ($126 ⫾ $49) for UC
as against NGN 17,908 ⫾ 8,549 ($116 ⫾ $55) for PC, P ¼ 0.1549. The
hospital care costs for UC and PC were NGN 10,302 ⫾ 5,657 ($67 ⫾
$37) and NGN 9,766 ⫾ 4,234 ($63 ± $27), P ¼ 0.4565, respectively,
while their auxiliary health care costs were NGN 4,060 ⫾ 1,675 and
NGN 8,687 ⫾ 2,365 ($56 ± $15), P o 0.0001, respectively. The total cost
per patient per year was NGN 50,452 ⫾ 35747 ($326 ± $231) for UC
and NGN 61,075 ⫾ 43763 ($394 ± $282), P ¼ 0.1009, for PC (Table 3).
Cost-effectiveness or cost-utility was NGN 78524.51 ($507) per
QALY for UC and NGN 80,098.36 ($517) per QALY for PC, while the
incremental cost and incremental QALY were NGN 10,623 ($69) and
0.12, respectively. Thus, the ICUR was NGN 88,525 ($571) per QALY.
Sensitivity Analysis
The cost-effectiveness plane that was obtained from a Monte
Carlo simulation with 1000 iterations showed that 93.8% of the
simulations were within the northeast quadrant, where the PC
intervention resulted in gain in QALY and cost, whereas 5.6% of
the simulations were in the southeast quadrant, where the PC
intervention resulted in gain in QALY and reduced cost. Only 0.5%
of the simulations were within the northwest quadrant, where
the addition of PC resulted in loss in QALY and increased cost.
The 1000 iterations produced an incremental QALY that ranged
from −0.022 to 0.293 and an incremental cost that ranged from
NGN −8,276.40 to NGN 28,294.27 (Fig. 1).
The mean NMB within a willingness to pay of NGN 0 to NGN
450,000 was greater in the PC intervention whatever the willingness to pay was. This result also revealed that 90% of PC
credibility interval was far above the mean of UC though the
interval overlapped with about 5% of the UC (Fig. 2).
The PC intervention led to incremental cost and incremental
QALY/effect of NGN 10,623 and 0.12 QALY gained, respectively,
with an associated ICUR of NGN 88525 per QALY gained. The
Table 2 – Baseline characteristics of the patients in PC and UC arms.
Demographic data
Mean age ⫾ SD (y)
Grouped age: 453 y, n (%)
Sex: male, n (%)
Level of education, n (%)
Primary school
Secondary school
University
Marital status, n (%)
Currently married
Widowed
Single
Occupation, n (%)
Self-employed
Employee
Retired
Smoking status: smoker, n (%)
Duration, mean ⫾ SD
Duration: ≥5 y, n (%)
Family history of diabetes, n (%)
Physical activity/exercise, n (%)
Comorbidities
Hypertension
Congestive heart failure
Ischemic heart disease
Arthritis
≥2 comorbidities, n (%)
Overnight hospitalization, n (%)
Emergency room, n (%)
Use of insulin, n (%)
Antidiabetic medications, n (%)
Other medications, n (%)
Daily aspirin
Diuretics
Antihypertensives
Lipid-lowering
Complications, n (%)
Myocardial infarction
Stroke
Foot ulcer
Blindness
Renal failure
UC (n ¼ 110)
PC (n ¼ 110)
52.8 ⫾ 8.2
81 (73.64)
49 (44.55)
52.4 ⫾ 7.6
75 (68.18)
44 (40)
3 (2.72)
71 (64.55)
36 (32.73)
6 (5.45)
63 (57.27)
41 (37.27)
37 (33.64)
71 (64.54)
2 (1.82)
46 (41.82)
63 (57.27)
1 (0.91)
37 (33.64)
35 (31.82)
38 (34.54)
34 (30.91)
4.5 ⫾ 2.2
62 (56.36)
71 (64.55)
18 (16.36)
34 (30.91)
42 (38.18)
34 (30.91)
21 (19.09)
4.8 ⫾ 2.8
71 (64.55)
62 (56.36)
23 (20.91)
P
0.708
0.373
0.495
0.406
0.409
0.611
0.043
0.378
0.215
0.214
0.387
60
11
7
37
72
9
1
17
103
(54.55)
(10.00)
(6.36)
(33.64)
(65.45)
(8.18)
(0.91)
(15.45)
(93.64)
73
15
8
43
81
7
2
13
107
(66.36)
(13.64)
(7.27)
(39.09)
(73.64)
(6.36)
(1.82)
(11.82)
(97.27)
0.073
0.404
0.789
0.400
0.187
0.604
0.561
0.432
0.195
43
71
98
23
(39.09)
(64.55)
(89.91)
(20.91)
57
84
78
14
(51.82)
(76.36)
(70.91)
(12.73)
0.058
0.055
0.0007
0.105
2
9
2
1
3
(1.82)
(8.18)
(1.82)
(0.91)
(2.73)
4
6
3
1
8
(3.64)
(5.45)
(2.73)
(0.91)
(7.27)
0.408
0.422
0.651
1.000
0.122
NGN, Nigerian naira; PC, pharmaceutical care; UC, usual care.
P o 0.05.
194
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 189–198
QALY value of PC was varied by ⫾10% (0.68625 and 0.83875) in the
sensitivity analysis, which made the ICUR moved from NGN
88,525 per QALY gained to upper and lower limits of NGN
252,173 per QALY gained and NGN 54,777 per QALY gained,
respectively. When a 3% and 6% adjustment rate of costs was
applied, the lower and upper extremes of ICURs were NGN 85,936
per QALY gained and NGN 88,799 per QALY gained from the base
value of NGN 88,525 per QALY gained.
Varying the base costs by ⫾20% (lower limit to upper limit) and
using these extreme values in simulations, the total intervention
costs center produced ICURs that ranged from NGN 47,426.41 to
NGN 126,544.5 per QALY gained. The antidiabetes cost center
produced ICURs that ranged from NGN 74,817.97 to NGN
104,206.6 per QALY gained. The ICURs moved from NGN 79,210.06
to NGN 98,038.14 per QALY gained when the base cost of the
antihypertensives cost center was varied. The antidiabetes medications plus antihypertensives produced ICURs that ranged from
NGN 70,020.39 to NGN 109,161.3 per QALY gained, and total drug
cost center produced ICURs that ranged from NGN 58,950.02 to
NGN 120,469.9 per QALY gained. The ICURs of hospital care cost
and auxiliary health care centers ranged from NGN 72,582.82 to
106,161.6 per QALY gained and NGN 74,454.77 to 105,414.0 per
QALY gained, respectively, when base costs of the cost centers
were varied. The ICUR was most sensitive to variation in QALY and
“total intervention-specific cost center” variable followed by that of
total drug cost center (Fig. 3). In the cost-effectiveness acceptability
curve, the probability that PC was cost-effective versus UC was 95%
at the threshold of NGN 250,000 per QALY gained and 52% at the
threshold of NGN 88,600 per QALY gained (Fig. 4).
Discussion
At the end of this period, the PC intervention resulted in an
incremental gain in QALYs and cost compared with the UC. This
economic evaluation demonstrates that PC is the most costeffective strategy for managing patients with type 2 diabetes if
the patients are willing to pay at least NGN 88,600 per QALY
gained. The addition of the PC intervention to UC, as noted in this
study, should be considered a highly cost-effective management
option for patients with type 2 diabetes because treatments
costing no more than £20,000 (NGN 4,761,905) to £30,000 (NGN
7,142,857) per QALY gained are generally considered to be costeffective [1,31]. This PC intervention also generated greater NMBs
when compared with UC; therefore, the addition of PC to UC
might be considered an appropriate management option for
patients with diabetes who have comorbidities where the probability or likelihood of drug-related problem is higher.
Cost-Effectiveness Plane
The cost-effectiveness plane showed that most of the simulations were within the northeast quadrant, where the addition of
PC to UC resulted in gain in QALY and cost, which indicated that
although the PC intervention generated more QALYs than did UC,
it was more costly. The 1000 iterations showed that the 95%
confidence interval of incremental QALYs and incremental cost
was wide. This wide range shows that there are uncertainties
surrounding both QALYs and cost. The magnitude of QALYs
gained, specific intervention cost, and cost of all drugs were
found to have affected the ICUR most. This provides avenues for
urgent intervention to reduce the cost of drugs used for the
management of diabetes and its comorbidities and an urgent
institution of intervention that will improve the quality of life of
patients with diabetes in Nigerian tertiary hospitals.
Net Monetary Benefit
Quality-Adjusted Life-Years
QALYs associated with PC were significantly higher than those
associated with UC after 12 months. This indicates that extending this study beyond 1 year could offer more benefits to patients
with diabetes in terms of QALYs gained. Some studies had
demonstrated that extension of PC beyond 1 year could offer
extra benefits to patients with diabetes [29,30].
The NMB approach provides a useful mechanism for identifying
which arm of the study is most cost-effective. The NMB of additional
PC over UC alone for a willingness to pay of NGN 0 to NGN 450,000,
the additional PC alternative, was associated with the greater mean
NMB whatever the willingness to pay was. It is interesting to note
that the lowest trough (NMB) of PC was far higher than the mean of
UC NMB. Addition of the PC intervention was found to be superior to
UC alone in all willingness to pay, even as the willingness to pay
Table 3 – Costs and QALY per patient per year at the end of the 12-mo follow-up period for UC versus PC (NGN
155 ¼ $1).
Cost per patient per year
General intervention and laboratory
cost
Specific intervention cost for PC
Total cost of intervention
Antidiabetes medications
Antihypertensives
Antidiabetes medications plus
antihypertensives
Other medications
Total drug cost
Hospital care cost
Auxiliary health care cost
Total cost per patient
QALY per patient per year
UC
PC
P
Mean cost
difference
16,519 ⫾ 7,905
17,369 ⫾ 6,673
0.4208
NA
16,519 ⫾ 7,905
9,703 ⫾ 4,632
6,625 ⫾ 4,691
16,328 ⫾ 5,086
7,345
24,714
7,808
5,155
12,963
⫾
⫾
⫾
⫾
⫾
2,651
11,655
4,183
2,619
7,549
NA
o0.0001
0.0033
0.0228
0.0004
NA
8,195
−1,895
−1,470
−3,365
NA
5,341.9–11,048
−3,150.1 to −639.92
−2,733.7 to −206.34
−5,209.3 to −1,520.7
3,243 ⫾ 2637
19,571 ⫾ 7514
10,302 ⫾ 5,657
4,060 ⫾ 1,675
50,452 ⫾ 35,747
0.6425 ⫾ 0.13
4,945
17,908
9,766
8,687
61,075
0.7625
⫾
⫾
⫾
⫾
⫾
⫾
1,687
8,549
4,234
2,365
43,763
0.15
o0.0001
0.1549
0.4565
o0.0001
0.1009
o0.0001
1,702
1,663
−536
8,627
10,623
0.1200
1,074.7–2,328.3
−3,960.2 to 634.16
−1,952.9 to 880.88
8,040.2–9,213.80
−2,088.5 to 23,335
(0.07–0.1601)
850
CI, confidence interval; NA, not applicable; PC, pharmaceutical care; QALY, quality-adjusted life-year; UC, usual care.
P ≤ 0.05; Negative cost differences indicate cost savings associated with the PC intervention.
95% CI per
patient
−1,228.1 to 2,928.1
195
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 189–198
35000.000
30000.000
Incremental cost (NGN)
25000.000
20000.000
15000.000
10000.000
5000.000
0.000
-0.050
0.000
-5000.000
0.050
0.100
0.150
0.200
0.250
0.300
0.350
-10000.000
-15000.000
Incremental QALYs
Fig. 1 – Cost-effectiveness plane. NGN, Nigerian naira; QALYs, quality-adjusted life-years.
increases. This result shows that if the willingness to pay is in the
range of NGN 0 to NGN 450,000, there is a net monetary gain or
saving of NGN 56,148 in 1 year. This amount is more than 3 months
salaries of a Nigerian low-income earner based on the current NGN
18,000 minimum wage [32]. Therefore, there is need for introduction
and exploitation of the PC intervention in Nigerian health facilities
because this is very cost-effective with enormous NMB.
Sensitivity Analyses
This study found out that a little variation in QALY gained in PC
to the tune of ⫾10% resulted in a tremendous increase and mild
lowering of the base ICUR, respectively. This result showed that
the economic burden placed on patients with diabetes by 10%
health deficit was enormous; therefore, interventions such as PC
that would be aimed at improving the quality of life of patients
and resolution/reduction of drug-related problems that would
ultimately reduce the cost of drugs would certainly reduce the
cost per QALY associated with diabetes.
These results indicate that the additional PC intervention has a
cost per QALY gained that is lower than that of UC. In the United
Kingdom, interventions appear to have a high chance of acceptance
by the National Institute for Clinical Excellence if their costeffectiveness is more favorable than approximately £30,000 per
QALY [1]. Several other studies had classified cost-effectiveness.
WHO-CHOICE classified interventions on the basis of the level of
400000
Net Monetary Benefits (NGN)
350000
PC
300000
UC
250000
200000
150000
0
50000 100000 150000 200000 250000 300000 350000 400000 450000
Willingness to pay (NGN)
Fig. 2 – Net monetary benefit of PC and UC at different levels of willingness to pay for a QALY. NGN, Nigerian naira; PC,
pharmaceutical care; QALY, quality-adjusted life-year; UC, usual care.
196
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 189–198
cost-effectiveness by convention as described in the literature [33–
35]. An intervention is cost saving when the intervention generates a
better health outcome and costs less than the comparison intervention. The intervention is cost neutral if the ICER is 0. The
intervention is very cost-effective when the ICER is more than 0 or
$25,000 or less per QALY or life-year gained (LYG) while the
intervention is cost-effective when the ICER is between more than
$25,000 to $50,000 per QALY or LYG. The intervention is marginally
cost-effective when the ICER is between more than $50,000 and
$100,000 per QALY or LYG, whereas an intervention is said to be not
cost-effective when the ICER is more than $100,000 per QALY or LYG.
WHO-CHOICE published in 2005 the cost-effectiveness threshold for different regions of the world. WHO-CHOICE suggested a
cost-effectiveness threshold based on gross domestic product
(GDP) per capita. An intervention that produces cost per QALY
gained of less than GDP per capita of the country is said to be very
cost-effective while an intervention with cost per QALY gained of
between one to three times the GDP per capita of the country is
cost-effective. An intervention with cost per QALY gained of
greater than three times the GDP per capita of the country is
not cost-effective. For AFRO D where Nigeria belongs, the costeffectiveness threshold ranges from $1,695 to $5,086 [7]. With a
conversion factor of NGN 155 ¼ $1, the threshold ranges from
NGN 262,725 to NGN 788,330. World Bank in 2010 published a GDP
per capita, considering purchasing power parity-current international $; for Nigeria, it is $2,381 (NGN 369,055:00) [34]. The
associated ICUR from this study was NGN 88,525 per QALY
($571.13/QALY) gained in the PC arm, which was far lower than
the GDP per capita of Nigeria in 2010 [36].
Based on the above facts, the PC intervention with an ICUR of
NGN 88,525 per QALY gained is very cost-effective although this
may still not be affordable for low-income earners in relation to the
NGN 18,000 minimum wage approved in Nigeria in 2011 because
NGN 88,525 is about 5 months’ salary of this group of Nigerians [32].
In probabilistic sensitivity analysis, based on the costeffectiveness acceptability curve, the PC dominated UC at the
threshold of NGN 88,600 per QALY gained and the probability of
PC being more cost-effective approached 95% at the threshold of
NGN 250,000 per QALY gained. Nevertheless, if a patient is willing
to pay NGN 400,000 per QALY gained, the probability that PC is
the most cost-effective option for managing patients with diabetes increases to 97%. In contrast, the probability that UC is the
most cost-effective option at the threshold of NGN 400,000 per
QALY gained approaches zero.
Studies of this kind must address inherent potential threats
to internal validity [37,38]. The major limitations of this study
300,000
252,173
250,000
Cost Per QALY (NGN)
200,000
150,000
126,545
120,470
109,161
104,207
100,000
106,162 105,414
98,038
88,799
85,936
74,818
79,210
72,583
70,020
74,455
58,950
50,000 54,777
47,426
0
Variables
Fig. 3 – Univariate sensitivity analysis of cost and utility variables on incremental cost-effectiveness ratio. NGN, Nigerian naira;
QALY, quality-adjusted life-year. A ¼ QALY (⫾10%); B ¼ Adjustment rate (3% and 6%); C ¼ Total intervention-specific cost (⫾20%);
D ¼ Cost of antidiabetes medications (⫾20%); E ¼ Cost of antihypertensives (⫾20%); F ¼ Cost of antidiabetes + antihypertensives
medications (⫾20%); G ¼ Total cost of drugs (⫾20%); H ¼ Hospital care cost (⫾20%); I ¼ Auxiliary health care cost (⫾20%).
197
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 189–198
1
Pharmaceutical Care
0.9
Probability cost-effecveness
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Usual care
0
0
50000
100000 150000 200000 250000 300000 350000 400000 450000
Treshold Willingness to pay for a QALY (NGN)
Fig. 4 – Cost-effectiveness acceptability curves. NGN, Nigerian naira; QALY, quality-adjusted life-year.
were missing data, selection bias, short period of study, attrition,
and consideration of only direct cost. Data on humanistic
outcome measures were self-reported; however, self-reported
data about diabetes status have been established to be both
valid and reliable [39]. We recommend that future research
studies of this kind address these limitations. This pharmaceutical intervention could be adopted for patients suffering from
other chronic diseases such as HIV, hypertension, asthma,
psychosis, epilepsies, and cerebrovascular and cardiovascular
diseases.
Conclusions
The PC intervention was very cost-effective among patients with
type 2 diabetes at the NGN 88,525 ($571.13) per QALY gained
threshold, although considerable uncertainty surrounds these
estimates. This study also revealed that cost incurred and QALYs
gained by patients in the PC group were higher than those of their
counterparts in the UC group. This indicates that the extra cost
paid for extra QALYs gained is worth it because it saves future
expenditures and improves the quality of life of patients.
The results of this study illustrate a convincing economic rationale for improving standards of care for patients with type 2 diabetes
through the PC intervention. This study provides further evidence
that the cost-effectiveness of interventions to reduce the burden of
diabetes-related complications compares favorably with that of other
accepted uses of health care resources. The results should be of
interest and used by other economists and health service researchers,
and in particular should be considered by decision makers when
considering the allocation of resources to diabetes care.
Cost-utility analysis thus increasingly facilitates the transparency of resource allocation processes. The usefulness of costutility analysis to decision makers explains the rapid expansion
in the utilization of cost-utility analysis over the last decade.
Acknowledgment
We acknowledge Health Utility Incorporated for granting and
awarding us HUI23S4En.40Q and HUI23.40Q.MNL.
Source of financial support: Funding for this project was
provided from Science and Technology Education Post Basic
(STEP-B) through the University of Nigeria. The views expressed
in this article are those of the authors, and no official endorsement by STEP-B is intended or should be inferred.
R EF E R EN C ES
[1] Rawlins M, Culyer A. National Institute for Clinical Excellence and its
value judgments. BMJ 2004;329:224–7.
[2] Torrance G. Measurement of health state utilities for economic
appraisal: a review. J Health Econ 1986;5:1–30.
[3] Drummond M, O’Brien B, Stoddart G, Torrance G. Methods for the
Economic Evaluation of Health Care Programmes. (2nd ed.). Oxford:
Oxford University Press, 1997.
[4] Akanji AO, Adetunji A. The pattern of presentation of foot lesions in
Nigerian diabetic patients. West Afr J Med 1990;9:1–4.
[5] American Diabetes Association. National Diabetes Fact Sheet. Available
from: http://www.diabetes.org/main/info/facts/factsna tl.jsp. [Accessed
December 15, 2002].
[6] McLarty DG, Swai ABM, Kitange HM, et al. Prevalence of diabetes and
impaired glucose tolerance in rural Tanzania. Lancet 1989;1:871–5.
[7] International Diabetes Federation. International Diabetes Federation:
diabetes atlas. Available from: http://da3.diabetesatlas.org/newsc269.
html. [Accessed November 15, 2011].
[8] Odili VU, Ugboka LU, Oparah AC. Quality of life people with diabetes in
Benin City as measured with WHOQoL-BREF. Internet J Law Healthcare
Ethics 2010;6(2). http//dx.doi.org/10.5580/18a.
[9] Nyenwe E, Odia O, Ihekwala A, et al. Type 2 diabetes in adult Nigerians:
a study of its prevalent and risk factors in Port Harcourt, Nigeria.
Diabetes Res Clin Pract 2003;62:177–85.
[10] Shaw J, Sicree R, Zimmet P. Global estimates of the prevalence of
diabetes. Diabetes Res Clin Pract 2010;87:4–14.
[11] Jaber L, Halapy H, Fenret M, et al. Evaluation of a pharmaceutical care
model on diabetes management. Ann Pharmacother 1996;30:238–43.
[12] Sadur C, Moline N, Costa M, et al. Diabetes management in a health
maintenance organization: efficacy of care management using cluster
visits. Diabetes Care 1999;22:2011–7.
[13] Majumdar SR, Guirguis LM, Toth EL, et al. Controlled trial of a
multifaceted intervention for improving quality of care for rural
patients with type 2 diabetes. Diabetes Care 2003;26:3061–6.
[14] Coast-Senior E, Kroner B, Kelley C, Trili L. Management of patients with
type 2 diabetes by pharmacists in primary care clinics. Ann Pharmacother 1998;32:636–41.
[15] Hawkins D, Bradberry JC, Cziraky MJ, et al. National Pharmacy
Cardiovascular Council treatment guidelines for the management of
type 2 diabetes mellitus: toward better patient outcomes and new roles
for pharmacists. Pharmacotherapy 2002;22:436–44.
198
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 189–198
[16] Husereau D, Drummond M, Petrou S, et al. ISPOR Health Economic
Evaluation Publication Guidelines-CHEERS Good Reporting Practices
Task Force. Consolidated Health Economic Evaluation Reporting
Standards (CHEERS)—explanation and elaboration: a report of the
ISPOR Health Economic Evaluation Publication Guidelines Good
Reporting Practices Task Force. Value Health 2013;16:231–50.
[17] Oyejide OO. Health Reseach Methods. Ibadan: Leniks Publishers, 1992.
[18] Mads-Haahr. Random Sequence Generator (1998–2011). Available from:
http://www.random.org/sequences/. [Assessed February 15, 2012].
[19] Hepler C, Strand L. Opportunities and responsibilities in pharmaceutical care. Am J Hosp Pharm 1990;47:533–43.
[20] National Health Insurance Scheme. NHIS Healthcare Providers Service
Price List. Abuja, Nigeria: Ministry of Health, 2005.
[21] Management Science for Health (MSH) and World Health Organization.
International Drug Price Indicator Guide. In: Frye JE, ed. Cambridge:
Management Science for Health, 2010.
[22] WHO-CHOICE. Cost-effectiveness thresholds (2005 International $); by
region. Available from: http://www.who.int/choice/costs/CER_levels/en/.
[Accessed February 15, 2012].
[23] Drummond M. Introducing economic and quality of life measurements
into clinical studies. Ann Med 2001;33:344–9.
[24] Furlong W, Feeny D, Torrance G. Algorithm for Determining HUI Mark 2
(HUI2)/Mark 3 (HUI 3) Health Status Classification Levels, Health States,
Single-Attribute Level Utility Scores and Overall Health Related Quality
of Life Utility Scores from HUI23.40Q Questionnaires. Dundas, Ontario,
Canada: Health Utilities Inc., 2000.
[25] Briggs A, Claxton K, Sculpher M. Decision Modelling for Health
Economic Evaluation. Oxford: Oxford University Press, 2006.
[26] Briggs B, Sculpher M, Buxton M. Uncertainty in the economic
evaluation of health care technologies: the role of sensitivity analysis.
Health Econ 1994;3:95–104.
[27] Stinnet A, Mullahy J. Net health benefits: a new framework for the
analysis of uncertainty in cost-effectiveness analysis. Med Decis
Making 1998;18:68–80.
[28] World Health Organization. Making Choices in Health: WHO Guide to
Cost Effectiveness Analysis. Geneva: World Health Organization, 2003.
[29] Cranor C, Bunting B, Christensen D. The Asheville Project: long-term
clinical and economic outcomes of a community pharmacy diabetes
care program. J Am Pharm Assoc 2003;43:173–84.
[30] Neto P, Marusic S, Júnior DP, et al. Effect of a 36-month pharmaceutical
care program on coronary heart disease risk in elderly diabetic and
hypertensive patients. J Pharm Pharm Sci 2011;14:249–63.
[31] McCabe C, Claxton K, Culyer A. The NICE cost-effectiveness threshold:
what it is and what that means. Pharmacoeconomics 2008;26:733–44.
[32] An act to amend the National Minimum Wages Act Cap. N61 Laws of
the Federation of Nigeria, 2004 to provide for a revised national
minimum waged and; for related matters. Available from: http://www.
aksjlegalresource.com/resource/Laws_of_the_Federation%
5CNATIONAL%20MINIMUM%20WAGE%20_AMENDMENT_%20ACT%
202011.pdf. [Accessed May 27, 2013].
[33] Klonoff D, Schwartz D. An economic analysis of interventions for
diabetes. Diabetes Care 2000;23:390–404.
[34] Laupacis A, Deeny D, Detsky A, Tugwell P. How attractive does a new
technology have to be to warrant adoption and utilization? Tentative
guidelines for using clinical and economic evaluations. Can Med Assoc
J 1992;146:473–81.
[35] Grosse S. Assessing cost-effectiveness in health care: history of the
$50,000 per QALY threshold. Value Health 2008;8:165–78.
[36] The World Bank. Data by country 2010: Nigeria. Available from: http://
data.worldbank.org/country/nigeria. [Accessed February 15, 2012].
[37] Campbell D, Stanley J. Experimental and Quasi-Experimental Designs
for Research. Chicago, IL: Rand McNally College Publishing Company,
1963.
[38] Cook T, Campbell D. Quasi-experimentation: Design and Analysis
Issues for Field Studies. Chicago, IL: Rand McNally College Publishing
Company, 1979.
[39] West J, Goldberg K. Diabetes self-care knowledge among outpatients at
a Veterans Affairs medical center. Am J Health-Syst Ph 2002;59:849–52.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 199–204
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Economic Burden of Cardiovascular Diseases in the Russian Federation
Anna Kontsevaya, PhD*, Anna Kalinina, PhD, Rafael Oganov, PhD
Department of Primary Prevention in Healthcare, National Research Center for Preventive Medicine, Moscow, Russia
AB STR A CT
Objectives: In the Russian Federation, cardiovascular disease (CVD) is
the primary cause of death and premature death; however, to date,
there have been no systematic cost-of-illness studies to assess the
economic impact of CVD. Methods: The economic burden of CVD was
estimated from statistic data on morbidity, mortality, and health care
resource use. Health care costs were estimated on the basis of
expenditure on primary, outpatient, emergency, and inpatient care,
as well as medications. Non–health care costs included economic
losses due to morbidity and premature death in the working age.
Results: CVD was estimated to cost Russia RUR 836.1 billion (€24,517.8
million) in 2006 and RUR 1076 billion (€24,400.4 million) in 2009. Of the
total costs of CVD, 14.5% in 2006 and 21.3% in 2009 were due to health
care, with 85.5% and 78.7%, respectively, due to non–health care costs.
Conclusions: CVD is a leading public health problem. We first
assessed the economic burden of CVD in Russia. Our results can be
used for planning investments in prevention programs and measures
for improving care for patients with CVD. Regular monitoring of the
economic burden of CVD in the future at the federal, regional, and
municipal levels will allow assessment of the dynamics of economic
burden, as well as the effectiveness of investments in the economy in
primary and secondary prevention. Because data are relatively
unavailable, there are important limitations to this study, which
highlight the need for more accurate CVD-specific information.
Keywords: cardiovascular disease, coronary heart disease, cerebrovascular diseases, cost-of-illness study, economic burden, Russia.
Introduction
Methods
In the Russian Federation, cardiovascular disease (CVD) is the
primary cause of death and premature death [1]. There has been,
however, no systematic cost-of-illness study to assess its economic impact. The World Bank calculated health care expenditures in two regions of Russia and extrapolated these data to the
entire country [2]. These regions, however, were not representative of the entire country because the calculations included only
health care costs. Another attempt to calculate the economic
burden of CVDs in Russia was made by the World Health
Organization. It calculated the economic burden for 2005 and
predicted a prognosis of burden for 2015 and 2030 of the most
prevalent noncommunicable diseases from a macroeconomic
perspective based on death rate [3].
The objectives of this study were to estimate the economic
costs of CVD in Russia, including health care costs and productivity loss, and to estimate the proportion of total CVD cost
attributable to coronary heart disease (CHD) and cerebrovascular
diseases as was estimated in the study of Leal et al. [4] in the
European Union (EU).
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Methodological Background
Cost-of-illness analyses involved the identification, measurement, and valuing of resources related to CVD in Russia in the
period 2006 to 2009. The calculation included health care costs
and costs outside the health care sector (productivity losses
associated with premature death or morbidity and disability
pensions). All expenditures were measured for the period 2006
to 2009 in the prices of the appropriate year. Additional file 1
includes the sources of information used for calculations.
The national currency rubles was converted to euros (€) by
using a weighted exchange rate for the period 2006 to 2009.
Epidemiological and health care utilization data were acquired
from the Ministry of Health of the Russian Federation and the
market research company COMCON from the published literature. Analysis was based on the International Statistical Classification of Diseases, 10th Revision (ICD-10) categories: CVD (ICD-10
category I00–I99), hypertensive diseases (ICD-10 category I10–
I15), CHD (ICD-10 category I20–I25), and cerebrovascular disease
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Anna Kontsevaya, Department of Primary Prevention in Healthcare, National Research Center for
Preventive Medicine, Petroverigski Lane, 10, 101990, Moscow, Russia.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.010
200
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 199–204
(ICD-10 category I60–I69). All sources of information used for the
calculations are listed in the Additional information file.
Health Care Expenditure
The following categories of CVD health care services were included
in the calculations: primary and outpatient care, accident and
emergency (A&E) care, hospital inpatient care, cardiosurgery and
percutaneous coronary interventions (PCIs), and medications. Cardiosurgery and PCI expenditures were calculated separately from
hospital inpatient care because they are financed through different
sources in Russia. Hospital inpatient care is paid by the health
insurance system, while cardiosurgery and PCIs are paid by direct
payment from the federal budget in the framework of the federal
program on high technology and costly medical care or by the
patient. Other types of activities related to the CVD were not
included because of the difficulties in locating information. Data
for A&E care and hospital inpatient care were received from the
Ministry of Health of the Russian Federation; data on primary and
outpatient care were obtained from several information sources;
and data on medications were received from annual pharmacoepidemiology surveys through the COMCON company and other
literature data.
Health Service Utilization
Primary and outpatient care
Primary care activities consisted of CVD-related visits to general
practitioners (GPs), as well as GP visits to patients’ homes. There is a
specific statistical form for all medical organizations that includes
data on the number of patients who visited this organization for
outpatient care during the year according to the ICD-10 categories.
This statistical form is centrally received in the Ministry of Health
and processed as a single form for the entire country. We received a
single form for the entire country for 2006 to 2009. The number of
visits for each outpatient was calculated on the basis of these
statistical forms and data from previous studies examining the
mean number of visits of patients with CVD during the year.
Hospital inpatient care
Inpatient care was estimated on the basis of the number of CVDrelated days in the hospital. There is also a specific statistical
form for all inpatient medical organizations that includes data on
the number of hospitalizations and number of hospital days in
this organization during the year according to the ICD-10 categories. This statistical form is also centrally received in the Ministry
of Health and processed as a single form for the entire country.
We received the statistical single form for inpatient care in Russia
for 2006 to 2009 and selected data on ICD-10 categories of interest.
A&E care
A&E care consisted of all CVD-related hospital emergency visits.
A 2009 special statistical form for all inpatient medical organizations included information regarding the number of hospital
emergency visits according to the ICD-10 categories. We selected
data on ICD-10 categories in 2009 and extrapolated these values
for 2006 to 2008.
Cardiosurgery and PCI
The main cardiosurgery institution in Russia, the Bakoulev Center
for Cardiovascular Surgery, centrally collects information from all
medical organizations involved in such interventions in Russia and
annually publishes a statistical yearbook. We selected data regarding the number of PCI, coronary artery bypass grafting, and some
other cardiosurgeries performed in Russia in 2006 to 2009.
Health care unit costs
Unit costs of an inpatient day, outpatient visit, and emergency visit
were obtained from the Ministry of Health. The official Web site
annually publishes information on the mean costs of inpatient days,
outpatient visits, and emergency visits in different specialties and
total expenditures in the framework of the program of the governmental guarantees of the medical care.
The costs of cardiosurgery and PCI paid directly by the federal
budget are published annually on the Web site of the Ministry of
Health in the description of the Federal Program on High
Technology and Costly Medical Care.
Expenditure on medication
There are no national sources of information regarding national
expenditures on medications in Russia. We used data from several
pharmacoepidemiology surveys made in Russia in 2006 to 2009 and
extrapolated these data for the entire country. The main source was
the databases of annual surveys conducted by the COMCON
company; other studies were used to identify patients with CVD
regularly taking medication for long periods as well as some other
data. Costs of medications were calculated on the basis of mean
prices during the studding years, including value added tax (VAT).
Non–Health Service Costs
Non–health service costs included productivity losses associated
with premature death and morbidity and disability pensions.
Because little information was found on informal care costs and
out-of-pocket expenses across the country, these costs were not
included in the calculations.
Estimation of productivity costs due to premature death during
working age
Productivity costs due to premature death during working age
included the gross domestic product (GDP) per employed person
related to CVD attributable to mortality.
The productivity loss from CVD-mortality was estimated by
calculating the following:
1. number of CVD-related deaths during working age (retirement
age is 60 years for men and 55 years for women);
2. number of remaining work years at the time of death (to
estimate the likely GDP that an individual who died would
have otherwise produced);
3. annual GDP per employed person; and
4. economic activity and unemployment rates.
Future GDP was not indexed in the main analysis, as the usual
discount rate of 3% to 3.5% is not reasonable for Russia. The
inflation rate was 9% in 2006, 11.9% in 2007, 13.3% in 2008, and
8.8% in 2009. The effects of indexation on productivity costs using
rates of 10% and 15% were studied through sensitivity analysis.
Estimation of productivity costs due to cardiovascular morbidity
Morbidity costs were defined as those associated with CVDattributable absence from work, estimated by multiplying the
number of certified days off work due to CVD by GDP produced in
one working day.
The number of CVD-related working days lost was obtained
from a special statistical form for all medical organizations that
included data on the number of disability days during the year
according to the ICD-10 categories. This statistical form is also
centrally received at the Ministry of Health and processed as a
single form for the entire country. We received a single form for
working days lost in Russia for 2006 to 2009 and selected data on
ICD-10 categories related to CVD.
201
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 199–204
Sensitivity analysis
We examined the effects of 20% changes in health care costs.
Because medication costs were estimated by using subjective
assumptions, the effect of 50% changes in these categories was
tested. We also assessed the effects of indexation on productivity
costs by using rates of 10% and 15%.
Results
Table 1 shows the average unit costs used for calculations and
those aggregated from different sources.
Health Care Costs
CVD accounted for more than 69.6 million hospital bed days in
Russia in 2006 and 68.0 million in 2009. The number of hospitalizations did not change, but the mean duration of hospital stay
decreased each year (from 15.0 to 14.1 days). The number of
hospital bed days for CHD was 27.1 million in 2006 and 26.3 million
in 2009, and for cerebrovascular diseases it was 17.6 and 17.2
million, respectively. CVD represented 479.2 hospital bed days per
1000 persons in 2009 (Table 2), 185.2 days for CHD, and 121.2 days
for cerebrovascular diseases. This parameter is higher than the
total for the EU [4] but comparable to some European countries.
The number of GP and outpatient visits was 1603.2 per 1000
persons in 2009, with CHD representing 300.9 visits for 1000
persons and cerebrovascular diseases representing 286.6 (Table 2).
CVD cost Russian health care systems approximately RUR 121
billion or €3558.5 million in 2006 and (Table 3) RUR 229.5 billion or
€5204.4 million in 2009. The major component of CVD-related
health care expenditure was inpatient care, which accounted for
RUR 52.7 billion (€1545.1 million) in 2006 and RUR 109 million
(€2470.8 million) in 2009, representing 43.4% and 47.5% of total
health care costs, respectively. CVD-related medication expenditure was also a large cost, representing 27% in 2006 (RUR 32.6
billion) of total costs and 21% in 2007 (RUR 47.6 billion) of health
care costs. The third largest component of CVD-related health care
expenditure was outpatient care, which represented 20.2% of the
total health care costs in 2006 and 21.7% of the total health care
costs in 2009. The other two cost components (emergency care and
cardiosurgery) accounted for 9.5% in 2006 and 10.0% in 2009 of
costs, with A&E representing the smallest component. The structure of the CVD costs in Russia was similar to that of the EU [4],
where inpatient care and pharmaceutical expenditure were major
components of CVD-related health care expenditure.
Non–Health Care Costs
Mortality losses were relatively high in Russia because of high CVD
death rates in the general population and in the working age
population. The number of CVD deaths per 1000 persons was several
times higher than in Europe [4]. There is a prominent difference in
the number of deaths in working-age men and women. The death
rate of working-age men (younger than 60 years) was 3.2 in Russia in
2009, while for working-age women it was 0.7. This explains the
significant difference in working year losses between men and
women (22.1 and 3.4 for 1000 persons in 2009). In the EU, working
year losses in men were also higher than in women [4], but to a
lesser degree than in Russia. CHD accounted for approximately half
of CVD mortality losses, while values for cerebrovascular diseases
were much lower, particularly regarding working-year losses.
CVD accounted for 2.1 million working years lost in 2006 and
1.7 million in 2009 owing to deaths during working ages. GDP
losses due to mortality in working ages were estimated to cost
approximately RUR 630 billion (€18,483.3 million) in 2006 and RUR
117.1 billion (€16,509.6) in 2009, respectively, after adjusting for
working status (Table 3).
There were 70.1 million working days lost in 2006 and 67.3
million in 2009 because of CVD morbidity. This represented a cost
of RUR 83.8 billion (€2458.4 billion) in 2006 and RUR 117.4 billion
(€2662.5 billion) in 2009 after adjusting for working status
(Table 3). The costs of disability pensions were rather small,
€17.7 million in 2006 and €23.9 million in 2009.
Total Costs of CVD
Overall, CVD was estimated to cost the Russian economy RUR
836.1 billion (€24,517.8 million) in 2006 and RUR 1076 billion
(€24,400.4 million) in 2009 (Table 3). Of the total costs of CVD,
14.5% in 2006 and 21.3% in 2009 were due to health care, whereas
non–health care costs were 85.5% and 78.7%, respectively. The
structure of costs is completely different from the EU data, where
the health care cost was the most significant [4].
Share of GDP (CVD)
The total costs of CVD were equal to 3.1% of the GDP of the
Russian Federation in 2006 and 2.8% in 2009, respectively
(Table 4).
Costs of CHD and Cerebrovascular Diseases
The total costs of CHD were RUR 339.7 billion (€9962.2 million) in
2006 (Table 4) and RUR 406.6 billion (€9220.4 million) in 2009. Health
care costs were RUR 58.6 billion (€1719.0 million) and RUR 103.9
billion (€2355.2 million) in 2009. The share of health care costs
relative to total costs increased from 17.3% in 2006 to 25.5% in 2009.
The total costs of cerebrovascular diseases were much
lower than those of CHD. Total care costs were RUR 139.2 billion
(€4083.3 million) in 2006 and RUR 184.5 billion (€4183.9 million) in
2009. Health care cerebrovascular costs were RUR 18.8 billion
(€551.9 million) in 2006 (Table 4) and RUR 38.6 billion (€874.1
million) in 2009. The share of health care costs compared with
total costs increased from 13.5% in 2006 to 20.9% in 2009.
Table 1 – Average unit costs in Russia in 2006–2009.
2006
GDP per employed person
Annual
Daily
Health care unit cost
GP and outpatient visit
A&E visit
Inpatient day
2007
2008
Ruble
Euro
Ruble
Euro
389,021
1,195
11,408.2
35.0
467,463
1,445
13,668.5
42.3
115
686
754
3.4
20.1
22.1
149
888
947
4.4
26.0
27.7
A&E, accident and emergency; GDP, gross domestic product; GP, general practitioner.
Ruble
2009
Euro
Ruble
Euro
584,338
1,837
16,053.2
50.5
562,500
1,745
12,755.1
39.6
181
1,110
1,232
5.0
30.5
33.8
219
1,387
1,602
5.0
31.5
36.3
202
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 199–204
Table 2 – Resource units per 1000 population in the Russian Federation in 2009.
Mortality losses
Deaths
All
Men
Women
Working years lost
All
Men
Women
Morbidity losses
Working days lost
Health care unit
GP and outpatient visits
Hospital emergency visits
Inpatient days
CVD
CHD
Cerebrovascular diseases
8.0
7.8
8.2
4.1
4.2
4.0
2.6
2.2
3.0
11.1
22.1
3.4
4.5
9.1
0.9
1.9
3.5
0.8
474.2
80.1
73.3
1603.2
47.9
479.2
300.9
5.1
185.2
286.6
10.6
121.2
CHD, coronary heart disease; CVD, cardiovascular disease; GP, general practitioner.
structures of the total costs were completely different in Russia
and the EU. A European study found that health care expenditure accounted for 61% of costs, but in our study the share of
health care expenditures was only 14.5% of total costs in 2006
and 21.3% in 2009. If our study included the cost of informal
care, health care costs would be even lower. Non–health care
costs (mainly due to premature death during working age)
account for 80% of total CVD costs in Russia.
Direct costs associated with CVD in the EU included costs for
hospital admissions (57%), costs for drug therapy (27%), and costs
for ambulatory care and A&E (16%) [4]. In general, the structure of
direct costs in our study was similar.
In our study, the share of health care costs for CVD relative to
total costs was 13.5% to 20.9%. In the study of 27 EU countries, it
was shown that total costs from stroke were 68.5% due to direct
costs and 31.5% due to indirect costs [5]. In the United States, the
economic burden of stroke includes 67% of direct health care
costs and 33% of losses in the economy due to premature
mortality and disability days [6]. Thus, in Russia, the structure
of the economic burden of CVD is different from that due to the
predominance of indirect costs.
Sensitivity Analysis
Varying total health care costs upwards and downwards by 20%
produced a variation of 1% in the baseline of total CVD-related
costs in the period 2006 to 2009. Our results did not vary
significantly when the assumptions used to derive medication
cost estimates were varied by 50%, resulting in changes of 1% in
total costs.
Future earning losses at a 10% discount rate were associated
with a reduction of 8% in costs, while a 15% discount rate was
associated with a reduction of 11% in costs.
Discussion
This is the first study to estimate the full burden of CVD,
including health care and non–health care costs in Russia. We
estimated that the total burden of CVD varied between €24.4
and €32.3 billion in the period 2006 to 2009. In a recent European
study, the estimated cost of CVD was €169 billion [4]. Because
the population of EU is several times higher than that of Russia,
the amounts were comparable. We found, however, that the
Table 3 – Cost of CVD in Russia in 2006–2009.
2006
Inpatient care
Primary and outpatient care
A&E care
Cardiosurgery and PCI
Medications
Total health care costs
GDP losses due to mortality in
working age
GDP losses due to morbidity
Disability pensions
Total non–health care costs
Total costs
2007
2008
2009
RUR
million
€
million
RUR
million
€
million
RUR
million
€
million
RUR
million
€
million
52,687.9
24,469.5
4,032.4
7,517.5
32,637.0
121,344.3
630,280.4
1,545.1
717.6
118.3
220.5
957.1
3,558.5
18,483.3
65,791.5
32,385.6
5,614.5
9,451.6
35,249.0
148,492.3
687,087.1
1,923.7
946.9
164.2
276.4
1,030.7
4,341.9
20,090.3
85,283.7
40,262.8
7,410.2
11,701.9
40,124.4
184,783.0
860,872.9
2,343.0
1,106.1
203.6
321.5
1,102.3
5,076.5
23,650.4
108,960.9
49,833.0
9,425.8
13,714.5
47,578.7
229,513.0
728,075.2
2,470.8
1,130.0
213.7
311.0
1,078.9
5,204.4
16,509.6
83,830.7
602.2
714,713.3
836,057.6
2,458.4
17.7
20,959.3
24,517.8
104,067.2
666.7
791,821.0
940,313.3
3,042.9
19.5
23,152.7
27,494.5
127,864.4
833.8
989,571.0
1,174,354.0
3,512.8
22.9
27,186.0
32,262.5
117,416.1
1,054.1
846,545.4
1,076,058.3
2,662.5
23.9
19,196.0
24,400.4
A&E, accident and emergency; CVD, cardiovascular disease; GDP, gross domestic product; PCI, percutaneous coronary intervention; RUR,
rubles.
874.1
20.9
2,355.2
25.5
5,204.4
21.3
Cere, cerebrovascular disease; CHD, coronary heart disease; CVD, cardiovascular disease; GDP, gross domestic product; RUR, rubles.
844.3
15.6
2,311.8
17.8
697.0
15.1
4,341.9
15.8
1,719.0
17.3
551.9
13.5
2,021.7
18.3
5,076.5
15.7
38,550.0
103,862.9
229,513.0
30,731.2
84,148.4
184,783.0
23,837.6
69,141.8
148,492.3
18,821.2
58,619.1
4,183.9
0.47
9,220.4
1.04
24,400.4
2.8
5,408.0
0.48
12,962.5
1.14
32,262.5
2.8
4,621.2
0.48
11,028.7
1.14
27,494.5
2.8
4,083.3
0.52
9,962.2
1.26
184,511.9
406,618.8
1,076,058.3
196,851.7
940,313.3
Total costs
RUR
836,057.6
million
€ million
24,517.8
3.1
Share of
GDP
(%)
Health care costs
RUR
121,344.3
million
€ million
3,558.5
14.5
Share of
total
costs
(%)
339,712.3
139,239.6
1,174,354.0
158,044.4
377,182.3
471,835.9
CHD
CHD
Cere
CHD
CVD
Cere
CHD
CVD
2006
Table 4 – Cost of CVD, CHD, and cerebrovascular diseases in Russia in 2006–2009.
CVD
2008
2007
Cere
CVD
2009
Cere
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 199–204
203
Several studies demonstrated that health care costs for
various CVDs differ significantly depending on the country, the
length of hospitalization, and other factors [7–9].
CHD was the most costly CVD (37.8% of total CVD costs).
According to estimates of economic burden in the United States,
CHD was also the most expensive disorder [10]. In a European study,
CHD accounted for 27% of total costs [4]. A UK study showed that
CHD and stroke costs were similar (29% and 27% of total CVD costs,
respectively) [11]. In some studies, it was shown that myocardial
infarction typically accounts for the largest share of CHD costs
owing to long hospitalization and costly interventions [12,13].
Studies examining the economic burden of diseases enable
comparisons between the burdens of different diseases, allowing
decision makers to prioritize limited research funds to areas with
the highest burden [14]. Furthermore, if such studies are performed at regular intervals, the impact of health policy decisions
can be measured. In Russia, these studies can be used to monitor
governmental programs involved in chronic disease prevention.
The results of calculations performed in this study can be
used to plan investments in prevention programs and improve
care to patients with CVD. Regular monitoring of the economic
burden of CVD in the future at the federal, regional, and
municipal levels will allow the assessment of the dynamics of
economic burden, as well as the effectiveness of investments in
the economy in primary and secondary prevention.
The next step of such studies is to estimate the economic cost
of risk factors of chronic disease. This will provide support for
prioritizing resources for prevention and public health [15].
To be in a better position to inform policy decisions aimed at
reducing the burden of disease, improved information regarding
epidemiology and accurate information regarding resource use
and unit costs is imperative. In Russia, there is lack of recent
reliable epidemiology data as well as gaps in the official statistics
on resource use.
Our results are likely underestimated. Some categories of
costs, such as costs of informal care, were not included because
of data limitations.
Despite these acknowledged and important data limitations,
this is the first study to quantify the burden of CVD in Russia.
Acknowledgments
We are grateful to Michail Khudjakov for useful contributions to
this project. The comments from three anonymous reviewers are
also acknowledged.Source of financial support: The authors have
no other financial relationships to disclose.
Source of financial support: The authors have no other
relationships to disclose.
R EF E R EN C ES
[1] Population data for 2006–2009 from Russian Federal Service of State
Statistics (Rosstat). Available from: http://www.gks.ru/wps/wcm/
connect/rosstat/rosstatsite/main/ [Accessed October 16, 2010].
[2] World Bank. Dying too young: addressing premature mortality and ill
health due to non communicable diseases and injuries in the Russian
Federation. 2005: 145. Available from: http://siteresources.worldbank.org/
INTECA/Resources/Dying_too_Young_Summary_UPDATED_Oct_19.pdf
[Accessed October 16, 2010].
[3] Abegunde DO, Mathers CD, Adam T, et al. The burden and costs of
chronic diseases in low-income and middle-income countries. Lancet
2007;370:1929–38.
[4] Leal J, Luengo-Fernández R, Gray A, et al. Economic burden of
cardiovascular diseases in the enlarged European Union. Eur Heart J
2006;27:1610–9.
[5] European Heart Network. European Cardiovascular Disease Statistics
2008. Brussels: European Heart Network, 2008. Available from: http://
204
[6]
[7]
[8]
[9]
[10]
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 199–204
www.herzstiftung.ch/uploads/media European_cardiovascular_disease
_statistics_2008.pdf [Accessed October 16, 2010].
Go AS, Mozaffarian D, Roger VL, et al. Heart disease and stroke statistics—
2008 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation 2008;117:e25–146.
Kauf TL, Velazquez EJ, Crosslin DR, et al. The cost of acute myocardial
infarction in the new millennium: evidence from a multinational
registry. Am Heart J 2006;151:206–12.
Tiemann O. Variations in hospitalisation costs for acute myocardial
infarction—a comparison across Europe. Health Econ 2008;17(1, Suppl.):
S33–45.
Epstein D, Mason A, Manca A. The hospital costs of care for stroke in
nine European countries. Health Econ 2008;17(Suppl.):S21–31.
Druss BG, Marcus SC, Olfson M, Pincus HA. The most expensive
medical conditions in America. Health Aff 2002;21:105–11.
[11] Luengo-Fernández R, Leal J, Gray A, et al. Cost of cardiovascular
diseases in the United Kingdom. Heart 2006;92:1384–9.
[12] Reinhold T, Lindig C, Willich SN, Brüggenjürgen B. The costs of
myocardial infarction—a longitudinal analysis using data from a large
German health insurance company. J Public Health 2011;19:579–86.
[13] Ioannides-Demos LL, Makarounas-Kirchmann K, Ashton E, et al. Cost
of myocardial infarction to the Australian community: a prospective
multicenter survey. Clin Drug Investig 2010;30:533–43.
[14] Gross CP, Anderson GF, Powe NR. The relation between funding by the
National Institutes of Health and the burden of disease. N Engl J Med
1999;340:1881–7.
[15] Scarborough P, Bhatnagar P, Wickramasinghe KK, et al. The economic
burden of ill health due to diet, physical inactivity, smoking, alcohol
and obesity in the UK: an update to 2006–07 NHS costs. J Public Health
(Oxf) 2011;33:527–35.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 205–209
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Cost for Treatment of Chronic Lymphocytic Leukemia in Specialized
Institutions of Ukraine
Olena Mandrik, PhD1,2,*, Isaac Corro Ramos, PhD3, Olga Zalis’ka, PhD1, Andriy Gaisenko, PhD4, Johan L. Severens, PhD2,3
1
Danylo Halytsky Lviv National Medical University, Lviv, Ukraine; 2Institute of Health Policy & Management, Erasmus University Rotterdam, Rotterdam, The
Netherlands; 3Institute for Medical Technology Assessment, Erasmus University Rotterdam, Rotterdam, The Netherlands; 4National Cancer Institute, Kyiv,
Ukraine
AB STR A CT
Objective: The aim of this study was to identify, from a health care
perspective, the cost of treatment for chronic lymphocytic leukemia in
specialized hospitals in Ukraine. Methods: Cost analysis was performed by using retrospective data between 2006 and 2010 from
patient-file databases of two specialized hospitals (145 patients).
Uncertainty was assessed by using bootstrapping and multivariate
sensitivity analyses. Linear regression analysis was used to analyze
whether patients’ characteristics are related to health care costs. In
addition, one-way analysis of variance (Welch test) and pairedsample t test were conducted to compare mean costs of treatment
between the two hospitals and mean expenses for drugs and inhospital stay. Results: The average annual cost for a patient’s drug
treatment is 2047 EUR. The cost of hospitalization was significantly
lower (t ¼ 5.026; significance two-tailed ¼ 0.000) and equal to 541 EUR
per person, resulting in total expenditures of 2589 EUR. Mean total
costs in the bootstrap analysis were equal to 2584 EUR (median 2576
EUR, 97.5th percentile 3223 EUR; 2.5th percentile 1987 EUR). The
regression analysis did not reveal a relation between patients’ characteristics and health care costs, although hospital choice was an
influential parameter (β ¼ −0.260; significance ¼ 0.002). Significant
difference in mean costs of two analyzed hospitals was also confirmed by one-way analysis of variance (Welch statistics 19.222, P ¼
0.000). Conclusions: Drug treatment comprises the largest portion of
total costs, but differences between hospitals exist. Because many
patients in Ukraine pay out of pocket for in-hospital drugs, these costs
are a high economic burden for patients with chronic lymphocytic
leukemia.
Keywords: chronic lymphocytic leukemia, cost of treatment, hematologic
malignancies.
Introduction
indicators of a patient’s life expectancy to positive values [6].
For example, in the United States for the time period 1999 to 2005,
the 5-year survival rate for leukemia was 82% (79% for CLL),
although in the time period 1975 to 1977, this indicator was close
to Ukrainian data—35% [7,8].
In-hospital medical care for patients with CLL is generally
provided in 35 hematologic departments, based in district hospitals (16), state city hospitals (12), oncologic dispensaries (4), and
specialized institutes of the National Academy of Medical Science
of Ukraine (3) [2,9]. To the latter group belong two hematologic
institutes and the National Cancer Institute, which is a leading
state institution additionally responsible for methodological and
scientific development in this clinical area. Treatment schemes
for patients with CLL are based on a clinical protocol that
proposes a number of treatment options for patients with CLL
and was first developed and approved under an order of the
Ministry of Health of Ukraine in 2010 [10]. State pharmaceutical
provision for adult oncologic patients is granted through the
national treatment program “Oncology” for the years 2011 to
2016, although governmental financing is insufficient and drug
treatment is usually paid out of pocket by patients [11].
Globally, there are approximately 7.4 million cancer deaths per
year, which is approximately 13% of deaths from all causes.
Because the population of many countries around the world is
aging, it can be expected that cancer incidence will increase [1].
Among oncologic diseases, chronic hematologic malignancies are
comparatively rare. In Ukraine in 2010 the officially registered
total morbidity rate for patients with diagnosed leukemia was 7.8
per 100,000 people, of which 39.3% did not live a year after
diagnosis [2]. Chronic lymphocytic leukemia (CLL) is the most
frequent form of leukemia in Western countries, and it accounts
for approximately 30% to 40% of all leukemias [3,4]. It is characterized by the clonal proliferation and accumulation of neoplastic
B lymphocytes in the blood, bone marrow, lymph nodes, and
spleen. Although the median age of patients at diagnosis is
higher than retirement age and so has no significant impact on
state productivity loss [5], the economic impact of CLL is significant due to long duration and high expenses related to treatment,
combined with low cure rates. Nevertheless, early diagnosis and
effective treatment of hematologic malignancies shift the
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Olena Mandrik, Danylo Halytsky Lviv National Medical University, 69 Pekarska Street, Lviv, Ukraine 79010.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.006
206
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 205–209
Although CLL has a significant impact on patients’ quality of
life [12,13], studies exploring economic costs and burden of
hematologic malignancies are relatively rare in Englishlanguage publications worldwide [14,15]. Possible reasons for this
lack of information appear to include the low incidence rate and
aged study population (over 60 years old), which make broad,
well-designed economic analyses a challenge for most researchers [5,16]. These few cost reviews identified cost drivers for CLL as
chemotherapy costs, intravenous immunoglobulin costs, transplantation costs, and costs associated with the differential staining cytotoxicity assay, with the main cost drivers related to the
treatment chosen [14–16].
The health care system itself, including organization of
medical care for oncologic patients, is going through a stage of
transformation. Changes include the implementation of a universal reimbursement system to begin in 2016, standardization of
medical help with enhanced control on follow-up of clinical
protocols, and more strict division between primary, secondary,
and tertiary levels of medical help. Considering that the major
recipients of the central state budget are specialized institutions
(tertiary level of help), the primary aim of this research was to
identify the cost of treatment for CLL in specialized hospitals in
Ukraine from a health care perspective and to understand
whether patient characteristics are related to these costs.
Methods
The study was conducted from a health care perspective,
accounting for direct medical costs to illustrate which costs will
be paid by the Ministry of Health after the health care system
transformation.
Analysis included data from databases of two specialized
hospitals—National Cancer Institute and State Institute of Hematology—that receive state financing through the national treatment program “Oncology.” These hospital databases were made
in the programs Access and Word for the purpose of data storage
and included all the information available in hard copies of
hospital cards; the data were typed into the hospital databases
retrospectively by qualified personnel (hospital assistants). Afterward, data from the two hospital databases were transferred into
Excel, and SPSS databases were created for the purpose of data
analysis.
The study population included all newly diagnosed and
relapsed patients with CLL (145 in total) who were hospitalized
during the period from 2006 to 2010 and whose data were
recorded in the electronic database. The information derived
from the hospital cards (excluding patients’ identification information) contained the following data: sex of the patient, age
during diagnosis and treatment, number of years a patient lives
with the disease, year of treatment, therapies prescribed and
duration of treatment, the number of hospitalizations per year,
and the duration of hospitalization. Stage of the patient’s disease
and health state on Eastern Cooperative Oncology Group (ECOG)
performance status were excluded from the factor list because
data on these parameters were frequently missing.
Costs related only to CLL diagnosis for the last observational
year were calculated. These costs included drug expenses and inhospital costs. The cost of diagnostics, medical procedures, hotel
services, and medical personnel is included in the integral inhospital cost, based on data of the economic department of the
National Cancer Institute. These costs reflect the approximate
costs for oncologic patients in a specialized hospital and are
equal to 16.3 EUR per patient-day [17]. Out-of-hospital health care
costs were not calculated because according to the clinical
protocol [10], the treatment of patients with CLL should be
conducted only in hospital. The average length of hospital stay
and drug costs were assessed by retrospective analysis of patient
file data.
To assess drug usage, daily defined doses and total amount
received during the year were recorded. To calculate drug costs,
we used a stepwise approach to determine an average price,
depending on the availability of information: tariff in governmental purchases (2010); price, registered in the Ministry of
Health; and distributors’ price.
Multivariate sensitivity analysis was conducted. Price deviations for the sensitivity test of all drugs were calculated by using
the minimum and maximum prices from the available sources
(hospital purchases, state registered prices, and distributors’
prices). There are no defined general tariffs for hospital stay in
Ukraine, which are relatively low in comparison to medical costs
in the European countries and may vary from 3.4 to 19.2 EUR [17–
19]. All statistical analyses were performed in IBM SPSS Statistics
20 (SPSS, Inc., Chicago, IL), and bootstrapping (1000 replications)
was performed in Microsoft Excel 2010. To analyze whether
choice of the hospital and age and sex of a patient have an
impact on total health care costs, logarithmic data transformation was performed on nonnormally distributed costs and a
linear regression analysis was applied. Because of frequently
missing data for the parameter “stage of the disease,” as a proxy
for disease progression we included “number of years a patient is
living with the disease” in the linear regression analysis. Based
on Cook’s distance (0.028571), which measures the effect of
deleting a given observation and so allows to define data points
with large residuals, we excluded six outliers to improve the
residuals plot and model validity. One-way analysis of variance
(Welch test) was conducted to compare mean costs of treatment
in the two hospitals involved (asymptotically F distributed).
Paired-sample t test was used to compare difference in mean
expenses for drugs and in-hospital stay.
Results
Overall, data of 113 patients from the first hospital (State
Hematology Institute) and of 32 patients from the second hospital
(National Cancer Institute) were analyzed. Patients were aged 40
to 85 years (mean age 62.9 years, mean age during diagnosis 60.3
years, SD 9.8 years). From the sample, 27.6% (40) of the patients
were newly diagnosed. Because of limited sample size, the
distribution of patients by sex was not equal in different age
groups, with the total proportion of men being equal to 60.7%
(88 men).
Values of the cost items (drugs) and cost deviations for the
sensitivity test are presented in Table 1. As can be seen in Table 1,
the highest cost per milligram was for fludarabine, vincristine,
and rituximab. Rituximab and fludarabine (if Fludara was prescribed) had the highest price per average daily dose, equal to
312.03 EUR for fludarabine and 237.93 EUR for rituximab. Because
drug expenditures depend not only on cost per item but also on
total volume used, we present cost-items utilization and characteristics of population using it in Table 2. Data are presented for
items that were used by more than 3% of patients. Cyclophosphamide, fludarabine, and vincristine were prescribed to most of
the patients. Characteristics of the study population using specific cost items showed a significant difference in the percentage
of men prescribed cyclophosphamide, mitoxantrone, and chlorambucil in comparison to other drugs. No significant difference in
the patients’ age was observed, although on average the age of
patients receiving alemtuzumab was lower and of those receiving
chlorambucil was higher. From Table 2 it also may be observed
that the injectable form of fludarabine is prescribed significantly
more than the oral form.
207
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 205–209
Table 1 – Price values of the drugs for the cost and sensitivity analyses.
Drugs*
Alemtuzumab (inj.)
Bleomycin (inj.)
Chlorambucil
Cyclophosphamide (Adriablastine)
Cyclophosphamide (other generics)
Dexamethasone (inj.)
Dexamethasone (po)
Doxorubicin (inj.)
Etoposide phosphate (po)
Fludarabine (Fludara inj.)
Fludarabine (Netran inj.)
Fludarabine (Netran po)
Fludarabine (other generics)
Methylprednisolone (inj.)
Methylprednisolone (po)
Mitoxantrone (inj.)
Prednisolone (inj.)
Prednisolone (po)
Rituximab (inj.)
Vincristine (inj.)
Vinblastine (inj.)
Base-case price
(EUR per mg)
Range used in sensitivity
analysis (EUR per mg)
0.4500
1.5100
0.0080
1.0600
0.0077
0.0200
0.0002
0.1300
0.1000
3.7400
0.7700
0.1500
2.7800
0.1100
0.0260
0.3800
0.0160
0.0072
2.0300
3.6700
0.6300
–
1.5100–1.6300
0.0080–0.2290
0.6700–1.4500
0.0010–0.0120
0.0040–0.0840
–
0.1300–0.4400
0.0980–0.1000
3.2200–3.7400
–
–
0.7600–3.3700
0.0140–0.1100
0.0220–0.0300
0.3800–3.8300
0.0130–0.0170
0.0072–0.0077
1.3500–2.9600
2.8600–4.0600
0.4000–0.6300
Source†
Tariff in governmental purchases
Distributors’ price
Tariff in governmental purchases
Price, registered in the MOH
Tariff in governmental purchases
Price, registered in the MOH
Price, registered in the MOH
Tariff in governmental purchases
Price, registered in the MOH
Tariff in governmental purchases
Tariff in governmental purchases
Tariff in governmental purchases
Price, registered in the MOH
Price, registered in the MOH
Price, registered in the MOH
Tariff in governmental purchases
Price, registered in the MOH
Price, registered in the MOH
Tariff in governmental purchases
Tariff in governmental purchases
Price, registered in the MOH
2010
2010
2010
2010
2010
2010
2010
2010
2010
2010
inj., injection; MOH, Ministry of Health; po, per os (by mouth).
Trade name is indicated if the product was prescribed specifically by it.
†
To value the use of the drugs from a health care perspective, we used a stepwise approach to determine an average price, depending on the
availability of information: tariff in governmental purchases (2010); price registered in the MOH; distributors’ price.
The average annual cost for a patient’s drug treatment is 2047
EUR. The average cost of in-hospital stay is 542 EUR per person,
resulting in total expenditures of 2589 EUR. Results indicate that
expenses for drugs significantly exceed hospitalization costs (t ¼
5.026; significance two-tailed ¼ 0.000).
Mean total cost in the bootstrap analysis was 2584 EUR
(median 2576, 97.5th percentile 3223 EUR; 2.5th percentile 1987
EUR). Sex of the patient, number of years a patient lives with the
disease, and age at the time of hospitalization had no significant
impact on health care costs per patient. Hospital choice (β ¼
−0.260; significance ¼ 0.002), however, was a strong determinant
of health care costs. One-way analysis of variance also showed a
significant difference in mean costs between the two hospitals
involved (Welch statistics 19.222, P ¼ 0.000).
The results of the multivariate sensitivity analysis showed
that in the best-case (lowest cost) scenario, the average annual
spending on drug treatment of a patient with CLL is 1659 EUR,
and in the worst-case scenario, it is 2332 EUR. The deviation of
drug costs does not exceed 12% on the negative side and 19% on
the positive side. The annual cost of hospitalization ranges from
251 to 597 EUR per person and depends on the type of hospital at
which a patient is treated.
Discussion
A literature review was conducted in the database PubMed to
explore whether our results were consistent with results from
studies in other countries and to understand whether factors that
impact the cost of other cancer conditions are similar to those
affecting CLL. The search was limited to a 10-year period of Englishlanguage articles studying multiple cancer conditions. The literature
review showed that the major factors influencing the cost of cancer
conditions are related to patients’ characteristics, such as stage at
diagnosis and stage at treatment, degree of comorbidity, age and
gender of a patient, and tumor site. Lal et al. [20], Longo et al. [21,22],
and Yabroff et al. [23] recorded increased costs due to higher stage
of the disease during treatment. Akushevich et al. [24], in a
retrospective analysis on oncologic patients in the United States,
determined that the highest costs exist in the period of treatment
immediately after diagnosis. Yabroff et al. [23] also recorded that
both the first stage and the last stage of the disease at the time of
treatment are associated with higher costs. The results of the
studies by Lal et al. [20] and Kuse et al. [25] demonstrated a
connection between the degree of comorbidity and treatment costs.
The impact of patient’s age on the cost of cancer was significant in a
number of studies, but differed in scale and type of impact [20–
22,24,25]. Some research described an impact of tumor site on total
costs of the diseases [21,22,24]. Yabroff et al. [23] showed that cost
for the treatment of localized diseases is lower, a conclusion
supported by Lai et al. [20] who noted that the highest costs were
for hematological malignancies among other types of cancer.
A limited number of economic analyses that describe factors
influencing the cost of CLL treatment specifically were found.
These studies showed a positive correlation between age and
cost of drug treatment [6,26]. Danese et al. [26] also concluded
that male gender is associated with higher cost for CLL drug
treatment.
Similar to Yabroff et al. [23], CLL phase-specific health care
costs for the US-Medicare population were found to have a Ushaped pattern over lifetime [27]. A study in the United States by
Lafeuille et al. [27] also reported significantly higher health care
costs for a CLL population compared to matched controls, mainly
because of the higher costs for physicians, caregivers, and
inpatient care. Besides inpatient hospital stay, pharmaceuticals
were the main cost drivers of CLL in a study in Germany [28],
where the cost of treatment per case is about twice as high as the
cost per case for highly prevalent diseases, such as chronic
obstructive pulmonary disease or diabetes. This study also
revealed that the average annual cost for patients with CLL from
208
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 205–209
Table 2 – Drug utilization related to patients’ population characteristics.
Drugs*
Patients
using the
drug (%)
Average age (SD) of
patients using this
item (y)
Men among ones
who are using this
drug
Alemtuzumab (inj.)
Chlorambucil (po)
Cyclophosphamide
(Adriablastine inj.)
Cyclophosphamide
(all brand names)
Dexamethasone
(inj.)
Dexamethasone (po)
Fludarabine (Fludara
inj.)
Fludarabine (po)
Fludarabine inj. (all
brand names)
Methylprednisolone
(po)
Mitoxantrone (inj.)
Prednisolone (inj.)
Prednisolone (po)
Rituximab (inj.)
Vincristine (inj.)
Mean
volume per
patient (mg)
n (%)
Prescriptions during
the first year of
treatment
10.3
7.6
4.8
57.90 ⫾ 9.25
74.45 ⫾ 4.61
64.29 ⫾ 5.85
9 (60.00)
5 (45.50)
6 (85.70)
0 (0.00)
6 (54.5)
4 (57.10)
87.14
4.76
7.24
66.2
62.95 ⫾ 9.13
59 (61.50)
29 (30.20)
2900.34
10.0
60.07 ⫾ 8.71
19 (65.50)
7 (24.10)
22.70
5.4
44.4
59.20 ⫾ 6.50
60.56 ⫾ 8.82
3 (60.00)
36 (57.10)
1 (20.00)
13 (20.60)
4.41
377.87
3.4
47.6
59.80 ⫾ 8.35
60.65 ⫾ 8.82
3 (60.00)
40 (58.00)
0 (0.00)
13 (18.80)
30.48
411.79
4.8
56.29 ⫾ 10.42
4 (57.1)
1 (14.3)
22.59
5.5
5.5
26.2
12.4
31.0
58.12
60.63
66.63
59.28
64.51
⫾ 6.31
⫾ 10.64
⫾ 8.07
⫾ 6.72
⫾9.57
7
5
26
13
34
(87.50)
(62.50)
(68.40)
(72.20)
(75.60)
0
2
14
5
17
(0.00)
(25.00)
(36.80)
(27.80)
(37.80)
4.35
224.63
179.37
258.62
1.66
inj., injection; po, per os (by mouth).
Trade name is indicated if a product was prescribed not by generic but by a trade name with a high frequency (for cyclophosphamide and
fludarabine).
the sickness fund perspective decreased with increasing age until
60 to 65 years, and thereafter increased.
Results of these US and German studies [27,28] differed from
the results of our analysis conducted in Ukraine, where the cost
of medical care is relatively low and the major expenses are drug
related. Similar to our results, a literature review conducted by
Stephens et al. [16] concluded that the cost of drug therapy is the
main driver for CLL treatment costs, significantly exceeding
hospitalization costs.
Previous research has suggested that the major factors
influencing the cost of cancer conditions are stage at diagnosis
and stage at treatment, degree of comorbidity, age and gender,
tumor site, and type of therapy received. Our study on a
Ukrainian sample from two specialized institutions, however,
showed that only hospital choice had a significant impact on the
cost of drug treatment. Possible explanations may be risk-patient
selection or difference in treatment practice within the hospitals.
High usage of injectable forms of drugs such as fludarabine and
dexamethasone also was observed in this study. Because no
health technology assessment agency currently exists, there are
no recommendations comparing injectable and oral forms developed in Ukraine. Nevertheless, the National Institute for Health
and Clinical Excellence [29] recommends giving preference to the
oral form of fludarabine because of its higher efficiency. Rationality of the use of the injectable form of fludarabine in CLL
treatment practice may be a potential topic for further research
in Ukraine.
Ukraine is a country with a post-Semashko model of the
health care system, and currently there is no state reimbursement system. Limited reimbursement for in-hospital treatment,
however, is provided under governmental programs for such
diseases as AIDS, tuberculosis, diabetes, cardiovascular diseases,
and cancer, among other diseases. These in-hospital state
purchases cover from 7% to 40% of oncologic patients’ needs
depending on the region and hospital type [30,31]. Major
expenses on drugs are covered by patients’ out-of-pocket payments. Thus, the high treatment cost of chronic conditions such
as CLL may be a significant economic burden, especially for
patients with low income.
The average annual cost of drug treatment for patients with
CLL is 2047 EUR, with the majority of costs paid out of pocket.
From December 1, 2011, the minimum annual subsistence level
in Ukraine is equal to 1155 EUR for people of working age and 920
EUR for those who are retired. These figures are lower than
annual costs of drug treatment for patients with CLL in Ukraine,
according to current clinical practice in specialized hospitals.
This may impose a significant economic impact of the disease on
vulnerable populations (e.g., elderly poor), taking into account
only limited governmental subsidy. With high costs for the
treatment of hematologic malignancies [15,26], and insufficient
reimbursement level for drug treatment in Ukraine [9], the treatment of CLL in specialized hospitals may be financially difficult
for economically unprotected patients because of high
therapy costs.
Implications
Our analysis indicates that there is likely to be a significant
difference in the practice of treating CLL within different hospitals of Ukraine, resulting in a significant deviation in drug
expenditures. Therefore, it is not clear whether treatment standards are being followed within the hospitals and whether the
schemes used are evidence based and rational. These issues
should be explored further in future studies.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 205–209
Limitations
Retrospective analysis allowed us to make an estimation of
treatment cost for CLL in specialized medical institutions of
Ukraine and to explore its correlation with patient characteristics. Our research, however, suffered from several limitations.
First, our research does not allow us to assess the economic
burden of CLL in Ukraine. It is expected that costs of treatment in
the current study may be higher than in the regional oncology
dispensaries because of larger state subsidiaries and patients’
expenditures on drugs.
Moreover, data on stage of the patient’s disease and health
state on Eastern Cooperative Oncology Group (ECOG) criterion
were missing and thus excluded from the factor list. It is possible,
however, that these factors may have an impact on CLL costs.
We conducted a linear regression analysis on logarithmictransformed costs data while excluding six outliers on the basis
of Cook’s distance. Disregarding these six observations may have
increased the significance of the statistical analysis and the
strength of the relation between the independent variable (hospital choice) and the dependent variable (health care costs).
Conclusions
Drug treatment comprises the largest portion of total costs, which
presumably may be a high economic burden for a patient with CLL
who is the major payer of treatment expenses in Ukraine. Costs of
drug treatment significantly depend on the type of hospital selected.
Acknowledgment
We thank Carter Mandrik, PhD, for help in manuscript editing.
Olena Mandrik is employed by MSD Ukraine and affiliated with
Erasmus University Rotterdam, Institute of Health Policy &
Management by means of a PhD-hospitality agreement. The
views expressed in this article are those of the authors and
should not be attributed to the authors’ employers.
Source of financial support: The authors have no other financial
relationships to disclose.
R EF E R EN CE S
[1] Dranitsaris G, Truter I, Lubbe MS, et al. Advances in cancer
therapeutics and patient access to new drugs. Pharmacoeconomics
2011;29:213–24.
[2] National Cancer Institute. National Cancer Register. 2011;12. Available
from: http://users.iptelecom.net.ua/ucr/eng/index_e.htm. [Accessed
December 3, 2011].
[3] Cheson BD, Bennett JM, Grever M, et al. National Cancer Institute
sponsored working group guidelines for chronic lymphocytic leukemia:
revised guidelines for diagnosis and treatment. Blood 1996;87:4990–7.
[4] Kalil N, Cheson BD. Management of chronic lymphocytic leukaemia.
Drugs Aging 2000;16:9–27.
[5] Redaelli A, Botteman MF, Stephens JM, et al. Economic burden of acute
myeloid leukemia: a literature review. Cancer Treat Rev 2004;30:237–47.
[6] Rozman C, Montserrat E. Chronic lymphocytic leukemia. N Engl J Med
1995;333:1052–7.
[7] American Cancer Society. Cancer facts and figures. 2010. Available
from: http://www.cancer.org/acs/groups/content/@nho/documents/
document/acspc-024113.pdf. [Accessed February 7, 2012].
[8] National Cancer Institute. Surveillance, Epidemiology, and End Results
Program, 1975–2005, Division of Cancer Control and Population
Sciences. 2008. Available from: http://seer.cancer.gov/csr/1975_2005/.
[Accessed February 7, 2012].
209
[9] [Achievements and challenges of the domestic haematology].
Ukraiinskiy medyuchnuy chasopyus. 2011. Available from: http://
www.umj.com.ua/wp-content/uploads/2011/06/Hematolodji.pdf.
[Accessed March 9, 2013].
[10] Order of the Ministry of Health of Ukraine #647 from 30.07.2010. About
approval of the clinical protocols on medical help provision for
patients on specialty “Hematology.”.
[11] Order of the Ministry of Health of Ukraine #769 from 13.09.2010.
About approval of the concept of pharmaceutical sector development
in health care branch in Ukraine for the years 2011–2020.
[12] Levin TT, Li Y, Riskind J, Rai K. Depression, anxiety and quality of life in
a chronic lymphocytic leukemia cohort. Gen Hosp Psychiatry
2007;29:251–6.
[13] Shanafelt TD, Bowen D, Venkat C, et al. Quality of life in chronic
lymphocytic leukemia: an international survey of 1482 patients. Br J
Haematol 2007;139:255–64.
[14] Goor KM, Schaafsma MR, Huijgens PC, van Agthoven M. Economic
assessment on the management of chronic lymphocytic leukaemia.
Expert Opin Pharmacother 2005;6:1179–89.
[15] Kasteng F, Sobocki P, Svedman C, Lundkvist J. Economic evaluations of
leukemia: a review of the literature. Int J Technol Assess Health Care
2007;23:43–53.
[16] Stephens JM, Gramegna P, Laskin B, et al. Chronic lymphocytic
leukemia: economic burden and quality of life: literature review. Am J
Ther 2005;12:460–6.
[17] Mandrik O, Zalis’ka O. [Assessment of pharmacoeconomic aspects
of chronic lymphocytic leukemia treatment in Ukraine].
Upravlinnya, economica, ta zabezpechenya yakosti v farmacii
2012;3:62–7.
[18] Gorlyachenko O, Shulgay AG, Gorlyachenko A, et al. [Cost of medical
help]. Novosti medicinyu I farmacii 2012;5(403): Available from:
http://www.mif-ua.com/archive/article_print/27173. [Accessed March
9, 2013].
[19] Reforming of the secondary medical help in Ukraine: basic
problematic and solution options]. Combined report on the project
EuropeAid/123236/C/SER/UA. EPOS Health Consultants/NI-CO/
ECORY. Available from: Shttp://www.eu-shc.com.ua/UserFiles/File/
SR_V04_ua.pdf. [Accessed March 9, 2013].
[20] Lal A, Bhurgri Y, Rizvi N, et al. Factors influencing in-hospital length
of stay and mortality in cancer patients suffering from febrile
neutropenia. Asian Pac J Cancer Prev 2008;9:303–8.
[21] Longo CJ, Deber R, Fitch M, et al. An examination of cancer patients’
monthly ‘out-of-pocket’ costs in Ontario, Canada. Eur J Cancer Care
(Engl) 2007;16:500–7.
[22] Longo CJ, Fitch M, Deber RB, Williams AP. Financial and family burden
associated with cancer treatment in Ontario, Canada. Support
Care Cancer 2006;14:1077–85.
[23] Yabroff KR, Lamont EB, Mariotto A, et al. Cost of care for
elderly cancer patients in the United States. J Natl Cancer Inst
2007;99:14–23.
[24] Akushevich I, Kravchenko J, Akushevich L, et al. Medical cost trajectories
and onsets of cancer and noncancer diseases in US elderly population.
Comput Math Methods Med 2011. http://dx.doi.org/10.1155/2011/857892.
[25] Kuse R, Colberg H, Marbé W, et al. Which factors render costcovering lump-sum charging difficult for the treatment of patients
with acute leukemias? Onkologie 2001;24:292–4.
[26] Danese M, Gleeson M, Reyes C, et al. Cost of chronic lymphocytic
leukemia (CLL) in Medicare patients. J Clin Oncol 2008;26(Suppl): abstr
17531.
[27] Lafeuille MH, Vekeman F, Wang ST, et al. Lifetime costs to Medicare of
providing care to patients with chronic lymphocytic leukemia. Leuk
Lymphoma 2012;53:1146–54.
[28] Blankart CR, Koch T, Linder R, et al. Cost of illness and economic
burden of chronic lymphocytic leukemia. Orphanet J Rare Dis 2013;8:32.
[29] Guidance on the Use of Fludarabine for B-Cell Chronic Lymphocytic
Leukaemia. NICE technology appraisal guidance TA29. London, UK:
National Institute for Health and Care Excellence, 2001. Available from:
http://publications.nice.org.uk/
guidance-on-the-use-of-fludarabine-for-b-cellchronic-lymphocytic-leukaemia-ta29/guidance. [Accessed March 9, 2013].
[30] [On efficiency of costs use by the Ministry of Health]. State finance
inspection of Ukraine. Available from: http://www.dkrs.gov.ua/kru/uk/
publish/article/75100;jsessionid=5305963B5762BAF2C3E27BEAC3E2A6F6.
[Accessed March 9, 2013].
[31] Adamchuk I. [The program exists, where is effectiveness?]. State
finance inspection of Ukraine. Available from: http://www.dkrs.gov.ua/
kru/uk/publish/printable_article/51727;jsessionid=BB368B7E53A135A1B
F1A7994BC73505E. [Accessed March 9, 2013].
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 210–217
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Costs of Medically Attended Acute Gastrointestinal Infections: The Polish
Prospective Healthcare Utilization Survey
Marcin Czech, PhD, MD, MBA1,*, Magdalena Rosinska, PhD, MD2, Justyna Rogalska, MSc2, Ewa Staszewska, MSc2,
Pawel Stefanoff, PhD, MD, MSc2
1
Faculty of Pharmacy, Department of Pharmacoeconomics, Medical University of Warsaw, Warsaw, Poland; 2Department of Epidemiology, National Institute of
Public Health–National Institute of Hygiene, Warsaw, Poland
AB STR A CT
Objectives: The burden of acute gastrointestinal infections (AGIs)
on the society has not been well studied in Central European
countries, which prevents the implementation of effective, targeted
public health interventions. Methods: We investigated patients of
11 randomly selected general practices and 8 hospital units. Each
patient meeting the international AGI case definition criteria was
interviewed on costs incurred related to the use of health care
resources. Follow-up interview with consenting patients was conducted 2 to 4 weeks after the general practitioner (GP) visit or
discharge from hospital, collecting information on self-medication
costs and indirect costs. Costs were recalculated to US dollars by
using the purchasing power parity exchange rate for Poland.
Results: Weighting the inpatient costs by age-specific probability
of hospital referral by GPs, the societal cost of a medically attended
AGI case was estimated to be US $168. The main cost drivers of
direct medical costs were cost of hospital bed days (US $28), cost of
outpatient pharmacotherapy (US $20), and cost of GP consultation
(US $10). Patients covered only the cost of outpatient pharmacotherapy. Considering the AGI population GP consultation rate, the
age-adjusted societal cost of medically attended AGI episodes was
estimated at US $2222 million, of which 53% was attributable to
indirect costs. Conclusions: Even though AGIs generate a low cost
for individuals, they place a high burden on the society, attributed
mostly to indirect costs. Higher resources could be allocated to the
prevention and control of AGIs.
Keywords: direct medical costs, direct nonmedical costs, gastrointestinal
infections, indirect costs, Poland.
Introduction
caused by other pathogens have not been well documented,
although recent evidence indicates that costs for AGIs caused by
different etiological factors can be similar [10].
Poland, located in Central Europe, with its 38.4 million habitants is
the 34th most populous country in the world and the 6th largest
country in the European Union. The gross national product per capita
is almost US $16,710 [11]. The overall quality of health care provision
nationwide, as judged by European standards, is regarded as being
high, which is reflected in the nation’s average life expectancy,
estimated at 71 years for men and 80 years for women [11]. Poland’s
health care system is based on an all-inclusive social insurance
system. An insured person and members of his or her family are
entitled to free health services offered by providers who have signed
contracts with the National Health Fund. The National Health Fund is
a state-owned third-party payer. It finances health services and
assures reimbursement of medicines. There is a rapidly growing
private sector consisting of private general practitioners’ (GPs’) and
specialists’ practices and rarely private clinics and hospitals, paid on
a fee-for-service or prepaid basis. The drugs reimbursement system
grants unrestricted access to essential drugs and different level of copayment for the remaining specialty medicines (depending on the
Acute gastrointestinal infections (AGIs) can be caused by viruses,
bacteria, and parasites, as well as through physical or chemical
intoxications. Symptoms and high incidence of AGIs may put a
substantial burden on patients and the health care system—from
medical, social, and economic perspectives. An estimation of the true
burden of AGI symptoms on the society, however, is difficult [1]. In
developed countries, AGIs are common but usually do not cause
severe disease. Sufferers frequently downplay its significance, and
doctors often do not trace the causes of individual cases. The
majority of AGIs in Poland are not treated at all or are treated with
rehydration and/or over-the-counter drugs [2,3]. For these reasons,
economic consequences of AGIs are not properly assessed and thus
their true burden is underestimated.
Available scarce evidence indicates that the societal cost of the
so-called mild gastrointestinal illnesses is considerably higher than
the costs associated with acute hospitalized cases [4]. To date,
high-quality evidence on burden and costs to the society has been
collected mostly for rotavirus-associated AGIs—in relation to both
ambulatory care [5,6] and hospital settings [7–9]. Costs for AGIs
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Marcin Czech, Faculty of Pharmacy, Department of Pharmacoeconomics, Medical University of Warsaw,
Zwirki i Wigury 81, 02-091 Warsaw, Poland.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.011
211
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 210–217
Anatomical Therapeutic Chemical [ATC] Classification System group,
disease, patient’s status, etc.).
Currently in Poland, no reliable cost data in health care system
are available [12]. These data are not recorded by health care
institutions, because GP practices are reimbursed for the number
of registered patients and hospital admissions are reimbursed on
the basis of discharge codes of the International Statistical Classification of Diseases, 10th Revision, not depending on the health care
procedures used. Consequently, there are no health care cost
databases in Poland that are similar to those in other developed
countries, such as the National Health Service database in the
United Kingdom [13].
The aim of the present article was to summarize direct
medical, nonmedical, and indirect costs related to the management of AGI cases to estimate its societal costs and to enable
better planning and monitoring of AGI prevention programs in
the future.
Methods
We estimated AGI-related resource use linked to GP consultations
and hospitalizations by using a prospective study conducted
between May 2008 and September 2009. We have described
detailed methods for study site selection and AGI cases recruitment in previous publication [3].
Studied Population
The Polish Prospective Healthcare Utilization Survey was performed in a randomly selected population sample served by 11
Table 1 – List of unit costs generated by AGI cases consulting GPs in Poland between May 2008 and September 2009.
Detailed cost categories
GP consultation
Laboratory testing
Complete blood count
Stool sample for microbiological
confirmation
Other (blood smear, electrolytes,
etc.)
Imaging diagnostics
(ultrasonography)
Pharmacotherapy (ATC)
Rehydration fluids (−)
Intestinal anti-infectives (A07A)
Antidiarrheal microorganisms
(A07F)
Other probiotics (−)
Other antidiarrheals (A07X)
Drugs for functional
gastrointestinal disorders (A03)
Analgesics and anti-inflammatory
products (N02B, M01A)
Antibacterials for systemic use (J01)
Other pharmaceuticals (A01A,
A02B, A07D, A09A, R05C, R06A,
V06D, no ATC code)
Materials used
Medical devices (syringes, needles,
ports, fluid transfusion sets)
Biological specimen containers
Other (gloves, swabs, disposable
towels, gowns, disinfectants,
soaps, etc.)
Emergency department visits
Specialist consultation
Transport (number of kilometers)
Cost per kilometer number of
kilometers
Home care (family, friends, service)
Absence from work (number of days
off work)
Cost per day off number of days
Unit cost per patient (US $)
Number
Total cost (US $)
Mean
95% CI
Median
95% CI
Paid by
payer†
Paid by
patient
385/385
14/385
10/385
10.25
19.26
3.52
9.64–10.86
13.91–24.62
3.02–4.03
8.14
19.46
3.74
8.05–8.21
12.40–26.52
2.83–4.65
3946.54
269.70
35.22
0.00
0.00
0.00
9/385
14.17
11.14–17.21
13.51
9.17–17.85
127.57
0.00
10/385
10.69
7.58–13.79
10.24
6.45–14.03
106.91
0.00
11/385
17.28
14.46–20.10
13.51
6.60–20.43
190.11
0.00
382/385
230/385
229/385
20.71
6.72
5.77
19.86–21.55
6.18–7.25
5.53–6.02
19.33
5.06
4.75
18.51–20.15
4.76–5.37
4.02–5.47
985.11
442.67
0.00
6924.65
1102.28
1322.01
144/385
8.30
8.03–8.57
8.83
–
0.00
89/385
127/385
8.97
7.59
8.14–9.84
7.48–7.70
8.03
7.53
–
–
0.00
0.00
798.83
963.81
81/385
5.12
4.36–5.89
4.59
4.37–4.81
41.81
373.22
82/385
4.02
3.28–4.77
1.74
0–3.58
67.32
262.42
60/385
8.02
6.68–9.35
7.35
6.85–7.85
159.89
320.94
68/385
10.63
8.82–12.44
8.47
3.32–13.62
273.42
449.19
66/385
0.34
0.17–0.51
0.09
0.04–0.14
22.30
0.00
1195.4
19/385
0.37
0.19–0.55
0.15
0.12–0.28
7.01
0.00
4/385
1.29
0.11–2.46
0.95
0–2.43
5.16
0.00
66/385
0.15
0.10–0.20
0.09
0.04–0.14
10.14
0.00
4/115
7/115
82/115
54.05
17.78
36.96
–
–
17.91–56.01
54.05
17.78
13.00
–
–
9.21–16.79
216.22
124.45
–
0.00
0.00
–
82/115
16.84
8.02–25.66
6.32
4.26–8.39
0.00
1380.85
34/115
127.98
64.37–29.42
62.89
36.11–89.66
4,351.30
54.05
10/115
6.10
10/115
511.48
3.89–8.31
330.91–692.04
6.00
503.09
3.03–8.97
262.75–743.44
–
5,114.77
–
0.00
AGI, acute gastrointestinal infection; ATC, Anatomical Therapeutic Chemical Classification System; CI, confidence interval; GP, general
practitioner; WHO, World Health Organization.
US $1 ¼ 1.85 PLN (WHO CHOICE).
†
State institution (National Health Fund or social insurance).
212
Table 2 – List of unit costs generated by AGI cases admitted to a hospital in Poland between May 2008 and September 2009.
Detailed cost categories
Unit cost per patient (US $)*
Total cost (US $)*
Mean
95% CI
Median
95% CI
Paid by payer†
504/504
504/504
503/504
497/504
384/504
310/504
479/504
494/504
308/504
385/504
325/504
149/504
467/504
351/504
78/504
42/504
50/504
288/504
251/504
178/504
96/504
109/504
84/504
504/504
492/504
163/504
156/504
39/504
55/504
136/504
179/504
4.74
832.63
49.74
4.03
3.27
4.51
2.34
6.57
3.24
2.76
4.21
5.13
3.7
13.68
5.83
13.99
8.55
13.19
32.29
18,00
13.19
10.06
30.21
15.57
5.72
3.09
1.80
2.43
1.68
1.36
15.08
4.41–5.06
782.84–882.43
47.36–52.10
3.86–4.19
3.14–3.40
4.29–4.72
2.24–2.44
6.17–6.98
3.10–3.38
2.62–2.89
3.03–5.39
4.47–5.79
3.54–3.86
12.89–14.46
5.22–6.43
10.83–17.15
7.43–9.67
12.1–14.28
28.89–35.69
16.63–19.37
11.42–14.96
8.29–11.84
24.43–35.98
13.98–17.15
5.25–6.18
2.58–3.6
1.45–2.15
1.98–2.87
1.30–2.06
1.08–1.65
13.06–17.09
4.00
688.84
47.84
4.05
3.24
3.62
2.16
4.59
3.24
2.16
2.16
3.24
3.78
9.73
4.86
9.73
8.65
10.27
24.32
20,00
9.73
5.41
23.94
9.70
4.25
2.19
1.10
2.22
1.17
0.63
11.73
3.86–4.14
644.99–732.70
44.46–50.95
3.96–4.15
–
2.08–5.16
–
–
3.21–3.28
1.63–2.70
2.05–2.27
2.49–3.99
3.26–4.30
9.26–10.20
–
9.33–10.13
5.48–11.82
8.18–12.36
23.16–25.49
19.15–20.85
8.51–10.95
2.88–2.52
18.61–29.26
8.70–10.71
3.69–4.81
1.92–2.47
0.87–1.32
1.57–2.87
0.83–1.51
0.35–0.90
10.41–13.05
–
419,647.46
25,015.72
2,002.53
1,254.95
1,396.70
1,120.80
3,248.30
999.24
1,062.27
1,369.04
764.65
1,728.10
4,800.70
454.35
587.57
427.57
3,798.95
8,104.74
3,204.39
1,266.35
1,096.91
2,537.09
7,798.49
2,813.16
503.55
280.50
94.66
92.15
185.37
2,698.90
–
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
338/504
3.34
2.74–3.95
1.71
1.51–1.90
1,130.20
0.00
71/504
504/504
504/504
17.78
30.26
3.89
15.22–20.33
25.99–34.53
3.59–4.19
16.22
12.42
2.98
15.42–17.02
10.72–14.13
2.70–3.25
1,262.16
15,250.69
1,959.83
0.00
0.00
0.00
475/504
1.23
1.11–1.36
0.92
0.88–0.97
586.27
0.00
Paid by patient
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 210–217
Number of bed days of hospitalization
Cost per bed day number of bed days
Laboratory testing
Complete blood count
Blood smear
C-reactive protein
Erythrocyte sedimentation rate
Electrolytes
Urea
Creatinine
Glucose
Gasometry
General urine examination
Stool sample for microbiological confirmation
Stool sample
Blood culture
Nasopharyngeal sample for microbiological confirmation
Other
Imaging diagnostics
Abdominal ultrasonography
Chest radiography
Electrocardiography
Other
Pharmacotherapy
Rehydration fluids
Intestinal anti-infectives (A07A)
Antidiarrheal microorganisms (A07F)
Other antidiarrheals (A07X)
Drugs for functional gastrointestinal disorders (A03)
Analgesics and anti-inflammatory products (N02B, M01A)
Antibacterials for systemic use (J01)
Other pharmaceuticals (A02A, A02B, A05B, A06A, A07C, A07D, A07E,
A11G, A12B, B05A, H02A, J05A, N03A, N06B, R05C, R06A, V06D,
V07A, no ATC code)
Specialist consultation
Materials used
Medical devices (syringes, needles, fluid transfusion sets)
Containers and test tubes (urine and other biological specimen
containers)
Number
213
ambulatories and an independent population served by 8 hospital units located in 6 Polish voivodships (14.4 million inhabitants,
38% of Poland’s population). During the study period, local
coordinators recruited patients meeting inclusion criteria. The
inclusion and exclusion criteria used in the present study were
compatible with the AGI definition proposed by the International
Collaboration on Enteric Disease Burden of Illness [14] and used
in the recent study of AGI prevalence in the community [15]. The
Ethical Committee at the National Institute of Public Health in
Warsaw reviewed and accepted the study protocol.
AGI, acute gastrointestinal infection; ATC, Anatomical Therapeutic Chemical Classification System; CI, confidence interval; WHO, World Health Organization.
*US $1 ¼ 1.85 PLN (WHO CHOICE).
†
State institution (National Health Fund or social insurance).
0.00
–
6850.80
6850.80
0.00
–
0.00
0.00
0.00
477,079.26
–
0.00
0.00
19,274.71
–
2,635.24
1,928.52
23,838.47
893.92–999.25
86.76–165.51
40.10–75.04
40.10–75.04
257.05–456.83
–
141.36–264.06
329.75–955.93
261.30–459.39
504/504
119/145
119/145
119/145
54/145
–
13/145
3/145
70/145
946.59
126.13
57.57
57.57
356.94
–
202.71
642.84
360.34
803.70
50.00
22.59
22.59
251.55
–
179.68
838.49
251.55
733.41–873.98
31.15–68.85
14.34–30.84
14.34–30.84
179.19–323.90
–
120.54–238.82
329.31–1130.18
161.06–342.03
0.00
0.00
967.55
6.44–8.30
7.37
1.62–2.24
19.66–26.91
1.93
504/504
Disinfectants and soaps
Other (gloves, swabs, cotton balls, pampers, disposable towels, gowns,
etc.)
Direct medical costs
Transport (number of kilometers)
Cost per kilometer number of kilometers
Direct nonmedical costs
Home care (family, friends, service)
Absence from work (only patients aged 18–65 y)
Due to hospitalization
After hospitalization
Indirect costs
501/504
23.29
0.70
0.56–0.85
11,737.04
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 210–217
Collection of Data on Costs
For each patient, local coordinators filled a standardized questionnaire on health care resource utilization that was developed
specifically for this study. The questionnaire recording information on GP visits consisted of 18 items, which were divided into
general demographic information, clinical presentation, use of
pharmaceutical preparations, diagnostic tests, materials, and
specialist consultations. For each health care resource used, a
monetary value was allocated. The hospital resource utilization
questionnaire consisted of 26 items, which were divided into
general demographic and clinical information as well as patient
management at the admission unit and separately at the hospital
ward. The local coordinators fill the information on monetary
cost incurred on the basis of internal unit prices.
Study coordinators interviewed all consenting patients 2 to 4
weeks after GP visit or discharge from hospital. A structured
telephone interview comprised 16 questions related to the
current occupation, effect of the disease on daily activities,
absence from work, and use of home care and transport in
relation to AGIs, further GP or specialist consultations, admissions to the hospital, and use of prescription and over-thecounter medications and diagnostic tests.
Cost Analysis
We obtained reference values for the calculation of costs from
official state publications for the period January to December 2009.
We summarized detailed use of health care resources separately for
GP consultations and hospital admissions. We collected data on
type, form, and quantity of pharmaceuticals, medical devices,
diagnostic procedures, laboratory tests, referrals to specialists and
emergency departments, as well as on hospitalizations and grouped
them by using a bottom-up microcosting approach [16]. GP consultation cost and hospital cost of bed day was obtained from a
query of studied health care facilities accounting departments by
using accounts breakdown [17], dividing cost items into personnel,
building, equipment, administration, food, cleaning, laundry, maintenance, and other costs. To calculate the cost of GP consultation,
we summed monthly costs pertaining to the maintenance of the GP
practice and divided them by the number of physicians working in
the health unit and their working time (cost per minute of GP
consultation). To calculate the cost of hospitalization bed day, we
divided the monthly costs of the hospital ward maintenance by the
number of patients admitted to the hospital ward. We obtained
costs of pharmaceuticals (ATC coding), medical devices, and other
medical materials from purchasing units in hospitals. Emergency
department costs were taken from an official diagnosis-related
group system. For the imputation of quantitative missing data, we
used the mean values from the current study data set or average
prices from participating health unit price lists. Missing values for
pharmaceutical costs were obtained from the national pharmaceutical formulary [18]. If only international names were provided by
respondents, then we would impute the values on the basis of prices
of the most commonly used generics. Transportation costs were
based on official mileage rates [19]. We calculated indirect costs
214
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 210–217
related to absence from work and time devoted to care by using a
human capital approach based on average wages lost [16]. We did
not take into account reduced productivity owing to presenteeism or
premature death.
We used three perspectives for cost calculations: household/
patient, public payer, and societal. We applied a standard cost
division into direct medical, nonmedical, and indirect costs [16].
We estimated separately costs of outpatient and inpatient care,
stratified by age group (o4, 5–18, 19–65, and 465 years). For each
cost category, we calculated the mean, median, and total cost. For
the mean and median, we computed confidence intervals by
using 1000 bootstrap percentile replications.
For the estimation of the overall cost of medically attended
AGI cases, we weighted the costs related to hospitalization by
age-specific proportion of patients referred to the hospital. These
figures were based on two recent Polish studies of AGI burden
[3,15]. All costs collected in the local currency (PLN) were
recalculated into US dollars by using the purchasing power parity
exchange rate for Poland [20]. All analyses and data manipulations were done in Stata version 10 (StataCorp LP, College Station,
TX) [21].
Estimating Population AGI Burden
For the estimation of AGI monetary costs in the society (total
national cost, TNC), we developed a stochastic model by using
@RISK 5.0 (Palisade Corporation, Ithaca, NY), a Monte Carlo
simulation add-in to Microsoft Excel:
TNC ¼
∑
pop n IGP n CGP þ pop n IHOSP n CHOSP
age group
where age group is either 0 to 17 and 18 years or more, pop is the
population, IGP is the rate of GP visits per person-year, IHOSP is the
hospital admission rate per person-year, CGP is the societal cost
per GP visit, and CHOSP is the societal cost per hospital admission.
All parameters are age-group specific.
Our study was designed to collect data on individual costs of
AGI cases at different levels, and it was not feasible to derive
incidence and follow-up data directly. Therefore, to obtain
population estimates of AGI-related GP visits and hospitalizations, we related our data to a population-based survey performed at the same time [15]. We selected this particular survey
owing to large sample size, similar time span, and geographic
coverage.
Annual rates of AGI-related GP visits and hospital admissions
were represented by beta distributions. The cost estimates per GP
visit and hospital admission were represented by log-logistic
distributions. The distribution parameters were estimated on
the basis of sample cost distributions by using maximum likelihood estimators. The national annual cost of medically
attended AGIs was calculated by using Monte Carlo simulation
with the above-mentioned probability distributions and the 2009
population estimate for Poland. The model was run for 150,000
iterations to stabilize the output distributions. The mean and 95%
uncertainty limits were reported.
Direct Medical Cost
Unit costs related to GP consultations of AGI cases are listed in
Table 1. Overall, direct medical costs amounted to US $34 per AGI
patient consultation. Prescribed pharmaceuticals, paid in 88% by
patients, were the main cost driver in relation to GP consultations. Particularly, pharmaceuticals of the A07 ATC group (antidiarrheal and intestinal anti-inflammatory/anti-infective agents)
were not reimbursed. However, an important share of antiinflammatory and analgesic drugs (M01A, N02B) and antibiotics
(J01) costs was reimbursed by the state. The next cost driver in
primary care were GP visits’ costs, equal to US $11, entirely paid
by the state payer. Laboratory testing and diagnostic procedures
were relatively cheap for an average AGI patient owing to their
rare utilization (US $0.7 and US $0.5, respectively) but substantial
in terms of unit cost values (US $19 and US $17, respectively).
Unit costs related to the hospital treatment of AGI cases are
listed in Table 2. Inpatient care generated on average US $947 of
direct medical costs, including US $833 of bed days (average cost
per bed day was US $176). The other cost drivers in inpatient care
were costs of laboratory diagnostics, materials, imaging diagnostics, and pharmacotherapy. All direct medical costs were covered
by the state-owned third-party payer.
Taking into account the summed costs generated by medically
attended AGI cases, direct medical costs were highest for patients
aged 5 to 18 years (Table 3). The excess cost was related mostly to
the higher cost of hospitalization bed day in pediatric wards than
in adult wards (US $226 vs. US $117), and substantially lower
probability of hospitalization among 18- to 64-year-old AGI cases.
Direct Nonmedical Costs
The only nonmedical direct cost considered in our study was the
cost of transportation in relation to health care services. In
outpatient care, the average cost of reaching the service (average
distance of 20 km) amounted for US $17 (Table 1). In relation to
inpatient care, this cost was higher (mean cost per patient was US
$58) (Table 2) owing to longer distance to the hospital and
common necessity of several trips to the hospital undertaken
by primary carers. The costs of transportation were higher for
persons younger than 18 years, which was especially marked in
relation to inpatient care-related costs (Table 3). All nonmedical
direct costs were covered by the patients.
Indirect Costs
The indirect costs assessed in our study related to home care
(average cost per hour was US $10 ) and absence from work in
relation to AGI (average cost of 1 day off work was US $84 ). In
relation to outpatient care, the average cost of home care of 34
patients (on average 13 hours of care) was US $128 and the
average cost of absence of 10 patients from work (on average 6
days off work) was US $511 (Table 1). In relation to inpatient care,
the average cost of home care of 54 patients was estimated at US
$357 and the average cost of absence from work of 13 patients
was US $351 (Table 2). Costs of home care were substantially
higher for persons younger than 18 years in outpatient care and
were highest for persons older than 65 years in relation to
inpatient care (Table 3). Costs of absence from work were almost
exclusively limited to patients aged 18 to 64 years and were by far
the most important cost driver in this age group overall.
Results
During the study period, valid questionnaires were collected on
385 GP visits, including 113 follow-up interviews, and on 504
hospital admissions, including 145 follow-up interviews. A
detailed description of the selection of study sites and study
population is presented elsewhere [3].
Cost Perspectives and Estimated AGI Societal Burden
The average cost of one medically attended AGI case in Poland,
taking into account patient management and flow of patients from
primary to secondary care observed in population-based studies, was
estimated at US $136 from the state-owned third-party payer, US $32
215
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 210–217
Table 3 – Average societal costs of medically attended AGI cases, by cost category and age group in Poland between May
2008 and September 2009.
Cost category (US $)*
Outpatient
Average cost of GP visit
Laboratory testing
Imaging diagnostics
Pharmacotherapy
Materials
Emergency department visit
Specialist consultation
Inpatient†
Average cost of hospitalization
Laboratory testing
Imaging diagnostics
Pharmacotherapy
Materials
Specialist consultation
Total direct medical costs
Transport related to GP visit
Transport related to hospitalization‡
Total direct nonmedical costs
Outpatient
Home care (family, friends, service) related to visit
Absence from work related to GP visit
Inpatient†
Home care related to hospitalization
Absence from work related to hospitalization
Total indirect costs
Total cost allocated to one studied patient
Of which direct
Of which indirect
0–4 y
5–18 y
19–64 y
65þ y
All age groups
10.94
0.89
0.16
21.19
0.06
0.00
0.67
10.15
0.49
0.69
19.35
0.09
3.86
0.36
9.62
0.78
0.73
20.71
0.05
3.28
0.00
10.42
0.00
0.76
21.18
0.01
0.00
0.00
10.25
0.70
0.49
20.54
0.06
1.88
0.32
38.56
1.97
0.35
0.50
1.56
0.15
77.00
12.89
2.27
15.16
57.20
3.68
0.79
0.77
1.36
0.21
99.00
10.59
3.39
13.98
4.35
0.36
0.14
0.15
0.21
0.01
40.39
12.80
0.06
12.86
22.56
1.29
0.99
0.63
0.74
0.02
58.60
1.51
0.49
2.00
28.11
1.68
0.54
0.52
1.02
0.08
66.19
12.01
1.60
13.61
45.66
0.00
51.66
8.98
16.52
147.37
10.48
0.00
38.31
44.48
4.77
0.00
50.43
142.59
92.16
50.43
5.62
0.00
66.26
179.24
112.98
66.26
0.81
2.64
167.34
220.59
53.25
167.34
8.97
0.00
19.45
80.05
60.60
19.45
4.49
1.06
88.34
168.15
79.81
88.34
AGI, acute gastrointestinal infection; GP, general practitioner; WHO, World Health Organization.
US $1 ¼ 1.85 PLN (WHO CHOICE).
†
Inpatient costs weighted by age-specific proportion of GP patients referred to the hospital.
‡
Based on the number of kilometers.
from household, and US $168 from societal perspective (Table 4). The
majority of this cost is paid by the state. The estimated societal cost
of medically attended AGI episodes amounted to US $2222 million
(95% uncertainty interval was US $734 to US $6367 million).
Discussion
The costs of medically attended AGI episodes in Poland were
examined in primary and secondary care, from state, household,
and societal perspectives, in terms of direct and indirect costs. We
did not assess separately costs attributed to AGI etiological agents
because the evidence from the literature showed that their direct
costs are in a similar range [10]. When weighting the inpatient costs
by age-specific probability of hospital referral by GPs, we estimated
the societal cost of a medically attended AGI case amounting to US
$168. In terms of direct medical costs, the main cost drivers were
cost of hospital bed days, cost of GP consultation (both paid by the
state-owned third-party payer), and cost of outpatient pharmacotherapy, covered in majority by patients. Overall, indirect costs
prevailed over direct costs (US $88 vs. US $80), which was mostly
pronounced among adults in the productive age, where 76% of the
costs were attributed to indirect costs. Based on the GP consultation
rate estimated in a parallel population-based telephone survey [15],
the estimated age-adjusted societal cost of medically attended AGI
episodes would be US $1681 million.
International comparisons of economic costs between countries
are difficult from the methodological point of view. The societal
costs of AGI cases depend on the disease prevalence, differences in
health-seeking behaviors, and health care system organization.
Historically, different AGI case definitions were used, different
currency exchange rates were applied, and different cost categories
were taken into account in the cost calculations. Estimates of AGI
costs referring to the general population are sparse. The societal
costs of medically attended AGI cases were estimated to be US
$3510 million in the United States [22], which, considering a
conservative assumption of 2% annual discounting until June 30,
2009, would give US $688 for a medically attended AGI case.
Another study estimated the societal costs of all AGI cases in
New Zealand at NZ $89 million [23], which would be equivalent to
US $359 per AGI episode, as of 2009. Straightforward comparisons
with developing countries are difficult because only estimates of
costs for AGI hospitalization costs were published. The societal cost
of rotavirus hospitalization ranged from US $36 in Vietnam [24], US
$77 in Uzbekistan [25], US $81 in one center in India [10] to US $87
in Kyrgyzstan [26]. The study on the hospital cost of bacterial
diarrhea in Thailand showed similar numbers: US $77 per inpatient
episode [8]. The above figures are much lower compared with our
estimate of US $1078 per hospital admission. These differences
reflect the important disparities in the cost of health care services
between the developed and developing world and are concordant
in great extent with World Health Organization estimates for
different World Health Organization regions [20].
216
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 210–217
Table 4 – Average costs of medically attended AGI cases, by cost category and cost perspective, in Poland between May
2008 and September 2009.
Cost category (US $)*
Outpatient
Average cost of GP visit
Laboratory testing
Imaging diagnostics
Pharmacotherapy
Materials
Emergency department visit
Specialist consultation
Inpatient†
Average cost of hospitalization
Laboratory testing
Imaging diagnostics
Pharmacotherapy
Materials
Specialist consultation
Total direct medical costs
Transport related to GP visit
Transport related to hospitalization†
Total direct nonmedical costs
Outpatient
Home care (family, friends, service) related to visit
Absence from work related to GP visit
Inpatient†
Home care related to hospitalization
Absence from work related to hospitalization
Total indirect costs
Total cost allocated to one studied patient
Of which direct
Of which indirect
Patient
Payer
Societal
0.00
0.00
0.00
17.99
0.00
0.00
0.00
10.25
0.70
0.49
2.56
0.06
1.88
0.32
10.25
0.70
0.49
20.54
0.06
1.88
0.32
0.00
0.00
0.00
0.00
0.00
0.00
17.99
12.01
1.60
13.61
28.11
1.68
0.54
0.52
1.02
0.08
48.21
0.00
0.00
0.00
28.11
1.68
0.54
0.52
1.02
0.08
66.19
12.01
1.60
13.61
0.47
0.00
37.84
44.48
38.31
44.48
0.00
0.00
0.47
32.07
31.60
0.47
4.49
1.06
87.87
136.08
48.21
87.87
4.49
1.06
88.34
168.15
79.81
88.34
AGI, acute gastrointestinal infection; GP, general practitioner; WHO, World Health Organization.
US $1 ¼ 1.85 PLN (WHO CHOICE).
†
Inpatient costs weighted by age-specific proportion of GP patients referred to the hospital.
In our study, indirect costs constituted 53% of all costs
attributed to AGI cases. The predominance of indirect costs over
direct costs generated by AGI cases was seen in all studies
performed in developed countries. Indirect costs represented
64% of total AGI-related costs in the United States [22], 70% in
Israel [6], and 87% in New Zealand [23].
In Poland, AGIs are perceived as low public health priority, which
is probably related to typical mild and self-resolving symptoms.
Control measures are limited to the investigation of foodborne
outbreaks, and Salmonella control program in poultry flocks mandated by the European Commission resolution since 2003 [27].
Rotavirus vaccination is also not considered as priority by the
national advisory group for immunization. Currently, the cost of
etiological investigation for selected pathogens (most commonly
Salmonella and Shigella) is free of charge only if performed in the
course of an outbreak investigation. As a result of this situation,
most AGI cases reported to surveillance have unknown etiology [15],
which prevents implementation of efficient interventions.
Our study indicates that AGIs place a high burden on the
society. Out-of-pocket spending for an average patient with AGI
in Poland seems to be relatively small, amounting only to US $32
for a medically attended case. Even coupled with indirect costs
associated with absence from work and home care equal to US
$88, the burden to the individual patient may be regarded as
relatively low as equivalent to 7% of mean monthly income of a
Polish citizen. Similarly calculated monetary burden on an
affected household was estimated to be 25% in Malaysia [28]
and 40% in Taiwan [29]. If we considered the impact of AGIs on
Poland’s economy, the ratio of estimated AGI societal cost to GDP
would be 0.3% and the ratio to the total expenditure per capita
would be equal to 4.3%. Considering the estimate of 81% of the
societal cost attributable to the public payer, AGIs clearly constitute a high burden on the national health care system. A
similar comparison performed in New Zealand found that the
ratio of foodborne illness cost to GDP was 0.1% [23].
Our study has a number of limitations. First, restricting the
study to health care facilities did not allow the estimation of
costs of AGI episodes not consulting GPs. The follow-up interviews provided some insight into self-medication practices and
indirect costs of home care. These data are however not sufficient
to extrapolate to the general population estimates of AGI cases
that were not medically attended. Second, noninclusion of thirdlevel reference hospitals led to the underestimation of our cost
estimates because it included only the costs generated by
uncomplicated AGI cases treated in general hospitals located
near patients’ residence. Cases with severe complications,
although rare, would lead to higher direct as well as indirect
costs of treatment. Third, we did not include private sector health
care units in our study. In the private practices and private
hospitals, patients are paying for each service. By not including
private health care units, we probably underestimated the AGI
management cost. However, the calculation of the actual costs
incurred in the health care units would lead to the approximation
of general health care costs, as the private sector units compete
for the same patients with the public sector health care and have
to calculate their services accurately.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 210–217
Conclusions
These data may imply that much higher resources could be
allotted to the prevention and control of acute gastroenteritis,
even when considering only their direct medical costs. For better
targeting of public health interventions, it would be necessary to
understand the relative importance of different transmission
routes as well as the relative burden placed by different pathogens in Poland. An appropriate prioritization of interventions, in
turn, would enable improvement in their cost-effectiveness.
These results can also impact health care planning, as the
prevention of nosocomial transmission of AGI pathogens may
lead to substantial savings.
Source of financial support: The Polish Prospective Healthcare
Utilization Survey was jointly funded by two European Commission Sixth Framework Programme projects—POLYMOD (SSP22CT-2004–502084) and the Network of Excellence MED-VET-NET
(FOOD-CT-2004-506122)—and cofinanced by the Polish Ministry of
Science and Higher Education.
R EF E R EN CE S
[1] Nathavitharana KA, Booth IW. Pharmacoeconomics of the therapy of
diarrhoeal disease. Pharmacoeconomics 1992;2:305–23.
[2] Szajewska H, Hoekstra JH, Sandhu B. Management of acute
gastroenteritis in Europe and the impact of the new recommendations:
a multicenter study. The Working Group on Acute Diarrhoea of the
European Society for Paediatric Gastroenterology, Hepatology, and
Nutrition. J Pediatr Gastroenterol Nutr 2000;30:522–7.
[3] Stefanoff P, Rogalska J, Czech M, et al. Antibacterial prescriptions for
acute gastrointestinal infections: uncovering the iceberg. Epidemiol
Infect 2012;15:1–9.
[4] Payment P. Epidemiology of endemic gastrointestinal and respiratory
diseases—incidence, fraction attributable to tap water and costs to
society. Water Sci Technol 1997;35:7–10.
[5] Flores AR, Szilagyi PG, Auinger P, Fisher SG. Estimated burden of
rotavirus- associated diarrhea in ambulatory settings in the United
States. Pediatrics 2010;125:e191–8.
[6] Stein M, Roisin H, Morag B, et al. The burden and cost of ambulatory
cases of rotavirus gastroenteritis in Central Israel. Isr Med Assoc J
2010;12:168–71.
[7] Osano BO, Wang’ombe JK, Rose W, et al. Cost analysis of care for
children admitted to Kenyatta national hospital with rotavirus
gastroenteritis. Vaccine 2011;29:4019–24.
[8] Riewpaiboon A, Intraprakan K, Phoungkatesunthorn S. Predicting
treatment cost for bacterial diarrhoea at a regional hospital in
Thailand. J Health Popul Nutr 2008;26:442–50.
217
[9] Lee BP, Azimi PH, Staat MA, et al. Nonmedical costs associated with
rotavirus disease requiring hospitalization. Pediatr Infect Dis J
2005;24:984–8.
[10] Mendelsohn AS, Asirvatham JR, Mkaya Mwamburi D, et al. Estimates of
the economic burden of rotavirus-associated and all-cause diarrhoea in
Vellore, India. Trop Med Int Health 2008;13:934–42.
[11] World Health Organization. WHO country profile for Poland. Available
from: http://www.who.int/countries/pol/en/ [Accessed July 28, 2011].
[12] Czech M. Pharmacoeconomic evaluation and its role in management in
health care in Poland. Hum Factors Ergon Manuf 2005;1:99–108.
[13] NHS reference costs 2010–11. Available from: http://data.gov.uk/
dataset/nhs-reference-costs-2010-11 [Accessed June 1, 2013].
[14] Majowicz SE, Hall G, Scallan E, et al. A common, symptom-based case
definition for gastroenteritis. Epidemiol Infect 2008;136:886–94.
[15] Baumann-Popczyk A, Sadkowska-Todys M, Rogalska J, Stefanoff P.
Incidence of self-reported acute gastrointestinal infections in the
community in Poland—a population-based study. Epidemiol Infect
2011;20:1–12.
[16] Drummond MF, Sculpher MJ, Torrence GW, et al. Methods for the
Economic Evaluation of Health Care Programmes (3rd ed). Oxford:
Oxford University Press, 2005.
[17] Act on Accounting of 29 September 1994. J Laws 1994, No 121, item 591.
[18] Medycyna Praktyczna. Pharmaceutical Index with Price List, issue
1/2009. Kraków: Medycyna Praktyczna, 2009.
[19] Decree of the Minister of Infrastructure of 25 March 2002 on the
conditions for specifying the amount and method for refunding costs
of the use of cars, motorbikes and mopeds not owned by the employer
for official purposes. Offi J Laws Oct 30, 2007, No 201, item 1462.
[20] World Health Organization. Tables of costs and prices used in WHOCHOICE analysis. Available from: http://www.who.int/choice/costs/en/
[Accessed November 13, 2011].
[21] StataCorp. Stata Statistical Software: Release 10. College Station, TX:
StataCorp LP, 2007.
[22] Garthright WE, Archer DL, Kvenberg JE. Estimates of incidence and
costs of intestinal infectious diseases. Public Health Rep
1988;103:107–16.
[23] Scott WG, Scott HM, Lake RJ, Baker MG. Economic cost to New Zealand
of foodborne infectious disease. N Z Med J 2000;113:281–4.
[24] Fischer TK, Anh DD, Antil L, et al. Health care costs of diarrhoeal
disease and estimates of the cost-effectiveness of rotavirus vaccination
in Vietnam. J Infect Dis 2005;192:1720–6.
[25] Isakbaeva ET, Musabaev E, Antil L, et al. Rotavirus disease in
Uzbekistan: cost-effectiveness of a new vaccine. Vaccine
2007;25:373–80.
[26] Flem ET, Latipov R, Nurmatov ZS, et al. Costs of diarrheal disease and
the cost-effectiveness of a rotavirus vaccination program in
Kyrgyzstan. J Infect Dis 2009;200(Suppl.):S195–202.
[27] Regulation no 2160/2003 of the European Parliament and of the Council
of 17 November 2003 on the control of salmonella and other specified
food-borne zoonotic agents. Off J Eur Union December 12, 2003, L 325/1.
[28] Chai PF, Lee WS. Out-of-pocket costs associated with rotavirus
gastroenteritis requiring hospitalization in Malaysia. Vaccine 2009;27
(Suppl. 5):F112–5.
[29] Kow-Tong C, Shiang-Fang F, Ren-Bin T, et al. Hospital-based study of
the economic burden associated with rotavirus diarrhea in Taiwan.
Vaccine 2007;25:4266–72.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 218–225
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Radiology Services Costs and Utilization Patterns Estimates in
Southeastern Europe—A Retrospective Analysis from Serbia
Mihajlo Jakovljevic,́ MD, PhD1,, Ana Rankovic,́ MD2, Nemanja Rančic,́ MD3, Mirjana Jovanovic,́ MD, PhD4, Miloš Ivanovic,́ PhD5,
Olgica Gajovic,́ MD, PhD6, Zorica Lazic,́ MD, PhD7
1
Faculty of Medical Sciences, Department of Pharmacology and Toxicology, University of Kragujevac, Kragujevac, Serbia; 2Radiology Diagnostic Service,
University Clinical Center, Kragujevac, Serbia; 3Medical Faculty, Centre for Clinical Pharmacology, Military Medical Academy, University of Defence, Begrade,
Serbia; 4The Psychiatry Clinic, University Clinical Center, Kragujevac, Serbia; 5The Faculty of Science, University of Kragujevac, Kragujevac, Serbia; 6The
Infectious Diseases Clinic, University Clinical Center, Kragujevac, Serbia; 7The Pulmonary Diseases Clinic, University Clinical Center, Kragujevac, Serbia
AB STR A CT
Objective: Assessment of costs matrix and patterns of prescribing of
radiology diagnostic, radiation therapy, nuclear medicine, and interventional radiology services. Another aim of the study was insight
into drivers of inappropriate resource allocation. Methods: An indepth, retrospective bottom-up trend analysis of services consumption patterns and expenses was conducted from the perspective of
third-party payer, for 205,576 inpatients of a large tertiary
care university hospital in Serbia (1,293 beds) from 2007 to 2010.
Results: A total of 20,117 patients in 2007, 17,436 in 2008, 19,996 in
2009, and 17,579 in 2010 were radiologically examined, who consumed
services valued at €2,713,573.99 in 2007, €4,529,387.36 in 2008,
€5,388,585.15 in –2009, and €5,556,341.35 in 2010. Conclusions: The
macroeconomic crisis worldwide and consecutive health policy measures caused a drop in health care services diversity offered in
some areas in the period 2008 to 2009. In spite of this, in total it
increased during the time span observed. The total cost of services
increased because of a rise in overall consumption and population
morbidity. An average radiologically examined patient got one frontal
chest graph, each 7th patient got an abdomen ultrasound examination, each 19th patient got a computed tomography endocranium
check, and each 25th patient got a head nuclear magnetic resonance.
Findings confirm irrational prescribing of diagnostic procedures and
necessities of cutting costs. The consumption patterns noticed should
provide an important momentum for policymakers to intervene
and ensure higher adherence to guidelines by clinicians.
Keywords: costs, interventional radiology, nuclear medicine, radiation
therapy, radiology diagnostics, utilization patterns.
Introduction
secondary and tertiary care hospitals. Health economic estimates
of radiation-mediated diagnostic and treatment procedures are
seldom reported in the literature. Of those that are available, most
deal with imaging diagnostics or radiotherapy in oncology on a
separate basis. This would be the first study to compare all these
examinations and interventions in a large-scale trial.
In Serbia, as a typical upper-middle income southeastern
European country, expensive high-tech services were centralized
to several tertiary facilities, the third largest of them being the
Clinical Center in Kragujevac. For this reason, we chose this
particular university clinic with approximately 1,300 beds, more
than 50,000 hospital admissions, and 400,000 outpatient examinations per year. Another problem lies in the fact that these
expensive high-tech services are nonrationally prescribed, which
also contributes to excessive consumption and spending from
the modest health budget. By analyzing the 3-year-long trend in
the consumption of services (the volume, i.e., the number of, the
frequency, or expenses), we are of the opinion that the key
Our time witnesses unseen contemporary advances in medical
technology and development of modern equipment in all
branches of medicine. A great number of new diagnostic and
therapeutic methods have come up in radiology as well. They are
very powerful, but their purchase price (which often amounts to a
few million euros, for, for example, computed tomography [CT]
and nuclear magnetic resonance [NMR]) limits broader usage and
replacement of existing appliances. Apart from diagnostic appliances, there has been a technological revolution in radiological
methods of intervention radiology and radiotherapy, the development and application of which have prospered in the last 10 years.
Parallel to the invention of such appliances and their procurement, a problem has been noted because their services are very
expensive and keep a constant burden on health funds [1–8].
These appliances represent mass consumers of health care
budgets worldwide. This is particularly the case when considering
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
* Address correspondence to: Mihajlo Jakovljevic, Faculty of Medical Sciences, Department of Pharmacology and Toxicology, University
of Kragujevac, Svetozara Markovica 69, 34 000 Kragujevac, Serbia.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.07.002
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 218–225
weakness of clinical practice is the lack of consistency in following guidelines for good clinical practice.
Based on these above-mentioned facts, interventions could
be made in the future aimed toward more rational prescription,
especially diagnostic methods, which would redistribute the
limited resources to the cases in which these are necessary for
proper treatment. Similar to our postulates, numerous studies
have shown an increasing trend of the unrealistic use/consumption of radiologic diagnostic methods [9–13], and it is
specially the case with new radiologic methods. The downside
of these studies, and thus of ours too, is the short time during
which the study was carried out (3–6 years), which will be
annulated by the future prolonging of the study time of this
study. We consider it necessary to establish a special organization whose purpose would be to monitor the prescription of
radiologic measures and mark the cumulative annual accepted
dosage of radiation given to patients, as was the case with
patients who took medicines in some countries, the Czech
Republic for instance [14].
The aim of this article was to establish whether radiologic
methods in diagnostics and therapy have been used rationally
during the last 4 years and to substantiate the need for making a
guideline for the application of radiologic methods in clinical
practice.
Methods
The 4-year-long retrospective analysis of total expenditure trends
of radiologic services from the spheres of the classical radiographic, high-tech imaging diagnostics, interventional radiology,
radiation therapy, and procedures of nuclear medicine during the
years 2007, 2008, 2009, and 2010 was provided by the database of
the Clinical Center in Kragujevac (Figs. 1 and 2). The regular
invoicing of services provided in the daycare service and to
hospitalized patients according to International Statistical Classification of Diseases, 10th Revision code of illnesses and the name/
surname/personal identification number resulted in a large
administrative database, which is regularly updated. By the
cooperation of the clinics and departments in charge, the preview
of the database was obtained.
Authors analyzed services that were used the most frequently
during the above-mentioned years (top 10 of the expenditure
volume) and the most expensive services (top 10 of the total
value of services), that is, those that by themselves take up 67% to
Fig. 1 – Division of core services.
219
95% (arithmetic mean ⫾ 1 or 2 SD) of the total value of services.
The population included in this research amounts to 600,000
inhabitants and is situated in central Serbia, in Sumadija, and the
clinical center in charge is the one in Kragujevac.
Results
More than 17,000 radiologically examined inpatients per year
have been noted in the clinical center in Kragujevac (Table 1).
During 2007, most of the services were provided to outpatients;
during 2008 and 2009, most of the services were provided to
inpatients; and in 2010, all the radiological diagnostic and
therapy services in nuclear medicine and interventional radiology were provided to inpatients exclusively. Total expenses
constantly increased during the 4-year period analyzed (Table 1).
Nine percent of total 4-year expenses belong to nuclear
medicine, 16% to radiotherapy, while the radiodiagnostic service
including interventional radiology spends 75% of the budget
intended for radiological services. The number of inpatients
constantly increased from 2007 to 2010 (Table 1), but the ratio
of patients who received one of the radiological services was
relatively reduced with periodic oscillations (44.04% in 2007,
34.55% in 2008, 35.55% in 2009, and 31.39% in 2010). In 2007, on
average every second patient hospitalized received one of the
radiological services, while in 2010, every third patient received
one of the radiological services.
The average price of radiological services per patient constantly
increased during this 4-year period analyzed—in 2007, 10,658.97
RSD (€134.89); in 2008, 20,516.81 RSD (€259.77); in 2009, 26,506.07
RSD (€283.67); and in 2010, 33,045.99 RSD (€316.08) (Fig. 3).
The total number of hospital admissions with radiologically
examined patients slightly reduced from 2007 to 2010 (Table 1).
Eighty-three percent of all the first hospitalizations at the Department of Radiology and Nuclear Medicine belong to radiodiagnostics with interventional radiology.
The number of patients examined and nuclear medicine
services provided in this period decreased (Table 1). Costs,
however, increased significantly although the number of patients
examined or services provided reduced, from 32,272,107.84 RSD
(€408,404.30) to 50,264,302.32 RSD (€481,597.20) in 2010, with a
slight decline in 2008 (Table 1).
The number of patients who received some of the radiodiagnostic services remained at about 15,000 during the 4-year period
(range of 17,000–19,000). The expenses of these services, however,
constantly increased, tripling from 2007 to 2010 (Table 1).
The number of patients who received some of the radiotherapy services, and the number of services provided as well,
had a slight increase (Table 1). This slight rise in the obtained
services volume follows the increase in cost (Table 1).
The services of nuclear medicine lowered their expenses from
15% in 2007 to 8.6% in 2010 per year. A similar fall was noticed in
radiotherapy, from 23% in 2007 to 12% in 2010. However, radiodiagnostic services mark a constant increase in the percentage
share from 62% in 2007 to 79.55% in 2010.
The total number of services given, including the repetitive
one, constantly increased (Table 1), With 81% of the services
provided belonging to the field of radiotherapy, 18% belonging to
radiology diagnostic services with interventional radiology, and
only 1% belonging to nuclear medicine. The total number of
nuclear medicine services provided was doubled in 2010 compared with that in 2007 (Table 1). In contrast to nuclear medicine,
the total number of radiology diagnostic services and radiotherapeutic services provided constantly grew (Table 1). The trend of
increase in the total number of radiology diagnostic services
provided from 2007 to 2010 was 30 times bigger (Table 1).
220
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 218–225
4,500,000
Total financial value
of spending (€)
4,000,000
3,500,000
3,000,000
Nuclear Medicine
2,500,000
Radiaon Therapy
2,000,000
Radiology Diagnosc Service
*(including intervenonal)
1,500,000
1,000,000
500,000
0
2007
2008
2009
2010
Fig. 2 – Expenditure comparison between three clinical branches.
The diversity of services offered suffered from a short recessional
drop in some areas but in total all the three groups (nuclear
medicine, radiology diagnostic, and radiotherapy) recorded an
increase of two to three times from 2007 to 2010 (Table 1).
A total of 3% of the patients received some of the radiotherapeutic services, 16% received nuclear medicine services,
while 81% of the patients received some other radiology diagnostic or emergency radiology services.
The upward trend in diversity is the largest in the field of
interventional radiology and classical radiographics, while the
trend shows a slight increase in nuclear medicine and radiation
therapy.
The services that are most used in nuclear medicine (Table 2)
are those that determine the thyroid gland hormones and
thyroid-stimulating hormone. We can notice that the number
of given services reduced by two times from 2007 to 2010. Most of
Table 1 – Display of all radiological services analyzed and respective costs incurred.
Number of patients treated
2007
2008
2009
2010
Total number of inpatients admitted (regardless of radiological
examination presence/absence)
Nuclear medicine
Radiology diagnostic service (including interventional)
Radiation therapy
Total
45,677
50,459
53,433
56,007
4,456
15,224
437
20,117
2,041
14,918
477
17,436
2,566
15,903
527
19,996
2,990
14,050
539
17,579
Number of single hospital admissions
Nuclear medicine
Radiology diagnostic service (including interventional)
Radiation therapy
Total
2007
5,193
18,320
457
23,970
2008
2,114
18,227
517
20,858
2009
2,715
19,231
563
22,509
2010
3,267
17,094
565
20,926
Total frequency of services provided (including repeated procedures)
Nuclear medicine
Radiology diagnostic service (including interventional)
Radiation therapy
Total
2007
18,145
53,416
2,975,493
3,047,054
2008
10,670
183,540
3,104,262
3,298,472
2009
18,684
928,777
3,196,789
4,144,250
2010
32,002
1,591,285
3,468,228
5,091,515
Diversity of services offered (number of different services available)
Nuclear medicine
Radiology diagnostic service (including interventional)
Radiation therapy
Total
2007
73
216
277
566
2008
159
488
266
913
2009
128
596
275
999
2010
169
658
348
1,175
2007
408,404.3
1,691,606.7
613,562.9
2,713,573.99
2008
339,539.4
3,352,212.6
837,489.5
4,529,387.36
2009
432,617.4
4,253,573.2
702,228.3
5,388,585.15
2010
481,597.2
4,424,357.8
659,968.8
5,556,341.35
Total financial value of services consumed (€)
Nuclear medicine
Radiology diagnostic service (including interventional)
Radiation therapy
Total
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 218–225
221
Fig. 3 – Average cost per patient treated and average cost per admission.
the expenses in nuclear medicine belong to services related to
determining the thyroid gland hormones (Table 2).
The greatest expenses of radiotherapy (Table 2) during the
analyzed 4-year period were for gentigrey. According to the
expenditure volume, gentigrey is the leader of radiotherapeutic
methods (Table 2).
According to the volume of services, nuclear medicine significantly decreased the prescription of its services from 2007
to 2010.
The most common radiotherapy service, gentigrey, shows the
application plateau at about 60 units per patient hospitalized
during the analyzed period, while the accelerator isocenter
technique reduced the volume of consumption, since every 8th
patient received that service, and in 2010, every 12th patient
received it. Team treatment in 2007 of the patient for radiation
therapy was significantly increased in volume. In 2007, every
501st patient received this service while in 2010 every 133rd
patient received it.
Table 2 shows that the lung RO in maximum expiration is the
most common method applied among the classical radiology
diagnostic (RO) methods. Its use has a clear trend of increase—an
increase of 3.5 times from 2007 to 2010. It means that almost
every patient who received some of the classical diagnostic
methods of radiology got at least one RO graphy of lungs in
maximum expiration, and out of all hospitalized patients every
4th patient got the lung RO graphy in maximum expiration in
2010, while in 2007 every 13th patient received it. The use of RO
graphy of the abdomen in the posterior-anterior (PA) position and
profile in the same period increased about 7 times.
Similarly, the consumption of classical methods of radiology
as the most widespread methods is the most expensive, and the
trend of consumption is evident, from 726,269.49 RSD (€9,190.96)
in 2007 to 3,278,406.48 RSD (€31,411.27) in 2010 (a jump of 4.5
times in consumption) (Table 2).
Imaging methods are great consumers, and the greatest
consumer is the CT-targeted imaging of particular organs accompanied with reconstructions, and the expenses of this method
increased 23 times during the 4-year period. Right afterwards
come the head and neck CT without contrast agents for which
there was an increase of 1.3 times.
In Table 2, we notice that imaging methods are prescribed very
often. Targeted CT of particular organs followed by reconstruction is
the most common method, and its frequency of usage represents a
clear trend of growth of 18.5 times during the period analyzed. It
means that every 3rd patient hospitalized received this service in
2010, while in 2007, it was every 52nd patient. It is interesting to
note that the use of the standard abdominal ultrasound examination (examination of liver, gall bladder, pancreas, spleen, and
kidney) decreased 1.8 times. The greatest leap (50 times) is noticed
for the head and neck CT without contrast agents.
Methods of interventional radiology are applied to a lesser
extent, and the most common method is invasive hemodynamics, followed by selective coronary angiography and cardiac
catheterization. Throughout the analyzed period, each patient
received the service of invasive hemodynamics. Percutaneous
transluminal angioplasty (PTA) revascularization (without stent
implantation) and endovascular treatment of intracranial aneurism are the largest consumers, and costs have a growing trend of
about three times during the period analyzed (Table 2).
The top 10 most expensive disorders to treat were recognized
by analyzing resource use and costs related to the patient's
diagnosis at hospital discharge. These are provided jointly for
radiology diagnostic examinations, nuclear medicine procedures,
and radiation therapy in oncology in Table 3.
222
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 218–225
Table 2 – Top 10 services according to the volume of unit consumption and according to treatment expenditures.
Radiodiagnostics
Unit consumption
Chest graph in the maximum expiration AP
Value of consumption (€)
9,598,728
RO graphy of the abdomen in the PA position and profile
399,191
The esophagus, stomach, and duodenum— double volume
control methods targeted shooting
RO graphy of knee joint, lower leg, ankle, or foot in two
directions
RO graphy pelvis—AP position
Irrigography double contrast medium
RO L and LC spine in two directions
RO graphy PA skull in profile position
351,846
281,978
196,383
184,345
183,697
Lung RO graphy in the D or L decubitus
156,841
380,503
Skull RO graphy in children
71,135
Target CT images with reconstruction of
some organs
PTA revascularization (without stent
implantation)
Endovascular treatment of intracranial
aneurysm
Coronarography—Catheterization
CT head and neck without contrast media
Selective coronary angiography
Invasive hemodynamics
CT of abdominal organs with contrast
medium
CT of abdominal organs without contrast
media
Percutaneous balloon angioplasty of
coronary artery catheter
1,650,655
1,531,675
1,414,823
885,123.9
524,185.6
312,544.9
291,353.5
21,2791
19,0800
56,910.3
Nuclear medicine
Unit consumption
Value of consumption (€)
TSH—tireotrophic homone RIA
Free T4 (FT4)—RIA
Free T3 (FT3)—RIA
Mikrosomic antibody (anti-TMS) (IRMA)
Serum prolactin (RIA) LTH
Cortisol determination (RIA)—method of incubation and
separation
Determination of insulin RIA
Tireoglobulin (RIA)
Titer thyroglobulin antibody
11,005
10,984
3,705
3,256
3,022
2,996
2,908
1,794
1,652
Scintigraphy of the whole body J-131
1,193
TSH—tireotrophic hormone RIA
Free T4 (FT4)—RIA
Free T3 (FT3)—RIA
Microsomic antibodies (anti-TMS) (IRMA)
Tireoglobulin (RIA)
Serum prolactin (RIA) LTH
Insulin determination RIA
Scintigraphy of the whole body J-131
Determination of cortisol (RIA)—method of
incubation and separation
Titer tireoglobulin antibody
170,801.1
144,161.3
110,124.7
108,182.5
63,723.5
59,135.47
51,449.5
47,230.93
40,631.77
26,918.96
Radiation therapy
Unit consumption
Gentigrey (in units)
Accelerator—isocentric technique
Supervoltage accelerator radiotherapy with the modified field
Supervoltage accelerator radiotherapy with wedge-shaped
filter
Determination of the airfield graphs
Determination of markers spelling
Team treatment for aerial treatment of the patient
Value of consumption (€)
11,581,975
19,732
12,224
8,028
Radiotherapy—accelerator leaning
After loading the applicator with the applications with
source intensity catheter
Intracavitary gynecological applications
20
Gentigrey (in radiation absorbtion units)
Determination of airfields spelling
Intracavitary gynecological applications
Team treatment for aerial treatment of the
patient
Determination of markers spelling
Accelerator—an isocentric technique
Supervoltage accelerator radiotherapy
with the modified field
Supervoltage accelerator radiotherapy
with wedge-shaped filter
Radiographic verification using selectron
20
Team treatment—selectron
6,320
1,334
1,250
579
1,439,235
26,566.63
21,739.46
19,564.45
5,199.80
735.70
463.62
436.10
203.54
142.78
AP, anterior-posterior; IRMA, Immunoradiometric assay; LTH, luteotropic hormone; PA, posterior-anterior; PTA, percutaneous transluminal
angioplasty; RIA, Radioimmunoassay; RO, radiology diagnostic; TMS, thyroid microsomal antibodies.
Discussion
In 2007, the largest portion of services was obtained outpatiently; in 2008 and 2009, services were equally obtained
outpatiently and inpatiently; and in 2010, radiological services
were mostly obtained inpatiently because it was noticed that
more strict prescription is important as a way to achieve
serious cost-containment, which is already less in comparison
to that in some other countries [15]. By that measure, the
number of radiological services provided decreased in 2010,
but because the price of these services multiplied, and some
more expensive methods started to be used, treatment
expenses tripled compared with those in 2007. We can notice
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 218–225
Table 3 – Top 10 most expensive diagnoses of
radiologically examined and/or treated patients
2007–2010 (€).
ICD
code
Z95
C50
C73
I21
I20
C01
I60
I67
I10
G81
Diagnosis
Value
consumed
Persons with potential health hazards
related to family and personal
history and certain conditions
influencing health status—Presence
of cardiac and vascular implants and
grafts
Malignant neoplasm of breast
Malignant neoplasms of thyroid and
other endocrine glands— Malignant
neoplasm of thyroid gland
Ischemic heart diseases—Acute
myocardial infarction
Ischemic heart diseases—Angina
pectoris
Malignant neoplasms of lip, oral cavity,
and pharynx—Malignant neoplasm
of base of tongue
Cerebrovascular diseases—
Subarachnoid hemorrhage
Cerebrovascular diseases—Other
cerebrovascular diseases
Hypertensive diseases—Essential
(primary) hypertension
Cerebral palsy and other paralytic
syndromes—Hemiplegia
1,365,643.22
973,285.91
920,159.01
664,869.96
653,568.84
460,959.67
442,914.85
424,082.16
410,427.79
373,981.90
ICD, International Statistical Classification of Diseases.
the disproportion between services volume and treatment
costs.
Radiology diagnostics and interventional radiology are the largest consumers of these three groups of services. This was expected
because these services are often prescribed, and the price of the
interventional radiology service was high, and so 75% of the budget
of radiology and nuclear medicine was spent for this purpose.
Because the number of hospitalized patients constantly
increased during the analyzed period, the expectations of cost
expenditure are justified. Although more stringent prescription of
these services is applied, so that the percentage share of patients
who receive radiological and nuclear services is decreased significantly, the leap in the expense of these services led to an
increase in total costs.
The average service price increased about three times per
patient and that is a great deal of the budget obtained for these
purposes. Although recession significantly affected the more
strict prescription, the services price leap led to the growth in
average service price per patient.
The number of patients hospitalized for the first time who
received some service was significantly decreased, and the
number of patients processed from all three groups decreased
as well. In contrast to reducing the volume of unit consumption,
the costs increased many times. Out of all services, the radiology
diagnostic service with interventional radiology recorded a constant increase in percentage cost share, from 62% to 79.55%. The
increase in expenses followed because of the increase in the
service volume of interventional radiology methods, which are
very expensive and their prices are constantly growing on the
world market. The increase in the number of various services
offered led to an increase in cost, because the constant promotion of institutions and the education of physicians widen the
223
service assortment available, and therefore covers the greater
range of diseases that can be treated.
Because the thyroid gland diseases are increasing in Serbia, it
is expected that the volume and the price of services used for the
thyroid gland examination will increase. This is most evident
when we look at nuclear medicine services among the top 10
services, and their volume and consumption—six to seven of the
top 10 services are connected with the examination of the thyroid
gland. It is believed that the main reason for the poor economic
status of the population (upper-middle gross national income per
capita), and therefore, poor diet, increasing stress, and environmental damage caused by the NATO bombing campaign of 1999.
In the last decade, a fivefold increase was marked in thyroid
gland and other endocrine gland diseases.
Malignant diseases, which require radiotherapy, are rising as
well. Thus, although the percentage share of expenditure of
radiotherapy services in relation to the entire annual consumption decreased, customer value of these services increased by
one third.
The most number of services was provided in the radiotherapy domain. That is to be expected when every session in
the treatment of some malignant disease means giving a larger
number of cytostatics several times during the treatment.
Almost every patient who received some of the diagnostic
methods of classical radiology got at least one lung RO graphy in
maximum expiration, and out of all hospitalized patients every
4th patient got the lung RO graphy in maximum expiration in
2010, while in 2007, every 13th patient got the lung RO graphy.
This meant that the number of hospitalized patients increased,
but the lung RO graphy in maximum expiration was prescribed
even more often in 2010—although it was expected because of its
stringent regulation—rarely an indication of recession. Therefore,
this method is the biggest consumer, and marks a jump of 4.5
times in the period analyzed.
The important data are that CT-targeted recordings of particular organs with reconstruction are the greatest consumers of
imaging methods and that their consumption increased 23 times,
which is affirmed by the data that the volume trend of this
method marked a clear growth of 18.5 times. The findings in the
literature were also similar [7,16].
The methods of interventional radiology did not record any
increase in the consumption volume, and the growth in the
prices of the services led to a consumption increase of about
three times. Although the opening of the catheterization hall in
Kragujevac should have led to a significant influx of patients in
this field and increase the service volume, it had not happened.
Because each 40th patient in our center received some kind of
a radiology service, it is considered as the nonrational use of
these diagnostic methods. The tendency of the growth in consumption from 2007 till 2008, which dropped in 2009, can be
explained by the means of the global economic crisis, and should
not be ascribed to the rationally prescribed radiologic procedures.
This is a devastating piece of information and must be taken with
a great caution and a guide of good practice should be made in
the future concerning the radiologic procedures and in that way
try to examine the prescription of the treatment and sanction
those who do not follow this guide. All this should reduce the
unnecessary waste of budget assets.
When compared with the Levin [17] study in Pennsylvania,
PA, of about 6 million health insurance holders, significantly
higher expenses were incurred in total on radiologic check-ups,
out of which 62% were nonradiologic expenses. It is a relatively
small consumption for these services, but it still represents an
enormous expense for our modest budget assets and it prevents
the allocation of necessary financial assets for other purposes.
Sunshine et al. [18] reported that during 1996 and 1997, large
financial resources were spent during the full-time workload at
224
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 218–225
radiology for about 320 radiologic procedures in the United States
(including the methods of radiologic oncology).
A great importance is given today to the methods of interventional radiology, which replace the classical surgery and greatly
shorten the postoperational recuperation of patients who are
hospitalized and in this way reduce the expenses as well. These
should be used more and more, and the fall in their use in our study
during 2009 was, although a partial explanation can be found in the
reduction of the health budget, due to the financial hardship.
Average expenses for radiologically examined patients
in all the three analyzed years were in the range from 10,000
RSD (€125) to 17,000 RSD (€218), and if this is multiplied with
the size of the population of a million residents, it is a great
expense.
The devastating fact is that irrespective of the discharge
diagnosis each patient gets a lung graph or every 10th gets a CT
of the endocranium on average. This is a dubious piece of
information when it is compared with other studies, as the one
by Bhargavan and Sunshine [19], in which out of the total 4176
radiologic procedures, 49.3% go to all the radiologic check-ups and
9.36% to all the CTs. When these two populations are compared,
this is a catastrophic information for our health care system and
speaks in favor of the human neglect and the great need for a
guide of good clinical practice with strict indications [7].
It is interesting to note that among the 10 most expensive
methods, RO of the lung graph ranked second, the reason for this
being its nonrational consumption, which must be stopped
because of not only huge expenses but also the unnecessary
radiation to which patients are exposed.
Many studies have shown that there has been a reduction in
the total number of radiologic check-ups. The number of offered
services using new methods (CT and NMR) has increased, which
means that they are often misused and that patients are
unnecessarily undergoing shots and radiation [20]. As it is a
well-known fact that the dosage during the CT treatment is 300
times higher than that during ordinary CT scan graph, here lies
the danger of the possible development of tumor, which originates from mutations caused by such radiation.
When compared with the American study from 2001, in which
out of the 4176 radiologic services there were 215 services of
interventional radiology [19], the percentage of the intervention
radiology service is modest in our case, but at the same time the
use of other conventional methods is very high.
The age distribution, as well as the sex distribution, only
confirms that today the presence of the disease is the same for
both sexes and that the older population is much more exposed
to the radiologic radiation. The devastating piece of information
is that one tenth of the patients who undergo the radiologic
check-up are younger than 18 years. With this population one
should be stricter when prescribing the radiology diagnostic
imaging (RDI) methods because of the impact of the radiation
on gonad cells.
Our study once again confirmed that the most frequent are
still the cardiovascular diseases and that their prevention is the
most important. A special attention should be paid to this piece
of information, and the guide for the prevention of such disorders
should be used.
Because CT ranked eighth among the top 10 services according
to the unit consumption, it is a worrying information because it is
not possible that all these patients have fulfilled the conditions for
certain indications in which these very powerful but at the same
time very harmful radiologic methods are prescribed. It is also very
important that the RDI of lungs is still the most often used of all
radiologic methods and that it is unbelievable that each patient gets
along a graph on average. It is an irresponsible prescription of this
RTG method, and strict instructions should be given to stop such
nonrational prescription.
Top 10 lists according to the total consumption have been
shown, which claim that the lungs RDI is ranked second and that
the remaining ranks are filled with CT methods and two intervention methods. Huge resource consumers are also the adjoining expenses of these radiologic procedures: RO films, contrasts,
and radioactive isotopes.
A special attention should be paid to lists of most expensive
disorders per clinical radiology branch-department. Nuclear medicine top 10 list of most expensive diagnosis reveals an anticipated structure, with endocrine disorders dominating. Oncology
radiation therapy services consumption exhibits unique morbidity rates of particular cancer types in the region. A very interesting finding in the domain of radiology diagnostics is for
cardiovascular and orthopedic surgical implantation procedures
and dealing with their complications, being placed at the very top
of the list. Other details can be studied in Table 3.
The trend of a constant rise in expenses as well as the volume
of services in radiology was marked with many services. We have
chosen only some of them and shown them on the graph, in
which a trend of growth can clearly be seen, with a slight
decrease during 2009 due to the reduction in the budget as a
result of the global economic crisis [21].
The data concerning the number of services provided, which
rises from year to year and is tripled every year, is a very good
indicator of how nonrationally the methods of radiologic diagnostics and therapy are used. At the same time, however, a
decrease in the application of the methods of interventional
radiology is discomforting, because the recently opened theatre
of interventional radiology should replace the big expenses of the
surgical blocks and contribute to the quick diagnostics and
treatment of these patients [22].
Conclusions
The total level of the value of the consumption of radiologic services
is comparable (higher/lower) with other tertiary clinics of similar
size among developing countries. However, the structure of the
consumption within the available hospital budget is not the best
and could be significantly improved by redirecting the nonrationally
spent resources toward priority domains. The suggested measure
for the improvement of the structure of the suggested diagnostic
and therapeutic intervention is the making of a local protocol of
good practice by which a desirable and suggested dynamics of the
frequency of doing certain check-ups in certain indications (pneumonia, stroke, tumors, etc.) is to be established. The investment of
resources for the implementation and application of such a guide
and a periodical (annual) estimation of following the protocol,
which was agreed upon, would be a good idea. Consumption
patterns should be transformed in respect to their quality and
upgraded, so that the decision is more significantly based on
evidence. The recession, macroeconomic events, and current cuts
in the budget for the health sector in Serbia will increase the
external pressure on management bodies and so such changes will
probably be inevitable in the coming years.
Acknowledgments
We express our gratitude to the Ministry of Science Education
and Technological Development of the Republic of Serbia for
grant number 175014, out of which this research project was
partially financed.
Source of financial support: This study was partially financed
by the Ministry of Science Education and Technological Development of the Republic of Serbia (grant no. 175014).
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 218–225
R EF E R EN CE S
[1] Goedert J. The hidden costs of radiology. Health Data Manag
2007;15:24–6.
[2] Clevert DA, Stickel M, Jung EM, et al. Cost analysis in interventional
radiology–a tool to optimize management costs. Eur J Radiol
2007;61:144–9.
[3] Golder W. Benefits, costs and analysis in diagnostic radiology:
definitions and glossary. Rofo 1999;170:73–9.
[4] Kelekis DA, Brountzos EN. How much does radiology contribute to
increasing health care costs? The situation in Greece. Acad Radiol
1996;3(Suppl. 1):S121–4.
[5] Gerhardt P. Diagnostic radiology: health care costs and need for therapy
relevant examination strategies. Röntgenpraxis 1994;47:129–38.
[6] Krotz D. Aetna outsources radiology services utilization review. Diagn
Imaging (San Franc) 1998;20:19.
[7] Khorasani R, Goel PK, Ma’luf NM, et al. Trends in the use of radiology
with inpatients: what has changed in a decade? AJR Am J Roentgenol
1998;170:859–61.
[8] Otero HJ, Ondategui-Parra S, Nathanson EM, et al. Utilization
management in radiology: basic concepts and applications. J Am Coll
Radiol 2006;3:351–7.
[9] Levin DC, Bree RL, Rao VM, Johnson J. A prior authorization program of
a radiology benefits management company and how it has affected
utilization of advanced diagnostic imaging. J Am Coll Radiol 2010;
7(1):33–8: (Erratum in: J Am Coll Radiol 2010;7:235).
[10] Maitino AJ, Levin DC, Parker L, et al. Nationwide trends in rates of
utilization of noninvasive diagnostic imaging among the Medicare
population between 1993 and 1999. Radiology 2003;227:113–7.
225
[11] Sunshine JH, Bushee GR, Mallick R. U.S. radiologists’ workload in 1995–
1996 and trends since 1991–1992. Radiology 1998;208:19–24.
[12] Zylak CJ, Gafni A. A methodologic overview of the evaluation of costs
and benefits in diagnostic radiology. Invest Radiol 1992;27:483–8.
[13] Heilman RS. Costs, benefits, and common sense in radiology.
Radiographics 1998;18:849–50.
[14] Biba V, Kubica Z, Strnadova J, et al. Drug consumption statistics in the
Czech Republic. Vestnik 1997;(Suppl.):1–68.
[15] Mecozzi B, Pancione L, De Intinis G, et al. Analysis of production
factors, costs, and process efficacy in the radiology department of a
local health agency in Italy. Radiol Med 2003;105:215–29.
[16] Nisenbaum HL, Birnbaum BA, Myers MM, et al. The costs of CT
procedures in an academic radiology department determined by an
activity-based costing (ABC) method. J Comput Assist Tomogr
2000;24:813–23.
[17] Levin DC. The practice of radiology by nonradiologists: cost, quality,
and utilization issues: Merrill C. Sosman Lecture. AJR Am J Roentgenol
1994;162:513–8.
[18] Sunshine JH, Burkhardt JH, Mabry MR. Practice costs in diagnostic
radiology. Radiology 2001;218:854–65.
[19] Bhargavan M, Sunshine JH. Utilization of radiology services in the
United States: levels and trends in modalities, regions, and
populations. Radiology 2005;234:824–32.
[20] Matin A, Bates DW, Sussman A, et al. Inapatient radiology
utilization: trends over the past decade. AJR Am J Roentgenol
2006;186:7–11.
[21] Jakovljevic MB. Resource allocation strategies in Southeastern
European health policy. Eur J Health Econ 2013;14:153–9.
[22] Rankovic A, Rancic N, Jovanovic M, et al. Impact of imaging diagnostics
on the budget – are we spending too much? Vojnosanit Pregl
2013;70:709–11.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 226–230
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Children Hospitalized for Varicella: Complications and Cost Burden
Ozden Turel, MD1,, Mustafa Bakir, MD2, Ismail Gonen, MD1,3, Nevin Hatipoglu, MD3, Cigdem Aydogmus, MD3,
Emine Hosaf, MSc4, Rengin Siraneci, MD3
1
Department of Pediatrics, Bezmialem Vakif University Faculty of Medicine, Istanbul, Turkey; 2Department of Pediatrics, Marmara University Faculty of Medicine,
Istanbul, Turkey; 3Department of Pediatrics, Bakirkoy Maternity and Children's Educational and Treatment Hospital, Istanbul, Turkey; 4Department of
Microbiology, Bakirkoy Maternity and Children's Educational and Treatment Hospital, Istanbul, Turkey
AB STR A CT
Objective: To evaluate the direct medical cost of hospital admissions
for patients with varicella (i.e., chickenpox) to assess the cost burden
of varicella from a health care perspective for ultimate use in health
economics studies in Turkey. Methods: Records of children hospitalized with varicella at the Bakirkoy Maternity and Children’s Hospital
between November of 2006 and June of 2011 were reviewed. Reasons
for hospitalization, types of varicella-associated complications, and
direct medical cost of hospitalization were noted. Patients with
underlying risk factors were excluded. Data obtained from one
hospital were used to estimate the national cost of the disease.
Results: During the 4.5-year study period, 234 patients were hospitalized with varicella. Of these cases, 48 (20%) children previously ill
with underlying cancers or chronic diseases were excluded from the
study. Ultimately, 186 previously healthy children (age range: 14 days
to 159 months, median age: 14 months) were included. The main
reasons for hospitalization were complications related to varicella
(79%), the most frequent of which was skin and soft tissue infections,
followed by neurological complications and pneumonia. The median
cost of hospitalization per patient was US $283, 50% of which was
attributed to medication costs. The annual cost for varicella hospitalizations in Turkey was estimated at US $396,200. Conclusions: A
significant number of healthy children are hospitalized for varicella
and associated complications. Descriptions of these complications
and their related costs provide important data for cost-effectiveness
studies for decisions about the inclusion of the varicella vaccine in a
childhood vaccination program.
Introduction
has also decreased in nonvaccinated groups, including adults and
infants who are too young to be vaccinated, thereby suggesting a
strong herd protection effect [7–10]. A widespread varicella vaccination program, however, has not yet been introduced everywhere,
especially in developing countries.
The health economics of a VZV immunization program is vital
for decisions on vaccine funding and has been studied in many
countries [14–17]. To facilitate the decision-making process
regarding the introduction of a vaccination program, each country needs to collect data on the incidence, complications, and cost
due to hospitalizations associated with the particular disease
under consideration. Surveillance of varicella complications is
also important to assess the potential impact of a vaccination
program. In Turkey, a few studies have evaluated complication
rates; however, knowledge about the cost of varicella hospitalizations is quite limited. A more detailed investigation of the cost
burden of varicella from a health care perspective can be
accomplished by collecting data about the number and cost of
hospitalizations in a tertiary care hospital in Istanbul, which can
then be extrapolated to the whole country to estimate the
national burden of this disease.
Varicella, otherwise known as chickenpox, that is, the primary
manifestation of a varicella zoster virus (VZV) infection, is generally mild; indeed, severe complications are seldom reported in
immunocompetent children [1]. Nevertheless, hospitalizations
due to varicella do occur in otherwise healthy children, thereby
producing an economic burden on the health care system [2].
Worldwide, the reported incidence of varicella-related hospitalizations involving children varies widely, that is, from 0.9 to 29.4/
100,000, depending on the geographic setting and hospital admission policies [3–5]. In Turkey, the exact incidence is unknown
because varicella is not on the list of tracked diseases; however,
estimates indicate that the rate is 6.3/100,000 [6].
A safe and effective vaccine against varicella was developed in
1970 and has been made a recommended part of childhood
vaccination programs in several countries. Countries with routine
childhood varicella vaccinations have seen a positive effect on
disease prevention and control [7–13]. In the United States, the
annual varicella-related hospitalization rate decreased from 0.5 per
10,000 in 1993 to 0.13 per 10,000 in 2001. The incidence of varicella
Keywords: child, complications, cost, varicella.
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflict of interest: The authors report no conflict of interest.
Address correspondence to: Ozden Turel, Department of Pediatrics, Division of Pediatric Infectious Diseases, School of Medicine,
Bezmialem Vakıf University, Adnan Menderes Bulvarı Vatan caddesi 34093 Fatih, Istanbul, Turkey.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.05.003
227
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 226–230
Methods
Table 1 – Varicella-related hospitalized patients
with underlying illnesses.
Study Design
Underlying illness
In this retrospective cohort study, we reviewed the records of
children with varicella who were admitted to the Bakirkoy
Maternity and Children’s Educational and Treatment Hospital
(BEH) between November of 2006 and June of 2011. In Istanbul, a
majority of patients are treated at secondary and tertiary care
hospitals [18]. BEH is one of the main tertiary referral centers for
pediatric patients in which 40,000 patients are hospitalized and
over 500,000 patients are examined annually [19]. The population
of children aged younger than 15 years in Istanbul is 3,455,049,
that is, one-fifth of the population under 15 in Turkey [20].
Data Collection
Enrollment in this study required a discharge diagnosis of
varicella or its associated complications as defined by the International Classification of Disease codes. Further detailed investigations
of medical records of patients hospitalized with varicella were
undertaken to avoid incorrect diagnoses. Data on the demographic features of the patients, their underlying conditions,
reasons for hospitalization, types of varicella-related complications, blood culture results, length of hospital stay, outcomes,
and costs were collected.
Hospital expenses noted in this study included the cost of the
prescribed drugs, doctor visits, nursing care, laboratory and
radiological diagnostic tests, bed stay, and other related charges.
The records of the patients logged hospital costs in the Turkish
lira and were converted into the US dollar.
Statistical Analysis
Version 16 of the Statistical Package for Social Sciences (SPSS for
Windows) was used for all statistical analyses. One-way analysis
of variance was used to compare continuous data among more
than two groups. Multiple comparisons were analyzed by using
Tukey’s honestly significant difference post hoc test, while
Pearson’s correlation test was used to assess the relationships
among continuous variables.
Results
Of the 684 children with varicella who were examined at BEH
during the study time period, 234 were hospitalized. Of these cases,
48 patients (i.e., 20%) had an underlying illness and were thus
excluded from the study (Table 1). Therefore, 186 previously healthy
children (i.e., 55.9% males) were included in this study. Considering
that BEH provides services to 15% of all children hospitalized in
Istanbul and 3% of all children hospitalized in Turkey, the annual
number of formerly healthy children younger than 15 years who are
hospitalized because of varicella in Turkey was estimated at 1400.
The median age of the patients with varicella in our study was
14 months (i.e., ranging from 14 days to 159 months). The highest
rate of hospitalization occurred in patients younger than age 3
years (i.e., 74.15%), of which 64.2% were younger than or equal to
age 1 year. The median length of hospital stay for this population
was 5 days. The majority of the cases were detected in the spring
and early summer months, with a peak in May (Fig. 1). The main
reasons for hospitalization were complications associated with
varicella (i.e., accounted for 79% of the admissions). Bacterial
superinfections involving the skin and soft tissues accounted for
32.6% of the admissions and were the most frequently observed
complications; this was followed by neurological complications
in 29.9% of the admissions and pneumonia in 21.7% of the
Malignancy
Acute lymphoblastic leukemia
Lymphoma
Rhabdomyosarcoma
Spinal tumor
Neuroblastoma
PNET
Metabolic
Type 1 diabetes mellitus
Congenital adrenal hyperplasia
Cystic fibrosis
Graves’ disease
Niemann Pick
Hematologic
Hereditary spherocytosis
Diamond Blackfan anemia
Thalassemia major
Thrombasthenia
Factor 7 deficiency
Chronic ITP
Primary immunodeficiency
SCID
IgA deficiency
CVID
Cyclic neutropenia
Other
Other
Holoprosencephaly
CMV hepatitis, hyperphenylalaninemia
Echinococcal cyst
Total
n
%
14
2
1
1
1
1
29
4.2
2.1
2.1
2.1
2.1
5
2
1
1
1
10.4
4.2
2.1
2.1
2.1
2
1
1
1
1
2
4.2
2.1
2.1
2.1
2.1
4.2
1
1
1
1
2
2.1
2.1
2.1
2.1
4.2
1
1
1
48
2.1
2.1
2.1
100
CMV, cytomegalovirus; CVID, common variable immunodeficiency; IgA, immunoglobulin A; ITP, idiopathic thrombocytopenic
purpura; PNET, primitive neuroectodermal tumor; SCID, severe
combined immunodeficiency.
admissions (Table 2). The children who were hospitalized for
pneumonia were younger than those hospitalized with neurological complications (P o 0.05) (Fig. 2). Most outcomes were
favorable with exception to one child needing thoracal tube
insertion to treat empyema, one with abduscent nerve paralysis,
and one with cellulitis that resulted in severe scarring.
The median cost of hospitalization per patient was US $283, 50%
of which was attributed to medication costs (Table 3). The costs for
physician visits were lowest among the cost categories listed in
Table 3 because revisits are not billed in accordance with current
Turkish regulations. A positive correlation was observed between the
total costs and the age of the patients (r ¼ 0.27, P o 0.001). Costs for
hematological complications were higher than the costs for any of
the other complications. For the five patients with the highest total
costs, three had hematological complications and received blood
transfusions; for the remaining two, one had septicemia and the
other child had severe cellulitis that necessitated 26 days of hospitalization. The direct annual cost for otherwise healthy children
hospitalized with varicella in Turkey was estimated to be US $396,200.
Discussion
Although varicella complications are believed to be rare in
immunologically healthy children, related hospitalizations have
228
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 226–230
35
Number of patients
30
25
20
15
10
5
0
Months
Fig. 1 – Number of hospitalized varicella cases according to months of the year.
been reported in children without any identified risk factors
[2,21]. Of the 24,488 varicella-related hospitalizations during
2000-2006 in the United States, 70% were among healthy persons
with no contraindications for vaccination [22]. Similarly, we
noted that 80% of the children with varicella who were hospitalized during the study period at BEH had no underlying
diseases. Carapetis et al. [2] reported that immunocompromised
patients with varicella were admitted earlier in their illness and
had lower complication rates than did otherwise healthy
patients. For this reason, we conducted a detailed examination
of hospitalizations for varicella only in previously healthy children. We estimated that hospitalizations due to varicella in
otherwise healthy children have an incidence of 7.7/100,000. A
Table 2 – Reasons for hospitalization in children
with varicella.
Cause
Complications
Skin, mucosa and soft tissue infections
Pyoderma
Cellulitis
Abscess
Cervical lymphadenitis
Stomatitis
Neurologic
Cerebellar ataxia
Convulsion
Meningoencephalitis
Papilloedema, optic neuritis
Pneumonia
Other
Hematologic (thrombocytopenia)
Sepsis
Upper respiratory tract infection
Arthritis
Gastrointestinal
Other causes
Ageo2 months
Fever
Generalized skin lesions
Poor feeding
Poor general appearance
Total
n (%)
147
48
25
12
5
2
4
44
19
17
5
1
32
23
7
8
4
2
2
39
21
8
3
3
2
186
(79)
(25.8)
(23.7)
(17.2)
(12.3)
(21)
(100)
multicenter study in Turkey also showed that 73.3% of hospitalized patients were previously healthy [23]. The estimated
overall incidence was 5.29 to 6.89/100,000 in all children aged 0
to 15 years; therefore, considering that hospitalized patients
constitute 1% of all cases of varicella, the overall incidence of
varicella was estimated as 466 to 768/100,000 [23].
The median age of children hospitalized with varicella-related
complications in the United Kingdom and Ireland was reported as
3 years [5]. In our study, children hospitalized for varicella tended
to be younger; indeed, 47.6% of the patients in our study were 1
year or younger. Seroprevalence studies in Turkey have shown
that after the decline in maternal antibodies present in the
bloodstream at birth, seropositivity rates for VZV were low until
the end of the first year (i.e., 16.6%) and gradually increased to
41.2% at age 5 years [24,25]. Protection of children younger than 1
year, however, can be achieved only by increased herd immunity
that is established via a widespread vaccination program because
the vaccine is not recommended in children younger than 1 year.
A multicenter study in France demonstrated a strong inverse
correlation between levels of circulating anti-VZV maternal antibodies in full-term infants and the occurrence of varicella complications in children who contract the disease at age younger
than 1 year [26]. During the first 3 months of life, maternal
antibodies against VZV protect most infants, and unless this
immunity is absent, newborns with mild chickenpox should not
require antiviral therapy [27]. In our study population, we
identified 21 patients aged 14 days to 2 months who were
hospitalized solely for varicella without any complications, and
most of them received antiviral therapy. This indicates that
many physicians still view the disease as a serious illness during
early infancy.
The majority of varicella complications identified in our study
were bacterial superinfections, which is similar to the findings of
previous studies [6,21]. Staphylococcus aureus and group A βhemolytic streptococci are the main pathogens responsible for
bacterial complications related to varicella [28,29]. Pathogens
were isolated from cultures of tissue samples in only five of our
study patients (i.e., two had coagulase-negative staphylococci
and one had α-hemolytic streptococci from blood cultures,
whereas one had methicillin-sensitive S. aureus from abscess
material, respectively). The low yield of microorganisms in our
study is likely caused by insufficient culturing practices that must
be improved in future studies and treatment of patients with
varicella.
Neurological complications were the second most common
complication in patients who were hospitalized for varicella in
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 226–230
229
P<0.05
(hematologic, sepsis, upper respiratory
tract infection, arthritis, gastrointestinal)
Fig. 2 – Distribution of complications according to age of the patients.
our study. Bozzola et al. [30] reported that although neurological
complications did not usually result in permanent sequelae, they
could lead to prolonged hospital stays and other indirect costs.
The indirect costs from absenteeism and loss of associated
productivity are more important than the direct health care costs
[31]. The proportion of indirect costs in the total cost of varicella
was reported to range from 42% to 98% in different sudies [32].
The highest percentage reported by Lieu et al. [33] was explained
by the inclusion of the cost of death or prolonged disability
resulting from the disease. Only direct costs were evaluated in
our study.
The costs of varicella-related hospitalizations may vary
according to the epidemiologic features, management strategies of the disease, and the type of complications. Zhou et al.
[17] estimated that in 2006, the direct per case medical cost of
hospitalization for uncomplicated varicella, varicella pneumonia, and varicella encephalitis was US $3,317, US $4,213,
and US $12,064, respectively. In Australia, the cost of hospitalization for children with varicella was reported to be US $3,272
per case in 2004 [2]. In Spain, hospitalization cost per case was
Table 3 – Total hospital costs (US dollar).
Variable
MED
PHY
LAB
NC
HOS
Total
Cost
Mean
SD
Median
Min
Max
206
7
60
28
106
407
469
12
80
30
73
552
108
4
30
20
88
283
0
0
0
0
0
15
4529
91
515
185
696
4886
MED, medications; PHY, physicians’ fees for examinations and
consultations; LAB, diagnostic tests and radiological examinations;
NC, nurse care; HOS: bed stay; SD, standard deviation.
US $5,113, with the total associated national cost excluding
symptomatic treatment at US $516,531 [34]. We estimated the
direct national cost of varicella-related hospitalizations as US
$396,200 in Turkey. Differences in health service fees such as
hospital and physician costs may explain why hospital care in
Turkey costs less than in the United States, Australia, and
Spain. For Brazil, a country with a socioeconomic background
similar to that of Turkey, Valentim et al. [14] reported that the
hospital cost for a case of varicella was US $439. In China,
inpatient care cost was US $640 per case [35]. In previous
studies, the costs of hospitalizations for varicella were significantly higher in children who had underlying diseases
[15,23]. Only otherwise healthy children were included in
our study.
Decisions on vaccine funding are often based on a number of
variables, such as immunogenicity of the vaccine and the costeffectiveness of the immunization program. The varicella vaccine
prevents the disease in 85% of immunized children but offers a
97% protection against its most severe forms [36]. Moreover, the
universal two-dose immunization has been shown to be costeffective in Western temperate countries [37]. In Brazil, the costeffectiveness of a universal vaccination program against varicella
was dependent on the vaccine price and the required number of
doses [14].
In Turkey, one dose of varicella vaccine costs US $45 and
the national birth cohort is 1.2 million. As a preventive
measure, a universal vaccination program will require a large
investment. In this study, however, we confirmed that hospitalization in varicella cases contributes to a significant cost
burden even in previously healthy children. Our estimates of
costs should contribute to future cost-effectiveness research in
Turkey.
Conclusions
In Turkey, a significant number of otherwise healthy children
have been hospitalized because of varicella and its complications.
Data gathered about hospitalization expenses provide important
230
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 226–230
knowledge for future cost-effectiveness studies. This information
will, therefore, assist policymakers in decisions about whether to
include the varicella vaccine in childhood vaccination programs.
Rates of complications may provide significant background
knowledge for evaluating the impact of vaccinations against
childhood varicella.
Source of financial support: The authors have no other
financial relationships to disclose.
R EF E R EN CE S
[1] Heininger U, Seward JF. Varicella. Lancet 2006;368:1365–76.
[2] Carapetis JR, Russell DM, Curtis N. The burden and cost of
hospitalized varicella and zoster in Australian children. Vaccine
2004;23:755–61.
[3] Liese JG, Grote V, Rosenfeld E, et al. The burden of varicella
complications before the introduction of routine varicella vaccination
in Germany. Pediatr Infect Dis J 2008;27:119–24.
[4] Dubos F, Grandbastien B, Hue V, et al. Epidemiology of hospital
admissions for pediatric varicella infections: a one-year prospective
survey in the pre-vaccine era. Epidemiol Infect 2007;135:131–8.
[5] Cameron JC, Allan G, Johnston F, et al. Severe complications of
chickenpox in hospitalized children in the UK and Ireland. Arch Dis
Child 2007;92:1062–6.
[6] Koturoglu G, Kurugol Z, Cetin N, et al. Complications of varicella in
healthy children in Izmir, Turkey. Pediatr Int 2005;47:296–9.
[7] Davis MM, Patel MS, Gebremariam A. Decline in varicella-related
hospitalizations and expenditures for children and adults after
introduction of varicella vaccine in the United States. Pediatrics
2004;114:786–92.
[8] Hambleton S, Phill D, Gershon AA. The impact of varicella vaccination
in the United States. Semin Pediatr Infect Dis 2005;16:38–43.
[9] Guris D, Jumaan AO, Mascola L, et al. Changing varicella epidemiology
in active surveillance–United States. 1995–2005. J Infect Dis 2008;197
(Suppl. 2):S71–5.
[10] Marin M, Watson TL, Chaves SS, et al. Varicella among adults: data
from an active surveillance project, 1995–2005. J Infect Dis 2008;197
(Suppl. 2):S94–100.
[11] Quian J, Rüttimann R, Romero C, et al. Impact of universal varicella
vaccination on 1-year-olds in Uruguay: 1997–2005. Arch Dis Child
2008;93:845–50.
[12] Cenoz MG, Catalán JC, Zamarbide FI, et al. Impact of universal
vaccination against chickenpox in Navarre, 2006–2010. An Sist Sanit
Navar 2011;34:193–202.
[13] Pozza F, Piovesan C, Russo F, et al. Impact of universal vaccination on
the epidemiology of varicella in Veneto, Italy. Vaccine 2011;29:9480–7.
[14] Valentim J, Sartori AM, de Soárez PC, et al. Cost-effectiveness analysis
of universal childhood vaccination against varicella in Brazil. Vaccine
2008;26:6281–91.
[15] Azzari C, Massai C, Poggiolesi C, et al. Cost of varicella-related
hospitalizations in an Italian pediatric hospital: comparison with
possible vaccination expenses. Curr Med Res Opin 2007;23:2945–54.
[16] Tseng HF, Tan HF, Chang CK. Varicella epidemiology and costeffectiveness analysis of universal varicella vaccination program in
Taiwan. Southeast Asian J Trop Med Public Health 2005;36:1450–8.
[17] Zhou F, Ortega-Sanchez IR, Guris D, et al. An economic analysis of the
universal varicella vaccination program in the United States. J Infect
Dis 2008;197(Suppl 2):S156–64.
[18] Istanbul Health Directorate. Istanbul health statistics 2010. Available
from: http://www.istanbulsaglik.gov.tr/w/anasayfalinkler/pano3.asp.
[Accessed April 10, 2013].
[19] Istanbul Health Directorate. Kanuni Sultan Süleyman Educational and
Treatment Hospital. Available from: http://www.istanbulsaglik.gov.tr/
w/sb/tedk/bakirkoy.asp. [Accessed April 10, 2013].
[20] Turkish Statistical Institute. Address-based population registration
system. Available from: www.tuik.gov.tr/VeriTabanlari.do?
vt_id=9&ust_id=5. [Accessed April 10, 2013].
[21] Grimprel E, Levy C, de La Rocque F, et al. Pediatric varicella
hospitalizations in France: a nationwide survey. Clin Microbiol Infect
2007;13:546–9.
[22] Lopez AS, Zhang J, Brown C, et al. Varicella-related hospitalizations in
the United States, 2000–2006: the 1-dose varicella vaccination era.
Pediatrics 2011;127:238–45.
[23] Dinleyici EC, Kurugol Z, Turel O, et al. The epidemiology and economic
impact of varicella-related hospitalizations in Turkey from 2008 to
2010: a nationwide survey during the pre-vaccine era (VARICOMP
study). Eur J Pediatr 2012;171:817–25.
[24] Alp H, Altinkaynak S, Ertekin V, et al. Seroepidemiology of varicellazoster virus infection in a cosmopolitan city (Erzurum) in the eastern
Turkey. Health Policy 2005;72:119–24.
[25] Savas S, Dallar Y, Arikan I, et al. Varicella-zoster virus seroprevalence
in children between 0-15 years old. Mikrobiyol Bul 2004;38:69–75.
[26] Pinquier D, Lécuyer A, Levy C, et al. Inverse correlation between varicella
severity and level of anti-varicella zoster virus maternal antibodies in
infants below one year of age. Hum Vaccin 2011;7:534–8.
[27] Lécuyer A, Levy C, Gaudelus J, et al. Hospitalization of newborns and
young infants for chickenpox in France. Eur J Pediatr
2010;169:1293–7.
[28] Lesko SM, O’Brien KL, Schwartz B, et al. Invasive group A streptococcal
infection and nonsteroidal antiinflammatory drug use among children
with primary varicella. Pediatrics 2001;107:1108–15.
[29] Peterson CL, Mascola L, Chao SM, et al. Children hospitalized for
varicella: a prevaccine review. J Pediatr 1996;129:529–36.
[30] Bozzola E, Tozzi AE, Bozzola M, et al. Neurological complications of
varicella in childhood: case series and a systematic review of the
literature. Vaccine 2012;30:5785–90.
[31] Preblud SR. Varicella: complications and costs. Pediatrics 1986;78:728–35.
[32] Soárez PC, Novaes HM, Sartori AM. Impact of methodology on the
results of economic evaluations of varicella vaccination programs: is it
important for decision-making? Cad Saude Publica 2009;25(Suppl. 3):
S401–14.
[33] Lieu TA, Cochi SL, Black SB, et al. Cost-effectiveness of a routine
varicella vaccination program for US children. JAMA 1994;271:375–81.
[34] Piqueras Arenas AI, Otero Reigada MC, Pérez-Tamarit D, et al.
Hospitalizations for varicella in the Hospital Infantil La Fe, Valencia,
Spain, 2001–2004. An Pediatr (Barc) 63:120–4.
[35] Da YP, Luo LY, Song LZ. Economic burden of inpatient of varicella in
Shandong, Gansu, and Hunan provinces, 2007. Zhongguo Yi Miao He
Mian Yi 2009;15:438–42.
[36] Gershon AA, Takahashi M, Seward J. Varicella vaccine. In: Plotkin
SA, Orenstein WA, eds., Vaccines (4th ed.). Philadelphia: WB Saunders,
2004.
[37] Flatt A, Breuer J. Varicella vaccines. Br Med Bull 2012;103:115–27.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 231–239
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
PATIENT-REPORTED OUTCOMES
The Methodological Challenges for the Estimation of Quality of Life in
Children for Use in Economic Evaluation in Low-Income Countries
Travor Mabugu1,*, Paul Revill2, Bernard van den Berg2
1
Clinical Research Centre, University of Zimbabwe, Harare, Zimbabwe; 2Centre for Health Economics, University of York, York, UK
AB STR A CT
Objectives: The assessment of quality of life (QOL) in children has been
underresearched in high- and low-income countries alike. This is partly
due to practical and methodological challenges in characterizing and
assessing children’s QOL. This article explores these challenges and
highlights considerations in developing age-specific instruments for children affected by HIV and other health conditions in Africa and other lowincome settings. Methods: A literature search identified works that have
1) developed, 2) derived utilities for, or 3) applied QOL tools for use in
economic evaluations of HIV interventions for children. We analyzed the
existing tools specifically in terms of domains considered, variations in age
bands, the recommended respondents, and the relevance of the tools to
African and also other low-income country contexts. Results: Only limited
QOL research has been conducted in low-income settings on either adults
or children with HIV. A few studies have developed and applied tools for
children (e.g., in Thailand, Brazil, and India), but none have been in Africa.
The existing methodological literature is inconclusive on the appropriate
width or depth by which to define pediatric QOL. The existing instruments
include QOL domains such as “physical functioning,” “emotional and
cognitive functioning,” “general behavior (social, school, home),” “health
perception,” “coping and adaptation,” “pain and discomfort,” “extended
effects,” “life perspective,” and “autonomy.” Conclusions: QOL assessment
in children presents a series of practical and methodological challenges. Its
application in low-income settings requires careful consideration of a
number of context-specific factors.
Introduction
many cases are simpler to address for adults. From a practical
viewpoint, it is simply less straightforward to ask children to
describe the domains that make up their QOL or to assess their
own QOL because they might find it hard to clearly communicate
their thoughts and feelings. They can also face difficulties in
distinguishing between good and ill health or well-being, and
permanent and transitory health problems. Approaches that
have been proposed to overcome these difficulties include using
innovative instruments to aid understanding (e.g., use of pictures
[5]), varying questions and tools by age (e.g., see Paediatric
Quality of Life Inventory [PedsQL] in Overview of Children’s QOL
Instruments section) and stage of development of children, and
delivering questions to adult proxies responsible for children
(such as parents or carers). Obviously, the responses of children
and their proxies can vary [6]. They have been shown to differ
most widely with respect to domains around social and emotional functioning, whereas they seem more similar within
physical activity, functioning, and symptoms domains [7].
Effectiveness research on child health interventions in lowincome countries has usually relied solely on intermediate
clinical markers (used primarily to inform narrow clinical decisions) instead of QOL measures, which aim at assessing the wider
The assessment of quality of life (QOL) is crucial to inform
comparisons of the effectiveness of health care and public health
interventions, comparative health research, performance measurement, purchasing decisions, and economic evaluations in
high- and low-income countries alike (see Smith et al. [1] for a
conceptual overview of the QOL literature). Much less progress
has been made in QOL assessment in children than in adult
health. The unique challenges of child health state assessment
have been highlighted elsewhere [2,3]. Ungar and Gerber [4]
explain that children start life as vulnerable infants and develop
toward independent persons. Throughout this process the child
experiences change in dependency relations with parents, relatives, friends, teachers, neighbors, and people from the community and encounters various types of health care workers such as
doctors, nurses, and community health workers. The nature of
the child’s relationships and his or her encounters with health
care workers are likely to significantly affect patterns of health
care use, which then impacts his or her health status.
The development of QOL instruments for children involves
tackling a series of practical and methodological issues that in
Keywords: children, HIV/AIDS, low-income, quality of life.
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflict of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Travor Mabugu, Clinical Research Centre, University of Zimbabwe, Parirenyatwa Hospital Annexe, Cnr J
Tongogara/Mazowe Street, Harare, Zimbabwe.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.07.005
232
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 231–239
impact of a disease on a child’s life and his or her social
environment (see Wilson and Cleary [8] on the conceptual
distinction between clinical and QOL outcomes). The usefulness
of these narrow clinical indicators to evaluate a range of medical
alternatives is limited. Instead, particularly when adopting a public
health approach, there is a need to understand the effect of health
care on a person’s QOL in whatever way this is defined. The
estimation of children’s QOL should therefore be a central priority.
In addition to the practical challenges in assessing children’s
QOL, there are considerable methodological challenges. This article
aims to contribute to the literature by debating methodological
challenges related to pediatric QOL measurement in low-income
settings. By doing so we challenge existing instruments based on the
scope and perspective of existing pediatric QOL tools. We also aim to
inform how pediatric QOL estimation can be used within economic
evaluation. Although the article has a general focus—contributing to
outcome research in general—special attention is paid to addressing
these challenges of QOL assessment for use in the economic
evaluation of health care, and especially HIV/AIDS interventions.
The article proceeds as follows. A literature search for QOL
assessment in children is outlined. Based on the literature search
considerations of the QOL domains, suggested respondents and
age bands are described. Next, methodological considerations for
transforming QOL outcomes for economic evaluation are presented
Screening
Idenficaon
Arcles idenfied
through PubMed
database
searching
(n = 393)
together with a general typology on the forms of economic
evaluation that QOL estimation can inform. The article concludes
after a general discussion on the scope of the implications of our
article and a reflection on our work compared with other studies.
Literature Search
Search
The literature search was conducted by the first author. He
followed the Preferred Reporting Items for Systematic Reviews
and Meta-Analyses guidelines. The primary search term used was
“quality of life” and the following secondary search terms were
used: “children,” “HIV,” “AIDS,” “instrument,” and “measurement.”
The PubMed and PubMed Central search engines were used. In
addition, the following selected journals were searched: Quality of
Life Research, Pharmacoeconomics, Journal of Health Economics, AIDS &
STD Patient Care, and Cost-Effectiveness and Resource Allocation.
Eligibility Criteria
Included articles 1) developed QOL instruments for, 2) derived
utilities (values) for, or 3) applied tools in economic evaluations of
Arcles idenfied
through PMC
database searching
(n = 159)
Arcles idenfied through selected Journals
Pharmaco-economics (n = 300)
Health Economics (n = 312)
Cost-effecveness & Resource Allocaon (n = 35)
AIDS Paent Care & STDs (n = 424)
Quality of Life research (n = 113)
Records aer duplicates removed and addion of informal search arcles
(n = 1495)
Eligibility
Records excluded aer
inial screening by tle
and abstract
(n = 1444)
Full-text arcles assessed
for eligibility
(n = 51)
Included
Full-text arcles excluded,
with reasons
(n = 35)
Studies included in
qualitave synthesis
(n = 13)
Fig. 1 – Flow chart of literature search events. Thirty-five articles were excluded for a number of reasons, with the main reasons
for exclusion being as follows: assessed the QOL of adults with HIV-infected children (13 articles), commentaries (3) or letters (4)
that highlighted measurement issues of QOL in HIV-infected children, compared the type of interview (researcher vs. selfadministered interviews) (2 articles), the tools used applied to children older than 18 years (8 articles), and reviews that
highlighted general measurement of health outcomes in both children and adults (5 articles). We ensured that tools used or
mentioned in these excluded articles were captured in the final 13 articles used in the analysis. QOL, quality of life.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 231–239
HIV interventions in children. Articles published in any language
other than English were excluded. The focus population was
children in low-income settings, but particular attention paid to
the QOL of children with HIV in Africa. See Figure 1 for more
details.
Overview of Children’s QOL Instruments
The overview revealed an appreciable body of work on QOL
research in HIV/AIDS conducted internationally [9], including
some studies in sub-Saharan Africa [10,11], as well as for children
[12–14], but not for children in Africa.
Table 1 shows all articles for which pediatric QOL tools for
HIV/AIDS have been developed or applied and names the tools
that were taken forward for qualitative analysis. The QOL instruments are either generic health status instruments (in which
scores can be compared for many diseases, e.g., Child Quality of
Life Scale [CQLS], PedsQL, and Quality of Life Assessment Questionnaire [QLAQ]) or HIV condition/disease-specific instruments
(which measure narrowly focused patient views on the impact of
HIV/AIDS). The literature search yielded a mixture of these tools
but mostly generic instruments.
The identified tools have all been developed for use primarily in
high- and upper-middle-income countries. A few studies have
applied these tools in middle-income countries such as Thailand,
Brazil, and India. Zambia is the only lower-middle income country
in Africa that has reported application of child utilities in economic
evaluation [15]. In this study, adult utilities were extrapolated for
use in children, and the authors highlight lack of child-specific QOL
weights appropriate to lower-income settings.
The dimensions or domains of the retrieved tools are summarized in Table 2. The number of domains varies from 4 (in
PedsQL 4.0) to 12 (Child Health Questionnaire 28). The most
common domains included in the tools were variants of physical,
emotional, and social functioning, and general health perception.
The CQLS contains leisure and family domains, and the QLAQ
has a measure of physical resilience. Despite requests we were
unsuccessful in obtaining access to the Thai Quality of Life for
HIV Infected Children tool, and therefore it was not included in
subsequent analyses.
Table 3 shows the origin of the tools considered in this
analysis. It also gives a summary of the key factors that were
considered in the elicitation of responses, how the summary
scores were obtained, and which age groups were considered in
the studies analyzed.
An important issue when assessing child QOL is the determination of an appropriate respondent—in particular whether it
should be a child and/or a proxy (parent or carer). Only the PedsQL,
Thai Quality of Life for Children, and Child Health Questionnaire 28
allow for both adult proxy responses and child self-responses. Other
instruments recommend either adult proxies (e.g., the QLAQ and
General Health Assessment for Children) or child self-response (e.g.,
the CQLS). Table 3 also contains a subanalysis on the age sensitivity
of the identified tools. We also note the width of age variation, and
this is further discussed below. Most instruments were developed
in what would be categorized as middle- to high-income countries
using standard World Bank classifications [16]. Application of these
tools in lower-income settings can raise a number of concerns
associated with the diverging socioeconomic and cultural contexts
of these settings, although some similarities can also be drawn
between these different economies. We believe that three primary
questions need to be addressed: 1) which dimensions and 2) which
age bands apply in the African context, and finally 3) whose
response is most appropriate to elicit (the child’s or a proxy’s).
233
Key Issues to Consider in Assessment of QOL in
Children
What Dimensions to Consider?
Table 4 presents the domains that have been used to assess child
QOL in existing instruments.
This table generally suggests, as observed above, that major
domains such as physical, psychological, and social functioning
and general health perception form basic and fundamental
pillars of a child’s QOL. Physical functioning is an assessment
of a child’s ability to perform daily tasks and includes sitting,
walking, running, and playing. It is about general motility.
Psychological functioning assesses aspects such as cognitive
and attentive abilities, emotions, personality, behavior, and interpersonal relationships. Social wellness assesses internal and
external engagement at all levels (such as at the micro level of
individual agency and the macro level of systems at school,
home, and other social structures). Finally, general health perception assesses sentiments relating to current health status. It is
a function of an individual’s perception of the extent of deviation
in health status from a desired or aspired level, and is often
affected by changes in health status over time.
In addition, other descriptors of QOL that need to be explored
for importance and relevance include domains such as life
perspective; autonomy, pain, and discomfort; extended effects;
and coping and adaptation strategies.
Whose Response to Consider?
It is inevitable that different individuals report QOL differently,
especially if they are proxy respondents. Various potential
choices exist for pediatric study populations. A key distinction
can be made between children themselves and adult proxies—
either a parent or caregiver, or a health care provider. The
literature highlighted here is inconclusive on this issue. A more
precise response can be obtained from self-responses, although
younger children face more cognitive difficulties in expressing
themselves than do older children. Therefore, because of developmental and cognitive changes that occur as the child grows
(see Is Age Sensitivity an Issue? section), it may be useful to
obtain either child and/or adult proxy responses depending on
the age of the child. Adult proxy responses represent close
substitutes for younger children who are not able to comprehend
and express themselves clearly. As children grow older, however,
their responses can then be considered.
Is Age Sensitivity an Issue?
A closely related consideration to determining the appropriate
respondent is to reflect childhood developments in QOL measurement based on a child’s age. A child’s life trajectory is typically
characterized by development in cognitive abilities and changing
dependency relations as the level of autonomy increases. The
studies highlighted in the review do not offer one consistent
approach to address the age-dependent elements of pediatric
QOL measurement (see Table 3).
If we consider the ecological model of influences on child
development [4], different systems exert differing effects on the
development of children and are each introduced at varying ages
(and with some degree of overlap). This results in unpredictable
impacts on the cognitive abilities of a child. Different health
systems in Africa have categorized children by age to inform
clinical practice. In Malawi, for example, children are grouped
into the following age groups: 0 to 2 years, 3 to 5 years, and 6 to 15
years. Appropriate tools should therefore be age sensitive to
reflect key developmental milestones that affect pediatric QOL.
234
Table 1 – Articles included for qualitative data analysis.
Title
AA
22
Oral health related
quality of life of
paediatric patients
with AIDS
Poor quality of life
among untreated
Thai & Cambodian
children without
severe HIV
symptoms
Health related quality
of life assessment
questionnaire for
children aged 5-11
years with HIV/AIDS:
cross-cultural
adaptation for
Portuguese language
Development of the
EQ-5D-Y—a child
friendly version of
the EQ-5D
AA
24
AA
25
AA
29
Authors
Year
Journal
Country
Massarente
et al. [33]
2011
BMC Oral
Health
Journal
Brazil
Bunupiradah
et al. [34]
2011
AIDS Care
Thailand,
Cambodia
Costa et al.
[14]
2011
Cad. Saúde
Pública,
Rio de
Janeiro
Brazil
Wille et al.
[31]
2010
Quality of
Life
Research
Germany,
Italy,
Spain,
and
Sweden
AA
17
Quality of life of
children living in
HIV/AIDS-affected
families in rural
areas in Yunman,
China
Xu et al. [13]
2010
AIDS Care
China
AA
18
Health-related quality
of life in HIVinfected children
using PedsQL™ 4.0
and comparison
with uninfected
children
Banerjee
et al. [35]
2010
Quality of
Life
Research
India
Objectives
Population
group
(sample
size)
Eligibility
criteria
Instrument
used
GNI per
capita
Setting
10–15 y (88)
Value
elicitation
OHR-QoL
11,420.00
MIC
1–11 y (294)
Tool
development
and value
elicitation
GCHA
8,360.00
and
2,230.00
UMIC/
LMIC
To assess cross-cultural
adaptation of the
QLAQ used to
measure the HRQOL
in Brazilian children
aged 5–11 y with HIV/
AIDS
5–11 y (35)
Tool
development
QLAQ
11,420
MIC
To develop a self-report
version of the EQ-5D
questionnaire for
younger respondents
(EQ-5D-Y), and test
comprehensibility of
children and
adolescents with
HIV/AIDS
To explore factors
influencing the
HRQOL of children
living in HIV-affected
families in rural
areas in Yuunan,
China.
Assess reliability and
validity of PedsQL 4.0
in children with HIV,
and the association
of HIV infection
treatment regimens
48 y
Tool
development
EQ-5D-Y
40,170.00/
32420.00/
31,440.00/
42,210.00
HIC (all)
8–17 y (225)
Values
elicitation
PedsQL 4.0
8,390.00
MIC
8–12 y (300)
Values
elicitation
PedsQL 4.0
3,620.00
LMIC
Assessment of oral
HRQOL of HIV/AIDSinfected children and
associated factors
To evaluate QOL in
untreated Thai and
Cambodian children
with HIV who do not
have severe HIV
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 231–239
Code
Value
elicitation
PedsQL 4.0
3,620.00
LMIC
8–16 y (292)
Value
elicitation
PedsQL 4.0
and ThQLC
8,360.00
MIC
To develop a reliable
and valid selfreported HRQOL
instrument for HIVinfected children in
Thailand
Z8 y (292)
Tool
development
ThQLHC
8,360.00
MIC
USA, Pourte
Rica
To examine the impact
of HIV disclosure on
pediatric QOL
45 y (395)
Values
elicitation
GHAC
48,820.00/-
HIC/
HIC
AIDS
Zambia
1–14 y (534)
Utilities
equated
VAS, TTO, SG
1,490.00
LMIC
Quality of
Life
Research
USA
Assess the CEA of
cotrimoxazole
prophylaxis in HIVinfected children in
Zambia
To expand the
psychometric
properties of the
Child Health
Questionnaire 28
(CHQ-28) parent
report short form as a
measure of wellbeing for children
with chronic illness
5–18 y(33)
Tool
development/
adjustment
CHQ-28
48,820.00
HIC
Quality of life and
psychosocial
functioning of HIVinfected children
Das et al. [36]
2010
Indian
Journal of
Paediatrics
India
AA
20
Health-related quality
of life of Thai
children with HIV
infection: a
comparison of the
Thai Quality of Life
in children (TlQLC)
with the PedsQL™
4.0 generic core
scales
Development of a
culturally
appropriate healthrelated quality of life
measure for HIVinfected children in
Thailand
Impact of disclosure of
HIV infection on
HRQoL among
children and
adolescents with
HIV infection
Cost- effectiveness of
cotrimoxazole
prophylaxis in HIVinfected children in
Zambia
Psychometrics of child
health questionnaire
parent short form
(CHQ-28) used to
measure quality of
life in HIV-infected
children in complex
anti-retroviral
therapy
Punpanich
et al. [37]
2010
Quality of
Life
Research
Thailand
Punpanich
et al. [38]
2010
Journal of
Peadiatrics
and Child
Health
Thailand
Butler et al.
[39]
2009
Paediatrics
Ryan et al.
[15]
2008
Byrue et al.
[40]
2005
AA
27
AA
26
AA
28
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 231–239
Z6 y(71)
AA
19
AA
21
and type of care
received on QOL
To assess the QOL and
psychosocial
problems of HIVinfected children, in
comparison to
children with cystic
fibrosis
Assess reliability and
validity of the ThQLC
instrument in
comparison with the
PedsQL 4.0 in
children on longterm HIV care
235
236
CEA, cost-effectiveness analysis; CHQ-28, Child Health Questionnaire 28; CQLS, Child Quality of Life Scale; EQ-5D, EuroQol five-dimensional; GHAC, General Health Assessment for Children; HIC,
high income country; HRQOL, health-related quality of life; LMIC, lower middle income country; OHR-QoL, oral health related-quality of life; PedsQL, Paediatric Quality of Life Inventory; QLAQ,
Quality of Life Assessment Questionnaire; QOL, quality of life; SG, standard gamble; ThQLC, Thai Quality of Life for Children; ThQLHC, Thai Quality of Life for HIV Infected Children; TTO, time
trade-off; UMIC, upper middle income country; VAS, visual analogue scale.
LMIC
11,420.00
CQLS
Tool
development
and
application
4–12 y (100)
To validate the scale of
children’s QOL in a
group of children
infected with HIV
receiving clinical care
in Brazil
2005
Ferreira et al.
[41]
AA
23
Validation study of a
scale of life quality
evaluation in a
group of paediatric
patients infected
with HIV
Ciência
Saúde
Coletiva
Brazil
Population
group
(sample
size)
Year
Authors
Title
Code
Table 1 – continued
Journal
Country
Objectives
Eligibility
criteria
Instrument
used
GNI per
capita
Setting
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 231–239
Table 2 – List of domains found in identified tools.
Number
Instrument
1
PedsQL4.0
2
ThQLC
3
ThQLHC
4
CQLS
5
QLAQ
6
EQ-5D-Y
7
GHAC
8
CHQ-28
Domains considered
Physical, emotional, social, and
school functioning
Physical, emotional, social, life, and
school functioning
HIV-targeted tool used to assess
symptoms associated with HIV in
children
Autonomy, leisure, functions, and
family
Physical, psychological, social/role
functioning, health perceptions,
health care utilization, and
physical resilience
Mobility, self-care, pain/discomfort,
anxiety/depression, and usual
activities
General health perception,
symptom distress, psychological
status, and physical functioning
Physical function; limitation to
social role related to emotional or
physical function; bodily pain;
general behavior; mental health;
self-esteem; general health
perception; change in health over
past year; impact on parental
time; impact on parental
emotions; family activities; and
family cohesion
CHQ-28, Child Health Questionnaire 28; CQLS, Child Quality of Life
Scale; EQ-5D-Y, EuroQol five-dimensional questionnaire for
younger respondents; GHAC, General Health Assessment for Children; PedsQL, Paediatric Quality of Life Inventory; QLAQ, Quality of
Life Assessment Questionnaire; ThQLC, Thai Quality of Life for
Children; ThQLHC, Thai Quality of Life for HIV-Infected Children.
This is an HIV-specific tool and could not access the tool for
comparison purposes.
Children aged 2 years or younger generally require the full-time
attention of the mother or another adult. The mother or carer is
then fully responsible for assessing the welfare of the child by
using means such as interpreting their expressions. Children
aged 3 to 5 years old have some increased autonomy and will
likely spend some time away from the major caregiver, be given
opportunity to play alone or with others, and will perhaps spend
time in day care activities. They also have an improved ability to
report ill health and discomfort, even though they may not be
able to describe this in detail. Children aged 6 to 12 years
experience life increasingly away from family members. They
are often introduced to formal education (although not always in
the African context), and the expectations placed on them are
somewhat higher—for instance, they may need to do homework
or help with housework. In the age range 13 to 15 years, children
have their final years of childhood. They start to experience
hormonal changes and transition to life as adults.
Using QOL Measurement in Economic Evaluation
After the outcomes of health care interventions have been
measured, using a context-sensitive QOL tool, there is a need to
237
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 231–239
Table 3 – Origins of tools, income settings, variation in respondents, and age sensitivity of tools.
Tool
Country originally developed (and tested)
Development
PedsQL
4.0
ThQLC
CQLS
QLAQ
EQ-5DY
GHAC
CHQ-28
Tested
Income
Respondent
Age sensitive
Self
Adult
proxy
Yes/
no
Specify
San Diego, CA
San Diego, CA
High income
Yes
Yes
Yes
2–18 y
Thailand
Brazil
Brazil
Germany, Italy, Spain,
and Sweden
USA
USA
Thailand
Brazil
Brazil
Germany, Spain, and
South Africa
USA
USA
Upper-middle income
Upper-middle income
Upper-middle income
High- and uppermiddle income
High income
High income
Yes
Yes
No
Yes
Yes
No
Yes
No
Yes
Yes
Yes
Yes
5–16 y
4–12 y
5–12 y
Z8 y
No
Yes
Yes
Yes
Yes
Yes
Z5 y
Z8 y
CHQ-28, Child Health Questionnaire 28; CQLS, Child Quality of Life Scale; EQ-5D-Y, EuroQol five-dimensional questionnaire for younger
respondents; GHAC, General Health Assessment for Children; PedsQL, Paediatric Quality of Life Inventory; QLAQ, Quality of Life Assessment
Questionnaire; ThQLC, Thai Quality of Life for Children.
value these outcomes if these are to be used within economic
evaluation.
Economic evaluations can guide decision makers on the
allocation of their budgets, whether this is in health care or more
specifically for HIV/AIDS interventions. There are various
approaches to conducting economic evaluations depending on
the scope and perspective of the study. Literature however points
broadly to two predominant forms of analyses: cost-utility (sometimes called cost-effectiveness) analyses and cost-benefit analyses [17]. The key challenge in all forms of economic evaluation is
to capture the impact of various interventions in affecting the
QOL of patients, or potential future patients, and their associated
costs. For pediatric evaluations, there is therefore a need to
measure the QOL impacts of pediatric interventions, using some
form of instrument, and then for these impacts to be valued.
These values can then be compared directly to costs to inform an
assessment of value for money.
There are then broadly two ways of incorporating QOL
estimation into economic evaluation, and the appropriate
approach to valuation relates to the form of analysis adopted:
(1) for cost-utility/cost-effectiveness analyses, effects of interventions have to be translated into “utilities,” and (2) for cost-benefit
analyses, it is necessary to translate effects of interventions into
monetary values. These two approaches are considered below.
Calculating Utilities for Use within Cost-Utility/Effectiveness
Analyses
Utilities value QOL on a scale in which 1 represents “full health”
and 0 represents “death.” This has two major advantages: first,
they enable comparisons across different interventions and
therapeutic areas, and second, they can enable reduced mortality
and reduced morbidity to be assessed concurrently.
The applied literature on the evaluation of pediatric interventions has often struggled to determine appropriate utilities in
both high- and low-income country settings. For instance, in an
evaluation of prophylaxis for HIV-infected children, Ryan et al.
[15] had to apply adult utility values to child health outcomes.
The authors explain that there were no utilities available for HIVinfected children and instead they have to rely on the use of adult
utilities. This clearly is a limitation of this and other pediatric
economic evaluations in HIV/AIDS.
Two of the most widely used methods to determine preference or utility scores are the standard gamble and time trade-off
[17,18]. These are choice or preference-based methods in which
standard gamble involves eliciting preferences from respondents
between health states that are uncertain, whereas time trade-off
involves making choices on the basis of the length of time spent
in alternative health states. Both approaches are relatively
expensive to undertake on a case-by-case basis. One alternative
is the visual analogue scale, which can be administered alongside
QOL measurement tools to inform utilities [19].
Table 4 – Description of new domains for QOL in
children.
Domain
Physical functioning
Emotional and
cognitive
functioning
General behavior
(social, school,
home)
Life perspective
Health perceptions
Copying/adaptation
Pain and discomfort
Extended effects
Autonomy
QOL, quality of life.
Description
Ability to sit, walk, run, play,
participate in general activities that
require use of body, general motility
Mental wellness (neurological
impairment, anxiety, depression),
emotions (sad, anger, fear,
happiness, self-esteem, confidence),
and cognitive (intellectual mind,
comprehension, etc.)
Behavior at home (e.g., family
participation and cohesion), school
(class participation, schoolwork,
group work, school absenteeism or
preseentism), and social/role
playing (among other kids)
Views on life events, life expectations
Overall perception of health (includes
perception of current health,
changes in health over the previous
period, health care utilization as a
proxy to overall health status)
Ability to cope or adapt to changes in
health or to changes not related to
health
Pain and discomfort
The external effects of caring for the
sick child, e.g., time, emotions,
burden, effort, and impact on your
current health and well-being
Examines the level of independence of
the child, self-reliance
238
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 231–239
Calculating Monetary Values for Use within Cost-Benefit
Analyses
A second means of assessing value from HIV/AIDS and health
care interventions is to assign monetary values to health outcome measures directly.
Monetary valuation can be assigned by using either revealed
preferences or stated preferences (SP) of subjects. The well-being
valuation method is a relatively new approach in the revealed
preferences tradition to calculate monetary valuations (see
Powdthavee and van den Berg [20] or Deaton et al. [21]). In the
absence of a functioning market for particular commodities, such
as is the case for health gains resulting from interventions, an SP
approach usually has to be adopted by an analyst.
One common way to elicit SP to calculate monetary values is
to use contingent valuation. This method elicits values by using
hypothetical questions contained in a survey [22,23]. Individuals
are asked to directly state their willingness to pay for medicines
or treatments associated with an illness, or for inclusion of
interventions in health insurance packages or reimbursement
lists [24–28].
The related approach of undertaking discrete choice experiments asks similar questions but over a whole range of attributes, including health domains and monetary outcomes,
enabling monetary values to be placed on the outcomes associated with the receipt of interventions [29].
Discussion and Conclusions
The idea of this article arose during the development of an
instrument for QOL assessment in children with HIV in subSaharan Africa. During the process of developing that instrument, it became clear that there were quite a few methodological
challenges that remained unresolved in the literature—in particular relating to choices of domains, recommended respondents,
and child age bands. It also became clear that these are of
relevance not only to the evaluation of HIV-related outcomes in
Africa but also to pediatric QOL estimation in low-income
countries more generally. This article therefore is not meant as
a research manuscript but instead has outlined some of the
methodological challenges. It is hoped this will encourage further
debate on methodological issues relating to QOL assessment in
children. The article has also outlined how QOL outcomes can be
valued for use within economic evaluation studies.
Identifying existing instruments on the basis of the literature
search suggested that quite a broad range of QOL domains should
be considered for inclusion in any instrument. These could be
labeled as “physical functioning,” “emotional and cognitive functioning,” “general behavior (social, school, home),” “health perception,” “coping and adaptation,” “pain and discomfort,”
“extended effects,” “life perspective,” and “autonomy.” We do
not claim that our search was exhaustive, for example the CHU9D
tool [42] did not appear in our search. However, the domains are
broadly similar.
This finding is particularly interesting because it suggests that
two of the most notable existing instruments (PedsQL and EuroQol five-dimensional questionnaire for young respondents [EQ5D-Y]) may have a focus that is too narrow to fully capture
children’s QOL. The PedsQL is considered one of the most
promising QOL instruments for children [13]. It includes domains
covering physical, emotional, social, and school functioning. It
does not, however, contain domains found in other tools such as
health perception, coping and adaptation, and pain and discomfort. The EQ-5D-Y stems from the well-known adult EQ-5D
questionnaire, which is the main outcome measure in costutility analysis [30]. It assesses childhood QOL by using domains
of mobility, self-care, pain/discomfort, anxiety/depression, and
usual activities. These match domains used for adults although it
is expected that the types of considerations within each may
change between adults and children—for instance, usual activities for adults will typically include work and leisure, whereas for
children they are more likely to focus around things such as
school and play.
The scope of both the PedsQL and the EQ-5D-Y may therefore
not be wide enough to capture what is really important for
children’s QOL. This has already notably been acknowledged,
for instance, by Willie et al. [31]. The implication, however, is that
the search for appropriate measurement tools for pediatric QOL,
even in high-income countries, is not a concluded task.
This article has also highlighted other considerations relating to
whether tools should be catered to particular age groups and the
related question of whose responses to consider (the child’s or an
adult proxy’s). As a result of developmental and cognitive changes
that occur as the child grows, we believe that catering tools to
different age groups is required. This also has implications on who
should respond to questions. Adult proxy responses represent close
substitutes for younger children who are not able to comprehend and
express themselves clearly. As children grow older, however, their
responses can be considered. Others have discussed QOL measurement in pediatric patients with HIV on the basis of demonstrating
adequate psychometric properties of existing instruments [12].
Finally, this article has outlined how pediatric QOL estimation
can be used within economic evaluation. This requires valuation
of the effects of interventions—either using utilities (for use in
cost-utility/effectiveness analyses) or assigning monetary values
(for use in cost-benefit analysis). The applied QOL literature has
tended to use adult health profiling instruments in child health
state assessment and/or adult preferences for health utilities.
This is a major limitation that risks leading to inappropriate
policy advice on the use of pediatric health care interventions.
Although there is a substantial literature on parents’ willingness
to pay even in low-income settings that we may not have
captured in our search, it is worth emphasizing that in this
literature it is generally parents who value children’s QOL. It
seems therefore fair to conclude that the more accurate informing of policy requires either both child-specific measurement and
valuation of QOL or better involvement of parents in valuation of
children’s QOL. When parents are going to value children’s QOL,
the analyst should consider basing QOL estimations on interdependent utility functions [32]. It also requires measurement and
valuation to be appropriately culturally sensitive [14]. Regardless
of the valuation approach, it is necessary to be sensitive to
cultural variations in the conception of QOL that will ultimately
affect the value put on the QOL estimated. Cultural variations
often correlate with levels of economic well-being, although this
is not necessarily always true. QOL estimation should therefore
be based on a culturally sensitive tool, and valuation should also
be sensitive to cultural differences and economic conditions.
In conclusion, QOL estimation in children presents a series of
practical and methodological challenges. Its application in lowincome countries requires careful consideration of a number of
context-specific factors. The article challenges existing instruments to capture a broader range of domains to assess pediatric
QOL for effectiveness research. We have also informed how
pediatric QOL outcomes can be valued for use within economic
evaluation studies.
Acknowledgment
We gratefully acknowledge the suggestions and comments of three
anonymous reviewers. We thank the Department for International
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 231–239
Development and the Medical Research Council Clinical Trials Unit
for supporting this work.
Source of financial support: This work was funded by the
Department for International Development, through the Medical
Research Council Clinical Trials Unit, as part of the ARROW:
Young Lives project and the Lablite Project.
R EF E R EN CE S
[1] Smith KW, Avis NE, Assmann SF. Distinguishing between quality of life
and health status in quality of life research: a meta-analysis. Qual Life
Res 1999;8:447–59.
[2] Halfon N, Nweacheck PW. Characterizing the social impact of asthma
in children. In: Weiss KB, Buist AS, Sullivan SD,eds., Asthma’s Impact
on Society, the Social and Economic Burden. New York: Marcel Dekker,
Inc, 2000.
[3] Sameroff A, Chandler M. Reproductive risk and the continuum of
caretaking casualty. In: Horowitz F,ed., Review of Child Development
Research. Chicago: University of Chicago Press, 1975.
[4] Ungar WJ, Gerber A. The uniqueness of child health and challenges to
measuring costs and consequences. In: Ungar WJ,ed., Economic
Evaluation in Child Health. New York, NY: Oxford University Press,
2010.
[5] De Civita M, Regier D, Alamgir AH, et al. Evaluating health-related
quality-of-life studies in paediatric populations: some conceptual,
methodological and developmental considerations and recent
applications. Pharmacoeconomics 2005;23:659–85.
[6] Britto MT, Kotagal UR, Chenier T, et al. Differences between
adolescents’ and parents’ reports of health-related quality of life in
cystic fibrosis. Pediatr Pulmonol 2004;37:165–71.
[7] Eiser C, Morse R. Quality of life measures in chronic diseases of
childhood. Health Technol Assess 2001;5:1–157.
[8] Wilson IB, Cleary PD. Linking clinical variables with health-related quality
of life: a conceptual model of patient outcomes. JAMA 1995;273:59–65.
[9] Skevington SM, O’Connell KA. Measuring quality of life in HIV and
AIDS: a review of the recent literature. Psychol Health 2002;18(3):331–50.
[10] Robberstad B, Olsen JA. The health related quality of life of people
living with HIV/AIDS in sub-Saharan Africa – a literature review and
focus group discussion. Cost Effect Res Alloc 2011;8:5.
[11] Jansen van Rensburg MS. Measuring the quality of life of residents in
SADC communities affected by HIV. AIDS Care 2007;21(9):1132–40.
[12] Garvie PA, Lawford J, Banet MS, West RL. Quality of life measurement in
paediatric and adolescent populations with HIV: a review of the
literature. Child Care Health Dev 2009;35:440–53.
[13] Xu T, Wu Z, Rou K, et al. Quality of life of children living with HIV/AIDSaffected families in rural areas in Yunnan, China. AIDS Care Psychol
Socio-Med Asp AIDS/HIV 2010;22:390–6.
[14] Costa LS, Dias de Oliveira Latorre MDR. Health related quality of life
assessment questionnaire for children aged 5-11 years with HIV/AIDS:
cross-cultural adaptation for the Portuguese language. CadSaude
Publica, Rio de Janeiro 2011;27:1445–9.
[15] Ryan M, Griffin S, Chitah B, et al. The cost-effectiveness of cotrimoxazole prophylaxis in HIV-infected children in Zambia. AIDS
2008;22:749–57.
[16] World Bank. How we classify countries [online]. Available from: http://
data.worldbank.org/about/country-classifications. [Accessed April 21,
2013].
[17] Drummond MF, O’Brien B, Stoddart GL, Torrance GW. Methods for the
Economic Evaluation of Healthcare Programmes (3rd ed.). New York,
NY: Oxford University Press, 2005.
[18] Torrance GW. Measurement of health utilities for economic appraisal:
a review. J Health Econ 1996;5:1–30.
[19] Parkin D, Devlin N. Is there a case for using visual analogue scale
valuations in cost utility analysis? Health Econ 2006;15:653–64.
[20] Powdthavee N, van den Berg B. Putting different price tags on the same
health condition: re-evaluating the well-being valuation approach.
J Health Econ 2011;30:1032–43.
239
[21] Deaton A, Forston J, Tortora R. Life (Evaluation), HIV/AIDS, and Death in
Africa. NBER Working Papers 14637. Cambridge, MA: National Bureau of
Economic Research, Inc., 2009.
[22] O’Brien B, Ganfi A. When do the ‘dollars’ make sense? Toward a
conceptual framework for contingent valuation studies in health care.
Med Decis Making 1996;16:288–99.
[23] Tsuchiya A, Williams A. Welfare economics and economic evaluation.
In: Drummond M, McGuire A,eds., Economic Evaluation in Health Care.
New York, NY: Oxford University Press, 2001.
[24] Donaldson C, Shackley P, Addalla M, Miedzybrodzka Z. Willingness to
pay for antenatal carrier screening for cystic fibrosis. Health Econ
1995;4:439–52.
[25] Liu J, James K, Hammitt JW, Liu JL. Mother’s willingness to pay for her
own and her child’s health: a contingent valuation study in Taiwan.
Health Econ 2000;9:319–26.
[26] Amin M, Khondoker F. A contingent valuation study to estimate
the parental willingness-to-pay for childhood diarrhoea and gender
bias among rural households in India. Health Res Policy Systems
2004;2:3.
[27] Pinto-Prades J, Loomes G, Brey R. Trying to estimate a monetary value
for the QALY. J Health Econ 2009;28:553–62.
[28] Van der Star SM, van den Berg B. Individual responsibility and healthrisk behaviour: a contingent valuation study from the ex ante societal
perspective. Health Policy 2011;101:300–11.
[29] Ryan M. Discrete choice experiments in healthcare. BMJ 2004;328:360.
[30] Richardson J, McKie J, Bariola E. Review and Critique of Health Related
Multi Attribute Utility Instruments. Research Paper 64, Centre for
Health Economics, Monash University, Melbourne, Australia, 2011.
[31] Willie N, Badia X, Bonsel G, et al. Development of the EQ-5D-Y: a child
friendly version of the EQ-5D. Qual Life Res 2010;19:875–86.
[32] Van den Berg B, Bleichrodt H, Eeckhoudt L. The economic value of
informal care: a study of informal caregivers’ and patients’ willingness
to pay and willingness to accept for informal care. Health Econ
2005;14:363–76.
[33] Massarente DB, Domaneschi C, Marques HHS, et al. Oral health-related
quality of life of paediatric patients with AIDS. BMC Oral Health
2011;11:2.
[34] Bunupuradah T, Puthanakit T, Kosalaraksa P, et al. Poor quality of life
among untreated Thai and Cambodian children without severe HIV
symptoms. AIDS Care. DOI:10.1080/09540121.2011.592815.
[35] Banerjee T, Pensi T, Banerjee D. HRQoL in HIV-infected children using
PedsQLTM 4.0 and comparison with uninfected children. Quality of Life
Research 2010;19:803–12.
[36] Das S, Mukherjee A, Lodha R, Vatsa M. Quality of Life and Psychosocial
Functioning of HIV Infected Children. Indian J Pediatr 2010;77:633–7.
[37] Punpanich W, Boon-yasidhi V, Chokephaibulkit K, et al. Health-related
Quality of Life of Thai children with HIV infection: a comparison of the
Thai Quality of Life in Children (ThQLC) with the Pediatric Quality of
Life Inventory™ version 4.0 (PedsQLTM 4.0) Generic Core Scales. Quality
of Life Research 2010;19:1509–16.
[38] Punpanich W, Hays RD, Detels R, et al. Development of a culturally
appropriate health-related quality of life measure for human
immunodeficiency virus-infected children in Thailand. Journal of
Paediatrics and Child Health 2011;47:27–33.
[39] Butler AM, Williams PL, Howland LC, et al. Impact of Disclosure of HIV
Infection on Health-Related Quality of Life Among Children and
Adolescents With HIV Infection. Pediatrics 2009;123:935–43.
[40] Byrne MW, Honig J. Psychometrics of child health questionnaire parent
short form (CHQ-28) used to measure quality of life in HIV-infected
children on complex anti-retroviral therapy. Quality of Life Research
2005;14:1769–74.
[41] Ferreira D, RomeroLealPassos M, dePaulaMottaRubini N, et al.
Validation study of a scale of life quality evaluation in a group of
pediatric patients infected by HIV. Ciência & Saúde Coletiva
2011;16:2643–52.
[42] Stevens K, Rathcliffe J. Measuring and Valuing Health Benefits for
Economic Evaluation in Adolescence: An Assessment of the Practicality
and Validity of the Child Health Utility 9D in the Australian Adolescent
Population. Value in Health 2012;15:1092–9.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 240–247
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
The Impact of Pharmaceutical Care Intervention on the Quality of Life of
Nigerian Patients Receiving Treatment for Type 2 Diabetes
Maxwell O. Adibe, PhD, MPharm, BPharm1,2,, Chinwe V. Ukwe, PhD, MPharm, BPharm1,2, Cletus N. Aguwa, PharmD1,2
1
Department of Clinical Pharmacy and Pharmacy Management, University of Nigeria, Nsukka, Enugu, Nigeria; 2Pharmacotherapeutic Group, Department of
Clinical Pharmacy and Pharmacy Management, University of Nigeria, Nsukka, Enugu, Nigeria
AB STR A CT
Objectives: To evaluate the impact of pharmaceutical care (PC)
intervention on health-related quality of life (HRQOL) of patients with
type 2 diabetes. Methods: This study was a randomized, controlled
study with a 12-month patient follow-up. The study protocol was
approved by the Research Ethical Committees of the institutions in
which this study was conducted. A total of 110 patients were
randomly assigned to each of the “intervention” (PC) and “control”
(usual care [UC]) groups. Patients in the UC group received the usual/
conventional care offered by the hospitals. Patients in the PC group
received UC and additional PC for 12 months. The HUI23S4EN.40Q
(developed by HUInc - Mark index 2&3) questionnaire was used to
assess the HRQOL of the patients at baseline, 6 months, and 12
months. Two-sample comparisons were made by using Student’s t
tests for normally distributed variables or Mann-Whitney U tests for
nonnormally distributed data at baseline, 6 months, and 12 months.
Comparisons of proportions were done by using the chi-square test.
Results: The overall HRQOL (0.86 ⫾ 0.12 vs. 0.64 ⫾ 0.10; P o 0.0001)
and single attributes except “hearing” functioning of the patients were
significantly improved at 12 months in the PC intervention arm when
compared with the UC arm. The HRQOL utility score was highly
negatively (deficit ≥10%) associated with increasing age (≥52 years),
diabetes duration (44 years), emergency room visits, comorbidity of
hypertension, and stroke in both PC and UC groups. Conclusion:
Addition of PC to UC improved the quality of life in patients with type
2 diabetes.
Keywords: HRQOL, patients with diabetes, pharmaceutical care
intervention, quality of life, usual care.
Introduction
standard treatment guideline to streamline the process of diabetes
management and what service the patients should receive [7].
Several research studies have been carried out on health deficit
associated with diabetes comorbidities. For instance, the work
done by Maddigan et al. [8] to assess the impact of comorbid heart
disease, stroke, and arthritis on HRQOL in people with diabetes in
the general Canadian population concluded that “The illness
burden experienced by individuals with diabetes is not only
associated with diabetes itself, but largely with co-morbid medical
conditions.” Also, Westaway [9] reported that chronic disease
status and comorbidities were more important determinants of
health and well-being than were ethnicity, age, language, gender,
and marital status. Quality of life (QOL) is also increasingly
recognized as an important health outcome in its own right,
representing the ultimate goal of all health interventions [10].
The health utilities index Mark 3 (HUI3) classification system
comprises eight attributes: vision, hearing, speech, ambulation,
dexterity, emotion, cognition, and pain—each with five or six
levels of ability/disability. Most of these attributes can be negatively affected by diabetes and its complications.
Pharmaceutical care (PC) is the direct, responsible provision of
medication-related care with the purpose of achieving definite
Chronic medical conditions can impact multiple dimensions of
health-related quality of life (HRQOL) [1]. Given that diabetes is
part of a metabolic syndrome that increases the risk of heart
disease and stroke [2], it is not uncommon for these conditions to
occur as comorbidities in individuals with diabetes. Because
comorbidities are prevalent in diabetes, it is unlikely that the
HRQOL deficits associated with diabetes would be limited to the
condition itself. Indeed, the presence and severity of complications or comorbidities have been associated with depression,
anxiety, and impairment on multiple dimensions of HRQOL in
diabetes [3]. The presence of cardiovascular complications as
comorbidity with diabetes also leads to deficit in HRQOL [4].
The national standardized prevalence rate of diabetes mellitus
in Nigeria is 2.2%, while the crude prevalence rate is 74% in those
aged 45 years and above who live in urban areas [5]. Global
estimates of the prevalence of diabetes for 2010 and 2030 showed
that the prevalence of diabetes in Nigeria in 2010 was 4.7% and
that it would be 5.5% in 2030 when compared with world
population [6]. The complex nature of diabetes management
prompted the Nigerian Ministry of Health to come up with a
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Maxwell O. Adibe, Department of Clinical Pharmacy and Pharmacy Management, University of Nigeria,
Nsukka, Enugu, Nigeria.
E-mails: [email protected]; [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.007
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 240–247
outcomes that improve a patient’s QOL [10]. The principal
elements of PC are that it is medication related; it is care that is
directly provided to the patient; it is provided to produce definite
outcomes; these outcomes are intended to improve the patient’s
QOL; and the provider accepts personal responsibility for the
outcomes [10]. It is also the determination of the drug needs for a
given individual and the provision of not only the required drug
but also the necessary services (before, during, or after treatment)
to ensure the optimally safe and effective drug therapy [11].
Diabetes is a disease that desperately needs more pharmacist
involvement. Pharmacists who are specialized in this growing
chronic condition can make a significant, positive impact on the
patient, the health care system, and themselves [12]. Health care
professionals are becoming increasingly aware of the need to
assess and monitor the QOL as an important outcome of diabetes
care. The QOL is an important outcome in its own right and
because it may influence the patient’s self-care activities, which
may consequently have an impact on the diabetes control [13].
Many PC programs have been established in various countries to
enhance clinical outcomes and the HRQOL. These programs were
implemented by pharmacists, with the cooperation of physicians
and other health care professionals. PC and the expanded role of
pharmacists are associated with many positive diabetes-related
outcomes, including improved clinical measures [14], improved
patient and provider satisfaction [15,16], and improved cost of
management [15,17]. The pharmacist can, therefore, in collaboration with physicians and other health care professionals, contribute to an improvement in the QOL of patients with diabetes by
informing and educating patients, answering their questions, and,
at the same time, monitoring the outcomes of their treatment
[18]. Such interventions, however, are not very common in
Nigeria. Therefore, this research was aimed at evaluating the
impact of the PC intervention on the QOL of patients with type 2
diabetes mellitus in a tertiary hospital setting in Nigeria.
Methods
Research Design
This study was a randomized, controlled, and longitudinal prospective study with a 12-month patient follow-up. The study
protocol was approved by the Research Ethical Committees of the
University of Nigeria Teaching Hospital, ItukuOzalla, and Nnamdi
Azikiwe University Teaching Hospital, Nnewi, in which this study
was conducted. These hospitals are tertiary hospitals that serve
as referral centers to most of the hospitals in Nigeria.
Patients with type 2 diabetes mellitus who fulfilled the
entrance criteria were identified and included in the study. The
inclusion criteria were as follows:
1. patients who were diagnosed with type 2 diabetes mellitus,
2. patients with type 2 diabetes who were receiving oral hypoglycemic and/or insulin therapy,
3. patients who provided written informed consent,
4. patients who expressed willingness to abide by the rules of
the study, and
5. patients who were certified fit for the study by their consulting
doctors.
Exclusion criteria were as follows:
1. patients who were diagnosed with type 1 diabetes (to avoid
complexity in the study scope),
2. patients who were younger than 18 years (they are legally
regarded as dependents and consequently they cannot take
decisions of their own),
241
3. patients who were pregnant (they are generally not allowed to
participate in a study of this nature by the institutions used
for the study), and
4. patients who expressed willingness to withdraw from the
study (participation is voluntary).
These criteria were according to the guiding principles of the
institutional review boards of the hospitals used in this study.
Following sample size determination, a sample size of at least 104
patients in each of the control and intervention groups was
required. Based on these data, to ensure sufficient statistical
power and to account for “dropouts” during the study, a target
sample size of 220 patients was recruited (110 patients from each
of the hospitals). The folders of the 110 selected patients in each
hospital were assigned numbers 1 to 110, which represented an
individual patient, and patients were randomly assigned to one
of two groups (intervention group or control group) on the basis
of the number on their folders by using online “random sequence
generator” [19] with sequence boundaries of 1 to 110 (boundaries
inclusive) set in a two-column format: the first column was priori
designated to the intervention group (55 patients) and the second
column to the control group (55 patients).
Patients in the usual care (UC) group received the usual/
conventional care offered by the hospitals, which included
hospital visits on appointment or on a sick day, consultations
with doctors, prescription of drugs and routine laboratory tests,
review of diagnosis and medications, refilling of prescriptions by
patients, and referral. This UC was offered with no education/
training of the patients on their diseases and drugs and without
empowerment of the patients to be fully involved in the selfmanagement of their illnesses. Patients in the PC group received
UC and PC for 12 months. This additional PC included a stepwise
approach: setting priorities for patient care, assessing patients’
specific educational needs and identification of drug-related
problems, development of a comprehensive and achievable PC
plan in collaboration with the patient and the doctor, implementation of this plan, and monitoring and review of the plan from
time to time [10]. The nurses collaborated with the pharmacist in
terms of organizing the patients and patients’ folders, taking
point-of-care testing, counseling the patients, and reinforcing the
information given to the patients during training sections. The
physicians provided the visitation/appointment schedule for the
patients, and prescription of laboratory tests. They were also
involved in the implementation of consensus strategies in managing drug-related problems in areas of changing, substitution,
and withdrawal of medications.
The educational/training program for the patients consisted
of four sessions of 90 to 120 minutes. The program covered the
following areas: diabetes overview and its complications, selfmonitoring blood glucose techniques and interpretation of
diabetes-related tests, medications and their side effects, lifestyle
modification, counseling, and effective interaction with health
providers. PC provided ground for the patients to monitor and
react to changes in their blood glucose levels, allowing them to
integrate their diabetes into the lifestyle they preferred.
Data Collection
The HUI23S4EN.40Q (developed by HUInc - Mark index 2&3)
questionnaire was used to assess the HRQOL of the patients.
HUI23S4EN.40Q questionnaires were interviewer-administered to
the patients in the intervention group and the control group at
baseline, 6 months, and 12 months.
The HUI3 classification system comprises eight attributes. It
defines 972,000 unique health states. Single-attribute scores of
morbidity are defined on a scale such that the worst level has a
score of 0.00 and the best level has a score of 1.00. Multiattribute
242
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 240–247
Table 1 – Baseline characteristics of the patients in PC and UC arms.
Demographic data
Mean age ⫾ SD (y)
Grouped age: 4 53 y, n (%)
Sex: Male, n (%)
Level of education, n (%)
Primary school
Secondary school
University
Marital status, n (%)
Currently married
Widowed
Single
Occupation, n (%)
Self-employed
Employee
Retired
Smoking status: Smoker, n (%)
Duration, mean ⫾ SD
Duration: ≥5 y, n (%)
Family history of diabetes, n (%)
Physical activity/exercise, n (%)
Comorbidities, n (%)
Hypertension
Congestive heart failure
Ischemic heart disease
Arthritis
≥2 comorbidities
Overnight hospitalization, n (%)
Emergency room, n (%)
Use of insulin, n (%)
Oral antidiabetic medication, n (%)
Other medications, n (%)
Daily aspirin
Diuretics
Antihypertensives
Lipid-lowering
Complications, n (%)
Myocardial infarction
Stroke
Foot ulcer
Blindness
Renal failure
UC (n ¼ 110)
PC (n ¼ 110)
52.8 ⫾ 8.2
81 (73.64)
49 (44.55)
52.4 ⫾ 7.6
75 (68.18)
44 (40)
3 (2.72)
71 (64.55)
36 (32.73)
6 (5.45)
63 (57.27)
41 (37.27)
37 (33.64)
71 (64.54)
2 (1.82)
46 (41.82)
63 (57.27)
1 (0.91)
37 (33.64)
35 (31.82)
38 (34.54)
34 (30.91)
4.5 ⫾ 2.2
62 (56.36)
71 (64.55)
18 (16.36)
34 (30.91)
42 (38.18)
34 (30.91)
21 (19.09)
4.8 ⫾ 2.8
71 (64.55)
62 (56.36)
23 (20.91)
0.043
0.378
0.215
0.214
0.387
60 (54.55)
11 (10.00)
7 (6.36)
37 (33.64)
72 (65.45)
9 (8.18)
1(0.91)
17 (15.45)
103 (93.64)
73
15
8
43
81
7
2
13
107
(66.36)
(13.64)
(7.27)
(39.09)
(73.64)
(6.36)
(1.82)
(11.82)
(97.27)
0.073
0.404
0.789
0.400
0.187
0.604
0.561
0.432
0.1954
P
0.708
0.373
0.495
0.406
0.409
0.611
43
71
98
23
(39.09)
(64.55)
(89.91)
(20.91)
57
84
78
14
(51.82)
(76.36)
(70.91)
(12.73)
0.058
0.055
0.0007
0.105
2
9
2
1
3
(1.82)
(8.18)
(1.82)
(0.91)
(2.73)
4
6
3
1
8
(3.64)
(5.45)
(2.73)
(0.91)
(7.27)
0.408
0.422
0.651
1.000
0.122
PC, pharmaceutical care; UC, usual care.
P ≤ 0.05
utility functions convert comprehensive health state descriptions
(i.e., vectors of one level for each attribute defined by a HUI
classification system) into preference measures of overall HRQOL.
The multiattribute scales of overall HRQOL are defined such that
the score for dead is 0.00 and the score for perfect health is 1.00.
Both HUI2 and HUI3 allow for negative scores of HRQOL that
represent health states considered worse than dead. The lowest
possible HRQOL scores are −0.03 for HUI2 and −0.36 for HUI3 [20].
Also collected were data on patients’ demographics characteristics, lifestyles, and medical conditions, as outlined in Table 1.
Because we used two hospitals, we initially made comparisons
of the groups (UC and PC) across the hospitals to determine their
similarity, or, more specifically, to uncover any problems related to
selection, history, or maturation effects. If the groups were found to
be essentially similar in these respects, we planned to combine the
groups for baseline, 6-month, and 12-month assessments of the
effects of PC. If major differences were identified, we planned to
analyze and report the group findings separately [21].
Data Analysis
Statistical analyses were performed by using the SPSS version 16.
An intention-to-treat approach was used. Two-sample comparisons were made by using Student’s t tests for normally distributed
variables or Mann-Whitney U tests for nonnormally distributed
data. Comparisons of proportions were done by using chi-square,
Fisher’s exact, or McNemar’s tests. The differences in PC and UC
were assessed at baseline, 6 months, and 12 months. An a priori
significance level of P less than 0.05 was used throughout.
Results
The medical and educational content of the training course was
rated positively by the 17 doctors and 29 nurses: the majority 38
(82.6%) rated the content as “excellent” and the remaining 8 rated
the content as “very good” or “good”; only 3 (6.5%) of them
o0.0001
0.0629
0.0014
o0.0001
o0.0001
o0.0001
o0.0001
o0.0001
o0.0001
0.0577
0.1853
0.1526
o0.0001
0.0629
o0.0001
o0.0001
0.005
0.0908
0.1947
0.7510
0.3516
0.2760
0.3333
0.008
0.77
0.86
0.89
0.81
0.81
0.85
0.91
0.71
0.10
0.13
0.21
0.27
0.19
0.12
0.12
0.12
0.81
0.90
0.92
0.82
0.79
0.87
0.89
0.67
Overall
HRQOL
Vision
Hearing
Speech
Ambulation
Dexterity
Emotion
Cognition
Pain
CI, confidence interval; HRQOL, health-related quality of life; PC, pharmaceutical care; UC, usual care.
P ≤ 0.05.
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
0.11
0.21
0.12
0.19
0.12
0.15
0.18
0.10
0.61 ⫾ 0.08
0.63 ⫾ 0.04
0.79
0.91
0.92
0.83
0.84
0.86
0.88
0.70
0.09
0.18
0.09
0.16
0.15
0.20
0.11
0.14
0.94
0.96
0.94
0.88
0.92
0.90
0.96
0.82
0.10
0.19
0.12
0.29
0.09
0.08
0.14
0.15
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
0.83
0.92
0.91
0.84
0.80
0.86
0.85
0.68
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
0.10
0.12
0.08
0.07
0.10
0.15
0.08
0.15
0.97
0.97
0.95
0.94
0.95
0.96
0.98
0.84
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
0.08
0.23
0.09
0.10
0.22
0.10
0.12
0.10
0.14
0.05
0.04
0.10
0.15
0.10
0.13
0.16
(0.1143–0.1657)
(−0.0027 to 0.1027)
(0.0157–0.0643)
(0.0753–0.1247)
(0.1008–0.1992)
(0.0639–0.1361)
(0.1008–0.1592)
(0.1239–0.1961)
o0.0001
o0.0001
0.1029
0.79 ⫾ 0.07
0.65 ⫾ 0.05
0.64 ⫾ 0.10
0.86 ⫾ 0.12
0.22 (0.1884–0.2516)
12 mo
6 mo
PC minus UC (95% CI)
PC (n ¼ 99)
UC (n ¼ 93)
PC (n ¼ 102)
UC (n ¼ 98)
PC (n ¼ 110)
UC (n ¼ 110)
12-mo change
12th month
6th month
Baseline
Outcomes
Table 2 – Changes in overall HRQOL and single attributes between the PC and UC groups after 12-mo follow-up.
Baseline
P
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 240–247
243
suggested little modification or changes. With the exception of
the number of participants taking hypertensive drugs and smoking, we found no other significant differences at baseline in both
UC and PC groups. The number of patients who completed the
study and whose data were analyzed at 6 months and 12 months
in UC and PC groups were 98 (8.09%) versus 102 (92.73%) and 93
(84.55%) versus 99 (90.0%), respectively (Table 1).
The overall HRQOL of the patients was significantly improved at 6
months and 12 months in the PC arm when compared with the UC
arm (0.79 ⫾ 0.07 vs. 0.65 ⫾ 0.05; P o 0.0001 and 0.86 ⫾ 0.12 vs. 0.64 ⫾
0.10; P o 0.0001, respectively). The following single attributes were
significantly improved in the PC arm over the UC arm at 6 months:
vision (0.94 ⫾ 0.10 vs. 0.79 ⫾ 0.09; P o 0.0001), dexterity (0.92 ⫾ 0.09
vs. 0.84 ⫾ 0.15; P o 0.0001), cognition (0.96 ⫾ 0.14 vs. 0.88 ⫾ 0.11; P o
0.0001), and pain (0.82 ⫾ 0.15 vs. 0.70 ⫾ 0.14; P o 0.0001). There was
no significant improvement in hearing, speech, ambulation, and
emotion attributes. At 12 months, there were significant improvements in all the single attributes except hearing (Table 2). Increasing
age (≥52 years, the overall mean age) had high negative impact
(deficit ≥10 %) on overall HRQOL in both PC and UC. A greater
percentage of patients had ages greater than the overall mean age of
about 52.6 years; this is reflected in more than one third of the
patients being retirees. Patients older than 52 years, that is, older
patients, had clinically significant lower QOL than did younger
patients in both UC and PC groups. A majority of the patients had
secondary education, and about a third were self-employed. Family
history of diabetes was reported by a majority of the patients in both
UC and PC groups. Those without a family history of diabetes had
higher QOL in both groups. This result was consistent with what was
expected, but it should be understood that some of the patients did
not know their families’ full medical history in detail. Diabetes
duration (44 years), emergency room visits, and comorbidity of
hypertension and stroke had a high negative impact (deficit ≥10%) on
the overall HRQOL in both PC and UC groups (Table 3). These factors
were also associated with a low single-attribute score recorded in
both arms of the study (Table 4). The changes in the overall HRQOL
score and single-attribute scores of patients in the PC group were
higher than in the UC group, which showed that the PC intervention
had an overriding effect over the patients’ characteristics than did
UC. The percentage changes (improvements) in the PC group were
higher than in the UC group (Table 3). Addition of PC to UC resulted
in a significant gain in quality-adjusted life-years (QALYs) (PC vs. UC)
(0.7625 ⫾ 0.15 vs. 0.6425 ⫾ 0.13; P o 0.0001), with 0.12 (0.07 to 0.1601;
95% confidence interval) QALY gained at the end of 12 months.
Discussion
Impact of the PC Intervention on the HUI3 Overall HRQOL
The overall HRQOL of the patients was significantly improved at 6
months and 12 months in the PC arm when compared with the
UC arm. All the single attributes were significantly improved in
the PC arm when compared with the UC arm at 12 months except
hearing. The PC intervention impacted positively on patients
assigned to it, and this change was both clinically (difference
≥0.03) and statistically significant [22,23]. PC interventions can
have a positive change on the overall HRQOL and single attributes, that is, how patients with diabetes are able to cope with
daily activities [24–26]. The improvement in the HRQOL may in
part be attributed to the increased contact of patients with
diabetes with the clinical pharmacist, but it is also likely to be
associated with improved adherence to lifestyle advice.
The results of this intervention were similar to that of a
prospective study on the impact of PC on QOL in patients with
type 2 diabetes mellitus that was conducted in a private tertiary
care teaching hospital in South India for a period of 8 months,
244
Table 3 – Impact of patients’ characteristics on overall utility score of the patients.
Patients’ characteristics
Overall HRQOL score after 12 mo (with characteristics minus without characteristics)
UC
Deficit
PC
Change† (%)
With condition
Without condition
With condition
Without condition
UC
PC
UC
PC
0.54
0.59
0.58
0.62
0.65
0.62
0.61
0.67
0.64
0.73
0.72
0.72
−0.08
−0.06
−0.04
−0.12
−0.05
−0.08
0.07 (13.0)
0.08 (13.6)
0.06 (10.3)
0.11 (17.7)
0.07 (10.8)
0.10 (16.1)
0.51
0.58
0.56
0.63‡
0.63‡
0.62
0.56
0.65
0.67
0.74‡
0.74‡
0.74
−0.12
−0.05
−0.06
−0.18
−0.09
−0.07
0.05 (9.80)
0.07 (12.1)
0.11 (19.6)
0.11 (17.5)
0.11 (17.5)
0.12 (19.4)
0.58
0.59
0.61
0.60
0.61
0.67
0.62
0.68
0.76
0.73
0.76
0.79
−0.02
−0.02
−0.06
−0.11
−0.08
−0.03
0.04 (6.9)
0.09 (15.3)
0.15 (24.6)
0.13 (21.7)
0.15 (24.6)
0.12 (17.9)
0.62
0.42
0.63
0.58
0.64
0.59
0.66
0.67
0.74
0.57
0.69
0.63
0.78
0.82
0.73
0.71
−0.02
−0.17
−0.03
−0.09
−0.04
−0.25
−0.04
−0.08
0.12
0.15
0.06
0.05
0.14
0.23
0.07
0.04
(19.4)
(35.7)
(9.5)
(8.6)
(21.9)
(39.0)
(10.6)
(6.0)
Note. Mean difference is the actual deficit associated with a particular condition (comorbidity, severity, and resource utilization). Negative values (−) indicate that the characteristic impacted
negatively (lower utility scores) on the patients. ER, emergency room; PC, pharmaceutical care; UC, usual care.
Deficit, utility scores of patients with the characteristic minus utility scores of patients without the characteristics within the UC or PC group.
†
Change, utility scores of patients in the PC group minus utility scores of patients in the UC group. “Deficit” and “change” values ≥0.03 or ≤0.03 are clinically significant.
‡
Noninsulin nonantidiabetes medications (NINM).
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 240–247
Sociodemographic
Age (≥52 y)
Family history of diabetes
Smoking status
Severity
Use of insulin and oral medications
Use of oral medication only
Diabetes duration 44 y
Resource utilization
Overnight hospitalization
Contact with physician in ER
Doctor visit more than 12 times
Comorbidity
Hypertension
Stroke
Eye problems
Number of medical conditions ≥2
Change in overall HRQOL score after
12 mo (PC minus UC)
245
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 240–247
Table 4 – Impact of patients’ characteristics on single-attribute utility score for HUI3.
Single attributes
PC
Mean difference
Sociodemographic
Vision
Hearing
Speech
Ambulation
Dexterity
Emotion
Cognition
Pain
Resource utilization
Smoking status
0.8681
0.0707*
0.9013
−0.0380
0.9962
0.0303
0.8928
0.0867*
0.8189
0.0433
0.9028
−0.0378
0.8157
0.0935*
0.7096
0.1008*
Overnight hospitalization
Family history of diabetes
0.7977
0.0302
0.9270
0.0063
0.9671
0.0110
0.8126
0.0257
0.8198
0.0629*
0.8746
0.0014
0.7215
0.0420
0.6283
0.0077
Contact with physician in ER
Age (≥52 y)
0.9295
0.9670
0.9890
0.9465
0.8494
0.9154
0.8606
0.7389
Physician visit
Vision
Hearing
Speech
Ambulation
Dexterity
Emotion
Cognition
Pain
Severity
0.8302
0.8522
0.9861
0.8909
0.7996
0.8948
0.8035
0.7059
0.8216
0.9378
0.9825
0.8688
0.8199
0.8884
0.7933
0.6799
0.8331
0.0487
0.9471
0.0419
0.9830
0.0249
0.8892
0.1541*
0.8475
0.1948*
0.8648
0.0217
0.8254
0.199*
0.7522
0.2923*
Duration of diabetes 4 4 y
Vision
Hearing
Speech
Ambulation
Dexterity
Emotion
Cognition
Pain
0.8319
0.9413
0.9886
0.9010
0.8531
0.8858
0.8179
0.7542
Comorbidity
Vision
Hearing
Speech
Ambulation
Dexterity
Emotion
Cognition
Pain
0.9047
0.9269
0.9914
0.9171
0.8790
0.9024
0.8400
0.7591
0.0315
0.0423
0.0254
0.1251*
0.0268
0.0407
0.1152*
0.1426*
Use of insulin
0.0377
0.023
0.0328
0.1568*
0.1417*
0.0252
0.1552*
0.2563*
Eye problem
0.1909*
0.0074
0.0388
0.1908*
0.1962*
0.0601*
0.2017*
0.2667*
PC
Mean difference
0.9750
1.0000
1.0000
0.9670
0.9880
0.9280
0.9760
0.8610
0.9190
0.9462
0.9300
0.9190
0.8138
0.9521
0.8352
0.7383
0.0445
0.0437
0.0557*
0.2488*
0.2005*
0.0855*
0.2886*
0.279*
Use of medication
0.1695*
0.0733*
0.0284
0.1479*
0.2128*
0.057*
0.244*
0.2406*
Heart disease
0.1228*
0.0185
0.0316
0.1082*
0.0327
0.0915*
0.1064*
0.1238*
PC
Mean difference
0.1686*
0.0534*
0.0234
0.1752*
0.0928*
0.0607*
0.1699*
0.1554*
≥ 12 visits
0.8653
0.8478
0.9911
0.9119
0.8403
0.8956
0.8424
0.7589
0.1278*
0.0436
0.0448
0.2098*
0.1266*
0.055*
0.2444*
0.3029*
0.8298
0.9399
0.9849
0.8883
0.8492
0.8932
0.7964
0.7158
Stroke
0.1091*
0.0654*
0.0821*
0.4262*
0.4371*
0.1332*
0.3598*
0.5748*
Note. Mean difference represents the single-attribute utility score for patients in the PC group minus patients in the UC group. Positive value
(þ) indicates high single-attribute utility score for the PC group (i.e., HRQOL gained).
ER, emergency room; HRQOL, health-related quality of life; HUI3, health utilities index Mark 3; PC, pharmaceutical care; UC, usual care.
Clinically significant for HUI3 single-attribute utility score (value ≥0.05).
which concluded that the PC program was effective in improving
the clinical outcome and QOL of patients with type 2 diabetes
mellitus [27]. Also, another 1-year observational study that
evaluated the QOL in patients at the Medical University of South
Carolina Family Medicine Clinic, who were followed by a clinical
pharmacist diabetes educator, showed that patients rated their
QOL high after the follow-up [28]. A 1-year study by Correr et al.
[29] concluded that pharmacotherapy follow-up of patients with
type 2 diabetes in community pharmacies can improve the
HRQOL and satisfaction of patients.
Impact of Patients’ Characteristics on Overall HRQOL
The health deficit imposed by old age cut across both groups, but
the patients in the PC group has a significantly higher QOL than
did those in the UC group, which showed that the intervention
had an overriding effect over the effect of age on QOL. Low
education and unemployment are associated with low income,
which consequently affects the QOL of patients. This result is
consistent with a study carried out in a Nigeria teaching hospital
by Issa and Baiyewu [30] who concluded that lower income, low
education level, and low-rated employment affect the QOL of
Nigerian patients with diabetes. The overall HRQOL scores of
nonsmokers in both UC and PC groups were greater than scores of
the smokers. This result was consistent with what was expected
but could be attributed to the fact that the PC group had more
female patients and there is sociocultural stigma associated with
females smoking in Nigeria. The comorbidities considered in the
study were eye problem, heart disease, and stroke. The respondents without eye problem had clinically significant higher QOL
than did those with eye problem, in both UC and PC groups. This
eye problem could be attributed to diabetes retinopathy, which is
a complication of diabetes. This is consistent with a study
performed in Canada [8] that concluded that the illness burden
experienced by individuals with diabetes was associated not only
with diabetes itself but largely with the comorbid medical condition. Lloyd et al. [31] concluded that the presence of even mild
diabetes complications had a significant impact on patients’ QOL.
Early diagnosis and treatment is essential to help prevent the
deterioration in the HRQOL of these patients.
246
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 240–247
A majority of the patients in this study had comorbidities.
Stroke imposed the greatest HRQOL deficit among all other
comorbidities in both PC and UC groups but more in the UC
group. The HRQOL deficit associated with stroke and hypertension
had been reported by Ekwunife et al. [32]. Their study demonstrated that patients with hypertension alone had a higher overall
mean utility score while hypertensive patients with stroke had the
lowest overall mean score. Other similar studies [2,9,33] revealed
that stroke and other comorbidities can impose considerable
health deficits on patients and concluded that the illness burden
experienced by individuals with diabetes is associated not only
with diabetes itself but largely with comorbid medical conditions.
The results of this study also conform to a study that found that
the presence of cardiovascular complications as comorbidity with
diabetes also leads to deficit in HRQOL [3]. This emphasized the
considerable public health impact that all these chronic conditions have on the HRQOL, particularly when they occur together.
This study has shown that diabetes complications have a
profound effect on the HRQOL of patients with type 2 diabetes.
Even the presence of mild diabetes complications has a significant
impact on the HRQOL. In this study, the following four parameters
were used to assess the severity of the disease: use of insulin and
diabetes medications, use of diabetes medication only, duration of
diabetes, and number of absenteeism from work or school in the last
1 year. Those who use a combination of insulin and oral antidiabetes medications and only oral antidiabetes medication had severe
health deficits when compared with patients using noninsulin
nonmedications in both UC and PC groups. This is consistent with
the results of a study by Rubin [34] who assessed the feelings and
complaints made by patients with diabetes using insulin and found
that these could impact negatively on their QOL. This shows that as
the disease progresses to a more severe situation, treatment option
changes from diet to the use of oral medication only and deteriorates to the use of insulin in combination with oral medication.
This study revealed that those with a duration of diabetes of
more than 4 years also had a considerable health deficit when
compared with patients with a duration of diabetes of 4 years or less
in both UC and PC groups. This result is consistent with a report from
the American Diabetes Association [35] that stated that the longer
the duration of diabetes the higher the chances of a patient
developing overt nephropathy, which, in turn, lowers the HRQOL of
the patients. To improve the HRQOL of patients with type 2 diabetes,
early diagnosis of the disease and aggressive management of risk
factors are necessary to prevent or delay the development of diabetes
complications. Resource utilization was assessed as overnight hospitalization, contact with doctor in emergency room, and number of
doctors’ visit in the past 1 year. These resources included physicians,
pharmacists, and nurses’ time, hospital bed space, and spent more
money. There was a considerable health deficit associated with
resource utilization factors in both UC and PC groups. Patients who
had overnight hospitalization and contact with physician or nurse in
emergency room in the past 12 months had more health deficit
when compared with patients who were not in their categories in
both UC and PC groups.
Despite the enormous burdens imposed by the above-outlined
patients’ characteristics on the overall HRQOL, the patients in the
PC group had significantly higher overall HRQOL scores than did
their counterparts in the UC group, implying that the PC intervention had an overriding effect over the debilitating powers of
the patients’ characteristics.
Impact of Patients’ Characteristics on the HUI3 Single
Attributes
Comorbid stroke and old age (≥52 years) were associated with a
relatively large mean difference between PC and UC groups in all
single attributes. This indicates that patients in the PC intervention
who were 52 years or older and had stroke benefited most in all
their single-attributes domains. Patients with diabetes in the UC
group who were with comorbid medical conditions had lower
utility scores on the anticipated diabetes-relevant single attributes
than did their counterparts in the PC group, which resulted in a
large mean difference. Based on the literature, it was anticipated
that diabetes would affect the vision, dexterity, ambulation, emotion, and pain attributes of the HUI3 [4,36,37]. The considerable
burden associated with diabetes and its comorbidities on these
specific single attributes was therefore not surprising. Association
of all sociodemographic, severity, comorbiditiy, and resource utilization factors, except family history, with the “pain” attribute was
clinically and statistically significant. Ischemic heart disease,
angina, and congestive heart failure are all associated with pain
[38,39]. This is consistent with a study conducted by Ekwunife et al.
[32] that demonstrated that the pain attribute had the lowest
health state utility score among Nigerian patients with hypertension alone, hypertension and heart failure, hypertension and
coronary heart disease, and hypertension and stroke. For the pain
attribute in this study, the utility score recorded by the patients
with stroke in the PC group far exceeded the score in the UC group
after 12 months, resulting in a large mean difference. This means
that the PC intervention was able to counter and override the
utility-reducing power of comorbidities within 12 months. It is
possible that some of the pain and discomfort (low score) found in
the UC group was not attributable to stroke, but rather to another
painful comorbid condition associated with diabetes and heart
diseases, such as peripheral vascular disease [40,41].
Functioning on the “ambulation” attribute was similar to that of
pain. All the sociodemographic, severity, comorbidity, and resource
utilization factors were significantly associated with ambulation
clinically. This means that the ambulation attribute of patients in
the PC intervention group was highly improved compared with
that of patients in the UC group. Also, the PC intervention had
overriding benefits over all these factors that impair the QOL. The
burden on the ambulation attribute associated with diabetes,
comorbidities, and other factors was quite visible, not peculiar to
one but all the factors. Clinically important burden on the emotion
attribute was apparent for comorbidities and older patients. The
largest emotional benefits of the PC intervention were observed in
patients with diabetes with stroke compared with the UC group.
These findings are consistent with a report of Maddigan et al. [8]
that demonstrated that diabetes and its comorbidities could impact
negatively on the overall utility state and single attributes.
QALYs associated with PC were significantly higher than those
of UC. This indicates that extending this study beyond 1 year could
offer more benefits to patients with diabetes in terms of QALYs
gained. Some studies had demonstrated that the extension of PC
beyond 1 year could offer extra benefits to patients with diabetes
[42,43]. This study lends support to the use of the HUI3 for the
evaluation of HRQOL as an outcome in PC intervention programs
for patients with diabetes, as suggested by Feeny et al. [20].
Although the patients’ characteristics affected the single
attributes of the patients’ HRQOL adversely, the intervention
not only had overriding effects of these patients’ characteristics
that lower the overall HRQOL but also improved the single
attributes of the QOL of these patients.
Limitations
Our study was subject to the following limitations, and the results
were interpreted in this light: Selection bias was a threat because
participation was voluntary though the groups were randomized.
It remains possible that patients who chose to participate in the
program may have differed in some important way from those
who did not participate, which could pose a threat to external
validity or generalizability. Given the difficulty in blinding in this
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 240–247
type of study, the clinicians involved in providing care and the
interviewers administering the HUI3 questionnaire were not
blinded to patient allocation. Recall bias was possible in this
study because HUI3 was a 4-week recall questionnaire. Attrition
bias or loss during follow-up was also a serious threat but was
avoided by using an intention-to-treat design. The data were selfreported; however, self-reported data about diabetes status have
been established to be both valid and reliable [44].
Conclusions
The addition of PC to UC resulted in improving the QOL of patients
with type 2 diabetes. The overall HRQOL of the patients was
significantly improved at 6 months and 12 months in the PC arm
when compared with the UC arm. Vision, cognition, and pain single
attributes were significantly improved in the PC arm over the UC
arm after 6 months and all single attributes except hearing were
significantly improved after 12 months. Stroke imposed the greatest
HRQOL deficit among all other patients’ characteristics in both PC
and UC arms but more in the UC arm. The results of this study
illustrate a convincing rationale for improving standards of care for
patients with type 2 diabetes through the PC intervention. Further
research, however, is needed to improve on the current PC
intervention strategies such that the recorded improvements in
HRQOL will be sustained for a very long time after an intervention.
Acknowledgment
We acknowledge Health Utility, Inc., for granting and awarding us
HUI23S4En.40Q and HUI23.40Q.MNL used in this study.
Source of financial support: Funding for this project was
provided by Science and Technology Education Post Basic
(STEP-B) through the University of Nigeria. The views expressed
in this article are those of the authors, and no official endorsement by STEP-B is intended or should be inferred.
R EF E R EN CE S
[1] Stewart M, Brown J, Boon H, et al. Evidence on patient-doctor
communication. Cancer Prev Control 1999;3:25–30.
[2] Beckman J, Creager M, Libby P. Diabetes and atherosclerosis: epidemiology,
pathophysiology, and management. JAMA 2002;287:2570–81.
[3] Peyrot M, Rubin R. Levels and risks of depression and anxiety
symptomatology among diabetic adults. Diabetes Care 1997;20:585–90.
[4] de-Visser C, Bibo H, Groenier K, et al. The influence of cardiovascular
disease on quality of life in type 2 diabetics. Qual Life Res
2002;11:249–61.
[5] Nyenwe E, Odia O, Ihekwala A, et al. Type 2 diabetes in adult Nigerians:
a study of its prevalent and risk factors in Port Harcourt, Nigeria.
Diabetes Res Clin Pract 2003;62:177–85.
[6] Shaw J, Sicree R, Zimmet P. Global estimates of the prevalence of
diabetes. Diabetes Res Clin Pract 2010;87:4–14.
[7] Federal Ministry of Health in collaboration with WHO, EC, DFID.
Standard treatment guidelines. Nigeria 2008:90–7.
[8] Maddigan S, Feeny D, Johnson J. Health related quality of life deficit
associated with diabetes and comorbidities in a Canadian National
Population Health Survey. Qual Life Res 2005;14:1311–20.
[9] Westaway M. Effects of ageing, chronic disease and co-morbidities on
the health and well being of older residents of Greater Tshwane. S Afr
Med J 2010;100:1–3.
[10] Hepler C, Strand L. Opportunities and responsibilities in
pharmaceutical care. Am J Hosp Pharm 1990;47:533–43.
[11] Polonsky WH. Understanding and assessing diabetes-specific quality of
life. Diabetes Spectr 2000;13:36–41.
[12] Davis TM, Clifford RM, Davis WA, Batty KT. The role of pharmaceutical
care in diabetes management. Br J Diabetes Vasc Dis 2005;5:352–6.
[13] Khan CR, Weir GC, King GL, Moses AC. Joselin’s Diabetes Mellitus. (14th
ed.). Philadelphia: Lippincott Williams & Wilkins, 2004.
[14] Jaber L, Halapy H, Fenret M, et al. Evaluation of a pharmaceutical care
model on diabetes management. Ann Pharmacother 1996;30:238–43.
247
[15] Sadur C, Moline N, Costa M, et al. Diabetes management in a health
maintenance organization: efficacy of care management using cluster
visits. Diabetes Care 1999;22:2011–7.
[16] Majumdar S, Guirguis L, Toth E, et al. Controlled trial of a multifaceted
intervention for improving quality of care for rural patients with type 2
diabetes. Diabetes Care 2003;26:3061–6.
[17] Coast-Senior E, Kroner B, Kelley C, Trili L. Management of patients with
type 2 diabetes by pharmacists in primary care clinics. Ann
Pharmacother 1998;32:636–41.
[18] Hawkins D, Bradberry JC, Cziraky MJ, et al. National Pharmacy
Cardiovascular Council treatment guidelines for the management of
type 2 diabetes mellitus: toward better patient outcomes and new roles
for pharmacists. Pharmacotherapy 2002;22:436–44.
[19] Mads-Haahr. Random Sequence Generator (1998–2011). Available from:
http://www.random.org/sequences/. [Accessed February 15, 2012].
[20] Feeny D, Furlong W, Torrance G, et al. Multi-attribute and singleattribute utility functions for the health utilities index Mark 3 System.
Med Care 2002;40:113–28.
[21] Cranor C, Christensen D. The Asheville Project: short-term outcomes of
a community pharmacy diabetes care program. J Am Pharm Assoc
2003;43:149–59.
[22] Horsman J, Furlong W, Feeny D, Torrance G. The Health Utilities Index
(HUI): concepts, measurements properties and applications. Health
Qual Life Out 2003;1:54.
[23] Grootendorst P, Feeny D, Furlong W. Health Utilities Index Mark 3
evidence of construct validity for stroke and arthritis in a population
health survey. Med Care 2000;38:290–9.
[24] Clifford R, Batty K, Davis T, et al. A randomised controlled trial of a
pharmaceutical care programme in high-risk diabetic patients in an
outpatient clinic. Int J Pharm Pract 2002;10:85–9.
[25] Clifford R, Batty K, Davis T, Davis W. Effect of a pharmaceutical care
program on vascular risk factors in type 2 diabetes (The Fremantle
Diabetes Study). Diabetes Care 2005;28:771–6.
[26] Posey L. Proving that pharmaceutical care makes a difference in
community pharmacy [editorial]. J Am Pharm Assoc 2003;43:136–9.
[27] Srirama S, Chack LE, Ramasamy R, et al. Impact of pharmaceutical care
on quality of life in patients with type 2 diabetes mellitus. JRMS 2011;16
(Special Issue):412–8: Available from: www.mui.ac.ir. [Accessed
December 16, 2011].
[28] Jennings DL, Ragucci KR, Chumney ECG, Wessell AM. Impact of clinical
pharmacist intervention on diabetes related quality-of-life in an
ambulatory care clinic. Pharm Pract 2007;5:169–73.
[29] Correr CJ, Pontarolo R, Souza RAP, et al. Effect of a pharmaceutical care
program on quality of life and satisfaction with pharmacy services in
patients with type 2 diabetes mellitus. BJPS 2009;45:809–17.
[30] Issa B, Baiyewu O. Quality of life of patients with diabetes mellitus in a
Nigerian teaching hospital. Hong Kong J Psychiatry 2006;16:27–33.
[31] Lloyd A, Sawyer W, Hopkinson P. Impact of long-term complications on
quality of life in patients with type 2 diabetes not using insulin. Value
Health 2001;4:392–400.
[32] Ekwunife OI, Aguwa C, Adibe M, et al. Health state utilities of a population
of Nigerian hypertensive patients. BMC Res Notes 2011;4:528.
[33] Sacco R, Boden-Albala B, Abel G. Race-ethnic disparities in the impact
of stroke risk factors: the northern Manhattan stroke study. Stroke
2001;32:1725–31.
[34] Rubin R. Diabetes and quality of life. Diabetes Spectr 2000;13:21.
[35] ADA. Standards of medical care in diabetes [Position Statement].
Diabetes Spectr 2011;34(Suppl.):S11–61.
[36] Maddigan S, Feeny D, Johnson J. A comparison of the Health Utilities
Indices Mark 2 and Mark 3 in type 2 diabetes. Med Decis Making
2003;23:489–501.
[37] Maddigan S, Feeny D, Johnson J. Construct validity of the RAND-12 and
Health Utilities Index Mark 2 and Mark 3 in type 2 diabetes. Qual Life
Res 2004;13:435–48.
[38] Mayou R, Blackwood R, Bryant B, Garnham J. Cardiac failure: symptoms
and functional status. J Psychosom Res 1991;35:399–407.
[39] Brown N, Melville M, Gray D, et al. Quality of life four years after acute
myocardial infarction: short form 36 scores compared with a normal
population. Heart 1999;81:352–8.
[40] Adler A, Stratton I, Neil H, et al. Association of systolic blood pressure with
macrovascular and microvascular complications of type 2 diabetes (UKPDS
36): prospective observational study. BMJ 2000;321:412–9.
[41] Belch J, Topol E, Agnelli G, et al. Critical issues in peripheral arterial
disease detection and management: a call to action. Arch Intern Med
2003;163:884–92.
[42] Cranor C, Bunting B, Christensen D. The Asheville Project: long-term
clinical and economic outcomes of a community pharmacy diabetes
care program. J Am Pharm Assoc 2003;43:173–84.
[43] Neto P, Marusic S, Júnior DP, et al. Effect of a 36-month pharmaceutical care
program on coronary heart disease risk in elderly diabetic and
hypertensive patients. J Pharm Pharm Sci 2011;14:249–63.
[44] West J, Goldberg K. Diabetes self-care knowledge among outpatients at a
Veterans Affairs medical center. Am J Health-Syst Ph 2002;59:849–52.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 248–253
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
An Audit of Diabetes-Dependent Quality of Life (ADDQOL) in Older
Patients with Diabetes Mellitus Type 2 in Slovenia
Eva Turk, MA, MBA1,, Valentina Prevolnik Rupel, PhD2, Alojz Tapajner, BSc3, Stephen Leyshon, RN, MSc1, Arja Isola, RN, PhD4
1
Det Norske Veritas, Healthcare & Biorisk, DNV Research and Innovation, Høvik, Norway; 2Institute for Economic Research, Ljubljana, Slovenia; 3Faculty of
Medicine, University of Maribor, Maribor, Slovenia; 4Faculty of Medicine, Institute of Health Sciences, University of Oulu, Oulu, Finland
AB STR A CT
Objective: This article reports a study to measure diabetes-dependent
quality of life (QOL) in older Slovenian patients with diabetes mellitus
type 2 (DMT2). Methods: A cross-sectional study of older (age ≥ 65
years) patients with DMT2 at outpatient diabetic centers was conducted in all regions in Slovenia. The Audit of Diabetes-Dependent
Quality of Life questionnaire was carried out between January and
May 2012. Statistical analysis was performed by using IBM SPSS
Statistics software, version 18.0. Results: After exclusion of noneligible respondents, a total of 285 respondents were included in the
analysis, which represented a 57% response rate. Lower QOL was
significantly connected to a heart attack episode (odds ratio 2.42; 95%
confidence interval 1.06–5.20) and to the perception of not having
diabetes under control (odds ratio 0.36; 95% confidence interval 0.18–
0.69). Eleven (3.9%) patients reported no impact of DMT2 on their QOL
at all, while in the remaining respondents, particular reference was
put to the effects on freedom to eat, dependency on others, and family
life. There was no significant difference between the older people
living in urban and rural areas. Conclusions: The findings of the
present study highlight the impact of DMT2 on QOL. DMT2 imposes a
personal burden on individuals. Information on the QOL of older
patients with diabetes is important to Slovenian policymakers and
family physicians to identify and implement appropriate interventions for achieving better management of diabetes and ultimately
improving the QOL of patients with diabetes.
Keywords: ADDQOL, DMT2, elderly, patient-reported outcomes, quality
of life.
Introduction
of Europe, the population in Slovenia is ageing and population
health improvement is an increasingly important component of
coordination and collaboration among patients and health care
providers [8,9].
Internationally, there has been a marked shift in thinking
about what health is and how it is measured [10]. Traditional
clinical ways of measuring health and the effects of treatment are
either accompanied by or even replaced by patient-reported
outcome measures (PROMs), which present an entirely subjective
report of the status of a patient’s health condition. Research has
shown that patients with diabetes are more concerned about
physical and social function, emotional and mental health, as
well as the burden of illness and treatments on daily life than
with clinical biomarkers such as hemoglobin A1c, blood pressure,
or lipid levels [11,12]. PROMs are thus meaningful and relevant
outcomes. Furthermore, there is evidence that when the healthrelated quality of life (HRQOL) of individuals with diabetes is
properly measured and the results are incorporated into health
care management, improvements in patient outcomes occur
[13,14]. Improvements in glycemic control and QOL, as well as
reduction in short-term complications including the incidence of
severe hypoglycemia, can be observed in combination of treatment and education of patients [15–17].
Diabetes is a chronic metabolic disease that can have a profound
impact on the health status and quality of life (QOL) of patients in
terms of physical, social, and psychological well-being [1–3].
Diabetes is now a global health concern: affecting both industrialized and transitioning countries. The number of people with
diabetes is increasing because of population growth, aging,
urbanization, and increasing prevalence of obesity and physical
inactivity. Diabetes mellitus (DM) currently affects about 285
million adults worldwide, and it is projected to rise to 366 million
in 2030 [4,5]. The most important demographic change to diabetes prevalence across the world appears to be the increase in
the proportion of people aged 65 years or older [4]. In Europe
alone, more than 50 million individuals are affected by diabetes,
90% of whom have diabetes mellitus type 2 (DMT2) [6].
Slovenia does not differ significantly from other European
Union countries with regard to the prevalence of diabetes. The
estimates of the National Institute of Public Health [7] amount to
approximately 125,000 patients with diabetes in Slovenia, which
is 6.3% of the population. Of these, 22.2% are aged 75 years or
older, and 16% are aged between 65 and 74 years, with the mean
age of patients with DMT2 being 65 years. Similarly, as in the rest
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflict of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Eva Turk, Healthcare Programme, DNV Research and Innovation, Veritasveien 1, 1322 Høvik, Norway.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.05.001
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 248–253
Many valid instruments to measure PROMs in diabetes have
been developed and are already used in industrialized countries
[18–20]. Among diabetes-dependent QOL measures, the Audit of
Diabetes-Dependent Quality of Life (ADDQOL) is a widely used
instrument [21–24]. In Slovenia, however, despite the high prevalence of diabetes, so far no studies evaluating patient-reported
outcomes, such as HRQOL, have been conducted. According to
the literature, less research was conducted on how various risk
factors influence the QOL of patients with diabetes [25–27]. In this
manner, the objective of the current study was to measure
diabetes-dependent QOL in the older Slovenian patients with
DMT2 and to assess its relationships with sociodemographic and
health factors.
249
A total of 500 patients with DMT2 were invited to participate
in the research. Of them, 391 agreed and after exclusion of
incomplete questionnaires, our sample included 285 patients
with DMT2. The response rate was 57%.
After informed consent was obtained, all prospective participants were given the questionnaire. Where assistance was
needed in completing the questionnaire, this was given by
medical students, who were trained in the use of the ADDQOL
questionnaire prior to the launch of this study.
Ethical Considerations
The study was approved by the National Medical Ethics Committee of the Republic of Slovenia. The data obtained through the
questionnaires were anonymous and based on participant
consent.
Methods
Statistical Analysis
Instrument
The ADDQOL consists of two overview items; one measures
generic overall QOL and a further 19 items are concerned with
the impact of diabetes on specific aspects of life. The 19 life
domains are as follows: leisure activities, working life, local or
long-distance journeys, holidays, physical health, family life,
friendships and social life, close personal relationships, sex life,
physical appearance, self-confidence, motivation to achieve
things, people’s reactions, feelings about the future, financial
situation, living conditions, dependence on others, freedom to
eat, and freedom to drink. These 19 domains ask the respondents
to evaluate how their life would be if they did not have diabetes.
The scales range from −3 to þ1 for 19 life domains (impact
rating) and from 0 to þ3 in attributed importance (importance
rating). A weighted score for each domain is calculated as a
multiplier of impact rating and importance rating (ranging from
−9 to þ3). Lower scores reflect poorer QOL. Finally, a mean
weighted impact score (ADDQOL score) is calculated for the entire
scale across all applicable domains [21,23,28]. Apart from the
perceived QOL, data on patients’ demographic characteristics,
duration of diabetes, and existing diabetic complications were
measured. The linguistic validation and cultural adaptation of
the original English ADDQOL into Slovenian version is described
elsewhere (E. Turk, V. Prevolnik-Rupel, A. Tapajner, et al., unpublished data, 2013).
Study Design and Participants
A cross-sectional study was conducted between January and May
2012 by using a structured questionnaire.
Patients from the 12 participating outpatient diabetic centers
were recruited by using the convenience sampling method. The
regions selected were defined by the Statistical office of the
Republic of Slovenia (E. Turk, V. Prevolnik-Rupel, A. Tapajner,
et al., unpublished data, 2013). For recruitment, we used the
largest outpatient center in each region in consideration that
patients were approximately half from urban and rural areas.
Each outpatient center recruited from 20 to 80 patients according
to region size and diabetes prevalence [29]. All the study patients
had an established relationship with the outpatient centers.
Patients who met our inclusion criteria were asked to participate
in this study. The inclusion criteria were as follows: physiciandiagnosed DMT2, noninsulin treatment, and age 65 years or
older. Patients who were diagnosed as suffering from type 1
diabetes, secondary diabetes, or gestational diabetes were
excluded. All patients were diagnosed by physicians in light of
diagnostic criteria recommended by the World Health Organization in 1999 [31].
The sample data were expressed as frequencies and percentages
for categorical variables or by mean values and SDs for continuous variables. Binary logistic regression analysis was used to
assess the influence of sociodemographic and health characteristics of patients with DMT2 on their QOL by using the ADDQOL.
The calculation included Wald chi-square, odds ratio (OR), 95%
confidence interval (95% CI), and P value. Nagelkerke’s R2 was
used to indicate goodness of fit. Patients were divided into two
groups by using the ADDQOL score by using quartiles; the first
group in the lower quartile was considered as having lower QOL.
Such a cutoff strategy was previously applied in the literature
[26,31]. Statistical analysis was performed with the SPSS 18.0
software (SPSS, Inc., Chicago, IL). A P value of less than 0.05 was
considered statistically significant.
Results
Sociodemographic characteristics of the studied population are
presented in Table 1. The age ranged from 65 to 84 years, with a
mean of 70.0 ⫾ 4.9 years. Among the 285 respondents, less than
half were female (135, 47.4%). The majority of the respondents
were married (191; 67.0%), owned their own house (171; 60.0%),
and lived in an urban area (243; 85.3%).
The body mass index (kg/m2) ranged from 16.9 to 53.0, with a
mean value of 29.6 ⫾ 5.0. A majority of the respondents have
been living with DMT2 for 11 years or more (56.5%), and many
had problems with hypertension (78.9%) and high cholesterol
(59.6%). More details about respondents’ health characteristics
are shown in Table 2. A vast majority of the respondents (230,
80.7%) reported to be satisfied with professional health support
provision, and 114 (40.0%) were of the opinion that their diabetes
was under control.
The ADDQOL score of 285 patients with DMT2 was calculated
in a range of −8.3 to 0.0 on a defined range from −9 to þ3. The
median ADDQOL score was calculated at −1.6, lower quartile
cutoff was calculated at −3.0, 213 (74.7%) patients with DMT2
reported an ADDQOL score of −3.0 or more, and 72 (25.3%)
patients had an ADDQOL score of less than −3.0 (lower QOL).
Eleven patients (3.9%) reported an ADDQOL score of 0, which
means that their QOL was not affected by diabetes at all.
Table 3 shows the logistic regression model results of the
predictors of QOL according to the ADDQOL score. Lower QOL was
significantly connected to a heart attack episode (OR 2.42; 95% CI
1.06–5.20). From a patient perspective, being of the opinion that
their diabetes was under control decreased the likelihood of a
lower QOL (OR 0.36; 95% CI 0.18–0.69). Living in a rural environment was not significantly connected to a lower QOL. Results in
Figure 1 show that only 13.6% of the patients without heart attack
250
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 248–253
Table 1 – Sample data.
N ¼ 285
Gender
Male
Female
Education
Primary education
Secondary education
College or higher
Marital status
Married, in partnership
Widowed
Divorced
Alone
Residence
Own house
Own apartment
Renting
Relatives
Nursing home
Monthly income in euro
≤365
366–730
731–1100
≥1101 or above
Region
≤200 per km2 (rural)
4200 per km2 (urban)
Age (y), mean ⫾ SD, range
Body mass index, mean ⫾ SD, range
%
150
135
52.6
47.4
87
163
35
30.5
57.2
12.3
191
71
11
12
67.0
24.9
3.9
4.2
171
92
11
8
3
60.0
32.3
3.9
2.8
1.1
36
162
60
27
12.6
56.8
21.1
9.5
42
243
70.0 ⫾ 4.9
29.6 ⫾ 5.0
was 54.0%) domains (Table 4). NA responses are important when
the reliability and construct validity of the ADDQOL instrument
are considered. In our study, because of the high NA response in
working life and sex life, reliability and instrument validity
calculation was possible only on 17 domains. Even then, the
sample size for reliability and construct validity was n ¼ 180
because holidays, family life, and personal relationship domains
in combination excluded n ¼ 105 respondents. The reliability of
the ADDQOL instrument according to Cronbach’s alpha was 0.91
(weighted impact score). For validation purposes, we used factor
analysis with forced one-factor solution. This condition was
imposed because the ADDQOL was intended to provide a single
summary score [21]. In the forced one-factor solution, all
domains with the exception of “freedom to drink” had factor
loadings of more than 0.4. Freedom to drink loaded with a value
of 0.285 into this factor. The forced one-factor solution explained
48.8% of the total variance.
Discussion
14.7
85.3
65–84
16.9–53.0
and being of the opinion that their diabetes was under control
reported a lower QOL.
The distribution of responses and the weights assigned to the
impact ratings are shown in Table 4. Diabetes had the greatest
impact on “freedom to eat” (mean impact rating: −1.5 ⫾ 1.0) and
the least impact on “people’s reaction” (mean −0.4 ⫾ 1.0).
“Dependence on others” was rated as the most important (mean
2.5 ⫾ 0.7), and “freedom to drink” was rated as the least
important to them (mean 1.2 ⫾ 1.0). After considering weighting,
“freedom to eat” remained as the most (mean −3.2 ⫾ 2.9) and
“people’s reaction” as the least (−0.8 ⫾ 1.5) affected QOL domains,
respectively.
The ADDQOL instrument includes five domains that respondents can choose not to score. If no answer is provided, the
ADDQOL score is calculated without these domains. Respondents
in this study showed less interest in working life (the “not
available” [NA] response was 76.5%) and sex life (the NA response
Table 2 – Health characteristics.
N ¼ 285
Duration of diabetes mellitus (y)
≤4
5–10
≥11
Hypertension
High cholesterol
Poor eye vision
Kidney dialysis
Foot amputation
Brain stroke
Heart attack
51
73
161
225
170
26
0
8
23
39
%
17.9
25.6
56.5
78.9
59.6
9.1
0.0
2.8
8.1
13.7
This study provides detailed information about diabetesdependent QOL and its assessment among older patients with
DMT2 in Slovenia by using the ADDQOL, which is a widely used
diabetes-specific scale in the literature [21–25,32,33]. To the best
of our knowledge, the current study is the first to measure the
HRQOL of older patients with DMT2 in Slovenia. Weighted
ADDQOL domain scores reliability was similar to that in previous
studies [26,27,32]. Structure validity results supported the onefactor scale structure of the ADDQOL, and the “freedom to drink”
domain was calculated as the only possible domain that may not
contribute to the ADDQOL instrument. The forced one-factor
solution explained 48.8% of the total variance, which was also
similar to that in previous studies [27,34]. The findings of the
present study highlight the impact of DMT2 on QOL. An interesting finding in the current study was that a few patients report no
impact of DMT2 on their QOL at all. In the rest, however,
particular emphasis was put on the impact of “freedom to eat,”
“dependency on others,” and “family life.” Consistent with earlier
studies [22–24,28,32], we found that the greatest negative impact
observed was for the domain “freedom to eat,” indicating the
strong influence of dietary restrictions on the QOL. Similarly, the
least affected domain was “people’s reaction.” Relative to the
overall negative effects of diabetes on the QOL, the effect of
specific sociodemographic and clinical factors was fairly modest
[25,31].
The results in the present study show that lower QOL was
significantly connected to the presence of additional health
problems (i.e., heart attack). Other studies show that the influence of comorbidities or health complications in diabetesdependent QOL seems unclear. Collins et al. [25] measured the
amount of diabetes complications and concluded that the
increased number did not result in lower QOL. Similarly, in
Chung et al. [26], increased microvascular complications showed
no association. Conversely, in Wang and Yeh [27], complications
resulted in lower QOL, yet comorbidities provided no association.
In concordance with Collins et al. [25], the study presented here
provided no association of demographic data on diabetesdependent QOL. Wang and Yeh [27], however, concluded that
more education has negatively affected the QOL, and Chung et al.
[26] found a positive association between older age and higher
QOL. In the study presented here, we also measured the influence
of diabetes duration on the QOL and found no association, which
was in concordance to the results of Wang and Yeh [27].
Respondents who reported that they managed their disease
well and have it under control showed a decreased likelihood of
lower QOL, which suggests the importance of self-management
251
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 248–253
Table 3 – Predictors for lower QOL according to the ADDQOL score (prediction model: χ 2 : 48.697, df ¼ 22, P o
0.001).
ADDQOL score
Living in rural areas (≤200 /km2)
Female
Age (y)
65–74 (old)
75–84 (old-old)
Body mass index
o25
≥25–o30
≥30
Education
Primary education
Secondary education
College or higher
Monthly income (euro)
≤365
366–730
731–1100
≤1101
Residence in own house
Single/divorced/widowed
Years of diabetes
≤4
5–10
≥11
Hypertension
High cholesterol
Poor eye vision
Foot amputation
Brain stroke
Heart attack
General health care satisfaction
Being of opinion to manage the disease
Higher
≥−3.0
n ¼ 213 (in %)
12.7
49.8
Lower
o−3.0
n ¼ 72 (in %)
20.8
40.3
Wald χ2
OR (95% CI)
P
2.38
1.54
1.96 (0.83–4.63)
0.66 (0.34–1.28)
0.123
0.215
72.3
27.7
75.0
25.0
0.32
1.00 (ref)
0.81 (0.38–1.71)
0.573
13.1
50.7
36.2
13.9
43.1
43.1
1.99
1.36
1.00 (ref)
0.48 (0.14–1.12)
0.56 (0.21–1.48)
0.156
0.244
29.6
57.7
12.7
33.3
55.6
11.1
0.01
0.73
1.00 (ref)
0.96 (0.46–2.01)
0.55 (0.14–2.16)
0.922
0.394
11.3
56.8
22.5
9.4
61.5
32.9
16.7
56.9
16.7
9.7
55.6
33.3
0.28
1.09
0.22
1.73
0.02
1.00
0.78
0.55
1.46
0.65
1.05
(ref)
(0.31–1.97)
(0.17–1.70)
(0.30–7.13)
(0.34–1.24)
(0.50–2.21)
0.598
0.296
0.639
0.189
0.888
17.8
27.7
54.5
77.9
56.8
7.0
1.4
9.4
11.7
82.2
45.1
18.1
19.4
62.5
81.9
68.1
15.3
6.9
4.2
19.4
76.4
25.0
2.10
0.48
0.02
2.24
3.61
1.34
3.23
4.53
0.10
9.25
1.00
0.50
0.75
0.95
1.64
2.53
2.62
0.28
2.42
0.89
0.36
(ref)
(0.20–1.28)
(0.33–1.69)
(0.44–2.02)
(0.86–3.15)
(0.97–6.61)
(0.51–13.38)
(0.07–1.12)
(1.06– 5.20)
(0.41–1.91)
(0.18–0.69)
0.147
0.487
0.888
0.134
0.057
0.247
0.072
0.036
0.757
0.002
Note. Nagelkerke R2 ¼ 0.215.
ADDQOL, Audit of Diabetes-Dependent Quality of Life; CI, confidence interval; OR, odds ratio; QOL, quality of life.
of the disease [16,36–37]. In the present study, despite the fact
that more than 50% of the patients have been living with DMT2
for 11 or more years, only 40% reported that they have their
disease under control. This finding suggests that work is needed
to increase patient empowerment and DMT2 self-management in
Slovenia.
Open access to primary health care is ensured for all health
care–insured individuals in Slovenia, in terms of both economic
and geographical accessibility. Despite this, differences between
regions in Slovenia have been reported, mainly due to an
inadequate distribution in the number of primary health care
personnel in some of the more remote rural areas, in which there
can be a lack of doctors [38]. Therefore, we hypothesized that
older patients with DMT2 living in rural areas would show a
lower QOL than would patients living in urban areas. The present
study, however, showed no connection with place of living and
QOL. This finding, together with the finding of the parallel
research on diabetes knowledge [39], implies that accessibility
to chronic disease care provision and information does not
depend on the place of living in Slovenia. This is in line with
the European Health Interview Survey study [29,40], which
implies that the use of primary care is relatively evenly spread
across socioeconomic classes in Slovenia. In addition to the QOL,
our results suggest that the older patients with DMT2 are
satisfied with the delivery of care.
Although the research presented here shows no difference in
QOL between rural and urban areas, this is not the same as
saying that both groups receive high-quality care or that it cannot
be improved. Health services should aspire to improve the QOL of
older patients with DMT2. This is going to become an increasingly important issue as the prevalence and economic burden of
diabetes among the Slovenian older population rises. A lack of
critical assessment at a system level may hamper the attempt to
improve the QOL. Hence, a well-translated and culturally adapted
disease-specific HRQOL measure such as the ADDQOL could
contribute to the more accurate assessment of the effectiveness
of disease management programs.
Study Strengths and Limitations
The study gives an overview of the self-perceived QOL of older
populations with DMT2. The main strength of the ADDQOL is that
it measures the QOL in various areas of people’s lives. This,
however, can have a consequence that not all areas are applicable to all respondents. As a result, some respondents did not
provide complete data for all ADDQOL domains, which may have
252
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 248–253
patients’ willingness to participate in the study, a response bias
might have occurred. A larger sample would provide more power
to detect significant relationships between study variables and
differences between groups.
In the literature, we were not able to find many studies that
researched the influence of various characteristics on diabetesdependent QOL. Generally, all studies included demographic data
and health problems together with some specific characteristics
of their research interest; in our case, this was urban or rural
living area. Among demographic data, age and gender were used
always; other data such as education, marital status, or employment (income) were used optional. Health problems were presented as sum variables called complications and once separately
as complications and comorbidities. In our study, we decided to
show the influence of some complications and comorbidities by
displaying the exact type of the disease (e.g., hypertension, heart
attack, and brain stroke). This can provide an added insight that
the existing literature is missing and lead to improved diabetesdependent QOL knowledge.
Fig. 1 – Percentage of patients in the lower ADDQoL quartile
(o3.0) according to heart attack (HA) status and their
perception of disease control. ADDQOL, Audit of DiabetesDependent Quality of Life; QOL, quality of life.
introduced unintended biases into the analyses. The reason for
the low response rate of the item “working life” might be a
reflection of the fact that all the responders were aged 65 years or
older and mainly retired.
The lack of a randomized sampling and use of a convenience
sampling limit the ability to generalize the results. Because of
Conclusions
The results of the current study are similar to findings in prior
research conducted in other countries. This study also demonstrates
that many of the factors related to diabetes-dependent HRQOL are
applicable regardless of the country and health care system.
DMT2 is of growing public health concern in Slovenia. It
imposes a personal burden on individuals and consumes a
significant portion of society’s scarce health care resources.
Information on the QOL of older patients with diabetes is therefore important to Slovenian policymakers and health workers. It
is essential in helping to identify and implement appropriate
Table 4 – Distribution of response (N ¼ 285) by impact and importance rating together with weighted
impact score.
Domain
Mean ⫾ SD
Not available response (%)
Impact rating
Leisure activities
Working life
Journeys
Holidays
Physical health
Family life
Friendship and social life
Personal relationship
Sex life
Physical appearance
Self-confidence
Motivation
People’s reaction
Feelings about future
Financial situation
Living conditions
Dependence on others
Freedom to eat
Freedom to drink
218 (76.5)
98 (34.4)
8 (2.8)
74 (26.0)
154 (54.0)
−1.1
−1.4
−1.3
−1.1
−1.2
−0.9
−0.9
−0.8
−1.1
−0.7
−0.8
−0.9
−0.4
−1.1
−0.6
−0.9
−0.6
−1.5
−0.9
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
1.0
1.0
1.0
1.0
1.0
0.9
1.0
1.0
1.0
0.9
1.0
1.0
0.7
1.0
0.9
0.9
0.9
1.0
1.0
Importance rating
1.9
2.1
1.8
1.8
1.9
2.4
2.0
2.3
1.8
1.5
2.0
1.9
1.5
1.9
2.0
2.1
2.5
1.8
1.2
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
0.7
0.7
0.8
0.8
0.7
0.6
0.8
0.7
0.9
0.9
0.7
0.7
0.9
0.7
0.7
0.7
0.7
0.9
1.0
Weighted impact score
−2.2
−3.0
−2.5
−2.0
−2.5
−2.3
−1.9
−1.9
−2.3
−1.4
−1.8
−2.0
−0.8
−2.5
−1.3
−2.0
−1.5
−3.2
−1.5
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
2.2
2.8
2.5
2.3
2.4
2.5
2.5
2.4
2.7
2.1
2.4
2.6
1.5
2.5
2.2
2.5
2.3
2.9
2.3
Notes. Impact rating (conditions without diabetes): −3, very much better; −2, much better; −1, a little better; 0, the same; þ1, worse.
Importance rating: 0, not at all important; 1, somewhat important; 2, important; 3, very important.
Weighted impact score ¼ impact rating (−3 to þ1) importance rating (0–3) ¼ −9 (maximum negative impact of diabetes) to þ3 (maximum
positive impact of diabetes).
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 248–253
interventions for achieving the better management of diabetes
and ultimately improving the QOL of patients with diabetes.
Acknowledgments
The authors thank Jelka Zaletel, MD, PhD, for her suggestions and
advice. We gratefully acknowledge the outpatient diabetic centers and the diabetic patients in the regions for their willingness
to participate in the study. E.T. and A.I. were responsible for the
study conception and design. E.T. performed the data collection.
E.T. and A.T. performed the data analysis and were responsible
for the drafting of the manuscript. A.I., S.L., and V.P.R. made
critical revisions to the manuscript for important intellectual
content. A.T. provided statistical expertise. A.I. and V.P.R. provided administrative, technical, or material support. S.L. edited
the manuscript. A.I. supervised the study.
Source of funding support: The project did not receive any
external funding.
R EF E R EN CE S
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[1] Cox WM, Blount JP, Crowe PA, Singh SP. Diabetic patients’ alcohol use
and quality of life: relationships with prescribed treatment compliance
among older males. Alcohol Clin Exp Res 1996;20:327–31.
[2] Boye KS, Yurgin N, Dilla T, et al. Health-related quality of life of patients
with type 2 diabetes mellitus in primary care in Spain: self-reported
and proxy assessment using the EQ-5D. J Med Econ 2007;10:41–58.
[3] Glasgow RE, Ruggiero L, Eakin EG, et al. Quality of life and associated
characteristics in a large national sample of adults with diabetes.
Diabetes Care 1997;20:562–7.
[4] Wild SH, Roglic G, Green A, et al. Global prevalence of diabetes:
estimates for the year 2000 and projections for 2030 response to
Rathman and Giani. Diabetes Care 2004;27:1047–53.
[5] Shaw J, Sicree R, Zimmet P. Global estimates of the prevalence of
diabetes for 2010 and 2030. Diabetes Res Clin Pract 2010;87:4–14.
[6] IDF. Diabetes atlas. Available from: http://archive.diabetesatlas.org/
content/europe. [Accessed July 17, 2012].
[7] Moravec Berger D, Zupanič T. Epidemiologija in prevalenca sladkorne
bolezni v Sloveniji. In: Lovšin D, ed., 1 nacionalna konferenca o
diabetesu 6–7. november 2008; Ljubljana: Zavod za izobraževanje o
diabetesu, 2008.
[8] Yong PL, Olsen LA, McGinnis JM. Value in Health Care: Accounting for
Cost, Quality, Safety, Outcomes, and Innovation: Workshop Summary.
Washington, D.C. National Academies Press, 2010.
[9] Ali MK, Weber MB, Narayan K. The global burden of diabetes. In: Holt
RIG, Cockram CS, Flyvbjerg A, et al., eds., Textbook of Diabetes. (4th ed.).
Oxford, UK: Wiley-Blackwell, 2010:69–84.
[10] Devlin NJ, Appleby J. Getting the Most Out of PROMs. London: Kings
Fund, 2010.
[11] Krumholz HM. Outcomes research: generating evidence for best
practice and policies. Circulation 2008;118:309–18.
[12] Barr J. The outcomes movement and health status measures. J Allied
Health 1995;24:13.
[13] Magwood GS, Zapka J, Jenkins C. A review of systematic reviews
evaluating diabetes interventions. Diabetes Educ 2008;34:242–65.
[14] Tapp RJ, O'Neil A, Shaw JE, et al. Is there a link between components of
health-related functioning and incident impaired glucose metabolism
and type 2 diabetes? Diabetes Care 2010;33:757–62.
[15] Aghamolaei T, Eftekhar H, Mohammad K, et al. Effects of a health
education program on behavior, hbA1c and health-related quality of
life in diabetic patients. Acta Medica Iranica 2005;43:89–94.
[16] Khanna A, Bush AL, Swint JM, et al. Hemoglobin A1c improvements
and better diabetes-specific quality of life among participants
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
253
completing diabetes self-management programs: a nested cohort
study. Health Qual Life Outcomes 2012;10:48.
Norris SL, Nichols PJ, Caspersen CJ, et al. Increasing diabetes selfmanagement education in community settings: a systematic review.
Am J Prev Med 2002;22:39–66.
Garratt A, Schmidt L, Fitzpatrick R. Patient‐assessed health outcome
measures for diabetes: a structured review. Diabetic Med 2002;19:1–11.
El Achhab Y, Nejjari C, Chikri M, Lyoussi B. Disease-specific healthrelated quality of life instruments among adults diabetic: a systematic
review. Diabetes Res Clin Pract 2008;80:171–84.
Watkins K, Connell CM. Measurement of health-related QOL in
diabetes mellitus. Pharmacoeconomics 2004;22:1109–26.
Bradley C, Todd C, Gorton T, et al. The development of an
individualized questionnaire measure of perceived impact of diabetes
on quality of life: the ADDQoL. Quali Life Res 1999;8:79–91.
Speight J, Woodcock A, Plowright R, Bradley C. Design of an
individualised measure of the impact of diabetes on the quality of life
of elderly people: the ADDQoL Senior. Qual Life Res 2003;12:1509.
Costa FA, Guerreiro JP, Duggan C. An audit of diabetes dependent
quality of life (ADDQoL) for Portugal: exploring validity and reliability.
Pharm Pract 2006;4:123–8.
Holmanová E, Žiaková K. Audit Diabetes‐Dependent Quality of Life
questionnaire: usefulness in diabetes self‐management education in
the Slovak population. J Clin Nurs 2009;18:1276–86.
Collins MM, O'Sullivan T, Harkins V, Perry IJ. Quality of life and quality
of care in patients with diabetes experiencing different models of care.
Diabetes Care 2009;32:603–5.
Chung J, Cho D, Chung D, Chung M. Assessment of factors associated
with the quality of life in Korean type 2 diabetic patients. Intern Med
(Tokyo, Japan) 2012;52:179–85.
Wang H-F, Yeh MC. The quality of life of adults with type 2 diabetes in
a hospital care clinic in Taiwan. Qual Life Res 2012;22:577–84.
Bradley C, Speight J. Patient perceptions of diabetes and diabetes
therapy: assessing quality of life. Diabetes/Metab Res Rev 2002;18
(Suppl. 3):S64–9.
SURS. Slovenske regije v številkah. Statistični urad Republike Slovenije:
Statistični urad Republike Slovenije, 2011.
Buzeti T, Gobec M. Health inequalities in Slovenia. Ljubljana, Slovenia:
National Institute of Public Health. 2012.
World Health Organization. Definition, diagnosis and classification of
diabetes mellitus and its complications. Geneva, Switzerland: WHO,
Department of Noncommunicable Disease Surveillance, 1999.
Trief PM, Wade MJ, Pine D, Weinstock RS. A comparison of healthrelated quality of life of elderly and younger insulin-treated adults with
diabetes. Age Ageing 2003;32:613–8.
Kong D, Ding Y, Zuo X, et al. Adaptation of the Audit of DiabetesDependent Quality of Life questionnaire to patients with diabetes in
China. Diabetes Res Clin Pract 2011;94:45–52.
Akinci F, Yildirim A, Gözü H, et al. Assessment of health-related quality
of life (HRQoL) of patients with type 2 diabetes in Turkey. Diabetes Res
Clin Pract 2008;79:117–23.
Wee HL, Tan CE, Goh SY, Li SC. Usefulness of the Audit of DiabetesDependent Quality-of-Life (ADDQoL) questionnaire in patients with
diabetes in a multi-ethnic Asian country. Pharmacoeconomics
2006;24:673–82.
Peyrot M, Rubin RR, Funnell MM, Siminerio LM. Access to diabetes selfmanagement education. Diabetes Educ 2009;35:246–63.
Chodosh J, Morton SC, Mojica W, et al. Meta-analysis: chronic disease
self-management programs for older adults. Ann Intern Med
2005;143:427–38.
Norris SL, Lau J, Smith SJ, et al. Self-management education for adults
with type 2 diabetes. Diabetes Care 2002;25:1159–71.
Albreht T, Turk E, Toth M, et al. Slovenia: health system review. In:
Health Systems in Transition (Vol. 11, No. 3). Copenhagen, Denmark:
World Health Organization, 2009.
Turk E, Palfy M, Prevolnik Rupel V, Isola A. General knowledge about
diabetes in the elderly diabetic population in Slovenia. Slovenian Med J
2012;81:517–25.
National Institute of Public Health. European Health Interview Survey
(EHIS). Ljubljana, Slovenia: National Institute of Public Health, 2007.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 254–258
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Patient-Reported Quality of Life During Antiretroviral Therapy in a
Nigerian Hospital
Azuka C. Oparah, PhD, Jeffrey S. Soni, PharmD, Herbert I. Arinze, PharmD, Ifeanyi E. Chiazor, PharmD
Department of Clinical Pharmacy and Pharmacy Practice, University of Benin, 300001, Benin City, Nigeria
AB STR A CT
Objectives: We assessed the reported quality of life of patients with
HIV/AIDS and explored the impact of patients’ sociodemographic
profile on the quality-of-life domains. Methods: Consenting outpatients who met criteria were consecutively selected in a secondary
health care facility in Benin City, Nigeria. Quality of life was determined in the nine domains of HIV/AIDS Targeted Quality of Life (HATQOL) instrument. Quality-of-life scores were computed on the scale of
0 to 100 and triangulated with a rated interval scale of 1 to 5 suited for
quantitative analysis. Association between rated scores and each
domain was explored by using Students’ t test and analysis of
variance at 95% confidence interval. Results: Out of the 403 patients,
82.1% were females; 147 (36.1%) belonged to the modal age group of 20
to 30 years; the mean age for grouped data was 39.2 years. About 239
(58.7%) were not married. Also, 338 (83.0%) earned below $1500 per
annum and 303 (74.40%) had basic education (grade 1 to grade 9). HATQOL scores indicated the following: overall function (89.96 ⫾ 5.62); life
satisfaction (91.94 ⫾ 3.62); health worries (87.06 ⫾ 4.28); financial
worries (81.00 ⫾ 3.95); medication worries (91.65 ⫾ 4.47); HIV mastery
(71.00 ⫾ 3.11); disclosure worries (27.50 ⫾ 7.57); provider trust (91.63 ⫾
1.96); and sexual function (70.25 ⫾ 3.51). Likert-type rated scores were
in agreement with HAT-QOL scores. Provider trust was associated
with gender, employment status, and educational level. Sexual
function was associated with gender and age (P o 0.05). Conclusions:
Patients reported satisfactory quality of life in the domains of overall
function, life satisfaction, health worries, financial worries, medication worries, HIV mastery, provider trust, and sexual function. Quality
of life was low in the domain of disclosure worries, indicating
concerns for discrimination and stigmatization. Age, level of education, and employment status had a strong impact on the quality of life
of patients with HIV/AIDs.
Introduction
well-being and includes aspects such as happiness and satisfaction with life as a whole. QOL relates both to the adequacy of
material circumstances and to personal feelings about these
circumstances with overall subjective feelings of well-being that
is closely related to morale, happiness, and satisfaction [3–5]. QOL
has recently been scientifically defined, and it has been considered synonymous with health status, functional status, psychological well-being, happiness with life, satisfaction of needs, and
assessment of one’s own life [6].
Several instruments for measuring QOL have been developed
and described [7]. A study on the QOL of PLWHA in São Paulo,
Brazil, reported that despite differences in sex, skin color,
income, and mental and immunological status, PLWHA have
better (physical and psychological) QOL than do other patients,
but lower quality in the social relationships domain [8]. A similar
study in South India also showed that patients had the worst QOL
in the social domain, indicating that the patients’ social contacts
and sexual activity were affected markedly to a great extent [9].
Fatiregun et al. [2], in their own study of PLWHA in Kogi State,
Nigeria, suggested that stigma and discrimination as well as
poor living conditions in the physical environment accounted
The HIV epidemic has resulted in history’s single sharpest
reversal in human development. Sub-Saharan Africa remains
the region most heavily affected by HIV, accounting for 67% of
all people living with HIV and for 75% of AIDS deaths in 2007. In
terms of disease burden, Nigeria rank second in sub-Saharan
Africa, behind South Africa, and third in the world, behind India
[1]. HIV/AIDS continues to contribute significantly to public
health problems in Nigeria. Although HIV infection was initially
limited to people with risky behaviors, such as commercial sex
workers, adolescents, youths, prisoners, and people with multiple sexual partners, the currently available evidence suggests
that this infection has permeated all strata of the Nigerian
population [2].
With the alarming increase in HIV/AIDS pandemic in developing countries, and the limited accessibility and availability of
highly active antiretroviral therapy (ART), majority of the people
living with HIV/AIDS (PLWHA) continue to suffer the disease,
with a serious impact on their quality of life (QOL). Quality of life
is a term that is popularly used to convey an overall sense of
Keywords: HAT-QOL, HIV/AIDS, Nigeria, quality of life.
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Azuka C. Oparah, Department of Clinical Pharmacy and Pharmacy Practice, University of Benin, 300001
Benin City, Nigeria.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.07.004
255
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 254–258
for lower QOL in the environment and social relationships
domain [2].
An investigation of the determinants of QOL of PLWHA identified gender, mental status, CD4 cell count, and stage of the disease
to be important factors associated with the QOL of patients [6]. In a
study of the QOL of HIV-infected Brazilians, the authors reported
that unemployment was associated with worse QOL in all the
domains measured, except the domain of spirituality [10]. Other
studies concluded that education, income, occupation, family support, and clinical categories were significantly linked to patients’
QOL [11,12]. In a sample of highly active ART-naive asymptomatic
HIV-infected subjects, high viral loads and low CD4 count were
significantly associated with poorer scores in the psychological and
social domains [13]. Self-reporting indicates that QOL is severely
compromised in PLWA in stages 3 and 4 and limitations in the four
domains of mobility, usual activities, pain/discomfort, and anxiety/
depression constitute major problems for PLWA [14]. Focus group
discussions have also revealed that QOL weights are strongly
correlated to disease stage. Furthermore, clinical experts consistently report that ART has a strong positive impact on the HIV/AIDS
Targeted Quality of Life (HAT-QOL) of patients, although this effect
appears to rebound in cases of drug resistance [15]. Cotton et al. [16]
tested six new items addressing personal and social domains of
religiosity and spirituality. The authors reported that many patients
had become more spiritual or religious by virtue of having HIV/
AIDS, and half of the patients believed that their spirituality/religion
was helping them to live longer.
In the setting of this study, patients with HIV/AIDS receive
free medications on a donor-funded project that has been in
existence for over 5 years. No investigation has been carried out
to determine the QOL of patients, which is an important treatment outcome. We assessed the reported QOL of patients with
HIV/AIDS and explored the impact of the patients’ sociodemographic profile on the QOL domains.
Methods
Design and Setting
This study used a cross-sectional descriptive design and was
carried out at Central Hospital Benin City, Edo State. The facility
has 432 bed spaces and an average daily outpatient attendance of
900. This hospital is a secondary public health care facility that
offers comprehensive HIV care services, including ART. HIV/AIDS
outpatients were accessed at the Antiretroviral Pharmacy, a
subunit of the main pharmacy in the hospital.
Data Collection
Data on patients’ QOL were gathered by using the HIV/AIDSTargeted Quality of Life (HAT-QOL) instrument [17]. The HAT-QOL
consists of 34 item stems in nine domains: overall function, life
satisfaction, health worries, financial worries, medication concerns, HIV mastery, disclosure worries, provider trust, and sexual
function. Response to the questionnaire items was anchored on a
scale of 5 to 1 as follows: all of the time ¼ 5, a lot of the time ¼ 4,
some of the time ¼ 3, a little of the time ¼ 2, and none of the time
¼ 1. Eight of the 34 items were negatively worded; these items
were reverse scored for analysis so that higher scores indicated
higher QOL in each domain.
Consenting outpatients, 18 years and above, visiting the facility
during the period of August 1 to 30, 2011, were consecutively
approached. A brief introduction of the questionnaire was given to
the participants, who were asked to complete the assessment.
Patients who were literate completed the questionnaire unaided,
while those unable to read were aided by an interpreter, until 403
participants were recruited within the targeted period. Sociodemographic characteristics of the patients were obtained from their
case notes.
Ethical Considerations
The ethical approval for this study was obtained from the Ethical
Committee of the Central Hospital, Benin City, Nigeria. Participants gave a verbal consent and were assured of the confidentiality of the information that they volunteered.
Data Analysis
Usable responses from the 34-item questionnaire with five-point
response scale were entered into a Microsoft Excel spreadsheet
and rechecked for accuracy prior to analysis.
Data were loaded into the Statistical Package for Social
Sciences software version 17.0 for descriptive statistical analysis.
All HAT-QOL domains were scored so that the final domain score
was transformed into a linear 0 to 100 scale, where 0 was the
worst score possible and 100 was the best score possible.
Obtaining this final transformed domain score was done in four
steps indicated below.
1. A value for all subject responses, item by item, as
noted below was imputed. Most item responses
were valued by using Code A, while others were
valued by using Code B.
Response Option
Code A
Code B
“All of the time”
1
5
“A lot of the time”
2
4
“Some of the time”
3
3
“A little of the time”
4
2
“None of the time”
5
1
2. Code A was used for the following items: 1b, 1c, 1d,
1e, 1f, 3a, 3b, 3c, 3d, 4a, 4b, 4c, 5a, 5b, 5c, 5d, 5e, 6a,
6b, 7a, 7b, 7c, 7d, 7e, 9a, and 9b; while
3. Code B was used for the following items: 1a, 2a, 2b,
2c, 2d, 8a, 8b, and 8c.
The mean score of items within each domain was used to
calculate the domain scores and subsequently transformed to a
0 to 100 scale by using the questionnaire scoring guide as follows:
Overall function: OVFXN100 ¼ (100/(306)) (OVFXN6)
Life satisfaction: LISAT100 ¼ (100/(204)) (LISAT-4)
Health worries: HEAWO100 ¼ (100/(204)) (HEAWO4)
Financial worries: FINWO100 ¼ (100/(153)) (FINWO-3)Medication worries: MEDWO100 ¼ (100/(25
5)) (MEDWO-5)
HIV mastery: HIVMA100 ¼ (100/(102)) (HIVMA-2)
Disclosure worries: DISWO100 ¼ (100/(255)) (DISWO-5)
Provider trust: PROTR100 ¼ (100/(153)) (PROTR-3)
Sexual function: SXFXN100 ¼ (100/(102)) (SXFXN-2)
Furthermore, the summated rated scores (range of 1–5) for each
domain were computed in terms of mean and the SD, and
Cronbach’s alpha. Association between rated scores and each
domain was explored by using Students’ t test and one-way
256
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 254–258
analysis of variance at 95% confidence interval with the aid of
GraphPad Instat version 3.0. The level of statistical significance
was set at P less than .05.
Results
Patients’ Demographic Characteristics
Of the 403 patients, 72 (17.7%) were males and 331 (81.3%) were
females. Majority of the patients (147, 36.1%) were aged 20 to 30 years,
and 239 (58.7%) were not married. Also, 338 (83.0%) earned below
$1500 per annum, and 303 (74.40%) had basic education (grade 1 to
grade 9). Other demographic characteristics are shown in Table 1.
Fig. 1 – Mean scores across the HAT-QOL domains. HATQOL, HIV/AIDS Targeted Quality of Life.
HAT-QOL Scores
The highest mean score was in medication worries (91.65),
followed by provider trust (91.63) and life satisfactions (91.94).
Disclosure worries had the least mean score (27.50) (Fig. 1).
Likert-Rated QOL Scores
Medication worries had the highest Likert mean score (4.67 ⫾
0.894; range 1–5), followed by provider trust (4.67 ⫾ 0.980).
Disclosure worries had the least Likert score (2.10 ⫾ 1.152). The
Cronbach’s alpha for the domains ranged from 0.580 to 0.954
(Table 2).
Impact of Sociodemographic Characteristics on HAT-QOL
Domains
Age and gender of the patients were associated with the sexual
function domain. Furthermore, level of education and
Table 1 – Sociodemographic data of patients.
Variable
Age (y)
Below 20
20–29
30–39
40–49
50 and above
Sex
Male
Female
Not indicated
Marital status
Not married
Married
Divorced
Widowed
Occupation
Unemployed
Employed
Monthly income (NGN 150 to $1)
Below 20,000.00
20,000–49,000.00
50,000–69,000.00
70,000–99,000.00
100,000.00 and above
Educational level
No formal education
Basic education (grades 1–9)
Post–basic education (grades
10–16)macmac
Number reporting, n (%)
5
73
147
98
84
(1.2)
(17.9)
(36.1)
(24.1)
(20.7)
72 (17.7)
331 (81.3)
4 (1.0)
239
106
22
40
(58.7)
(26.0)
(5.4)
(9.9)
46 (11.3)
361 (88.7)
338
51
9
5
4
(83.0)
(12.5)
(2.2)
(1.2)
(1.0)
33 (8.1)
303 (74.4)
71 (17.4)
employment status were associated with the provider trust
domain (P o 0.05). Details of the statistically significant associations between the sociodemographic factors and HAT-QOL
domains are presented in Table 3.
Discussion
The validity of HAT-QOL has been demonstrated [18–20]. No
studies, however, have been conducted in this southern part of
Nigeria. The instrument produced high internal consistency in
the various domains, indicating the validity of the HAT-QOL
instrument in a Nigerian setting, bearing the cultural sensitivity
of QOL instruments.
Patients in this cross-sectional study reported a high QOL in
all the domains except the domain of disclosure worries. This is
consistent with the study carried out in Kogi State, Nigeria, where
disclosure worries was also found to be low [2]. This could be a
result of fear of stigmatization and discrimination among
patients with HIV/AIDS. Stigmatization is an inhibiting factor to
the uptake of HIV counseling and testing and other preventive,
care, and support services. Stigmatization and discrimination
contribute inadvertently to the transmission of HIV/AIDS among
unsuspecting sexually active people. The fear of violence as a
consequence of HIV/AIDS status disclosure also contributes to
HIV transmission. HIV stigma has a significantly negative and
constant impact on life satisfaction QOL for people with HIV
infection. In the absence of any intervention to address and
reduce stigmatization, individuals will continue to report poorer
life satisfaction, evidenced by reduced living enjoyment, loss of
control in life, decreased social interactivity, and decreased
perceived health status [19].
Table 2 – Likert-rated scores and Cronbach’s alpha
for the HAT-QOL domains.
Domain
Likert score
Cronbach’s α
Overall function
Life satisfaction
Health worries
Financial worries
Medication worries
HIV mastery
Disclosure worries
Provider trust
Sexual function
4.56
4.68
4.48
4.24
4.67
3.84
2.10
4.67
3.81
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
0.824
0.816
0.633
0.694
0.704
0.871
0.790
0.580
0.954
0.936
0.905
1.070
1.317
0.894
1.555
1.152
0.980
1.759
HAT-QOL, HIV/AIDS Targeted Quality of Life.
257
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 254–258
Table 3 – Impact of sociodemographic characteristics on HAT-QOL domains.
Mean score ⫾ SD
Calculated value
P
Sex
○ Male
○ Female
4.433 ⫾ 1.206
4.690 ⫾ 0.956
t ¼ 1.967
0.0499
Sexual functions
○ Male
○ Female
4.450 ⫾ 1.375
3.675 ⫾ 1.831
t ¼ 3.388
0.0008
Health worries
Employment status
○ Employed
○ Unemployed
4.525 ⫾ 1.014
4.150 ⫾ 1.401
t ¼ 2.229
0.0264
○ Employed
○ Unemployed
4.710 ⫾ 0.889
4.146 ⫾ 1.541
t ¼ 3.628
0.0003
Level of education
○ Basic education
○ Post–basic education
4.357 ⫾ 1.241
3.793 ⫾ 1.146
t ¼ 3.451
0.0006
Provider trust
○ Basic education
○ Post–basic education
4.705 ⫾ 0.936
4.320 ⫾ 1.360
t ¼ 2.820
0.0050
Life satisfaction
Age (y)
○ 20–29
○ 30–39
○ 40–49
○ 50 and above
4.382
4.700
4.692
4.875
⫾
⫾
⫾
⫾
1.259
0.871
0.850
0.528
F ¼ 6.169
0.0004
3.450
3.685
4.160
4.025
⫾
⫾
⫾
⫾
1.145
1.624
1.268
1.508
F ¼ 4.370
0.0048
2.130
4.190
3.705
2.970
⫾
⫾
⫾
⫾
1.490
1.518
1.817
1.968
F ¼ 26.767
0.0001
HAT-QOL domain
Provider trust
Provider trust
Financial worries
HIV/AIDS mastery
Variable
○ 20–29
○ 30–39
○ 40–49
○ 50 and above
Sexual function
○ 20–29
○ 30–39
○ 40–49
○ 50 and above
HAT-QOL, HIV/AIDS Targeted Quality of Life.
HAT-QOL scores were high in the following domains: health
worries, medication worries, HIV/AIDS mastery, provider trust,
overall function, financial worries, life satisfaction, and sexual
function. This is in line with the studies conducted in Philadelphia, PA [18], although a similar study carried out in the Niger
Delta region in Nigeria revealed that there was impairment in the
QOL except in medication worries and sexual function domains
in PLWHA in that region [19].
Income assessment did not affect any of the HAT-QOL
domains. This also supports the finding in a study carried out
in three hospitals in Thailand [21]. This is mainly due to
similarities in some sociodemographic characteristics, free availability of and adherence to ART, and accessibility to other HIV/
AIDS care components.
Provider trust and sexual function were affected by gender.
Gender refers to the expectations or norms within a society about
the roles and responsibilities that are appropriate for women and
men. Research has shown that the gender-based imbalance in
power found in the economic and social spheres of life is
reflected in sexual relationships. Women often have less control
over the nature and timing of sex and the practice of protective
behaviors. A woman’s ability to practice safer sex may be
influenced by her ability to communicate openly about sex with
her partner, the power dynamic in their relationship, or how
much the partner believes in traditional gender roles. Beliefs or
norms about masculinity and femininity often encourage men to
have multiple partners and women to be passive and ignorant
about matters of sexuality and reproduction. Gender, therefore,
affects both women’s and men’s risk of HIV and other sexually
transmitted infections [23].
Employment status affected health worries and provider trust
domains. Level of education affected financial worries and
provider trust, while life satisfaction, HIV/AIDS mastery, and
sexual function were affected by age. Positive changes tended
to be associated with higher educational levels and greater
income, while negative changes were inversely related to health
status and level of optimism [12,13,22,23].
Implications of the Findings
Patients reported high QOL in all domains, except disclosure
worries, indicating their concern for stigmatization and discrimination in the society. There is need for governmental and
nongovernmental agencies to heighten advocacy toward reducing stigmatization and discrimination of PLWHA to encourage the
uptake of HIV/AIDS care and reduce its spread.
258
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 254–258
The QOL was found not to be associated with income. The
reason may probably be due to free care the patients received
through donor agencies. That would remain so if free care is
sustained. Bearing donor fatigue, Nigerians need to build local
capacity for long-time sustainability of free HIV/AIDS care at
community levels.
Provider trust has implications for patient care. Patients with
a lower level of trust in their care provider are more likely to
report that requested or needed services are not provided.
Understanding this relationship may lead to better ways of
responding to patient requests that preserve or enhance patient
trust, leading to better outcomes.
Study Limitations
This study used a cross-sectional design, which could not
account for changes in the QOL of PLWA over time such as a
longitudinal survey would provide. Some respondents were aided
to complete the survey, and the possibility of bias cannot be
completely eliminated, despite efforts. Patients were consecutively recruited during the study period; a randomized selection
process would yield a more generalizable result.
Conclusions
Patients reported satisfactory QOL in the domains of overall
function, life satisfaction, health worries, financial worries, medication worries, HIV mastery, provider trust, and sexual function.
The QOL was low in the domain of disclosure worries, indicating
concerns for discrimination and stigmatization. Age, level of
education, and employment status had a strong impact on the
QOL of patients with HIV/AIDS.
Source of financial support: The authors have no other
financial relationships to disclose.
R EF E R EN CE S
[1] United Nations Joint Programme on HIV/AIDS (2008). Report on the
global AIDS epidemic. Geneva, UNAIDS. Available at: www.unaids.org/
en/media/unaids/contentassets/dataimport/pub/globalreport/2008/
jc1510_2008globalreport_en.pdf. [Accessed August 30, 2013].
[2] Fatiregun AA, Mofolorunsho KC, Osagbemi KG. Quality of life of people
living with HIV/AIDS in Kogi State, Nigeria. Benin J Postgrad Med
2009;2:2–10.
[3] Lesserman J, Perkins DO, Evans DL. Coping with the threat of AIDS: the
role of social support. Am J Psychiatr 1992;149:1514–20.
[4] Namir S, Wolcott D, Fawzy F, Alumbaugh M. Implications of different
strategies for coping with AIDS. In: Temoshack L, Baum A,eds.
Psychological Perspectives on AIDS. Hillsdale, NJ: Erlbaum, 1990.
[5] Rabkin JG, Remien R, Kattoff L, Williams JB. Residence in adversity
among long time survivors of AIDS. Hosp Comm Psychiatr
1993;44:162–7.
[6] Elisabete C, Morandi Dos Santos, Ivan FJ, Fernanda L. Quality of life of
people living with HIV/AIDS in São Paulo. Brazil: Rev Saúde Pública
2007;741(supp.2):647.
[7] McDowell I, Newell C. Measuring Health: A Guide to Rating Scales and
Questionnaires. (2nd ed.). New York: Oxford University Press, 1996.
[8] Razera F, Ferreira I, Bonamigo R. Factors associated with health-related
quality of life in HIV infected Brazilians. Int J STD 2008;19:519–23.
[9] Basavarj KH, Navya MA, Rashmi R. Quality of life in HIV/AIDS. Ind J Sex
Transmission 2010;31:75–80.
[10] Razera F, Ferreira I, Bonamigo R. Factors associated with health-related
quality of life in HIV infected Brazilians. Int J STD 2008;19:519–23.
[11] Wig N, Lekshmi R, Pal H, et al. The impact of HIV/AIDS on the quality of
life. Ind J Med Sci 2006;60:3–12.
[12] Odili VU, Oparah AC, Usifoh SF, Ikhurionan IB. Determinants of quality
of life in HIV/AIDS patients. West Afr J Pharm 2011;22:42–8.
[13] Chandra PS, Gandhi C, Satischchandra P, et al. Quality of life of HIV sub
type c infections among asyptomatic subjects and its association with
CD4 counts and viral load—a study from India Qual Life Res
2006;15:1597–605.
[14] Janeen H, Jennifer J, Emilou M, et al. The health-related quality of life of
people living with HIV/AIDS. Informa Healthcare 2004;26:371–6.
[15] Robberstad B, Olsen JA. The health related quality of life of people
living with HIV/AIDS in sub-Saharan Africa—a literature review and
focus group study. Cost-Effect Res Alloc 2010;8:5. Available from: http://
www.resource-allocation.com/content/8/1/5). [Accessed February 28,
2012].
[16] Cotton S, Puchalski CM, Sherman SN, et al. Spirituality and religion in
patients with HIV/AIDS. J Gen Intern Med 2006;21(Suppl):S5–13.
[17] Holmes WA, Shea JA. Performance of a new HIV/AIDS Targeted Quality
of Life (HAT-QOL) instrument in asymptomatic sero-positive
individuals. Qual Life Res 1997;6:561–71.
[18] Holmes WC, Ruocco JE. Test-retest evaluation of HAT-QoL and SF-36 in
an HIV-seropositive sample. AIDS Care 2008;20:1084–92.
[19] Greeff M, Uys LR, Wantland D, Makoae L. Perceived HIV stigma and life
satisfaction among persons living with HIV infection in five African
countries: a longitudinal study. Int J Nurs Stud 2009: (NS-1559:12).
[20] Ekott FA, Bassey JU, Etukumana AE. Quality of life in people living with
HIV/AIDS in Niger Delta Region, Nigeria. J Ment Health 2010;19:211–8.
[21] The HIV/AIDS Survey Library HIV research domains: gender and sexual
relationships. Available from: www.popcouncil.org/Horizons/
ORToolkit/AIDQuest/topics/gender.html. [Accessed February 29, 2012].
[22] Updegraff JA, Taylor SE, Kemeny ME, Wyatt GE. Positive and negative
effectives of HIV infection in women with low socioeconomic
resources. Pers Soc Psychol Bull 2002;28:382–94.
[23] Marie P, Fabienne M, Patrizia MC, France L. Health-related quality
of life in French people living with HIV in 2003: results from the national
ANRS-EN12-VESPA Study. The VESPA Study Group. 2007;21(Suppl.):
S19–27.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 259–263
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
CLINICAL OUTCOMES STUDIES
Clinical Burden of Invasive Pneumococcal Disease in Selected Developing
Countries
Namaitijiang Maimaiti, PhD1,2,, Zafar Ahmed, PhD1,3, Zaleha Md Isa, PhD2, Hasanain Faisal Ghazi, PhD1,2,
Syed Aljunid, PhD1,3
1
International Institute for Global Health, United Nations University, Kuala Lumpur, Malaysia; 2Department of Community Health, Faculty of Medicine, UKM;
International Centre for Case-Mix and Clinical Coding, UKM Medical Centre
3
AB STR A CT
Objective: To measure the clinical burden of invasive pneumococcal
disease (IPD) in selected developing countries. Methods: This is an
extensive literature review of published articles on IPD in selected
developing countries from East Asia, South Asia, Middle East, subSaharan Africa, and Latin America. We reviewed all the articles
retrieved from the knowledge bases that were published between
the years 2000 and 2010. Results: After applying the inclusion,
exclusion, and quality criteria, the comprehensive review of the
literature yielded 10 articles with data for pneumococcal meningitis,
septicemia/bacteremia, and pneumonia. These selected articles were
from 10 developing countries from five different regions. Out of the 10
selected articles, 8 have a detailed discussion on IPD, one of them has
s detailed discussion on bacteremia and meningitis, and another one
has discussed pneumococcal bacteremia. Out of these 10 articles, only
5 articles discussed the case-fatality ratio (CFR). In our article review,
the incidence of IPD ranged from less than 5/100,000 to 416/100,000
population and the CFR ranged from 12.2% to 80% in the developing
countries. Conclusions: The review demonstrated that the clinical
burden of IPD was high in the developing countries. The incidence of
IPD and CFR varies from region to region and from country to country.
The IPD burden was highest in sub-Saharan African countries followed by South Asian countries. The CFR was low in high-income
countries than in low-income countries.
Introduction
high, with an estimated 600,000 to 800,000 adult deaths each year
from pneumococcal pneumonia, meningitis, and sepsis [8]. The
vast majority of its victims come from the world’s poorest
countries, and half of them are children younger than 5 years.
It is extraordinary, in view of these facts, that pneumococcal
disease remains a relatively unknown disease and does not have
a higher place on the agenda of the international community [3].
Compared with other diseases affecting the developing world,
determining the incidence of pneumococcal disease is relatively
difficult [9]. This is due to a number of factors including the
difficulties involved in stringent laboratory testing and sample
collection and the unavailability of quality surveillance data in
developing countries. This lack of epidemiological evidence has
likely contributed to a gross underappreciation of the economic,
clinical, and human burdens imposed by pneumococcal disease
and hindered public health planning and decision making in
developing countries [3]. Most of the researchers have done the
review of invasive pneumococcal disease (IPD) for patients
belonging to a particular age group, for example, pediatric age
group, and limited to a single region. We extended our review of
Streptococcus pneumoniae is a major public health problem causing
meningitis, bacteremia, pneumonia, and acute otitis media [1,2].
S. pneumoniae is the most common cause of pneumonia worldwide, causing approximately 36% of all childhood pneumonias.
S. pneumoniae can cause potentially life-threatening lung infections including severe pneumonia, which hinders the movement
of oxygen into the bloodstream, potentially resulting in death
from respiratory failure [3]. Invasive pneumococcal infections
often prove rapidly fatal, even where good medical treatment is
readily available. In developed countries, up to 20% of people who
contract pneumococcal meningitis die; however, in the developing world, mortality is closer to 50%, even among hospitalized
patients [4]. In one study from Gambia, 48% of children who
contracted pneumococcal meningitis and reached hospital did
not survive [5]. The World Health Organization (WHO) estimated
that 1.6 million people die from pneumococcal disease every year
[6], with 0.7 million to 1 million being children younger than 5
years [7]. The pneumococcal disease burden among adults is also
Keywords: clinical burden, developing countries, East Asia, IPD, Latin
America, Middle East, South Asia, sub-Saharan Africa.
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
* Address correspondence to: Namaitijiang Maimaiti, International Institute for Global Health, United Nations University, UKMMC, Jalan
Yaacob Latif, Bandar tunrazak, Cheras 56000, Kuala Lumpur, Malaysia.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.07.003
260
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 259–263
IPD to all age groups on the basis of availability of data and
extended to five different regions in the world. The developing
countries are defined as a nation with a low living standard,
underdeveloped industrial base, and low human development
index relative to other countries [10]. This will help us better
understand the disease burden in all the age groups in selected
developing countries in different parts of the world. We measured the clinical burden of IPD in selected developing countries
by geographical and economical point of view. Clinical burden is
a burden that differs from genetic burden mainly in the added
component of morbidity; a trait that is either clinically or
genetically lethal may be grossly disabling [11].
Methods
We conducted a detailed review of published articles on IPD.
Literature searches were conducted by using the PubMed database, Google scholar, and The Lancet, and were limited to articles
written in English. We selected 10 developing countries from five
regions, namely, East Asia (Malaysia and Thailand), South Asia
(Bangladesh and Nepal), Middle East (Saudi Arabia and Qatar),
sub-Saharan Africa (The Gambia and Mozambique), and Latin
America (Cuba and Peru). The World Bank categorizes the
countries by income into five different categories: low-income
economics, lower-middle-income economics, upper-middleincome economics, high-income economics, and high-income
Organisation for Economic Co-operation and Development members [12]. The main criterion for the selection of these countries
from the designated regions was that they should belong to same
economic status (as defined by the World Bank) and preferably be
neighbors. When selected countries did not have an equivalent
neighbor, we selected the next nearest country in the region with
the same economic status.
The search term combinations used to search the knowledge
base included pneumococcal pneumonia, epidemiology, incidence rate, clinical burden AND pneumococcal meningitis, epidemiology, incidence rate, clinical burden AND pneumococcal
septicemia, epidemiology, incidence rate, and clinical burden. We
started with reviewing the abstracts of these articles published
between the year 2000 and 2010 to find out which of the studies
met our inclusion criteria and then reviewed only those fulllength articles that complied with our inclusion criteria. The
knowledge base searches were conducted from March 1, 2011, till
July 31, 2011. We included only published articles and articles
that described the clinical burden of pneumococcal disease with
the quantitative data of the selected developing countries in the
different regions. If there was more than one article in each
country, the most comparable study or the study with a high
quality of methodology was reviewed. We excluded case reports,
reports, and special populations, such as reports for patients with
HIV/AIDS. We also excluded articles published before 2000 and
after 2010, or articles from developed countries. We accepted
each author’s definition and methods except for the exclusion
criteria mentioned.
Two individuals (Namaitijiang Maimaiti and Zafar Ahmed
[NM and ZA]) independently screened the titles and abstracts of
each citation retrieved from the search term and identified the
articles for full review. The decision to review an article in detail
was based on the content of the abstract and whether the article
described the clinical burden of pneumococcal disease. If the
reviewed article included both clinical and economic burdens of
pneumococcal disease, we still included this article in our study,
but reviewed only the clinical burden of disease for our study.
Initially, the search process using the search term yielded 51
articles. Finally, 22 articles remained to be reviewed in detail. If
more than one article from the same country was identified that
described the clinical burden of pneumococcal disease, we selected
the article with the later publication date with same quality or
better quality methods compared with previous studies. The
detailed review of the selected 22 articles was carried out by MN
and ZA in a team, and finally they selected 10 articles that fulfilled
all the selection criteria. Out of these 10 selected articles, 8 were
full-text original publications and 2 were review articles, and the 10
articles selected were then used for our analysis. We gathered
information such as study design (retrospective/prospective study),
study period, study location (urban/rural), study population and
age of sample, incidence rate, and case-fatality ratio (CFR) (Fig. 1).
Results
The list of the selected articles from 10 developing countries
belonging to five different regions is presented in Table 1.
Among the 10 articles selected, 8 articles had a detailed
discussion on IPD, 1 discussed meningitis, and 1 discussed
pneumococcal bacteremia. Out of the 10 articles selected, only 5
articles described the CFR (Table 2). The studies included in this
review were carried out in a variety of settings. Seven studies
were carried out in an urban seating; two studies were carried out
in rural areas, whereas only one study was carried out in both
urban and rural settings.
51 were identified
22 were read and abstracted
29 from full text did not meet inclusion
criteria, published before 2000 or after 2010,
duplicate reference, repeated data or poor
quality.
10 Full text article, review
Fig. 1 – Flow diagram for the process of review of the literature. *HI, high income; LI, low income; UMI, upper-middle income.
261
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 259–263
Table 1 – List of the 10 selected articles and data of publication.
No.
1
2
3
4
5
6
7
8
9
10
Title of selected articles
Year of
publication
Epidemiology of invasive pneumococcal infection in children aged five years and under in Saudi
Arabia: a five-year retrospective surveillance study
Acute bacterial meningitis in Qatar
Invasive pneumococcal disease among children in rural Bangladesh: results from a populationbased surveillance
Invasive pneumococcal disease in Kanti Children’s Hospital, Nepal, as observed by the South
Asian Pneumococcal Alliance Network
Invasive pneumococcal diseases among hospitalized children in Lima, Peru
Bacterial meningitis in children and adolescents: an observational study based on the national
surveillance system
Invasive pneumococcal disease in children o5 years of age in rural Mozambique
Efficacy of nine-valent pneumococcal conjugate vaccine against pneumonia and invasive
pneumococcal disease in The Gambia: randomised, double-blind, placebo-controlled trial
Overview of the disease burden of invasive pneumococcal disease in Asia
Incidence of pneumococcal bacteremia requiring hospitalization in rural Thailand
Table 3 shows that the IPD incidence ranged from less than 5
per 100,000 to 416 per 100,000. In general, the incidence rate of
IPD varies directly with the respective country’s income category,
with higher income countries having a low incidence rate and
low-income countries having a high incidence rate. The only
country that shows an exception is the Saudi Arabia, where the
overall incidence rate for IPD (bacteremia and meningitis) is 17.4
per 100,000 population; if we look at the incidence rate for
meningitis alone, 4/100,000, it is higher than that in Qatar, 2.4/
100,000.
Figure 2 also shows that the incidence rate in low-income
countries is way higher than the incidence rate in countries
among other income categories. We selected low-income countries from two different geographic regions. The data show that
the incidence rate of IPD in the low-income countries of subSaharan Africa (Gambia and Mozambique) is considerably higher
than the incidence rate of IPD in the low-income countries from
South Asia (Bangladesh and Nepal).
Reference
2010
[13]
2006
2009
[14]
[15]
2009
[16]
2010
2005
[17]
[18]
2006
2006
[19]
[20]
2010
2009
[21]
[22]
When we look at the incidence rate of IPD according to the
geographic region, we found that the incidence rate in East Asian
and Latin American countries was not very different from each
other—less than 10 per 100,000 for both the regions. The incidence rate of IPD in middle eastern countries was slightly higher
than that in East Asian and Latin American countries, but it was
still less than 20 per 100,000.
Among the five regions, the CFR was reported in four regions
—East Asia, South Asia, Middle East, and Latin America. Based on
available data, the highest CFR was recorded in South Asia (80%),
followed by East Asia (33.3%), Latin America (27%), and Middle
East (12.2%). Concluding from these published evidences, we can
say that the incidence rate is high in low-income developing
countries such as Mozambique, Gambia, Bangladesh, and Nepal,
and this could be due to inequality, inequitable distribution of
wealth, inadequate provision of health care services, lack of
human resource and human capacity to manage IPD effectively,
and also lack of latest technology in the health care system.
Table 2 – Summary of epidemiology of invasive pneumococcal infections in selected developing countries by
geographic area and per-capital income.
Region
Country
Country
category
Study design
Study period
Study
location
Study
population
Age of
sample
East Asia
Malaysia
Urban
216
o5 y
2005–2009
Rural
23,853
2004–2007
Rural
22,378
All age
groups
o5 y
Nepal
Saudi Arabia
Qatar
Mozambique
Low- income
High- income
High- income
Low-income
2004–2007
1999–2003
1998–2002
2001–2003
Urban
Urban
Urban
Urban
2,528
82
64
10,702
o5 y
o5 y
o12 y
o5 y
Gambia
Low-income
2000–2003
Urban
17,437
o3 y
Peru
Upper-middle
income
Upper-middle
income
Retrospective
study
Population-based
surveillance
Population-based
surveillance
Retrospective study
Retrospective study
Retrospective study
Prospective
surveillance
Clinics: controls in a
vaccine trial
Retrospective study
2004–2006
Bangladesh
Upper-middle
income
Upper-middle
income
Low-income
2006–2008
Urban
101
o5 y
1998–2003
Both
1,023
o18 y
Thailand
South Asia
Middle east
Sub-Saharan
Africa
Latin
America
Cuba
Retrospective study
CFR, case-fatality ratio; IPD, invasive pneumococcal disease; IR, incidence rate; PP, pneumococcal pneumonia.
262
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 259–263
Table 3 – The IR and CFR of pneumococcal disease by region, country, and economic level of countries
Region
Middle east
East Asia
Latin America
South Asia
Sub-Saharan Africa
Country category
High-income
High-income
Upper-middle
Upper-middle
Upper-middle
Upper-middle
Low-income
Low-income
Low-income
Low-income
income
income
income
income
Country
Saudi Arabia
Qatar
Malaysia
Thailand
Peru
Cuba
Bangladesh
Nepal
Mozambique
Gambia
IR of IPD/100,000
17.4
2.4
8.6
5.7
7.7
6
86
52.4
416
388
CFR
12.20%
NA
33.30%
NA
22%
27%
NA
80%
NA
NA
CFR, case-fatality ratio; IPD, invasive pneumococcal disease; IR, incidence rate; NA, not applicable/available.
Single disease IR.
When we look at the CFR, it was high in low-income countries
(80%) and upper-middle-income countries (33.3%, 31.7%, and 22%)
than in high-income countries (12.2%). Even though the incidence
rate was higher in high-income countries than in upper-middleincome countries, the CFR was low in these high-income countries than in middle-income countries. This can be explained
with the argument that generally high-income countries have
good capacity to provide excellent quality health care to their
population by using latest technology and hire properly qualified
professionals to operate health facilities.
We recommend that future studies should be carried out with
large sample size, be conducted in both urban and rural areas
among all age groups, and should report the CFR.
This study included five different regions covering both urban
and rural settings and critical morbidity. In this study, most of
the articles covered all age groups. A limitation of this study was
that it was conducted only in 10 developing countries and the
result may not reflect the situation in other developing countries.
Conclusions
Discussion
The clinical burden of the disease was slightly different among
the neighboring countries within the same region in this review.
These differences in the disease burden between neighboring
countries even though they belong to the same economic level
may be due to differences in the health care system, human
resource, budget for health care expenditures, health insurance
coverage, climate, and so on. Collectively, these differences can
also be due to the objectives of the study, study design, different
sample demographic, location, and other factors. One of the
limitations of the study is that only two countries were selected
from each region.
Our review confirmed that the clinical burden of IPD was high in
developing countries. The incidence rate and the CFR of IPD were
different from region to region and country to country. The
disease burden was high in sub-Saharan African countries
followed by South Asian countries. The CFR was low in highincome countries than in low-income countries. The disease
burden was not significantly different among the countries that
belong to the category of upper-middle-income countries. The
review found that generally the disease burden was high in rural
areas than in urban areas, which can be explained by the inequity
in health care facilities available in rural areas. Studies from
Middle Eastern countries and Peru had a very small sample size;
therefore, because of the small sample size, these studies may
Fig. 2 – Incidence rate of IPD by income. IPD, invasive pneumococcal disease.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 259–263
not reflect the true disease burden of these countries. Only three
studies were conducted among all age groups, and the other
studies were carried out only among children.
The WHO reported that 1.6 million people die every year
from pneumococcal disease, with 0.7 million to 1 million of
these being children younger than 5 years. The pneumococcal
disease burden among adults is also high, with an estimated
600, 000 to 800, 000 adult deaths each year from pneumococcal
pneumonia, meningitis, and sepsis [7]. The review done is in
accordance with the WHO report that says that the pneumococcal disease burden is high among children as well as among
adult populations.
Source of financial support: This research project was funded
by GlaxoSmithKline SdnBhd.
R EF E R EN CE S
[1] WHO position paper: pneumococcal vaccines. Wkly Epidemiol Rec
1999;74(23):177–83.
[2] O’Brien KL, Wolfson LJ, Watt JP, et al. Burden of disease caused by
Streptococcus pneumoniae in children younger than 5 years: global
estimates. Lancet 2009;374:893–902.
[3] Preventing pneumococcal disease: report from the All-Party Parliamentary
Group on Pneumococcal Disease Prevention in the Developing World.
2008. Available from: www.appg-preventpneumo.org.uk.
[4] World Health Organization. Pneumococcal vaccines. Wkly Epidemiol
Rec 2003;14:110–9.
[5] Goetghebuer T, West TE, Wermenbol V, et al. Outcome of meningitis
caused by Streptococcus pneumoniae and Haemophilus influenza type b in
children in The Gambia. Trop Med Int Health 2000;5:207–13.
[6] World Health Organization. Pneumococcal conjugate vaccine for
childhood immunization: WHO position paper. Wkly Epidemiol Rec
2007;82(12):93–104.
[7] World Health Organization. WHO Report: Bi-regional Meeting on
Prevention of Childhood Pneumonia and Meningitis by Vaccination,
Kuala Lumpur, Malaysia, 30–31 March 2006. Manila, Philippines:
World Health Organization Regional Office for the Western Pacific,
2006.
263
[8] Chun first initial, 2 more authors, et al. Estimation of otitis media
disease and cost burden in Korea. Presented at 12th Western Pacific
Congress on Chemotherapy and Infectious Diseases [WPCCID], 2010.
[9] Madhi SA, Kuwanda L, Cutland C, et al. The impact of a 9-valent
pneumococcal conjugate vaccine on the public health burden of
pneumonia in HIV-infected and -uninfected children. Clin Infect Dis
2005;40:1511–8.
[10] Sullivan A, Sheffrin SM. Economics: Principles in Action. Upper Saddle
River, NJ: Pearson Prentice-Hall, 2003.
[11] Drugs.com. Clinical burden. Available from: http://www.drugs.com/
dict/clinical-burden.html.
[12] The World Bank. Country and landing groups. Available from: http://
data.worldbank.org/about/country-classifications/country-and-lending-groups#East_Asia_and_Pacific.
[13] Memish ZA, El-Saed A, Al-Otaibi B, et al. Epidemiology of invasive
pneumococcal infection in children aged five years and under in Saudi
Arabia: a five-year retrospective surveillance study. Int J Infect Dis
2010;14(8):e708–12.
[14] Elsaid MF, Flamerzi AA, Bessisso MS, Elshafie SS. Acute bacterial
meningitis in Qatar. Saudi Med J 2006;27:204–10.
[15] Arifeen SE, Saha SK, Rahman S, et al. Invasive pneumococcal disease
among children in rural Bangladesh: results from a population-based
surveillance. Clin Infect Dis 2009;48(Suppl. 2):S103–13.
[16] Shah AS, Knoll MD, Sharma PR, et al. Invasive pneumococcal disease in
Kanti Children’s Hospital, Nepal, as observed by the South Asian
Pneumococcal Alliance Network. Clin Infect Dis 2009;48(Suppl. 2):S123–8.
[17] Ochoa TJ, Egoavil M, Castillo ME, et al. Invasive pneumococcal diseases
among hospitalized children in Lima, Peru. Rev Panam Salud Publica
2010;28:121–7.
[18] Dickinson FO, Pérez AE. Bacterial meningitis in children and
adolescents: an observational study based on the national surveillance
system. BMC Infect Dis 2005;5:103.
[19] Roca A, Sigaúque B, Quintó L, et al. Invasive pneumococcal disease in
childreno5 years of age in rural Mozambique. Trop Med Int Health
2006;11:1422–31.
[20] Cutts FT, Zaman SM, Enwere G, et al. Efficacy of nine-valent
pneumococcal conjugate vaccine against pneumonia and invasive
pneumococcal disease in The Gambia: randomised, double-blind,
placebo-controlled trial. Lancet 2005;365(9465):1139–46.
[21] Bravo LC. Overview of the disease burden of invasive pneumococcal
diseases in Asia. Vaccine 2009;27(52):7282–91.
[22] Baggett HC, Peruski LF, Olsen SJ, et al. Incidence of pneumococcal
bacteremia requiring hospitalization in rural Thailand. Clin Infect Dis
2009;48(Suppl. 2):S65–74.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 264–266
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
HEALTH POLICY ANALYSIS
Capacity Building for HTA Implementation in Middle-Income Countries:
The Case of Hungary
Zoltán Kaló, MSc, MD, PhD1,2,, József Bodrogi, MSc3, Imre Boncz, MSc, MD, PhD4,5, Csaba Dózsa, MSc, PhD6,
Gabriella Jóna, MD, MSc7, Rita Kövi, MD7, Zsolt Pásztélyi, MD, MSc8, Balázs Sinkovits, MSc9, on behalf of ISPOR Hungary Chapter
1
Health Economics Research Centre, Eötvös Loránd University (ELTE), Budapest, Hungary; 2Syreon Research Institute, Budapest, Hungary; 3Independent Senior
Health Economist, Budapest, Hungary; 4Faculty of Health Sciences, Institute for Health Insurance, University of Pécs, Pécs, Hungary; 5Faculty of Economics,
Health Economics and Health Technology Assessment Research Centre, Corvinus University of Budapest, Budapest, Hungary; 6Health Care Faculty, University of
Miskolc, Miskolc, Hungary; 7National Institute for Quality- and Organizational Development in Healthcare and Medicines, Budapest, Hungary; 8Railway Health
Services, Budapest, Hungary; 9AstraZeneca, Budapest, Hungary
AB STR A CT
Objectives: Middle-income countries often have no clear roadmap for
implementation of health technology assessment (HTA) in policy
decisions. Examples from high-income countries may not be relevant,
as lower income countries cannot allocate so much financial and
human resources for substantiating policy decisions with evidence.
Therefore, HTA implementation roadmaps from other smaller-size,
lower-income countries can be more relevant examples for countries
with similar cultural environment and economic status. Methods: We
reviewed the capacity building process for HTA implementation in
Hungary with special focus on the role of ISPOR Hungary Chapter.
Results: HTA implementation in Hungary started with capacity building at universities with the support of the World Bank in the mid 90's,
followed by the publication of methodological guidelines for conducting health economic evaluations in 2002. The Hungarian Health
Economics Association (META) - established in 2003 - has been
recognized as a driving force of HTA implementation. META became
the official regional ISPOR Chapter of Hungary in 2007. In 2004 the
National Health Insurance Fund Administration made the cost-effectiveness and budget impact criteria compulsory prior to granting
reimbursement to new pharmaceuticals. An Office of Health Technology Assessment was established for the critical appraisal of
economic evaluations submitted by pharmaceutical manufacturers.
In 2010 multicriteria decision analysis was introduced for new
hospital technologies. Conclusion: The economic crisis may create
an opportunity to further strengthen the evidence base of health care
decision-making in Hungary. In the forthcoming period ISPOR Hungary Chapter may play an even more crucial role in improving the
standards of HTA implementation and facilitating international collaboration with other CEE countries.
Keywords: capacity building, HTA implementation, ISPOR Hungary
Chapter, middle-income countries.
Introduction
countries such as the United Kingdom (i.e., the “National Institute
for Health and Clinical Excellence experience”) may not be relevant
because lower income countries cannot allocate so many financial
and human resources for substantiating health policy decisions
with evidence. In addition, the size of the country matters; the
smaller a country is the more limited facilities for preparing full
HTA reports it has. Therefore, HTA implementation roadmaps
from other smaller size, middle-income countries can be more
relevant, especially if the country is from the same geographical
and cultural environment with similar economic status.
One of the most important questions of the HTA implementation roadmap is whether capacity building should come first
or whether mandatory HTA requirement in the reimbursement process can induce the necessary background knowledge.
The health status of the population in middle-income countries
is usually worse than in high-income countries. Because health
care resources are scarcer in these countries, the societal cost of
inappropriate pricing and reimbursement decisions of new
health care technologies is even higher. Implementation of
health technology assessment (HTA) in the decision-making
process may alleviate this problem. Key success factors for HTA
implementation, however, are building human resource and
financial capacities, establishing a transparent decision-making
process, and implementing robust HTA methodology.
Middle-income countries often have no clear roadmap for HTA
implementation. Examples from high-income and resource-rich
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflict of interest: Authors have no conflicts of interest.
* Address correspondence to: Zoltán Kaló, Faculty of Social Sciences, Eötvös Loránd University, Pázmány Péter sétány 1/a, H-1117
Budapest, Hungary.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.002
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 264–266
Both options are feasible and have their advantages and
disadvantages.
Previous publications have already addressed major steps of
HTA implementation in Hungary [1–3]. Because HTA implementation in Hungary started with capacity building, our objective
was to summarize the role of International Society of Pharmacoeconomics and Outcomes Research (ISPOR) Hungary Chapter in
human resource development. The Hungarian example for HTA
capacity building can be especially relevant for other Central
Eastern European countries.
Hungary has 9.94 million inhabitants, and its gross domestic
product per capita was €10,016 in 2011 [4]; therefore, the country
has passed the World Bank’s threshold of high-income countries. A
major part of the HTA implementation, however, happened in the
years when Hungary was classified as a middle-income country.
Step 1: HTA Capacity Building
In the mid-1990s, two new academic centers, the School of Public
Health at the University of Debrecen and the Health Care
Management Training Center at Semmelweis University, were
established. Tutors received World Bank scholarships to study
health economics, health care management, epidemiology, biostatistics, public health, and other related sciences in distinguished international academic centers. Many graduates of the
World Bank program left the country, but many of them stayed or
even returned after some years of international experience.
Smaller scale educational projects also contributed to HTA
capacity building. The TUDOR program at the University of Szeged
was established to facilitate the application of evidence-based
medicine in Hungary. The program was sponsored by the British
Department for International Development, Know How Fund [5].
By 2000, the number of trained professionals with a thorough
understanding of HTA reached 50.
Several Hungarian universities (Corvinus University of Budapest, University of Debrecen, University of Pécs, and University of
West Hungary) introduced training programs in economic evaluation of medical technologies for undergraduate students. The
first postgraduate course with a major focus on economic
evaluation and economic modeling was introduced by the Faculty of Social Sciences at the Eötvös Loránd University in 2007.
By 2010, the number of trained professionals with personal
experience in HTA research or appraisal exceeded 200.
Step 2: Methodological Guidelines
Standardization of economic evaluations is a necessity prior to
mandating the use of economic evaluation in policy decisions.
The Hungarian methodological guidelines for conducting economic evaluation of health care interventions were published in
2002 [6]. These guidelines covered all health care interventions;
therefore, they were not specific for pharmaceuticals and not
limited to reimbursement questions. The intention was to update
the guidelines every 2 years; however, the Ministry of Health did
not implement any revision before 2013.
Step 3: Scientific Organization
The Hungarian Health Economics Association (Magyar Egészséggazdaságtani Társaság [META]) was established in 2003. The
founders aimed at establishing an independent organization to
discuss major health economic and health policy issues at
monthly meetings. Since 2003 META has been organizing 8 to
10 meetings a year. Each 2-hour monthly meeting is dedicated to
a particular research, policy, or methodological topic, with an
advocate (or researcher), an opponent, and a moderator. Pricing
and reimbursement policy of new health technologies and
265
methodological standards of HTA research have been discussed
at several meetings over the years.
In 2006, META was one of the main organizers of the highly
successful 6th European Congress of Health Economics in Budapest [2]. In 2007, META became the official Hungarian Chapter of
the ISPOR. ISPOR Hungary Chapter has a strong commitment to
facilitate international collaboration with other Central Eastern
European countries.
In 2007, ISPOR Hungary Chapter organized its first 1-day annual
national congress in health economics. Since 2010, the 2-day congress with more than 200 participants has an international plenary
session with invited speakers from other ISPOR regional chapters.
The achievements of the ISPOR Hungary Chapter have been
recognized by policymakers. Several former ministers and state
secretaries gave lectures at the annual health economics congress.
Moreover, in 2010, the State Secretary of Health invited META to
establish the Management and Health Economics Section of the
Professional Health Care College (the Advisory Board of the Health
Care Secretariat of the Ministry of Human Resources). META also
gained official recognition on behalf of other professional medical
societies by joining the Association of Hungarian Medical Society
(Magyar Orvostársaságok és Egyesületek Szövetsége [MOTESZ]).
In 2012, ISPOR Hungary Chapter and Eötvös Loránd University
launched a 1-week summer university course with the title of
“Implementation of HTA in CEE countries,” attended by participants from 10 countries.
In 2013, ISPOR Hungary Chapter has 114 members, with 350
professionals visiting local meetings in 2012.
Step 4: Compulsory HTA in Policy Decisions
In 2004, the Hungarian National Health Insurance Fund Administration made the cost-effectiveness and budget impact criteria
compulsory prior to granting reimbursement to new pharmaceuticals. The Ministry of Health established a Department of Health
Technology Assessment (HTA Department) at one of its background institutes for the critical appraisal of economic evaluations
submitted by pharmaceutical manufacturers. The summary
appraisal prepared by the HTA Department is taken into account
by the National Health Insurance Fund Administration in reimbursement decisions. The single HTA process for patented outpatient pharmaceutical products is described in Fig. 1.
Experience of the first 6 years of the Hungarian fourth hurdle
indicated that the quality of economic evaluations submitted in
reimbursement dossiers was rather heterogeneous. In 2009, a
working group was set up to develop a policy-relevant, detailed
Hungarian critical appraisal checklist to improve the quality of
pharmacoeconomic evaluations submitted for single HTA in
pharmaceutical reimbursement applications. The critical
appraisal checklist has been published recently [7].
In 2010, new multicriteria decision analysis was introduced
for new hospital technologies, mainly for medical devices.
Recent Activities and Further Steps
ISPOR Hungary Chapter has been growing continuously. The
society has a balanced membership structure from the academia,
public, and private sectors. It has a Young Professional Unit with
an age limit of 35 years. Young professionals are especially active
in driving changes by establishing working groups in several
policy and research areas. The chapter established a student unit
for graduate, postgraduate, and PhD students in 2012.
ISPOR Hungary Chapter has coordinated the revision of
methodological guidelines for conducting economic evaluation
of health care interventions. The revised guidelines were published in March 2013. The new guidelines indicate explicit costeffectiveness thresholds for Hungary (two to three times of gross
266
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 264–266
Fig. 1 – Pricing and reimbursement process with single HTA for patented outpatient pharmaceutical products. HTA, health
technology assessment.
domestic product per capita). According to the guidelines, these
thresholds are considered as a tool (reference point), and so they
are not used as mandatory criterion for reimbursement. In the
new guidelines, the discount rate for both costs and health gains
has been reduced from 5% to 3.7%.
Yet, health policy and major reimbursement decisions are still
not fully transparent in Hungary. The National Health Insurance
Fund Administration has implemented serious cost-containment
measures for pharmaceuticals in recent years; therefore, budget
impact has become the most important element for reimbursement decisions with mandatory financial risk-sharing agreements.
The Young Professional Unit of ISPOR Chapter Hungary has
established a working group to develop proposals for improving
the transparency and evidence base of pharmaceutical pricing
and reimbursement decisions.
Conclusions
The global economic crisis significantly influenced the Hungarian
economy. Public health care resources are highly limited, and
they are not sufficient to maintain the current health care
infrastructure and the publicly funded benefit package; therefore,
health care financing and provision have to be restructured.
There is growing pressure on policymakers to justify their major
policy decisions. The economic crisis may create an opportunity
to strengthen the evidence base of health care decision making in
Hungary [8].
ISPOR Hungary Chapter has played a crucial role in HTA
implementation. In the forthcoming period, the chapter will be most
likely to play an even more crucial role in improving the standards of
HTA implementation and facilitating international collaboration with
other countries in the Central Eastern European region.
Source of funding: The authors have no other financial
relationships to disclose.
R EF E R EN C ES
[1] Gulácsi L, Boncz I, Drummond M. Issues for countries considering
introducing the “fourth hurdle”: the case of Hungary. Int J Technol
Assess Health Care 2004;20:337–41.
[2] Boncz I, Dózsa C, Kaló Z, et al. Development of health economics in
Hungary between 1990-2006. Eur J Health Econ 2006;7(Suppl.1):4–6.
[3] Gulácsi L, Orlewska E, Péntek M. Health economics and health
technology assessment in Central and Eastern Europe: a dose of reality.
Eur J Health Econ 2012;13:525–31.
[4] Central Statistical Office of Hungary. Per capita gross domestic product
(GDP) (1995-), 2012. Available from: http://www.ksh.hu/docs/eng/
xstadat/xstadat_annual/i_qpt016.html. [Accessed November 18, 2012].
[5] EBM TUDOR 2012. Available from: http://tudor.szote.u-szeged.hu/
webeng/what/whatbe.php. [Accessed November 18, 2012].
[6] Szende Á, Zs Mogyorosy, Muszbek N, et al. Methodological guidelines for
conducting economic evaluation of healthcare interventions in
Hungary: a Hungarian proposal for methodology standards. Eur J Health
Econ 2002;3:196–206.
[7] Inotai A, Pékli M, Jóna G, et al. Attempt to increase the transparency of
fourth hurdle implementation in Central-Eastern European middle
income countries: publication of the critical appraisal methodology.
BMC Health Serv Res 2012;21:332.
[8] Kaló Z, Boncz I, Inotai A. Implications of economic crisis on health care
decision-making in Hungary: an opportunity to change? J Health Pol
Outcomes Res 2012;1:20–6.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 267–272
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
What Influences Recommendations Issued by the Agency for Health
Technology Assessment in Poland? A Glimpse Into Decision
Makers’ Preferences
Maciej Niewada, MD, PhD1,2,, Małgorzata Polkowska, MS2, Michał Jakubczyk, PhD2,3, Dominik Golicki, MD, PhD1,2
1
Department of Experimental and Clinical Pharmacology, Medical University of Warsaw, Warsaw, Poland; 2HealthQuest Consulting Company, Warsaw, Poland;
Institute of Econometrics, Warsaw School of Economics, Warsaw, Poland
3
AB STR A CT
Objective: This study aimed to evaluate the factors that are associated with positive (supporting public funding) and negative recommendations of the Agency for Health Technology Assessment in
Poland. Methods: Two independent analysts reviewed all the recommendations publicly available online before October 7, 2011. For each
recommendation, predefined decision rationales, that is, clinical
efficacy, safety, cost-effectiveness, and formal aspects, were sought,
either advocating or discouraging the public financing. In the analysis,
we used descriptive statistics and a logistic regression model so as to
identify the association between predefined criteria and the recommendation being positive. Results: We identified 344 recommendations—218 positive (62.8%) and 126 negative (37.2%). Negative
recommendations were better justified and also the comments were
less ambiguous in accordance with the recommendation (except for
clinical efficacy). In general, the specified criteria supported the
decision (either positive or negative) in 209 (60.8%), 107 (31.1%), 124
(36.0%), 96 (27.9%), and 61 (17.7%) recommendations, respectively, and
ran contrary to the actual decision in the remaining ones. Threshold
values for either cost-effectiveness or budget impact distinguishing
positive from negative recommendations could not be specified. The
following parameters reached statistical significance in logistic regression: clinical efficacy (both explicitly positive and explicitly negative
evaluations impacted in opposite directions), lack of impact on hard
end points, unfavorable safety profile, cost-effectiveness results, and
formal shortcomings (all reduced the probability of a positive recommendation). Conclusions: Decision making of the Agency for Health
Technology Assessment in Poland is multicriterial, and its results
cannot be easily decomposed into simple associations or easily
predicted. Still, efficacy and safety seem to contribute most to final
recommendations.
Introduction
independent entity playing a key role in reimbursement decision
making. The most important role of the AHTAPol is to prepare
recommendations for and support decision making by the Ministry of Health on financing health care services from the public
budget. The AHTAPol assesses and appraises all medical technologies, drugs, devices, and other services that are claiming
public funding. The role of the AHTAPol covers the assessment
and appraisal of the HTA reports including systematic review of
clinical findings, economic evaluation, and budget impact analysis, majority of which are submitted by the pharmaceutical
industry. Assessment is provided by a team of analysts and based
on the Polish HTA guidelines (first issued in 2007 and reviewed in
2009) [1]. Appraisal is completed by the Consultative Council
(transformed into the Transparency Council with the beginning
of 2012), a team of highly qualified and experienced specialists,
and the president of the AHTAPol. Final judgment is made in the
specific context of the alternative options available, social
Health technology assessment (HTA) agencies play a vital role in
the decision-making process, whether or not to reimburse given
health technologies. These agencies are expected to be guided by
medical, economic, and ethical criteria and to account for limited
resources and sometimes limited evidence regarding the profile
of assessed technologies. Therefore, there are many possible
drivers for the final decision.
The aim of the current scientific project was in general to
detect the criteria that can be considered important for the
Agency for Health Technology Assessment in Poland (AHTAPol),
and in particular to try to find the characteristics of HTA reports
that are associated with positive and negative recommendations.
The AHTAPol was established in 2005 by the Ministry of
Health as a first of its kind of institution in Central and Eastern
Europe. Since 2009, the AHTAPol is defined as a legal and
Keywords: decision making, health technology
incremental cost-effectiveness ratio, reimbursement.
assessment,
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Maciej Niewada, Department of Experimental and Clinical Pharmacology, Medical University of Warsaw,
Krakowskie Przedmieście 26/28, 00-927 Warsaw, Poland
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.05.002
268
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 267–272
consequences, health care delivery organization implications,
national health priorities, and social and ethical aspects. Based
on AHTAPol recommendations, reimbursement decisions are
made by the Ministry of Health following negotiations with
pharmaceutical industry representatives.
Recommendations issued by the AHTAPol have evolved over
time. The new types of recommendations (i.e., conditional,
temporal, combined, and others [2]) were introduced. Legal background of recommendations has also changed and currently
statements by the Consultative Council and final recommendations by the President of the AHTAPol are issued [3].
The current article can be located in the line of research
established by a classical article of Devlin and Parkin [4] and a
study by Towse [5]. Devlin and Parker analyzed past decisions
made by the National Institute for Health and Clinical Excellence
(NICE) in the United Kingdom to determine factors that were
associated with positive decisions, and in particular the threshold level for the incremental cost-effectiveness ratio (ICER). In
their study, they managed to detect that the threshold level
probably lies above approximately 35,000 GBP. The study of
Devlin and Parker motivated subsequent articles. And so,
Tappenden et al. [6] tried to identify the preferences of the
members of NICE Appraisal Committees by using a
questionnaire-based study. They concentrated more on the
ethical issues, that is, on the impact of such variables as baseline
quality of life or age of the beneficiaries. Dakin et al. [7]
introduced multinomial approach to these kinds of studies,
accounting for conditional approval by NICE.
Analysis—Data Extraction and Interpretation
Methods
Clinical criteria
Material
The analysis covered all recommendations and statements of the
Consultative Council of the AHTAPol issued following two separate regulations (the Ordinance of the Minister of Health dated
September 10, 2009, and the Act on Healthcare Services Financing
From Public Funds) and available on the official Web site of the
agency (http://www.aotm.gov.pl) before October 7, 2011. It may be
somewhat misleading that we call “a recommendation” both the
text published by the AHTAPol and the final conclusion thereof.
We do not call the latter “a decision” because this is made only by
the Ministry of Health and need not agree with the AHTAPol
recommendation. At the same time, we decided to analyze
recommendations, not decisions, because the decisions are not
accompanied by any justifications and thus would be difficult to
spot any regularities.
Only recommendations’ texts were analyzed, neither HTA reports
nor critical appraisal, which in most cases were not available on
the official Web site of the AHTAPol. For every recommendation,
the following data were extracted: medical technology being
evaluated, medical therapeutic area in which the technology
reimbursement was appealed, and the year of issuing the
recommendation. Different types of AHTAPol recommendations
(e.g., supporting or rejecting funding, conditional, temporary,
and combined) were redefined into statements of limited or
no financing technology (negative recommendations) or ones
supporting financing or increase in funding (positive recommendations).
Each recommendation was evaluated independently by two
researchers by using predefined criteria listed below (language
specialist with experience in auditing of HTA reports and HTA
specialist). Disagreements were resolved by discussion. For every
recommendation (positive or negative), it was classified whether
the final recommendation was supported or discouraged by each
criterion. Table 1 presents the data interpretation. Consistency
was found if for a particular criterion positive and negative
findings were reported and explicitly referred to support positive
and negative recommendations, respectively. For other situations, we interpreted the criterion as not reflected in the final
judgment. Following pilot analysis of the AHTAPol recommendations, clinical, economic, and formal criteria used to judge final
statements were distinguished. In a few recommendations,
rationales used to judge final statement could not be classified
as the above-listed criteria and were not defined separately.
The importance of general relative efficacy and safety over
comparators in decisions’ reasoning by the Consultative Council
was recognized in pilot analysis. Thus, the clinical criteria were
further split into three subcriteria: the efficacy (benefit over the
comparator used in the analysis—an active treatment or placebo), safety, and the impact of the technology on clinical hard
end points (which were treated separately as anticipated significant driver of clinical decision making). Hard end points were
defined following reviewed Polish guidelines for HTA as clinically
significant end points, playing an important role in a given
disease, that is, deaths, cases or recoveries, quality of life, adverse
effects (divided into serious and nonserious), or medical events
[1]. The issue of difference between the efficacy, studied in
clinical trials, and effectiveness, observed in real life settings,
was not taken up explicitly in any recommendation; thus, it was
not addressed in our analysis.
Economic criteria
Inclusion and exclusion criteria
All recommendations and statements were included with the
exception of collective recommendations for dental interventions
(covering not a single technology but a group of technologies).
Some recommendations were explained either poorly or not at all
—then the recommendation was excluded altogether.
Economic criteria were also further split into two subcriteria:
cost-effectiveness and the impact on the payer’s budget. The
evaluation of the technology’s cost-effectiveness significance was
based on values for cost per quality-adjusted life-year (QALY)
or life-year gained (LYG) reported in the recommendations
and assumed cost-effectiveness threshold of three times
the gross domestic product per capita (∼83,239 PLN) [7]. The
Table 1 – Data interpretation for predefined criteria determining the AHTAPol recommendations.
Positive recommendations
Criterion
Positive data (consistent
impact on final judgment)
Negative data (not
driving final judgment)
AHTAPol, Agency for Health Technology Assessment in Poland.
Negative recommendations
Negative data (consistent
impact on final judgment)
Positive data (not
driving final judgment)
269
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 267–272
cost-effectiveness interpretation of technology profitability by the
Consultative Council was analyzed independently of the statistical analysis of cost per QALY or LYG values reported in
recommendations.
The extent of the medical technology reimbursement impact
on the payer’s budget and its influence on the recommendations
were also analyzed. Because there was no clear definition of a
large or small impact on the budget, we could extract only the
Consultative Council perception and interpretation of budget
impact data and analyze independently estimated annual mean
payer’s expenses reported in recommendations.
Formal criteria
Another group of rationales recognized in pilot analysis, which
influenced the issuing of the positive or negative recommendation by the AHTAPol Consultative Council, was of legal or
procedural nature and encompassed failure to meet requirements of the AHTAPol for HTA reports or deficiencies of the
reports submitted to the agency, submission of reports after the
deadline, the discrepancy between expert opinions and analyses
submitted by pharmaceutical industry, submission for unauthorized (off-label) use, and so forth.
Statistical Analysis
We compared frequencies of values for binary criteria and means
for continuous ones between positive and negative recommendations. Mean values were compared with unpaired t test with
separate variance estimation. We performed multivariate analysis by using logistic regression to determine the association
between criteria values and final recommendation. In this
approach, we redefined the binary criteria to denote that a given
criterion is favorable or unfavorable for technology (rather than
agrees or not with the final recommendation). Because for many
recommendations, there was neither a favorable nor an unfavorable comment, for each criterion we introduced two binary
variables that were used in econometric modeling—denoting,
respectively, that a favorable, unfavorable comment is presented
in the recommendation text. Obviously, both these variables
could be equal to zero, but not to one for any single recommendation. While building the logistic regression model, we removed
individually variables that were not statistically significant
(α ¼ 0.05).
Results
Three hundred forty-four recommendations were analyzed. We
identified 218 positive (62.8%) and 126 negative (37.2%) recommendations. Eighty-nine recommendations addressed oncological technologies, mostly drugs. Other most common
recommendations were identified for psychiatry, cardiology,
neurology, rheumatology, and diabetes (32, 24, 20, 19, and 17
recommendations, respectively). Only 14 recommendations were
dedicated to technologies authorized for rare diseases.
Clinical efficacy, impact of hard end points, safety, costeffectiveness, and formal issues were explicitly discussed by
the Consultative Council in 238 (69.2%), 169 (49.1%), 155 (45.1%),
140 (40.7%), and 47 (13.7%) recommendations, respectively. From
this perspective, clinical issues seem to be more important than
economic ones. Exact ICER values could be found for 106 (30.8%)
recommendations, while exact budget impact estimates could be
found for 193 (56.1%) recommendations.
In general, we observed that in the recommendations texts
the elements that support the final recommendation were
explicitly stated more often. The clinical efficacy, impact of hard
end points, safety, cost-effectiveness, and formal issues have
been explicitly pointed by the Consultative Council to justify 209
(60.8%), 107 (31.1%), 124 (36.0%), 96 (27.9%), and 61 (17.7%)
recommendations, respectively. However, 29 (8.4%), 62 (18.0%),
31 (9.0%), 44 (12.8%), and 36 (10.5%) recommendations were made
against the results reported on the clinical efficacy, impact of
hard end points, safety, cost-effectiveness, and formal issues,
respectively. This data are presented in Table 2.
In addition, these numbers are presented separately for
positive and negative recommendations in Figures. 1 and 2. The
following interpretations can be made. First, Consultative Council
tries explicitly to give rationale especially for negative recommendations—the percentage of recommendations accompanied
with a comment on criterion that backs this recommendation up
is higher for negative recommendations (stacked columns higher
in Fig. 2), except for clinical efficacy. This type of reasoning is
more often presented for positive recommendations. Second, the
Consultative Council is more reluctant to present arguments
opposing the final opinion in case of negative recommendations
(stacked columns consisting mostly of the “agree” part). In other
words, in the case of positive recommendations, the Council
tones down the message more often. Again, the clinical efficacy is
Table 2 – Summary of the impact of different criteria on positive and negative AHTAPol recommendations.
Criteria
Clinical
Efficacy
Impact on hard end points
Safety
Economic
Cost-effectiveness
Budget impact
Other
Formal issues
Positive recommendations (n ¼ 218)
Negative recommendations (n ¼ 126)
n (%)*
n (%)†
n (%)‡
n (%)§
141 (64.7)
59 (27.1)
67 (30.7)
9 (4.1)
43 (19.7)
21 (9.6)
68 (54)
48 (38.1)
57 (45.2)
20 (15.9)
19 (15.1)
10 (7.9)
40 (18.3)
18 (8.3)
41 (18.8)
8 (3.7)
56 (44.4)
17 (13.5)
3 (2.4)
4 (3.2)
35 (10.2)
61 (17.7)
0
0
AHTAPol, Agency for Health Technology Assessment in Poland.
Number (%) of positive recommendations that referred directly to the favorable profile of technologies in the analyzed criterion.
†
‡
§
Number (%) of positive recommendations issued despite the explicitly referred lack of benefit (advantage) of the technology for the analyzed
criteria.
Number (%) of negative recommendations referred to unfavorable profile of the technology for the analyzed criteria.
Number (%) of negative recommendations issued despite the explicitly referred benefit (advantage) of the technology for the analyzed
criteria.
270
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 267–272
Fig. 1 – Number of references to single criteria for positive recommendations. BI, budget impact; CE, cost-effectiveness.
an exception— here, it is more uniformly in concordance with the
final recommendation for positive ones than for negative ones.
Third, the criteria that are most ambiguous for positive
recommendations (dashed and solid parts almost equal in
stacked columns) are the impact on hard end points and costeffectiveness—among all the positive recommendations that
mention these criteria in only 57.8% and 49.4%, respectively,
these criteria support the final recommendation. For example, for
clinical efficacy in 94% of the cases it is presented for a positive
recommendation, it is also favorable for the technology. What is
quite intuitive is that formal issues are almost exclusively
brought forward in a fashion unfavorable for a technology, and
much more frequently in negative recommendations. Except for
that, it is cost-effectiveness that is usually casting an unfavorable
light on the technology.
We also calculated how much the relative frequencies of
favorable or unfavorable comments differ for positive or negative
recommendations. We calculated the odds ratio (OR) for a criterion
being favorable between positive and negative recommendations
(by definition identical to the OR of a recommendation being
positive between favorable and unfavorable criterion). Clinical
efficacy changes most between two types of recommendations
(OR ¼ 53.3) followed by safety and cost-effectiveness (OR ¼ 18.2
both), and then impact on hard end points (OR ¼ 3.5). The
composition for the formal issues criterion barely differs.
More in-depths analysis of economic findings, either costeffectiveness (mean cost per QALY or LYG) or budget impact
(mean annual payer’s spending) values, showed no difference in
mean values reported for positive and negative recommendations (P ¼ 0.9045 for ICER and ¼ 0.3868 for budget impact). No
clear relation between cost-effectiveness (Fig. 3) and budget
impact (Fig. 4) values and positive and negative recommendations rates was observed. This is the first signal, before the
logistic regression, that threshold values either for costeffectiveness or for budget impact distinguishing positive and
negative recommendations cannot be specified.
Finally, a logistic regression model showed joint associations
between analyzed criteria and final recommendation (being positive; Table 3). Clinical efficacy seems to be the most important
variable—impacting both when explicitly stated to be favorable or
Fig. 2 – Number of references to single criteria for negative recommendations. BI, budget impact; CE, cost-effectiveness.
271
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 267–272
Fig. 3 – Distribution of ICER values for positive and negative
recommendations. ICER, incremental cost-effectiveness
ratio.
unfavorable. The remaining variables have predictive power only
when presented in an unfavorable fashion.
Discussion
In deciding whether or not to recommend the reimbursement of
technology from the public budget, the AHTAPol Consultative
Council seems to be considering many aspects of the problem.
Although the AHTAPol has adopted a standardized approach and
continuously publishes key documents online, in case of individual decisions, various aspects of the assessment were weighted
differently. It is still therefore a challenge to define a clear pattern
of decision making by the AHTAPol.
The most important criterion found to be used as a rationale
to back up the decisions by the AHTAPol Consultative Council
was clinical efficacy. It is called for most often in the recommendations texts (69.4%). It is illustrated by the number of
positive recommendations guided by the proved efficacy and
negative ones resulting from the lack of it, which altogether
accounted for 61% of all recommendations. Safety and costeffectiveness contributed less often to justification of the final
decisions. To be granted with positive recommendation, technology should prove unequivocal efficacy. Almost twice less often,
the impact on hard end points was supportive for the recommendations. Hard end points are hierarchical, which was
reflected in the content of the recommendations. It is likely that
the Consultative Council of the ATHAPol appreciated information
on mortality more than on the quality-of-life benefit, but it needs
further studies.
Another important criterion for reasoning the recommendations was safety for either negative or positive ones. The economic profile of the technology, that is, cost-effectiveness and
budget impact, was less often used to justify the final decisions.
The cost-effectiveness seems not to be a sine qua non criterion
when issuing positive recommendations, because in 41 cases,
even though the council referred to the reports of an unfavorable
cost-effectiveness profile, the recommendation was positive.
Such a nonstrict use of cost-effectiveness data was also reported
for NICE health technology coverage decisions [8]. NICE seems to
perceive cost-effectiveness as secondary to clinical efficacy, and
profitability is considered only if the technology has passed a
clinical effectiveness hurdle.
We failed to detect any empirical threshold value for costeffectiveness and budget impact analyses that would separate
positive and negative recommendations. Similar approach for
NICE decisions was based on modeling and produces inconsistent results [9,10]. Studies indicated that NICE makes use of some
form of cost-effectiveness threshold but expressed concern about
its basis and its use in decision making [4]. In addition, sensitivity
analysis around cost-effectiveness represents a challenge in
making it accessible to those making decisions [11].
In our study, we follow most closely the approach by Devlin
and Parker, trying to detect factors associated with positive
recommendation, and among others, to find the impact of ICER
on AHTAPol’s decisions. Both the explanatory variables and
interpretation, however, differ from above-mentioned studies.
And so we dispose over not the original reports but the recommendations issued by the AHTAPol. These recommendations
usually contain information on the interpretation by the
AHTAPol whether a technology was shown in the original HTA
report to offer benefits in each of the analyzed criteria. More
importantly, these descriptors are available only when AHTAPol
decides to include it in the recommendation text. Therefore, the
characteristics of the HTA report are often missing, and most
probably not missing in random. Therefore, we would hardly
interpret our analysis as finding predictors of AHTAPol’s decisions, but rather finding regularities in the way its decisions are
rationalized.
Table 3 – Logistic regression modeling results:
Variables associated with positive
recommendation.
Variable; only statistically
significant predictors are listed
(all P values o0.005)
Fig. 4 – Distribution of budget impact values for positive and
negative recommendations.
Clinical efficacy established
No evidence for impact on hard end
points
Lack of clinical efficacy
Safety ambiguous or unfavorable
No cost-effectiveness
Formal shortcomings
Odds ratio for positive
recommendation
12.29
0.251579
0.213248
0.170623
0.187739
0.21062
272
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 267–272
Further efforts on promoting systematic consideration of all
decision criteria and the underlying evidence are necessary. A
decision support framework can allow a consistent approach to
appraising health care interventions and evidence-based
resource allocation as shown by incorporating multiple-criteria
decision analysis in HTA to support transparent and systematic
appraisal of health care interventions [12–15].
Limitations
Our analysis was based only on recommendation texts (predominantly on recommendation justification section); thus, indirect
conclusions on recommendation determinants could not be fully
objective. It should be emphasized that the recommendations
available on the Web site of the AHTAPol do not have a unified
structure that would allow a thoroughly consistent analysis.
Although the Consultative Council certainly evaluated all the
aspects of submissions, in the recommendations’ texts, presumably selected and the most important elements determining the
decision were reported. For this reason, our evaluation of recommendations is subject to some subjectivity, which we tried to
limit as far as possible by developing a common interpretation
driven by both language and technical interpretation of the
available documents. Our analysis, because of formal grounds,
defined by available information (the text of individual recommendations issued by the Consultative Council or the opinion of the
president), could not focus on all aspects that influenced issuing a
particular opinion, but only on those that appeared explicitly in the
recommendations’ text. We analyzed not only justification sections, but all the recommendations’ document content; therefore,
some arguments not raised in the justification but emphasized as
important could be also identified and analyzed.
In assessing the technology efficacy it was taken into account
whether the Consultative Council found it satisfactory, regardless
of whether the effectiveness of the technology was compared
with placebo or an active comparator. Comparable efficacy profile
of the assessed intervention over its comparators, depending on
individual circumstances, might have been assessed by the
Council separately, that is, negatively, as an intervention that
did not add anything and represented only a burden on the
budget of the payer, or positively, as an intervention of similar
efficacy to others used for the same therapeutic indication.
Difficulties in determining the significance of a technology’s
cost-effectiveness resulted from alternative ways of expressing
the ICER, other than the cost of the QALY or the LYG, which made
interpretation practically impossible.
Conclusions
The decision-making process of the AHTAPol Consultative Council is a multifaceted and a multicriterial one, and its results
cannot be easily decomposed into simple associations nor can be
easily predicted. Still, efficacy and safety profile seem to contribute most to final recommendations.
Source of financial support: These findings are the result of
work supported by the Medical University of Warsaw, Warsaw
School of Economics, and HealthQuest Consulting Company.
The views expressed in this article are those of the authors,
and no official endorsement by the Medical University of
Warsaw or Warsaw School of Economics is intended or should
be inferred.
R EF E R EN C ES
[1] Guidelines for conducting Health Technology Assessment (HTA),
version 2.1, April 2009. Available from: http://www.aotm.gov.pl/assets/
files/wytyczne_hta/2009/09.06.29_wytyczne_HTA_eng_MS.pdf.
[Accessed December 15, 2012].
[2] Agency for Health Technology Assessment in Poland (AHTAPol). http://
www.archiwum.aotm.gov.pl/pliki/edu/AOTM%20zarz%2020.pdf.
[Accessed December 7, 2012].
[3] Załącznik nr 2 do Zarządzenia nr 20 Dyrektora Agencji Oceny
Technologii Medycznych z dnia 27 marca 2007 r.
[4] Devlin N, Parkin D. Does NICE have a cost-effectiveness threshold and
what other factors influence its decisions? A binary choice analysis.
Health Econ 2004;13:437–52.
[5] Towse A. What is NICE’s threshold? An external view. In: Devlin N,
Towse A,eds., Cost Effectiveness Thresholds: Economic and Ethical
Issues. London: King’s Fund/Office for Health Economics, 2002:
chapter 2.
[6] Tappenden P, Brazier J, Ratcliffe J, et al. A stated preference binary
choice experiment to explore NICE decision making.
Pharmacoeconomics 2007;25:685–93.
[7] Dakin HA, Devlin NJ, Odeyemi IA. “Yes”, “no” or “yes, but”? Multinomial
modelling of NICE decision-making. Health Policy 2006;77:352–67.
[8] Agency for Health Technology Assessment in Poland (AHTAPol). http://
www.aotm.gov.pl/assets/files/rada/uchwala_rk_aotm_56_16_2008_
sunitynib_Sutent.pdf. [Accessed December 7, 2012].
[9] Williams I, Bryan S, McIver S. How should cost-effectiveness analysis
be used in health technology coverage decisions? Evidence from the
National Institute for Health and Clinical Excellence approach. J Health
Serv Res Policy 2007;12:73–9.
[10] Schlander M. The use of cost-effectiveness by the National Institute for
Health and Clinical Excellence (NICE): no(t yet an) exemplar of a
deliberative process. J Med Ethics 2008;34:534–9.
[11] Williams I, McIver S, Moore D, et al. The use of economic evaluations in
NHS decision-making: a review and empirical investigation. Health
Technol Assess 2008;12:1–175.
[12] Andronis L, Barton P, Bryan S. Sensitivity analysis in economic
evaluation: an audit of NICE current practice and a review of its use and
value in decision-making. Health Technol Assess 2009;13:1–61.
[13] Tony M, Wagner M, Khoury H, et al. Bridging health technology
assessment (HTA) with multicriteria decision analyses (MCDA): field
testing of the EVIDEM framework for coverage decisions by a public
payer in Canada. BMC Health Serv Res 2011;11:329.
[14] Miot J, Wagner M, Khoury H, et al. Field testing of a multicriteria
decision analysis (MCDA) framework for coverage of a screening test
for cervical cancer in South Africa. Cost Eff Resour Alloc 2012;10:2.
[15] Goetghebeur MM, Wagner M, Khoury H, et al. Bridging health
technology assessment (HTA) and efficient health care decision making
with multicriteria decision analysis (MCDA): applying the EVIDEM
framework to medicines appraisal. Med Decis Making 2012;32:376–88.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 273–278
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
A Framework for Applying Health Technology Assessment in Cyprus:
Thoughts, Success Stories, and Recommendations
Panagiotis Petrou, MBA, PhD, PhDc, Michalis A. Talias, PhD
Healthcare Management Programme, Open University of Cyprus, Nicosia, Cyprus
AB STR A CT
Objectives: Health care decision making, assessment, and procurement of medicines is a complex, human resource–demanding, and
time-consuming process. A thorough evaluation of all factors involved
is necessary to optimize the process. The objective of this study was
to describe and analyze the current stage of health technology
assessment (HTA) in Cyprus. Methods: Literature research and private communication with all involved parties and competent authority. Moreover, data, decisions, and recommendations of the Drug’s
Committee were used. Results: Cyprus is a latecomer in this field.
HTA has entered a growing phase after the 2007 reform. It has not
reached its full potential, and the current state is applicable only to
the public sector, because of the nonexistence of a national health
system. Therefore, this poses both a great challenge and a great
barrier considering maximization of the value of money spent and
health access equity. Conclusions: There is definitely enough space
and clear necessity for further dissemination, and early successes
indicate that steps should be taken toward the introduction of an HTA
procedure that will cover both private and public sectors. The
introduction of a national health system will further enhance the
uptake of HTA, optimize the process, and use the common knowledge
strategy for evidence-based decision making.
Health Care Sector in Cyprus
2. Lack of prescribing control due to the nonexistence of an
interface management system. The system was launched in
2010, but it is still not fully operational.
3. No direct contribution of beneficiaries—Exploitation of moral
hazard.
4. Policy susceptible to colloquial evidence especially regarding
new expensive products.
5. Pharmaceuticals in the private sector are regulated only at the
price level.
6. There is a duplication of high-cost hospital services in Cyprus,
which have high running cost but are not fully utilized.
7. The above remark is augmented by the low value of the public
sector perceived by beneficiaries. This was an undisputable
finding of a recent study [6] that examined the value for
money regarding beneficiaries of the public sector. Under the
hypothesis that all health care systems want to gain more
health for the same amount of money, the perceived value of
the health system was assessed. The most important finding
is presented in Fig. 1.
8. Preventive programs are underfunded. Preventive programs
apply usually to beneficiaries, while the financial burden of
many diseases is entirely shifted to the MOH.
9. There are no quality indicators. As a result, the MOH cannot
assess any health policy, and consequently arbitrary decisions
are taken regarding abortion or carryover of them.
Currently in Cyprus, two fragmented systems run in a parallel,
overlapping, and competitive manner with clear disparities
among them: public sector and private sector. This situation is
caused by the absence of a national health system. The Ministry
of Health (MOH) is the provider, regulator, and payer of public
sector beneficiaries. Public health care is highly centralized, and
the policymaking process takes place at the macro (ministerial)
level. There are five major categories of beneficiaries [1] according
to income and employment status. It is essential to underline
that 85% of the total population is entitled to free public medical
care, without any direct or indirect contribution. As a result,
moral hazard [2] has been prominent and was expressed by
overuse and misuse of medicines in the pharmaceutical sector. In
contrast, private sector’s patients pay the full amount out of
pocket, unless they are covered by an optional private insurance.
Health care costs in Cyprus account for 6% of gross domestic
product [3], which pushes Cyprus to the European low segment.
The rate of increase in costs in the health care sector outpaces
almost all other European Union (EU) countries [4] primarily
because of the following reasons [5]:
1. An aging population that has an increasing life expectancy,
with concurrent increased morbidity.
Keywords: Drug’s Committee, evaluation, HTA, pharmaceuticals,
private sector, public sector.
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
* Address correspondence to: Panagiotis Petrou, Healthcare Management Programme, Open University of Cyprus, P.O. BOX 12794, 2252.
Latsia, Nicosia Cyprus.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.016
274
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 273–278
Fig. 1 – Estimated savings from free public health care by age per household member (as percentage of household income). For
beneficiaries to free public health care of the age group 30 to 50 years, “no perceptible benefit is realized from access to free of
charge public medical care.” This partly explains the fact that although 85% of the total population is a beneficiary of free
medical care, Cyprus has one of the highest out-of-pocket contributions in the European Union, along with the higher prices
of the private sector.
The pharmaceutical market has some unique and inherent
traits that make it quite hard to interpret and definitely tell it
apart from other products with regard to their market analysis.
The pharmaceutical market posseses an unrivalled demand and
supply feature. There is a three-tier demand structure in which
the recipient (patients) of the products consumes but has little, if
any at all, involvement in the decision-making process. Moreover, the prescriber of the product is perceived as the customer,
but does not consume the product. Another feature is that the
cost does not represent the production cost. This is quite
illustrative in generic products that have, in certain cases such
as in Italy, half of the original product’s price, and they are still
profitable. The price of the product is set to offset the research
and development expenses and in certain cases, such as in
France, is set at a premium to reward innovation.
Governments worldwide and health agencies have applied
specific and strict legislation to the pharmaceutical market to
ensure that
1. Life-saving products are available; health systems should not
be exploited by industry.
2. Good manufacturing processes are safeguarded along
the way.
3. The unique demand and supply does not hinder the control
role of health agencies regarding product availability [7].
HTA in Cyprus
Many authors have described HTA in a detailed manner [8]. In
Cyprus, HTA appeared as a term of reference of the Drug’s
Committee in early 2000 as a tool to address uncontrolled
increase in expenditure through rationalization of the decisionmaking process [9]. Terms of reference were updated and
enriched in 2007, allowing further flexibility and introduction of
more complex and legally demanding schemes. HTA is performed through the Drug’s Committee, which falls under the
MOH (Pharmaceutical Services). We must highlight the participation of Health Insurance Organization in the Stakeholder
Forum and the participation of Pharmaceuticals Services at the
Joint Action 2 of European Network of Health Technology
Assessment .
The successful use of tendering, however, led to significantly
low prices for the public sector, which distorted the need for a
sustaining and rational decision-making process (Fig. 2).
Goals of HTA in Cyprus
According to the terms of reference, HTA should reach the
following goals [9,10]:
1. Constantly upgrade, change, and improve clinical guidelines.
Currently,
guidelines
exist
in
the
majority
of
therapeutic areas.
2. Define performance indicators and assess effectiveness of
medicines.
3. Limit the use of newly launched technologies to therapeutic
areas for which there is sufficient documentation of efficacy
and safety.
4. Reevaluate high expenditure monopoly medicines that contribute disproportionately to the overall cost.
5. Categorize evidence deficit in areas in which certain technologies are destined and ways to fill this.
6. Disinvestment.
Criteria for Inclusion of a Medicine in the Formulary
The Drug’s Committee decides on the reimbursement (or not) of a
product. It assesses drug request on the basis of five main pillars:
1. Prevalence and epidemiology of the disease (prioritization of
resource allocation).
2. Comparative effectiveness according to common practice.
3. Economic evaluation, primarily budget impact analysis and to
a lesser degree substantial cost-effectiveness studies (no
inclusion of indirect data).
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 273–278
275
Fig. 2 – SWOT analysis of health technology assessment in Cyprus. Source: Authors approach. HTA, health technology
assessment; MOH, Ministry of Health; NHS, National Health System; SWOT, strength, weakness, opportunities, and threat;
WTP, willingness to pay.
4. Appraisal of medicine by other HTA agencies such as the
National Institute of Clinical Excellence.
5. Existing competitive medicines in the formulary.
The breadth and quality of data are assessed. As in other small
countries, the goal is to foster best practices instead of developing
ones. Because assessment is a context-free and context-sensitive
issue [11], transferability of data may be flawed because of
1. Demographic heterogeneity.
2. Costs. Difference in pricing, reimbursement rates, and
between the negotiating power of health prices will lead to
cost divergence between countries.
3. Health care practices/different efficiency factor between
health systems.
4. Cultural differences and social values between different populations [12].
The Committee assesses medicines on the basis of several
criteria (Table 1) [13]:
Health outcomes measures include cost per quality-adjusted
life-years, life-years gained, progression-free and overall survival,
and disease-specific measures such as Psoriasis Approach Severity
Index and American College of Rheumatology Index. The number
needed to treat approach was implemented in the assessment of
smoking cessation products. This was also implemented for
competitive medicines that have a significant price difference (e.
g., different Anatomic Therapeutic Chemical 3 categories) in order
to enhance competition in the tendering process.
In 2007, the ministerial decision [13] enabled the formation of a
therapeutic algorithm based on the outcome of the tender. Therapeutically equal products competed and instead of eliminating
the losers, these were designated as second- and third-line therapy,
respectively. This occurred in certain therapeutic categories for
which there is enough documentation that tolerance to relevant
medicines is limited and therefore, there is strong possibility that
patients may need to switch treatment. The case of the anti–tumor
necrosis factor agents was a landmark because the contribution of
all stakeholders (MOH, physicians, and patients) to the HTA process
has led to a mutually beneficial outcome, and as a result every year
the list for anti–tumor necrosis factor agents has still available slots.
Major therapeutic categories that got into this scheme include
aromatase inhibitors, adjuvant immunosuppressive treatment,
such as mycophenolic acid, antidepressants, antiepileptic agents,
and erythropoetins. Before the assessment, companies are
allowed to provide further supporting materials regarding the
efficacy and estimated cost of their products, which adds to the
transparency of the process.
Another concern of the Drug’s Committee is the possible offlabel use of expensive products, due to the overuse of medicines
in Cyprus as addressed earlier. This may lead to a reduction in
health benefits associated with the use of a product, due to the
uncertainty associated with its off-label use. Therefore, the risk of
off-label use is counterbalanced by the requirement of preapproval. This was the primary reason for the rejection of ranibizumab, despite recommendations by the National Institute for
Health and Clinical Excellence [14].
HTA implementation has not always been very successful in
broadening the scope and in certain cases overlooked one
intrinsic factor, interrelation to health policy [15]. The MOH
implemented a public campaign to create awareness among the
public for the prevention of cervix cancer; however, the Drug’s
Committee did not approve the only available vaccine. Moreover,
as the complexity factor of the therapeutic regimen increases,
such as in immunosuppressive ones, assessment can be complicated and lengthy. A prominent example was assessment of the
mammalian target of rapamycin inhibitors whose dosage is
highly personalized and budget impact of each product is
276
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 273–278
Table 1 – Criteria for assessment.
Criteria
Importance
Disease prevalence
Guidelines of the National Institute
for Health and Clinical Excellence
Efficacy data
Budget impact analysis
Off-label use
Cost-effectiveness
Impact on spending for other
medical interventions
Comments
Major
Major
Easy to assess
Transferability of data has to be checked for major
divergences
Clinical effectiveness must be assessed
May conflict with cost-effectiveness approach
Major
Dominant
Medium (unless specific trend
documented)
Major (difficult to apply for each
medicine a country-specific
study)
Medium
Difficult to assess, may compromise actual medical need.
Interface management will address this issue
A basic economic analysis is performed
Incorporation of nonpharmaceutical interventions. Interface
management may enable control. Difficult to assess
Source: Terms of Reference, Drug’s Committee.
determined by other factors, such as adjuvant immunosuppressive treatment.
Prescription Guidelines and Preapprovals
The Drug’s Committee has successfully implemented controlled
prescription of certain medicines. The majority of guidelines specify
treatment line, exceptions, patient profile, further requirements such
as serum marker levels, expected duration of treatment, and indication of failure. Statins were one of the most successful examples. The
introduction of prescription guidelines for statins (including preapproval for high-potency statins) concomitantly with the introduction
of generic simvastatin was successful in avoiding the “reallocation of
demand,” as observed in Belgium (2007–2011: Daily defined dose
consumption increased by 52%, cost decreased by 48%).
Similar guidelines were elaborated for the prescription of all
oncology medicines, insulin glargine, rosiglitazone, cinacalcet,
and darbeopetin alpha. For the majority of these products, a
preapproval is also necessary, usually with the obligation for the
submission of relevant laboratory documentation. The details of
the patients are filed.
Indication-based guidelines were elaborated for angiotensin II
receptor blockers. Different protocols were compiled for hypertension, congestive heart failure, and diabetic nephropathy.
In oncology medicines with significant uncertainty and high
cost, due to the lack of effectiveness data, the Drug’s Committee
has made exceptions for the compassionate use of cancer drugs
in a small target group population in which benefits may not be
sufficiently captured. Criteria are as follows:
1. Patient’s life expectancy is less than 24 months.
2. There is sufficient data that the treatment will extend life for
at least an additional 3 months compared with current
treatment.
3. There is no alternative treatment with equal effectiveness
available.
4. The target group is a small patient population [16].
Managed Entry Agreements in Cyprus
Risk sharing [17] has not been implemented and price volume
agreement has been applied only scarcely, mainly due because
of human resources required for monitoring. Pemetrexed gained
another indication of malignant pleural mesothelioma,
in addition to the one existing for non–small cell lung cancer.
Because of comparative effectiveness among all available
treatments for non–small cell lung cancer, an approach was
set up that consisted of three scenarios. The prices incorporated expected efficacy of the product and net benefit for the
company.
The addition of a new indication of deferasirox and the
increase in the daily dose to 40 mg/kg led to the managed entry
agreement of dose capping agreement between Novartis and
the MOH. On the basis of this agreement, the MOH reimburses
daily doses up to 30 mg/kg (average 2160 mg daily per patient)
while additional dosage burdens the company. The company is
obliged to provide free goods to the MOH, based on dose agreements. Currently, 38 patients are registered in this scheme, which
will last for 3 years, and data will be revised every 6 months to
check for deviations.
Further Potential for HTA in the New Financial Era
Cyprus has recently signed a memorandum of understanding with
Troica (term used to define the committee consisting of the International Monetary Fund, the European Union, and the European Central
Bank) to secure a life-sustaining bailout of 10 billion euros. As Troica’s
primary target is to enhance the efficiency of governance, one
important prerequisite for the health sector is the implementation
of HTA for the 10 costlier pharmaceutical and medical equipment.
This provision will further leverage shift toward an integrated HTA
system, and several approaches are currently being considered, such
as the introduction of two HTA formats, according to the estimated
budget impact (light and full version), an approach currently implemented in many countries such as The Netherlands [18].
Challenges for HTA in Cyprus
The current product mix of these two fragmented markets does
not allow dissemination of policies in the private sector, negatively affecting health equity. Moreover, the Drug’s Committee
should be authorized to act proactively regarding the assessment
of new medicines.
An important limitation is the lack of an official willingnessto-pay threshold in Cyprus [19] regarding economic evaluation.
Finally, the conflicting role of the MOH, which in the case
of HTA, it assesses, appraises, and procures as well, is a
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 273–278
major negative aspect of the system, compromising objective
appraisal.
Discussion
HTA should be incorporated in the context of a general health
policy because health policy defines the optimum use of a
medicine to deliver superior results. Health policy should cover
both sectors because ultimately all chronic patients will be
beneficiary of the public sector and prevention of discrimination
must be pursued.
Incorporation of other stakeholders (university, patient associations) may enrich the decision-making process and lead to a
context for more informative outcomes, diverting the conflicting
role of the MOH.
A detailed cost-effectiveness analysis must be implemented,
at least for the cost driver categories, which also needs to be
regulated with regard to technical parameters, such as time
horizon, modeling methods, discounting, and type of economic
evaluation (budget impact, cost-effectiveness analysis). This was
also recommended by the Internal Audit department of Cyprus
[20].
Uncertainty needs to be addressed in the context of HTA in all
its expressions because it affects the impact of the decision and
sensitivity analysis must cover a range of assumptions.
Because of the variety of statistical approaches such as
metanalysis or mixed treatment comparison, an accepted format
should be elaborated taking into consideration particularities of
certain therapeutic categories. Moreover, the rating of evidence is
crucial and we suggest that GRADE classification should be
adapted because it provides an excellent grading of evidence [21].
The scope of the assessment should be broad and must be
able to compare interdisciplinary interventions (i.e., medical vs.
pharmaceutical interventions) [22,23].
The existing coverage system does not make provision for
different levels of reimbursement, because all diseases are
considered to have the same utility. Consequently, the reimbursement levels are the same for all diseases. A classification of
strength of evidence will enable better forecasting, resource
allocation, and demand control.
We observed that reduction in the pharmaceutical growth
rate seems to have coincided with the introduction of HTA
methods; however, this has to be verified. The normalization of
the annual pharmaceutical growth rate to less than 3% is a good
first sign and at least partially is attributed to HTA implementation, which is gaining ground in Cyprus [24], as seen in Table 2.
We must however interpret this reduction in the increase rate
with caution because Cyprus’s economy entered a prolonged
recession period in 2008 [25]; therefore, this reduction can be
attributed to a tighter control of expenditures, rather than to the
ability to control expenses as such.
Table 2 – Output of Drug’s Committee.
Year
2010
2009
2008
Number of
HTAs
performed
11
24
8
Total number of appraisals
(including resubmissions)
116
116
165
HTAs, health technology assessments.
Source: Annual Report of Ministry Of Health, 2010.
277
The Drug’s committee evaluates each product only at request.
We believe that this leads to lack of symmetry of the system.
More importantly, this does not allow comparison among treatments and does not elaborate an overall approach to a disease,
which could be achieved by assessing other intervention methods. Several therapeutic interventions run in parallel and comparative effectiveness has not been documented. Periodical
assessment of guidelines and disease management should be
established and also assessment of medicines by the Drug’s
Committee would minimize reactions from the industry and
subsequent exerted pressures.
Finally, newer approaches must be incorporated taking into
account clinical uncertainty. The introduction of risk-sharing methods, price volume agreements, and managed entry agreement will
further optimize the HTA context in Cyprus. HTA is emerging in
Cyprus’s health care sector. There is an imperative need, and the
current financial era does support further dissemination. The small
size of the country and market fragmentation hinders its full uptake.
Sources of financial support: The authors have no other
financial relationships to disclose.
R EF E R EN C ES
[1] Golna C, Pashardes P, Allin S, et al. Health Care Systems in Transition:
Cyprus. Copenhagen: WHO Regional Office for Europe on behalf of the
European Observatory on Health Systems and Policies, 2004.
[2] Sachs JD. Macroeconomics and Health: Investing in Health for
Economic Development. World Health Organization WHO Library
Catalogue-in-Publication Data World Health Organization. Geneva,
Switzerland: World Health Organization, 2001.
[3] Eurostat [online]. Health care statistics, 2009. Available from: http://
epp.eurostat.ec.europa.eu/statistics_explained/index.php/
Healthcare_statistics. [Accessed September 17, 2011].
[4] Sammoutis G, Paschalides C. Will the sun shine over Cyprus’s national
health system? Lancet 2011;377:29.
[5] Antoniadou M. Can Cyprus overcome its health care challenges? Lancet
2005;365:1017–20.
[6] Andreou M, Pashardes P, Pashourtidou N. Cost and value of health care
in Cyprus. Policy paper. Nicosia: University of Cyprus, 2010. Available
from: http://www.ucy.ac.cy/data/ecorece/DOP02-10. [Accessed August
18, 2011].
[7] Kanavos P, Vandegrift M. Health policy versus industrial policy in the
pharmaceutical sector: the case of Canada. Health Policy
1997;41:241–60.
[8] Battista RN, Hodge NJ. The development of health care technology
assessment: an international perspective. Int J Technnol Assess Health
Care 1995;11:287–300.
[9] Petrou P. PHIS Hospital Pharma Report; 2009. Available from: http://
phis.goeg.at/downloads/hospitalPharma/PHIS%20Hospital%20Pharma
%20Report%202009%20Cyprus.pdf. [Accessed September 6, 2011].
[10] Ministry of Health. Pharmaceutical reform in Cyprus. 2006. Available
from: http://www.cyprus.gov.cy/moi/pio/pio.nsf/All/BEFF3C4943A9C69
6C2257127003D66A8?OpenDocument&print. [Accessed September 11,
2011].
[11] Velasco Garrido M, Gerhardus A, Røttingen JA, Busse R. Developing
health technology assessment to address health care system needs.
Health Policy 2010;94:196–202.
[12] Goeree R, He J, O’Reilly D, et al. Transferability of health technology
assessments and economic evaluations: a systematic review of
approaches for assessment and application. Clinicoecon Outcomes Res
2011;3:89–104.
[13] Drug’s Committee, Ministry of Health. Terms of reference drug’s
committee. Available from: http://www.moh.gov.cy/moh/phs/phs.nsf/
dmlphcomm_gr/dmlphcomm_gr?OpenDocument. [Accessed June 10,
2013].
[14] Appraisal of macular degeneration (age-related)—ranibizumab and
pegaptanib (TA155). 2008. Available from: http://www.nice.org.uk/
TA155. [Accessed December 17, 2011].
[15] Drummond M, Kanavos P, Sorenson C. Ensuring Value for Money in
Health Care: The Role of Health Technology Assessment in the
European Union. Denmark: World Health Organization on behalf of the
European Observatory on Health Systems and Policies, 2008.
[16] National Institute for Health and Clinical Excellence. Appraising end of
life medicines. 2008. Available from: http://www.nice.org.uk/media/
26E/43/Endoflifemedicines.pdf. [Accessed December 17, 2011].
278
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 273–278
[17] Espín J, Rovira J, García L. Experiences and Impact of European Risk
Sharing Schemes Focusing on Oncology Medicines. Grenada, Spain:
Andalusian School of Public Health, 2011.
[18] Stolk AE, de Bont A, van Halteren AR, et al. Role of health technology
assessment in shaping the benefits package in The Netherlands. Expert
Rev Pharmacoecon Outcomes Res 2009;9:85–94.
[19] Dorte Gyrd-Hansen. Willingness to pay for a QALY. Health Econ
2003;12:1049–60.
[20] General audit recommendations for rational medicines procurement. 2011. Available from: http://pdf.politis-news.com/pdf/
pdf?-A=257761,pdfview.html&-V=pdfarchivefullpaper. [Accessed
November 17, 2011].
[21] Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus
on rating quality of evidence and strength of recommendations. BMJ
2008;336:924–6.
[22] Culyer AJ, Lomas J. Deliberative processes and evidence-informed
decision making in healthcare: do they work and how might we know?
Evidence Policy 2006;2:357–71.
[23] Garrido VM, Kristensen FB, Nielsen CP, et al. Health Technology
Assessment and Health Policy Making in Europe: Current Status,
Challenges and Potentials. Denmark: World Health Organization, 2008.
[24] Cyprus Ministry of Health. Annual health report, Nicosia; 2010.
Available from: http://www.moh.gov.cy/moh/moh.nsf/page09_gr/
page09_gr?OpenDocument. [Accessed June 10, 2013].
[25] Statistical Service of the Republic of Cyprus. Latest figures: GDP growth
rate, 1st quarter 2010. Available from: http://www.cystat.gov.cy/mof/
cystat/statistics.nsf/All/ADAE80767589A7B9C22579AE0029E929.
[Accessed June 8, 2012].
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 279–283
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Systematic Review of Economic Evaluation Literature in Ghana: Is Health
Technology Assessment the Future?
Emmanuel Ankrah Odame, MBCHB, MPH
Ghana College of Physicians and Surgeons, Accra, Ghana
AB STR A CT
Objectives: In many countries, such as Ghana, there is an increasing
impetus to use economic evaluation to allow more explicit and transparent health care priority setting. An important question for policymakers
in low-income countries, however, is whether it is possible to introduce
economic evaluation data into health care priority-setting decisions.
Methods: This article systematically reviewed the literature on economic
evaluation on medical devices and pharmaceuticals in Ghana published
between 1997 and 2012. Its aim was to analyze the quantity, quality, and
targeting of economic evaluation studies that relate to medical devices and
pharmaceuticals and provide a framework for those conducting similar
health technology assessment reviews in similar contexts. Results: The
review revealed that the number of publications reporting economic evaluations was minimal with regard to medical devices and pharmaceuticals.
Conclusions: With the introduction of the National Health Insurance
Scheme since 2004 policymakers are confronted with the challenge of
allocating scarce resources rationally. Priority setting therefore has to be
guided by a sound knowledge of the costs of providing health services. The
need for economic evaluation is thus important. More costing studies were
found; there were very few cost-effectiveness analysis studies. If economic
evaluation is useful for policymakers only when performed correctly and
reported accurately, these findings depict barriers to using economic
evaluation to assist decision-making processes in Ghana; hence, there is
a need for an independent health technology assessment unit.
Background
although the principles were considered relevant by HTA producers
and users, the level of application was uniformly low.
Although resource allocation for health and demand for new
health technologies have increased in many low-income countries, robust decision-making mechanisms have not developed in
parallel. Decisions are often driven by experience, thrust of donor
agencies, and lobbying pressure [6]. For example, a report from
Peru noted that decisions on the human papilloma virus vaccine
at the local level were mainly driven by local political pressure
rather than scientific evidence [11]. In Rwanda, the government
had allocated a disproportionately large amount of funds for HIV/
AIDS than for malaria and other greater perceived needs, because
donor grants were specifically allocated for HIV/AIDS [22]. Likewise in India, sustained single-point focus on poliomyelitis
eradication using supplementary immunization (owing to World
Health Organization and global pressure) has critically weakened
the routine immunization program with other childhood vaccines [20].
Commercial pressure is also a major force skewing the
decision-making process in developing countries; this is especially
relevant for newer vaccines, expensive drugs, devices, and equipment [24]. For example, current immunization recommendations
of the Indian Academy of Pediatrics were produced by expert
consensus at a meeting sponsored by a multinational company.
Modern health care is cost and technology intensive and expects
value for money, creating demand for evidence-based practice
and health technology assessment (HTA). In high-income countries, HTA is often done in specialized HTA institutions. In
developing countries, however, HTA is often lacking, despite
the apparent need. Therefore, health care decisions are often
subjective. Improved understanding of the practice of evidencebased medicine (EBM) in many developing countries, and organizations such as the Cochrane Collaboration, now facilitate
evidence-informed decisions [5].
HTA is the scientific process of evaluating health technologies
(pharmaceuticals, vaccines, surgical procedures, medical equipment and devices, etc.) to facilitate informed decisions by stakeholders: health care providers, payers, consumers, regulators,
policymakers, and so on [1]. In high-income countries, HTA is a
formal discipline undertaken by trained professionals to guide
stakeholders, including governments, to make decisions on the
basis of sound scientific principles. Most resource-poor settings lack
formal HTA mechanisms; in such settings, health care decisions are
often based on no evidence and are more subjective. A recent
survey [9] evaluating the use of key HTA principles [2] reported that
even in the few developing countries in which HTA is being used,
Keywords: economic evaluation, Ghana, medical devices, pharmaceuticals.
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The author has indicated that he has no conflicts of interest with regard to the content of this article.
Address correspondence to: Emmanuel Ankrah Odame, Ghana College of Physicians and Surgeons, 54 Independence Avenue, Ridge,
Accra, Ghana.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.07.006
280
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 279–283
Not surprisingly, the stated objective of the recommendations was
to produce guidance for three products recently marketed by the
company in India [15].
Although poorer countries should be more careful in spending
money, the opposite often happens in most instances. Health care
systems sometimes successfully negotiate lower pricing for pharmaceuticals, but public health programs end up paying more than
the negotiated prices [16]. Such observations corroborate the
argument of Chalkidou et al. [6] that in many developing countries,
“health services and technologies purchased with public funds are
selected through idiosyncratic processes that often have little to
do with systematic analysis of their potential health benefit or
value for money.” In many developing countries, “expert-based”
guidance is used as a surrogate for robust methods, perhaps in
good faith [12]. A group of “experts” prepares a “consensus” statement on a given health technology. The procedure for selecting
the experts and the processes used to reach consensus are seldom
described [15]. In developing countries, physicians often base their
“advice” on nonscientific considerations, particularly the influence
of the pharmaceutical industry [21]. Material provided by pharmaceutical manufacturers is reported as the most frequently used
resource by many physicians, with prescribing decisions influenced by training activities sponsored by pharmaceutical companies and visits by sales representatives [21].
In Ghana, a National Health Insurance Scheme (NHIS) was
established in 2004, and the Ghana Diagnostic Related Group
provider payment mechanism has been fully implemented, and
although there are a number of challenges, the payment system
is functioning well and generally accepted by providers. It has
not, however, succeeded in containing costs, particularly for
outpatient services, with outpatient claims now accounting for
70% of total NHIS claims and 30% of total costs of the NHIS.
Furthermore, between 2007 and 2009, the average outpatient cost
per claim increased by nearly 50% from US$3.47 to US$5.06.
Without a control of the rapid rise in service delivery costs of
the NHIS, in addition to mobilization of more revenue, the
scheme will not be sustainable [29].
Meanwhile, to date there is no institution that does cost
-effectiveness analysis of the pharmaceuticals that are part of
the benefit package. There is no evidence base guiding the drugs
that are part of the benefit package.
Clearly, health care decisions by all stakeholders in Ghana are
often highly subjective. There is an urgent need to bring in
objectivity, reproducibility, and transparency. HTA as a scientific
process of evaluating health technologies (pharmaceuticals, vaccines, surgical procedures, medical equipment and devices, etc.)
to facilitate informed decisions by stakeholders—health care
providers, payers, consumers, regulators, policymakers, and so
on—can address this need. Hence, the need of this systematic
review to critically evaluate the evidence base of costeffectiveness analysis of medical devices, vaccines, pharmaceuticals, and surgical procedures.
Methods
Literature searches were carried out in November 2012 by using
the following keywords: “Ghana” and “economic evaluation” or
“cost-minimisation” or “cost-effectiveness” or “cost-utility” or
“cost-benefit.” The search was performed by using the following
databases: PubMed, EMBASE (Ovid), and Academic Search Elite
(EbscoH). It included all published and unpublished literature
available between January 1, 1997, and October 31, 2012. Inclusion
criteria were all economic evaluations on medical devices and
pharmaceuticals including vaccines.
All identified abstracts were reviewed by the first author.
Studies were excluded if they did not present both the costs
and the outcomes of a study, or if they were an editorial or
methodological article. Studies were also rejected if they were not
applied to a Ghanaian context and all other economic evaluations
apart from medical devices, vaccines, and pharmaceuticals. All
remaining articles were reviewed by using their full-text formats
and classified according to: 1) the type of evaluation, 2) the type of
intervention, and (3) the body system affected by the particular
health problem.
Published articles were grouped by type of evaluation, and
were considered to be: 1) a partial economic evaluation if only
either the costs or the outcomes of a single intervention were
compared; 2) a cost-minimization analysis if costs of different
interventions were compared with evidence of equal ease burden
in terms of disability-adjusted life-years; 3) a cost-effectiveness
analysis if health outcomes were presented in intermediate
terms, for example, disease prevented; (4) a cost-utility analysis
if health outcomes were expressed in terms of quality-adjusted
life-years or disability-adjusted life-years; and 5) a cost-benefit
analysis if health outcomes were measured in monetary units.
Only those studies that did economic evaluation in relation to
medical devices, pharmaceuticals, and vaccines were considered.
The quality of studies was measured in two different ways.
First, studies were appraised on their adherence to specific
methodological and reporting practices based on published recommendations [7,8]. These included: 1) clearly indicating the
study perspective; 2) description of comparator(s); 3) use of
discounting methods if the costs and/or outcomes were from a
study period of more than 1 year; 4) reporting the incremental
cost-effectiveness ratio (ICER) rather than the average costeffectiveness ratio; 5) performing uncertainty analysis on the
results; and 6) disclosing funding sources.
Results
A total of 50 abstracts were identified from the search of both
published and unpublished material (Fig. 1).
Of these, 45 abstracts were initially excluded because costs
and outcomes were not mentioned simultaneously and because
they disclosed funding sources. Seven articles were reviewed in
full-text format. From the review of seven full-text articles, four
articles were found not to be relevant because they were not
economic evaluations of medical devices and pharmaceuticals
including vaccines. The culmination of this was three economic
evaluations, one looked at vaccines and two on malaria
management.
In terms of where they were published, international peer
review was the source and an international person was the
principal author for two of them, with a Ghanaian author as the
principal author for one of them. All three economic evaluations
were cost- effectiveness analysis. Two of them used an ICER.
International standards recommend that economic evaluation studies should extend (through modeling) over a time period
that is long enough to capture the full costs and consequences of
an intervention. The funding sources were all from international
nonprofit organizations.
There was no significant relationship between the source of
funding and the quality of the report (using chi-square and
Fisher’s exact tests, where appropriate). See Table 1.
Discussion
HTA is the scientific process of evaluating health technologies
(pharmaceuticals, vaccines, surgical procedures, medical equipment and devices, etc.) to facilitate informed decisions by stakeholders: health care providers, payers, consumers, regulators,
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 279–283
281
50 abstracts were identified
45 abstracts were excluded
5 full articles were reviewed
2 articles were identified from references of the initially
searched articles
3 economic evaluation publications were
finally reviewed
4 articles excluded after reviewing their full
texts
Fig. 1 – Flow diagram of systematic review.
policymakers, and so on [1]. A medical device is a product used
for medical purposes in patients, diagnosis, therapy, or surgery.
Pharmaceutical products achieve their principal action by pharmacological, metabolic, or immunological means. However, medical devices act by other means such as physical, mechanical,
thermal, physicochemical, or chemical [28]. A survey of experts,
conducted by Daar et al. [29], clearly showed the greatest need of
new technology for affordable, simple diagnosis of infectious
diseases in developing countries.
From the systematic review it is evidently clear that very
scanty technology assessment has been done for medical devices
and pharmaceuticals. In Ghana, currently with the NHIS the cost
of medicines is a major cost driver, but evaluation of medicines as
part of the benefit package is done without a robust evidence
base. In this age in which there are multiple solutions and
generally a wide spectrum of possibilities and strategies for most
health care problems, the need for the evaluation of the technology and of the relevant alternative technologies available has
never been greater. Second, as the health sector has limited
budgets, HTA becomes one of the most important tools used to
contain the increasing cost of health care without compromising
safety. HTA should form the basis for health technology policies
[2]. Prioritizing resource allocation and the need for new medical
technologies are also increasing in developing countries [3].
Health insurance programs are emerging and expanding in subSaharan Africa [6]. Besides, health insurance programs are emerging and expanding more and more in this region [27]. But decisions can easily be influenced by experience, thrust of donor
agencies, and lobbying pressure for new technologies, for example, from commercial organizations. This can lead to inappropriate use of technologies, which do not address health needs, and
inefficient use of resources. Pichon-Riviere et al.’s [9] survey about
the usage of HTA methods in a resource-poor setting suggested
that principles of HTA were seen as relevant, but there was a lack
of application. South Africa, as a middle-income country, however, has planned strategies for HTA. Yet, the implementation of such a national HTA framework has been slow. The lack
of skills related to HTA is also a critical problem. Although there is
a growing base of skills in the running of clinical trials, there are
very few health economists trained in applying HTA methodologies. In Ghana, there was some training on pharmacoeconomics organized by the Ministry of Health, but the follow-up
activities have not been sustained.
The emergence and spread of EBM was expected to address
relevant needs, through building local capacity for using systematic reviews and influencing policymakers to make evidencebased decisions. However, EBM has major limitations in that it
focuses on generating evidence of efficacy alone. In addition, it
fails to factor in local needs and contexts for transferability of
evidence generated in different health care settings [5]. The
development process of HTA is usually expensive and time
consuming [6], which poses problems in many resource-poor
settings. In Ghana, as in other low- and middle-income African
countries, policymakers have in recent years come under increasing pressure to justify resource allocation decisions in the health
sector [26]. The number of economic evaluation studies in Ghana
is quite low, however, especially in the area of medical devices,
vaccines, and pharmaceuticals, in contrast to Canada, the United
Kingdom, or The Netherlands [17] where economic evaluation has
been formally accepted for use in policy decision making. In
addition, this review found that the majority of economic evaluation studies performed in Ghana between January 1, 1997, and
October 31, 2012, were vulnerable to bias because of the quality of
the evidence used. Poor reporting quality limits the usefulness of
economic assessment in the making of policy decisions.
Evidence-based policy according to scientific methods can
reduce costs and improve the outcome for patients. Such analysis
thus provides an ethical way of evaluating new health technologies. Therefore, the participation of health care authorities in
Ghana is crucial. There is an urgent need to bring in objectivity,
reproducibility, and transparency for local health policy makers
in Ghana.
The review indicates that cost-effectiveness analysis was the
only study type for economic evaluations performed in a Ghanaian setting for the study period. This is comparable to findings
in other settings [13,14], probably because the approach is
relatively straightforward. It compares costs with outcomes
measured in natural units, such as per life saved, per case
detected, and per pain- or symptom-free day.
This is in contrast to cost-utility analyses (which require more
assumptions on health-state preferences [18]) and cost-benefit
approaches (which face difficulties and controversy in applying a
monetary value to human life [19]).
Cost-effectiveness analysis can be very useful when different
health care interventions are not expected to produce the same
outcomes. This type of study alone, however, cannot handle
questions of efficiency of resource allocation when such decisions have to be made across different health problems [23].
In both technical and resource-allocation terms, there is a
need to encourage and support the undertaking of cost-utility or
cost-benefit analyses by academics and health researchers
because these evaluation types are better able to assist decision
makers in judgments about the allocation of resources across
health care programs.
In addition, this review highlights that serious attention needs
to be given to the quality of the reporting and information used in
282
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 279–283
Table 1 – Type of economic evaluation and their characteristics.
Type of evaluation by year of
publication
Type of uncertainity analysis
Use of
ICER
Discount
rate used
Cost-effectiveness analysis published
in 2010
Cost-effectiveness analysis published
in 2012
Cost-effectiveness analysis published
in 2012
Probabilistic sensitivity analysis
ICER used
3%
One-way and multiway sensitivity
analysis
One-way and multiway sensitivity
analysis
ICER used
3%
ICER not
used
3%
Funding source
International funding
agency
International funding
agency
International funding
agency
ICER, incremental cost-effectiveness ratio.
economic analyses performed in Ghanaian settings. The advantages of using systematic reviews of clinical effects are twofold
[4]. First, a more precise estimate can be obtained from combining outcome data from a number of studies. Second, by using
results from studies carried out in a range of settings (assuming
that these studies are sufficiently homogenous to be comparable), the estimate can be applied to a more general patient
population with different baseline risks, rather than specifically
to the narrow population selected for individual economic evaluations as was observed in the studies trial.
It is noteworthy that there were two serious methodological
pitfalls that were commonly found in the Ghanaian economic
evaluations. The first was the lack of calculation of an ICER in one
of the studies. This study reported an average cost-effectiveness
ratio, that is, total cost divided by total effect for the interventions
being compared. The report of an average ratio may lead to
dangerously flawed conclusions and may limit attempts to make
direct comparisons between interventions because an average
ratio implies the comparison of each alternative with an intervention that incurs no costs and no effects [10]. Without calculating and presenting ICERs, it is possible for readers to be
misguided by the results and to conclude that the new technique
was simply fourfold more expensive than the standard test.
Combined with the use of poor-quality evidence for estimating clinical effects, this could seriously undermine confidence in
the findings of these economic evaluations and their ability to
inform health care resource-allocation decisions.
Among the studies that applied discounting, the review also
found that the discount rate used was 3%. There is still no
international agreement, however, on how to deal with future
costs and benefits.
Major debate continues about whether it is justified to
discount health benefits, and if so, what discount rate should
be used and whether it should be different from that used for
monetary costs [3]. Nevertheless, this review of the literature on
economic evaluations performed in Ghanaian settings has shown
limited evidence of economic evaluation publications with regard
to medical devices and pharmaceuticals.
It is important to point out the limitations of this study. There
is no national database for health care publications in Ghana.
This study searched the literature in international databases and
only included literature published in English and also unpublished literature in some of the university campuses. The use of
these databases also meant that abstracts, conference proceedings, Ghanaian publications, masters and doctorate theses, and
articles presented at symposia or seminars were not included in
the search results. The number of publications available, however, can be used as a proxy to reflect the research capacity in
this field in Ghana, and this review has shown that the quantity,
quality, and targeting of economic evaluation studies is not yet
adequate to meet the needs and concerns of decision makers in
Ghana. Current studies support the establishment of HTA institutions in low- and middle-income countries [25].
Conclusions
This review demonstrates an urgent need for a comprehensive
and systematic method for conducting economic evaluations for
medical devices and pharmaceuticals, especially in an era of
increasing cost in the Ghanaian health care system and also
dwindling donor support to the health sector. HTA of medical
devices and pharmaceuticals seems the most viable option to
ensure cost-containment, which is the main agenda of the NHIS,
and ultimately save the health care system.
Acknowledgment
I thank Amanda Odame for her secretarial support.
Source of financial support: The author has no other financial
relationships to disclose.
R EF E R EN C ES
[1] Anonymous. Health technology assessment. Available from: http://
www.inahta.org/HTA/. [Accessed November 28, 2012].
[2] Drummond MF, Schwartz JS, Jonsson B, et al. Key principles for the
improved conduct of health technology assessments for resource
allocation decisions. Int J Technol Assess Health Care 2008;24:244–58.
[3] Cairns J. Discounting in economic evaluation. In: Drummond M,
McGuire A, eds. Economic Evaluation in Health Care: Merging Theory
with Practice. Oxford: Oxford University Press, 2001.
[4] Drummond M, Pang F. Transferability of economic evaluation results.
In: Drummond M, McGuire A, eds. Economic Evaluation in Health Care.
Oxford: Oxford University Press, 2001.
[5] Mathew JL. KNOW ESSENTIALS: a tool for informed decisions in the
absence of formal HTA systems. Int J Technol Assess Health Care
2011;27:139–50.
[6] Chalkidou K, Levine R, Dillon A. Helping poorer countries make locally
informed health decisions. BMJ 2010;341:c3651.
[7] Drummond MF, Jefferson TO. Guidelines for authors and peer reviewers
of economic submissions to the BMJ. BMJ 1996;313:275–83.
[8] Drummond M, Sculpher M, Torrance G, et al. Method for the Economic
Evaluation of Health Care Programmes. (3rd ed.). Oxford: Oxford
University Press, 2005.
[9] Pichon-Riviere A, Augustovski F, Rubinstein A, et al. Health technology
assessment for resource allocation decisions: are key principles relevant
for Latin America? Int J Technol Assess Health Care 2010;26:421–7.
[10] Drummond M, Sculpher M. Common methodological flaws in
economic evaluations. Med Care 2005;43(7, Suppl.):5–14.
[11] Piñeros M, Wiesner C, Cortés C, Trujillo LM. HPV vaccine introduction
at the local level in a developing country: attitudes and criteria among
key actors. Cad Saude Publica 2010;26:900–8.
[12] Kahveci R, Meads C. Analysis of strengths, weaknesses, opportunities,
and threats in the development of a health technology assessment
program in Turkey. Int J Technol Assess Health Care 2008;24:235–40.
[13] Lee KS, Brouwer WB, Lee SI, et al. Introducing economic evaluation as a
policy tool in Korea: will decision makers get quality information? A
critical review of published Korean economic evaluations.
Pharmacoeconomics 2005;23:709–21.
[14] Neumann P. Using Cost-Effectiveness Analysis to Improve Health Care:
Opportunities and Barriers. Oxford: Oxford University Press, 2005.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 279–283
[15] Indian Academy of Pediatrics Committee on Immunization.
Recommendations on immunization, 2008. Ind Pediatr 2008;45:635–48.
[16] Seoane-Vazquez E, Rodriguez-Monguio R. Negotiating antiretroviral
drug prices: the experience of the Andean countries. Health Policy Plan
2007;22:63–72.
[17] Oliver A. Health economic evaluation in Japan: a case study of: one
aspect of health technology assessment. Health Policy 2003;63:197–204.
[18] Coast J. Is economic evaluation in touch with society’s health values?
BMJ 2004;329:1233–6.
[19] Tan-Torres E, Baltussen R, Adum T, et al. Making CHOICES in Health:
WHO Guide to Cost-Effectiveness Analysis. Geneva: World Health
Organization, 2003.
[20] Mittal SK, Mathew JL. Polio eradication in India: the way forward. Ind J
Pediatr 2007;74:153–60.
[21] Vancelik S, Beyhun NE, Acemoglu H, Calikoglu O. Impact of
pharmaceutical promotion on prescribing decisions of general
practitioners in Eastern Turkey. BMC Public Health 2007;7:122.
[22] Republic of Rwanda, Ministry of Financial Planning, Ministry of Health.
Scaling up to achieve the health MDGs in Rwanda. 2006. Available from:
www.oecd.org/dataoecd/34/39/37759625.pdf. [Accessed May 16, 2009].
283
[23] Teerawattananon Y, Russell S, Mugford M. Systematic review
of economic evaluation literature in Thailand—are the data good
enough to be used by policy makers? Pharmacoeconomics
2007;25:467–79.
[24] Mathew JL. Pneumococcal vaccination in developing countries:
where does science end and commerce begin? Vaccine 2009;27:
4247–51.
[25] Howitt P, Darzi A, Yang G-Z, et al. Technologies for global health.
Lancet 2012;380:507–35.
[26] Doherty J, Kamae I, Lee KKC, et al. What is next for pharmacoeconomics and outcomes research in Asia? Value Health 2004;7:
118–32.
[27] Carapinha JL, Ross-Degnan D, Wagner AK. Health insurance systems in
five sub-Saharan African countries: medicine benefits and data for
decision making. Health Policy 2011;99:193–202.
[28] European Commission. COUNCIL DIRECTIVE 93/42/EEC of 14 June 1993
concerning medical devices.
[29] Daar AS, Thorsteinsdóttir H, Martin DK, et al. Top ten biotechnologies
for improving health in developing countries. Nat Genet
2002;32:229–32.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 284–289
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Dossier System as a Practical Tool for Compiling Reimbursement Lists
Maria V. Sura, MD, PhD*, Vitaly V. Omelyanovskiy, MD, PhD
National Center for Health Technology Assessment, Moscow, Russia
AB STR A CT
The article describes the procedure for preparing reimbursement lists
with the “Dossier” automated system at the regional level. Basic
advantages and characteristics of the system, procedures for filling out
the application form (dossier) for the inclusion of the drug into reimbursement lists, and the algorithm of expert evaluation are presented.
Keywords: “Dossier” automated system, health technology assessment
(HTA), reimbursement lists.
Introduction
health care system can serve as prototypes of such HTA bodies.
Unified algorithms, criteria, and decision-making rules as well as
evidence-based medicine and clinical and economic analysis
principles that were developed within the framework of the
Dossier automated system represent an important platform for
incorporating the HTA methodology in the activities of the
aforementioned expert organizations.
The project of preparing formulary lists of pharmaceutical drugs
at various levels, from formularies for particular medical organizations to federal lists, has been going on in Russia for nearly 20
years. These lists are made by different expert bodies and serve a
number of functions: provide the basis for state regulation of
prices, determination and adjustment of wholesale and retail
surcharges, drug reimbursement, and state supply orders in
health care.
At present, the system of drug selection at the federal level is
the most transparent and well regulated [1–3], while that at the
level of medical organizations is the least developed. As for the
rules and procedures for making reimbursement lists at the
regional level, in practice there are no standardized requirements
or criteria for drug evaluation: the principles and rules of drug
evaluation, decision-making criteria, and requirements concerning the information to be submitted vary to a considerable extent
(if such mechanisms exist at all) [4]. The need to standardize the
requirements for expert evaluation and submission of information about the drug, the decision-making criteria, the implementation of the principles of evidence-based medicine, and clinical
and economic analysis (pharmacoeconomics), as well as the need
to make the process of decision making for the inclusion of
particular drugs in the regional reimbursement lists more transparent, has led to the creation of an automated system called
“Dossier” for the preparation of formulary lists.
This system was originally created as a potential health
technology assessment (HTA) tool for compiling reimbursement
drug lists on the regional level. If and when necessary, however,
such a system can be easily adapted for the purpose of compiling
formulary lists on the federal and hospital/medical organization’s
levels.
Currently, there is no accredited official HTA body in Russia.
Formulary commissions working on the different levels of the
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Dossier: An Automated System for the Preparation of
Reimbursement Lists
Dossier, an automated system for preparing reimbursement lists
(hereafter referred to as “the system”), was developed in the
Research Center for Clinical and Economic Evaluation and Pharmacoeconomics of the N.I. Pirogov Russian National Research
Medical University in 2010. It is designed to help formulary
commissions of the ministries/departments of health in federal
subjects of Russia in their task of preparing reimbursement lists
of pharmaceuticals (formulary lists) on the regional level. The
online system Dossier allows the user to enter, update, store, and
evaluate information about the medicines when they are submitted for inclusion into reimbursement lists. The entire cycle of
evaluation is thus supported by the system, from the initial
application to the final decision of the formulary commission to
grant or refuse inclusion of the drug in the reimbursement list.
Three basic principles were taken into account during the
development of the Dossier system: the principle of objective
evaluation, the principle of intellectual support, and the principle
of informatization, or the information technology principle.
The principle of objective evaluation presumes a multilevel
assessment of pharmaceutical drugs that includes three main
stages: technical evaluation, scientific evaluation (by nonstaff
chief specialist of the regional Department/Ministry of Health
in different fields of medicine), and the final decision to include
(or not to include) the drug into the reimbursement list.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
* Address correspondence to: Maria V. Sura, National Center for Health Technology Assessment, POB 88, Moscow 117335, Russia.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.012
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 284–289
Technical evaluation is a preliminary stage of processing the
application (drug dossier), which includes verifying whether
the data submitted by the applicant meet all relevant requirements. Scientific evaluation is the stage of assessment by a
chief specialist. This task calls for expertise in a particular
medical field and a good working knowledge of modern therapies and management of the disease in question in real-life
clinical practice and its current financial burden. The evaluation includes an analysis of data on clinical efficacy and costeffectiveness of the drug, real-life management of patients
covered by the regional reimbursement scheme, the cost of
the new medication, and so on.
The first two stages of the evaluation take place on the Web
page of the automated system. At the third stage, members of the
formulary commission meet and discuss the issue with chief
specialists, and after their meeting, the decision of the commission is posted on the Internet page.
The principle of intellectual support refers both to the applicant
(usually members of the pharmaceutical industry) and to health
care authorities (regional ministries or departments of health,
formulary commissions). To implement this principle, the application form contains sections that encourage the applicant to
take a more rational approach to the task of determining
prioritized indications, target patient groups, and the requested
reimbursement. For example, the system asks to specify the
patient group, the number of patients covered by the regional
reimbursement scheme, the cost of medical treatment of the
disease in question, and the estimated cost to the budget of
purchasing the drug if it is included in the list for the prioritized
group. Thus, the applicant (when requesting that the drug be
included in the list), the chief specialist of the Department of
Health (when providing recommendations regarding the expediency of adding the new medication to the reimbursement list),
and the representative of the Ministry/Department of Health
(when deciding whether to include the drug in the list) focus on
the actual costs and the actual patients who are already
reimbursed for the treatment of this particular medical condition. Thanks to this approach, it becomes possible to provide a
clear justification of the need to include the new drug and the
additional financial costs that it entails, or, alternatively, of the
need to reject the application. The essential considerations that
determine whether the drug will be included in the list within
the framework of the Dossier system are the following: a
reasonable (scientifically proven) reduction in costs for the
budget, an adequate (effective and economically expedient)
deployment of the available resources, and a clear recognition
of the existing situation with regards to the provision of
medications to patients covered by the regional reimbursement
scheme.
The principle of informatization presumes that the submission,
evaluation, and scientific analysis of the dossier (application for
inclusion in the list) take place online, while the wizard presents
prompts (requirements) that guide the user in the process of
filling in the blanks. Besides, informatization of the entire process
of adding a new drug to the reimbursement list makes the final
decision more transparent because the system allows users to
track and time each stage of evaluation and decision making and
to view comments by experts. Working online makes it possible
to ensure interregional cooperation of formulary commissions,
and the involvement of experts from various fields (more than 25)
enables an interdisciplinary approach to selecting the drugs to be
included in the reimbursement lists.
At the moment of publication, the automated system consisted of four independent modules, one for each of the following
regions: Moscow region, Sverdlovsk region, Khanty-Mansi autonomous area, and Samara region. More regional modules can be
added to the system as required.
285
Registration and User Levels in the Dossier Automated
System
To start using the system, it is first necessary to choose the right
region and section and to register. Depending on the accessibility
of particular databases and the procedures used for evaluating
the submitted applications and for their revision/modification
(changing application status), there are five possible user levels:
applicant, technical expert, scientific expert, member of the
formulary commission, and regional Ministry/Department of
Health. The lowest level of access to databases of the system is
the applicant level, and the level of regional Ministry/Department
of Health is the highest.
Access to various databases of the automated system, authorization to edit/moderate applications (dossiers), menu types, and
other options that depend on the user level can be changed
(adjusted) depending on the needs of particular regions and
medical organizations.
Status of the Applications Stored in the Dossier Automated
System
Depending on the stage of evaluation, the application (dossier)
stored in the automated system is assigned a particular status
(moderation). The system specifies by whom (the name of
applicant and expert), when (time), and how (comments on the
evaluation and its result) the application was submitted and
evaluated and at which time it was assigned a particular status.
The status of an application can be modified in the “Moderation”
section. The system supports eight different statuses, for example, “under evaluation,” “passed” or failed” technical or scientific
evaluation, “approved” or “rejected” by the formulary commission, and others. The entire cycle of expert evaluation is thus
fully transparent, from the submission of the initial application
until the final decision of the formulary commission.
Application Form for Inclusion in the Regional List and Its
Completion
The application (dossier) for inclusion of the drug in the regional
reimbursement list contains more than 30 sections that may be
divided into three main blocks: general information about the
medication, or the passport block (name and pharmaceutical
group of the drug, etc.); study results, or the evidence block
(results of clinical and pharmacoeconomic studies); and the
regional specification block (the number of patients entitled to
reimbursement in the region, local cost of pharmacotherapy for
this disease, as determined by the system based on the submitted
International Statistical Classification of Diseases, 10th Revision [ICD10] code, high-priority groups of patients who will be receiving
this medication, including the number of patients, the cost of
therapy, etc.) (Table 1). The application thus contains as much
information as possible not only about the drug but also about
the current situation with supply of reimbursed medications in
the region. Submitting an application for inclusion in the list and
evaluating submitted applications, applicants, chief specialists,
and members of the formulary commission together determine
the strategy of drug promotion and its potential/optimal place
(“niche”) in the system of drug reimbursement.
Now we describe in greater detail how to fill out the application and some sections that are particularly problematic in terms
of the frequency of mistakes made by the applicants. The
sections that seldom give rise to questions and mistakes are
mentioned only briefly.
Sections 1 to 8, 10 to 13, and 30 may be called the “passport”
part of the application that provides a description of the drug: its
name, pharmaceutical form, pharmacotherapeutic group, group
of Anatomical Therapeutic Chemical classification, grounds for
286
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 284–289
Table 1 – “Dossier” automated system application form: three main blocks and their 30 sections.
Passport block
Evidence block
Regional specification block
1. International Nonproprietary
Name (INN)
14. Advantages of a new drug
2. Trade name
15. Clinical efficacy and safety
(clinical trials)
15А. Level of evidence
9. Should the drug be covered from
the regional budget under the
government resolution regulating
reimbursement at the regional
level?
17. Clinical trials at the regional level or
real-practice data
18. Priority patients’ groups (age,
severity and duration of disease,
complications)
19. Other indications (not for
reimbursement list)
20. Cost of 1-y treatment
21. The number of patients entitled for
reimbursed drugs in the region
22. Share and number of patients
eligible for prescribing of a new
drug
23. Share and number of priority
patients’ groups eligible for a
new drug
24. Current standard treatment
practice
25. Presence of drug in the Standards of
Care, Clinical Guidelines
26. Current cost of medications for
treating of patients for whom the
proposed drug is prescribed
27. Exclusion of other drugs
3. Original/generic
4. Included in other reimbursement
lists or not
5. Manufacturer
6. About the applicant
16. Pharmacoeconomics (costeffectiveness of the drug)
7. Pharmacotherapeutic group
8. Anatomical Therapeutic Chemical
(ATC) code
10. Indications for which the drug is
reimbursed
11. Dosage
12. Dosage forms
13. Course of treatment/lifelong
prescription
30. References
inclusion in the list, and so on, as well as information about the
applicant. In section 6 (“About the applicant”), it is important to
include a personal e-mail address (of the person in charge of
completing the application) without any spelling mistakes in the
address. Failure to do so will result in the applicant not being able
to stay in touch with the database and be informed about the
progress of evaluation and its results.
Section 9 asks to specify the disease (or the disease within a
particular target social category) that is the indication for prescribing the drug, in accordance with Decree № 890 of the Russian
government of July 30, 1994.
In this section, the applicant has to specify whether the drug
can be included in the regional reimbursement list for a particular social or nosological category (a particular medical condition,
independently of a social category), or both.
Section 14 of the application asks the applicant to describe the
advantages of the new drug that justify its inclusion in the
reimbursement list. The applicant should list the main features
of the medication that make it preferable to, and different from,
other medications with the same indications that are already on
the reimbursement list. Apart from clinical advantages, it is
important to specify the economic advantages of the drug in this
section (if any), as well as convenience of use (a more convenient
regimen, tablets vs. injections, etc.), good compliance with treatment, inclusion in international and Russian treatment guidelines or standards of medical care, and so on. Any advantages of
28. Cost for inclusion of a drug into the
reimbursement list for priority
patients’ groups
29. Training of physicians
the drug have to be very specific (preferably numbered) and
supported by evidence from well-designed clinical and pharmacoeconomic studies, approved standards, guidelines, and
instructions.
Section 15, “Clinical efficacy and safety,” requires filling out a
separate form (abstract) describing the clinical studies that are
referred to in this section. Priority is given to systematic reviews,
meta-analyses, and randomized controlled double-blind clinical
trials. Abstracts should contain a description of study design,
duration of follow-up, number of patients included in the study,
comparators, study results, level of evidence for the efficacy of
the drug, and so on.
At present, many applicants consciously (exaggerating the
quality of the study on purpose) or unconsciously (because of
technical errors or poor knowledge) make mistakes in this
section. The problem is that incorrect data that contradict the
clinical trials submitted for evaluation are entered in the form; for
example, randomization is claimed when it was not performed,
and a control group is mentioned where there was none, or the
level of evidence for drug efficacy is exaggerated. The data
presented in the abstract are always compared with the actual
results of the clinical trial in the course of evaluating the
application; therefore, the expert will not be led astray by such
incorrect statements.
In section 15A, the applicant has to specify the level of
evidence proving the clinical efficacy of the drug, that is, the
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 284–289
highest level achieved in clinical trials submitted for evaluation.
A list of possible levels of evidence is given in the application.
Section 16 contains data on the cost-effectiveness of the drug.
As with clinical efficacy and safety, the applicant has to fill out a
separate form—an abstract of the pharmacoeconomic study. The
abstract should include a description of study design, the type of
pharmacoeconomic analysis, effectiveness criteria, costs, effectiveness, cost/effectiveness ratio, etc. As in the clinical efficacy
and safety section, applicants often make mistakes here or
misinterpret the results of pharmacoeconomic studies. Often,
the results of these studies do not correspond to the data
included in the abstract. Thus, costs may be incorrectly described
as direct and indirect, whereas only direct costs were considered
in the study, effectiveness criteria may be described incompletely
or incorrectly, incremental cost-effectiveness ratio may be mentioned while it was never calculated in the study, and so on.
In section 17, experience in the region, the applicant has to list
all clinical trials that were conducted in medical organizations of
the region for which inclusion in the reimbursement list is
considered. In practice, however, such studies are very uncommon. Accordingly, the applicant will normally simply list all
medical organizations that have already used this drug. A
complete lack of experience using the drug in this particular
region or by the chief specialist may become the ground for
rejection of the application for inclusion into reimbursement
lists, even if the chief specialist provides a formal approval. There
have already been such cases.
In section 18, the applicant is asked to define prioritized
groups of patients who will receive the drug for a particular
indication(s) as far as this is possible (these are the patients who
will be particularly likely to need this medication if it is included
in the reimbursement list). Prioritized patient groups, or segments, may be defined by age, severity and duration of the
disease, complications, comorbidities, and so on. This section is
completed together with a chief specialist who is a member of
the formulary commission. In practice, it may be hard for the
applicant (who is usually a representative of the pharmaceutical
company that manufactures the drug) to define the prioritized
group(s) of patients. First, the applicant is not well informed
about all the characteristics of regional reimbursed patients.
Second, the applicant’s a priori position is that all patients to
whom such treatment is indicated and who are entitled to
reimbursement should be given this particular drug. This reasoning, however, fails to take into account the types of therapy
currently in use and the results of comparative analysis of the
efficacy, safety, and cost-effectiveness of alternative therapies
already on the list.
In section 19, the applicant has to indicate whether the drug
may be used for other indications than the one chosen as the
main justification for inclusion in the list (section 9 based on ICD10 and section 10 based on instructions for use). In other words, is
there a possibility that there will be a “leakage” of the drug for the
treatment of other medical conditions or forms of disease if it is
included in the list?
Section 20 asks the applicant to estimate the cost of treatment
with the new medication per patient per year. The cost should be
calculated for the categories listed in section 9 in accordance with
ICD-10. The source of price information, the day when calculations were performed, and the steps of cost estimation should be
provided in the field for text comments.
Sections 21 to 23 are devoted to quantitative analysis of
patients on the regional reimbursement list for the given ICD10 code.
The number of patients on the regional reimbursement list for
the given ICD-10 code (section 21) is determined automatically,
thanks to an inbuilt database of regional reimbursement (regionspecific), which contains information about the number of
287
prescriptions for the previous year according to the statistics of
regional ministries/departments of health. Sections 22 and 23 are
concerned with “quantitative prioritization” of patients entitled
to reimbursement. The percentage and number of patients who
may be prescribed the reimbursed drug are estimated in the
former and the percentage and number of patients who should
have priority in drug supply are estimated in the latter. Once the
percentages are entered, the number of such patients is calculated automatically: in the former case (section 22) as a share of
the total number of patients entitled to reimbursement in the
region (section 21), and in the latter case (section 23) as a share of
the total number of patients entitled to reimbursement in the
region who may be prescribed the evaluated drug (section 22).
High-priority patient groups are always determined by the chief
specialist or members of the formulary commission. The applicant, if this is a representative of a pharmaceutical company,
does not have enough competence or authority to make such
decisions.
In section 24, the applicant has to describe the existing
approaches to the management of patients with this disease in
the region. Two approaches are possible when filling out this
section. Either the chief specialist may provide this description or
(the more formal approach) a query may be made for the
appropriate ICD-10 code to collect data about the prescription of
drugs to patients entitled to reimbursement over the last year.
This information can be found in the automated system itself as
a separate database. If the second approach is used, it is
important to filter the results obtained from the database,
selecting from the list only those drugs that are directly relevant
to the therapy of this particular disease (have the same indications as the evaluated drug). The reason for this is that the
database search yields information about all the medications
dispensed to reimbursed patients under a particular ICD-10 code
(which often refers to symptomatic treatment, therapy for exacerbations or comorbidities, etc.).
In section 25, the applicant submits information about the
inclusion of the evaluated drug in Russian and foreign regulations governing drug acquisition and prescription, such as the list
of vital and essential drugs, standards of medical care, national
and foreign treatment guidelines, and so on.
Section 26 of the application describes the cost of managing
patients with the medical condition for which the evaluated drug
is indicated. This section also contains a database listing the cost
of prescribed medications for different ICD-10 codes over the
last year.
Section 27 of the application discusses the possible need to
exclude or limit the use of other medications previously included
in the regional list in case the new drug is added. This section is
particularly relevant in case of regional budget deficit and should
be completed by a chief specialist on the basis of a detailed
analysis of the use of medications in the framework of the
regional reimbursement scheme.
In section 28, the applicant has to estimate the cost
of including the evaluated drug in the list for prioritized reimbursement groups. The following formula should be used for
this calculation: the cost of treating one patient with this
medication for 1 year (s. 20) the number of patients entitled
to reimbursement in the region who have a high priority for the
provision of this medication (s. 23)
savings due to exclusion/
limited use of other medications (s. 27). This calculation should
be repeated for all conditions (ICD-10 codes) mentioned in the
application.
Section 29 describes the need for special training of physicians
concerning the prescription of the new drug in their practice.
Overall, because the new medication will be prescribed and used
in ambulatory practice, any possible difficulties with dosing,
administration, special training, and so on have to be minimal.
288
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 284–289
The applicant has to list all the sources of information used as
references in the completed application in the bibliography
(section 30). Bibliography can be arranged alphabetically or
in the order of citation and should include the author(s) of
each study (surname and initials), study/regulation title, and
publisher.
The application can be completed in several sessions. Incomplete applications can be saved as drafts and continued later at a
convenient time.
Before previewing and submitting the application, the system
performs a preliminary check. If the checkup is not successful,
the user is prompted to fill in the incorrect or missing fields.
According to experts and representatives of the regional
ministries of health, the Dossier automated system is convenient for use and allows one to make evidence-based decisions
taking fully into account all the preparation’s characteristics,
subgroup analysis results, and capacities of the local budgets. On
average, the full expert cycle of one preparation (from the
moment of the application’s submission in the system to the
formulary commission’s final decision) takes 1.5 months. It can
take longer than that if there are mistakes in the Dossier
application form (which may lead to the return of the application
for updating) and depending on the frequency of the formulary
commissions’ meetings, which make a final decision on listing
(in some regions, formulary commissions’ meetings take place
not more than once a quarter). Special trainings are arranged for
experts and potential applicants (trainings for experts are conducted separately) before the implementation of the Dossier
system in the practical activities of the regional ministries of
health. Following completion of the theoretical section of the
training, the participants are offered to work in the system
online independently, for example, to enter application and
perform evaluation of the drug.
The number of new drugs that receive positive assessment by
experts of formulary commissions and are listed may vary
significantly depending on the region, number, and composition
of people entitled to the regional pharmaceutical benefits programs, financial situation in the region as well as the backlog of
applications submitted for listing. For example, in the Moscow
region, the number of drugs that were included in reimbursement lists during 1 year did not exceed 15 names. The total
number of submissions through the Dossier system, however,
was approximately twice as large.
Algorithm of Decision Making Based on an Analysis of the
Application (Dossier) for Inclusion in the Reimbursement List
The decision to include or not to include the evaluated drug in
the list has to be reached by the chief specialist and members of
the formulary commission on the basis of a multicriteria analysis
of the data pertaining to the evaluated drug, the number of
reimbursed patients in the region, the high-priority groups to
whom the drug is indicated, the cost of therapy, and so on. Thus,
an expert making a decision has to have an intimate knowledge
of both WHAT is recommended for inclusion in the list and FOR
WHOM this drug is added to the list (as regards the number of
patients and the cost). The stages of expert evaluation of
submitted data leading to the final decision are presented in
Fig. 1.
Adherence to this algorithm, a careful consideration of all
pros and cons of adding the drug to the list, and a stepwise
evaluation will allow us to optimize the purchase of medications,
choosing from a large number of drugs suggested for inclusion in
limited lists the most expedient in terms of both clinical and
economic characteristics and the real demands of pharmacological therapy and the budget resources.
Fig. 1 – Algorithm for evaluation of the application (dossier)
for inclusion in the regional reimbursement list.
Conclusions
The creation of the Dossier automated system is an initiative of
the Research Center for Clinical and Economic Evaluation and
Pharmacoeconomics of the N.I. Pirogov Russian State Medical
University with the support of several regional ministries and
departments of health. Today, the actual implementation of this
automated system is a unique experience of only a handful of
regions. The system, however, is ready for implementation and, if
necessary, can be adjusted for use in every federal subject of
Russia.
The advantages of setting up an automated system for the
compilation of regional lists are self-evident. For instance, we can
mention the following: a formal procedure for the completion of
the application form (dossier) and expert evaluation; a stepwise
evaluation and multidisciplinary approach; decision making
based on a comprehensive, multicriteria analysis of data on the
new drug; transparency of all decisions (every stage of the
evaluation is reflected in the system, including the grounds for
every decision and the time of decision); a shift to electronic
documentation, which can save time and money (the applications can be completed, evaluated, and edited online); and an
opportunity to process data on the new medication analytically,
considering pooled data on pharmacotherapeutic and Anatomical Therapeutic Chemical groups, nosological categories, and so
on. Furthermore, the system has a significant educational component that helps the user build a logical chain to promote the
new drug on the market and helps the chief specialist to
determine whether there is a real need to add this drug to the
list, as well as to estimate the requisite purchase size and budget
expenses.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 284–289
If such a system was implemented in federal subjects,
this could provide standard algorithms of expert evaluation,
decision-making criteria, and requirements concerning the data
to be submitted with the application for inclusion in the lists. As
more data on different medications are accumulated in regional
automated systems, a single “databank” can be created with the
help of both experts and applicants that will include comprehensive information about the efficacy, safety, and economic acceptability of reimbursed drugs, as well as the cost and quantitative
characteristics of pharmacotherapies of various medical conditions.
Source of financial support: The authors have no other
financial relationships to disclose.
R EF E R EN CE S
[1] Sura M V, Omelyanovsky V V. Evolutsiya sistemy expertizy pri
formirovanii Perechnya zhiznenno neobhodimyh i vazhneyshih
289
lekarstvennyh preparatov. Meditsinskiye tehnologii. Otsenka i
vybor 2011;3(5):30–3. In Russian [Sura MV, Omelyanovsky VV. The
evolving assessment system for preparing the list of vital and
essential pharmaceuticals. Medical Technologies. Assessment and
Choice.]
[2] Prikaz Ministerstva zdravoohraneniya i sotsialnogo razvitiya Rossiyskoy Federatsii ot 27.05.2009 g. No276n “O poryadke formirovaniya
proekta Perechnya zhiznenno neobhodimyh i vazheyshih lekarstvennyh sredstv”. In Russian [Order 276n of the Ministry of Health and Social
Development of the Russian Federation of 27.05.2009, “On preparing a
list of vital and essential pharmaceuticals”.]
[3] Proekt prikaza Ministerstva zdravoohraneniya i sotsialnogo razvitiya Rossiyskoy Federatsii ot 28.04.2011 g. “O poryadke formirovaniya proekta Perechnya zhiznenno neobhodimyh i vazheyshih
lekarstven- nyh sredstv”. In Russian. [Draft order of the Ministry of
Health and Social Development of the Russian Federation of
28.04.2011, “On preparing a list of vital and essential pharmaceuticals”.]
[4] Omelyanovsny V V. Avxentyeva M V. Soldatova I G, et al. Klinicheskaya i ekonomicheskaya exper- tiza pri formirovanii perechney
lekarstvannyh sredstv. Meditsinskiye tehnologii. Otsenka i vybor
2010;1:28–31. In Russian. [Pharmaco-economic evaluation for preparing
formulary lists of pharmaceuticals. Medical Technologies. Assessment
and Choice.]
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 290–299
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Impact of the Pharma Economic Act on Diffusion of Innovation and
Reduction of Costs in the Hungarian Prescription Drug Market (2007–2010)
Rok Hren, PhD, MSc, IHP (HE)
University of Ljubljana, Ljubljana, Slovenia
AB STR A CT
Objective: In this study, we examined the impact of the Pharma
Economic Act, which was introduced in Hungary in 2007. Methods:
We used detailed data on the Hungarian prescription drug market,
which had been made publicly available by the authorities. We
evaluated the effect of the Pharma Economic Act on both dynamic
and static efficiencies and also on equity, which has been historically
a controversial issue in Hungary. We analyzed the overall prescription
drug market and statin and atorvastatin markets; as a proxy for
determining dynamic efficiency, we examined the oncology drug
market for some specific products (e.g., bortezomib) and the longacting atypical antipsychotic drugs market. Results: There is no
denying that the authorities managed to control the overall prescription drug costs; however, they were still paying excessive rents for offpatent drugs. Examples of oncology and long-acting atypical antipsychotic drugs showed that the diffusion of innovation was on percapita basis at least comparable to G-5 countries. While the share of
out-of-pocket co-payments markedly increased and the reimbursement was lowered, the concurrent price decreases often meant that
the co-payment per milligram of a given dispensed drug was actually
lower than that before the Act, thereby benefiting the patient.
Conclusions: It appears that strong mechanisms to control volume
rather than price on the supply side (marketing authorization holders)
contained the drug expenditure, while offering enough room to strive
for innovation. Making data on prescription drug expenditures and
associated co-payments publicly available is an item that should be
definitely followed by the surrounding jurisdictions.
Introduction
physicians, pharmacists, and patients) and one group on the
supply side (e.g., MAHs). If we use a simple equation that
expenditure for prescription drugs E is
In this study, we examined the impact of the Pharma Economic
Act (PEA) [1], which was introduced in Hungary in 2007. The
motivation to analyze this particular legislative measure within a
single midsize country is in its comprehensiveness and unique
approach toward marketing authorization holders (MAHs). Moreover, the recent economic crisis is making such a “laboratory of
cost-containment tools” attractive for authorities and payers not
only among the Central and Eastern European (CEE) countries but
also among Western (e.g., European Union [EU]-15) jurisdictions.
The final reason for the analysis is the availability of detailed data
on the Hungarian prescription drug market, which are provided
to the public by the authorities [2].
Controlling prescription drugs spending is fraught with difficulties, and particularly notable are examples of the GP Fundholding scheme in the United Kingdom [3–5] and prescribing drug
budgets [5–7]. The difficulties arise because of multifactorial
reasons for pharmaceutical expenditure growth, which are usually addressed only partially with national prescription drug
policies. Typically, the authorities/payers introduce two groups
of policies: one group on the so-called demand side (e.g.,
Keywords: austerity measures, cost-containment, prescription drug
spending.
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
E ¼ ∑pi Vi
where pi is the price of a given drug and Vi is its corresponding
volume, then supply-side policies target mostly the price side of
the equation, while demand-side policies work mostly on the
volume side of the equation. For example, international reference
pricing [8] has been widely embraced by authorities/payers in CEE
jurisdictions as a particular supply-side tool geared toward
regulating prices of primarily branded drugs; while such an
approach may indeed result in a short-term reduction in prescription drug expenditure, the long-term effects may be ambiguous because the volume of prescribed drugs may readily
expand.
Hungarian authorities/payers with PEA took a somewhat
different approach and included—among a plethora of other
measures—strong mechanisms to control volume on the supply
side, with MAHs (i.e., branded and generic firms) being responsible to cover any overshoot of the preagreed volume. In that
sense, PEA is unique among CEE jurisdictions; however, at its
Conflicts of interest: The author has indicated that he has no conflicts of interest with regard to the content of this article.
Address correspondence to: Rok Hren, University of Ljubljana, Jadranska 19, 1000 Ljubljana, Slovenia.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.013
291
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 290–299
conception there were serious warnings that the austerity act
may jeopardize innovation and could even lead to withdrawals of
breakthrough branded drugs [9]. Accordingly, we are here specifically evaluating the effect of the PEA on both dynamic and static
efficiencies [10–12], which present well-known and fundamental
microeconomic framework, and also on equity, which has been
historically a controversial issue in Hungary (see, e.g., [13–15]).
Throughout this article, we will present data in local currency—
Hungarian forint (HUF), with the conversion rate (as of March 28,
2013) of 1 euro (EUR) for 304.24 HUF, which means that 1 billion
HUF equals approximately 3.3 million EUR.
expenditure stood at 6% in 2006, with public pharmaceutical
expenditure of 25%; the compound annual growth rate (CAGR) in
the period 1998 to 2006 was 13%.
Although these data appear intuitively supportive of austerity
measures, it is noteworthy that population health of Hungary has
been severely lagging behind the EU-15 jurisdictions; for example, life
expectancy at birth is 73.3 years versus 80.9 years in Sweden, infant
mortality rate is 5.82 versus 2.56 in Sweden, and coronary heart
disease death rate is 169 versus 30 in France [18]. Hungary’s population
health indicators are among the worst even among CEE countries.
Cost-Containment Tools of the PEA
Policy Drivers of the PEA
It is impossible to consider the PEA outside of the wider package
of the Hungarian governmental austerity measures aimed at
reducing the then (in 2006) whopping budget deficit of 10.1% of
the gross domestic product (GDP) to the Maastricht level of 3% of
the GDP by 2010, with special focus on reducing the health care
spending by 0.9% of the GDP by 2009. While Hungary spent 8.3%
of the GDP on health care in 2006 (vs. 8.9% of the Organisation for
Economic Co-operation and Development [OECD] members' average), the prescription drug spending appeared a convenient costcontainment target because of (1) its high proportion in the total
health care expenditure (in 2006, 26.7% vs. 15.5% of the OECD
average) [16] and (2) the fifth highest growth rate in the 1998 to
2003 period among the OECD members, following Ireland, Korea,
the United States, and Australia, which all are jurisdictions with
substantially higher GDP per capita than Hungary [17].
National Health Insurance Fund-Országos Egészségbiztosítási
Pénztár (NHIF-OEP) data [2] reveal that total public health
Cost-containment measures had not started in 2007: The NHIFOEP already in 2003 mutually agreed with the pharmaceutical
industry to introduce the so-called claw-back system (firms
required to cover overspending the budget for reimbursable
drugs) and later negotiated with the industry the price freeze of
drugs in exchange for allowing new products to enter the
reimbursement list. The industry in fact under the claw-back
regime paid back to the NHIF-OEP the amount of 20 billion HUF
(66 million EUR) and 23 billion HUF (76 million EUR) in 2005 and
2006, respectively. The PEA thus built on this general claw-back
system by bringing additional broad cost-containment policies
described below to the prescription drug market.
Supply-side control—MAHs
In this review, we will focus on cost-containment policies applied
to the MAHs. The first bold measure was an introduction of 12%
statutory rebate/payback on reimbursed expenditure for both
branded and generic firms, which substantially reduced the profit
800,000
700,000
600,000
500,000
Actual
Actual minus 12% rebate
400,000
Projected /GR 6%/
Projected /CAGR 03-06/
300,000
200,000
100,000
0
2003
2004
2005
2006
2007
2003
2004
Actual
228,050 257,370
Actual minus 12% rebate
Projected /GR 6%/
Projected /CAGR 03-06/
year-on-year Actual change, including rebate
13%
Savings 07-10 = Actual minus Projected /GR 6%/
12% rebate 07-10
mio HUF
2008
2005
316,196
23%
2009
2006
364,250
364,250
364,250
364,250
15%
2010
2007
295,845
260,344
386,105
425,784
-29%
90,260
35,501
2008
310,973
273,656
409,271
497,713
5%
98,298
37,317
2009
332,929
292,978
433,828
581,794
7%
100,899
39,951
2010
350,416
308,366
459,857
680,078
5%
109,441 398,898
42,050 154,820
Fig. 1 – Public prescription drug expenditure in Hungary between 2003 and 2010 [2], shown in million HUF. Two counterfactual
curves are displayed: one for 6% annual growth rate and one for the CAGR of 16.9% for the period 2003 to 2006. CAGR,
compound annual growth rate; GR, growth rate; HUF, Hungarian forint.
292
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 290–299
400,000
350,000
300,000
250,000
Public expenditure
Public expenditure minus 12% rebate
200,000
Social welfare "co -pay"
Co-payment
150,000
100,000
50,000
0
2003
2004
2005
mio HUF
Public expenditure
Public expenditure minus 12% rebate
Social welfare "co-pay"
Co-payment
Share of co-payment as of total
prescription drug market, rebate
included
year-on-year change of co-payment
2006
2007
2008
2009
2003
228,050
2004
257,370
2005
316,196
15,789
73,695
15,893
82,412
23%
23%
12%
2010
18,320
95,956
2006
364,250
364,250
19,062
104,634
2007
295,845
260,344
18,943
119,205
2008
310,973
273,656
17,835
113,137
2009
332,929
292,978
17,986
115,782
2010
350,416
308,366
18,326
116,201
22%
16%
21%
9%
30%
14%
28%
-5%
27%
2%
26%
0%
Fig. 2 – Public prescription drug expenditure along with out-of-pocket co-payment and “co-payment” covered by the social
welfare for disadvantaged population in Hungary between 2003 and 2010 [2], shown in million HUF. HUF, Hungarian forint.
of the firms. The 12% rebate could be proportionately decreased
by the firm’s voluntary price decrease (e.g., 10% price decrease
would lead to 10.8% rebate) or could be partially or from 2010
onwards fully waived because of the firm’s local R&D investment.
As of July 1, 2011, this rebate was due to financial crisis increased
to 20% for the duration of 3 years [19].
Second, the PEA required firms to pay sales representative
registration fee in the amount of 5 million HUF (about 16,500 EUR)
per pharmaceutical sales representative per annum, thereby increasing firms’ operational costs and consequently decreasing their
profits; the registration fees remained in power in terms of monthly
fees (416,000 HUF) despite the opposition of the pharmaceutical
industry; more so, as of July 1, 2011, the fee has been increased to an
annual level of 10 million HUF (33,000 EUR). Particularly domestic
generic firms reacted to these fees by employing sales representatives as market researchers, while the branded firms reduced the
size of their sales force; the overall number of sales representatives
in Hungary fell from 2700 to 2300 because of the PEA [20].
Third, price-volume agreements based on ad-hoc negotiations
between the branded firm and the NHIF-OEP were put in place
already in 2003; on the request of the pharmaceutical industry,
these agreements have stayed confidential despite the fact that
the NHIF-OEP in principle preferred transparency. Any overspending beyond the agreed-upon expenditure (which in most cases
also includes wholesalers’ and pharmacists’ margins, and value
added tax [VAT]) would need to be paid by the firm to the NHIFOEP. Price-volume agreements were implemented on three levels:
1. a single product per 1 year;
2. a single product per 3 years (e.g., 20%, 35%, and 45% of the
3-year expenditure within the first, second, and the third year,
respectively, although the overall expenditure remained
crucial);
3. multiple products, sometimes even with multiple indications
(e.g., biologics), where the payback is calculated by the market
share of an individual drug within the group.
Price-volume agreements have been applied by the NHIF-OEP
to new entries since 2003 and essentially constituted the microcontrol of the overall prescription drug budget.
In addition to these policies controlling the expenditure, that
is, the “volume,” well-known price controls were strengthened by
the PEA, such as internal (“therapeutic”) reference pricing, quarterly revised by the two-step “iterative” bidding process of the
firms. As of July 1, 2011, the bidding has been blind, precluding
firms to resubmit their prices.
Demand-side control
Among other demand-side policies, the PEA decreased reimbursement levels from 30% to 25%, 60% to 55%, and 90% to 85%,
while to 100% reimbursement level the mandatory prescription
fee in the amount of 300 HUF (1 EUR) was attached, with the
maximum total prescription fee per patient per annum of 16,667
HUF (55 EUR). Socially disadvantaged patients (approximately 5%
of the population) had prescription fees and co-payments covered by the special welfare fund set up by the government.
Assessment of the PEA Cost-Containment Tools in Practice
Overall prescription drug market
We first examined the overall prescription drug market by simply
assessing the change in the public expenditure and co-payments
293
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 290–299
30,000
25,000
20,000
Atorvastatin
Other statins
15,000
Simvastatin
10,000
Total statin market (C10AA)
5,000
0
2006
2007
2008
2009
2010
Public expenditure for statins minus 12% rebate
mio HUF
2006
2007
Atorvastatin
9,682
8,236
Other statins
6,440
3,992
Simvastatin
9,311
6,662
Total statin market (C10AA)
25,433
18,890
year-on-year change
-26%
Simvastatin + ezetimibe
0
183
0
0
Atorvastatin + amlodipine
2008
10,943
3,018
5,162
19,123
1%
1,096
305
2009
12,698
3,719
4,110
20,528
7%
1,585
546
2010
13,247
4,766
2,849
20,862
2%
1,861
627
Fig. 3 – Public expenditure, including 12% rebate, for statins (ATC4 group C10AA) between 2006 and 2010 [2], shown in million
HUF. HUF, Hungarian forint.
made by patients and to this end, we perused NHIF-OEP data,
which are publicly available [2]. Co-payments are directly linked
to equity because of their regressive nature. Figure 1 shows public
prescription drug expenditure in Hungary between 2003 and 2010
[2]; for the sake of assessing the impact of the PEA, two counterfactual curves were derived: one for 6% annual growth rate and
one for the CAGR of 16.9% for the period 2003 to 2006, which
clearly indicates that the public expenditure on prescription
drugs was unsustainable and that reforms were indeed mandatory. The effect of the PEA cost-containment policies due to price
reductions is evident, and the cummulative amount “saved” in
the period 2007 to 2010 was 399 billion HUF (1.3 billion EUR) if we
take conservative 6% annual growth curve as the comparator. We
modeled additional savings of the NHIF-OEP by the lump sum
12% rebate because the data on specific measures had not been
available before 2010: our estimated cummulative amount in
various forms of paybacks/rebates was for 2007 to 2010 155 billion
HUF (510 million EUR) (Fig. 1), thus bringing the grand total
savings to 554 billion HUF (1.81 billion EUR). In fact, our “12%rebate” estimate turns out to be conservative if we compare for
2010 our lump sum 12% rebate of 42 billion HUF with publicly
available 2010 figures [2], which indicate that the overall OEPNHIF revenue amounted to 50.4 billion HUF (165 million EUR): The
12% rebate was actually 30.5 billion HUF (due to reductions
arising from R&D investment discussed above), the total sales
representative fees were 11.8 billion HUF, the wholesaler fees
were 0.55 billion HUF, the price-volume agreements for products
totaled 8.8 billion HUF, and there was no overall claw-back.
The public expenditure for prescription drugs decreased by
18.8% in 2007 versus 2006 or by 29% (Fig. 1) if we take into account
our lump sum 12%-rebate estimate; our calculation is close to the
Business Monitor International report [21], which appraised that
in 2007 versus 2006 the NHIF-OEP prescription drug expenditure
contracted by 30.5%. As shown in Figure 1, the prescription
Table 1 – Public expenditure along with out-of-pocket copayment for statins (ATC4 group C10AA) between 2006
and 2010 [2].
Total statin market
Public expenditure
Social welfare “co-pay”
Co-payment
Total market, no rebate
Public expenditure minus 12% rebate
Share of co-payment as of total statin drug market, rebate included (%)
Year-on-year change of co-payment (%)
Note. Values are in million HUF unless otherwise indicated.
HUF, Hungarian forint.
2006
2007
2008
2009
2010
25,433
13
3,789
29,235
25,433
13
21,466
354
4,879
26,699
18,890
20
29
21,731
464
4,639
26,834
19,123
19
5
23,327
551
5,560
29,438
20,528
21
20
23,707
647
6,373
30,727
20,862
23
15
294
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 290–299
25,000
20,000
Public expenditure
15,000
Projection /trend 04 - 06/
Social welfare "co -pay"
10,000
Co-payment
Public expenditure minus 12% rebate
5,000
0
2003
2004
2005
2006
2007
2008
2009
2010
atorvastatin
mio HUF
Public expenditure
Projection /trend 04-06/
Social welfare "co-pay"
Co-payment
Public expenditure minus 12% rebate
Share of co-payment as of total
atorvastatin market, rebate included
year-on-year change of co-payment
2003
2,267
2004
3,685
2005
6,496
0
908
0
612
0
845
29%
14%
-33%
12%
38%
2006
9,682
9,682
6
1,492
9,682
2007
9,359
12,681
126
1,836
8,236
2008
12,435
15,679
217
2,362
10,943
2009
14,430
18,678
317
3,402
12,698
2010
15,053
21,677
389
3,935
13,247
13%
77%
18%
23%
17%
29%
21%
44%
22%
16%
Fig. 4 – Public expenditure along with out-of-pocket and social welfare co-payments for atorvastatins (ATC5 group
C10AA05) between 2006 and 2010 [2], shown in million HUF. A counterfactual curve based on the trend in the period 2004 to
2006 is displayed. HUF, Hungarian forint.
expenditure grew moderately in the period between 2007 and
2010, with a CAGR of 5.8%. The major reason for the 18.8%
decrease was extensive price cuts of both branded and generic
drugs and an increase in co-payment of patients (see below). The
price cuts throughout 2007 were truly extensive; in April 2007
alone, the prices of 1000 drugs were reduced on average by 16%
[2]. While the overall claw-back on total public expenditure
remained in place even after 2007, it has never been realized: it
seems that all other measures in the view of the NHIF-OEP
sufficiently contained the costs.
Figure 2 shows the public prescription drug expenditure along
with out-of-pocket co-payment and “co-payment” covered by the
social welfare for disadvantaged population in Hungary between
2003 and 2010. The out-of-pocket co-payment increased by 14%
in 2007 versus 2006, and with public expenditure falling, its share
of the total prescription market surged sharply (30% in 2007 vs.
21% in 2006). As shown in Figure 2, the absolute increase in copayments in 2007 was around 15 billion HUF (50 million EUR),
which is less than that estimated by the Business Monitor
International report [21], putting the range at 20 to 50 billion
HUF. An increase in co-payment was due to two factors: (1)
decreasing the reimbursement level for a number of drugs and (2)
a 300-HUF prescription fee. Co-payment was relatively stable
from 2008 onwards and even slightly declined both in absolute
value and in its relative share; in 2010, co-payment amounted to
26% of the total prescription drug expenditure, which is still a
substantial figure [2].
Social welfare “co-payments” remained stable since 2005,
being around 4% of the total prescription drug expenditure [2].
Opposition against co-payments is considerable in Hungary, and
in 2008 public with large majority rejected on the referendum copayments for physicians’ visits and in-patient stays although copayments for prescription drugs were not voted on. In the survey
of 2007, 11% of the people stated that prescription drug fees of 300
HUF, which are capped at an annual maximum of 16,667 HUF (55
EUR), would make otherwise fully reimbursed drugs unaffordable
[15]; recent local data (in the period from April 2010 through
March 2011) however showed that 97% and 75% of the population
pay less than 6,667 HUF (22 EUR) and 1,667 HUF (5.5 EUR) per
annum, respectively.
Specific prescription drug markets
In the previous section, the NHIF-OEP data provide strong evidence that the PEA resulted in cost-containment of the public
expenditure for prescription drugs in Hungary. Cost-containment
itself, however, may not lead to efficiencies either in the prescription drug market, let alone in the overall health care market. As is
well known, cost-containment may stifle innovation, thereby
reducing the dynamic efficiency; however, price reductions of
off-patent drugs and their generic versions may be insufficient,
allowing their prices to remain way above the marginal cost of
production and thus promoting static inefficiencies [11,22,23].
Statins (Anatomical Therapeutic Chemical [ATC4] group
C10AA)
To assess static efficiency, we examined a group of statins (ATC4
group C10AA), and within this group in more detail atorvastatins
(ATC5 group C10AA05). Statins are suitable proxies for highly
genericized therapeutic group at the time of introduction of
the PEA.
Figure 3 shows the public expenditure for statins, taking into
account the 12% rebate. Similarly as the overall market, the statin
market contracted by 26% in 2007 (vs. 2006) and grew modestly at
the CAGR of 2.9% in the period 2008 to 2010. Statins accounted for
8.0% of the total public expenditure in 2006 and 6.3% in 2010. In
2006, simvastatin and atorvastatin were the dominant molecules,
about equal in public expenditure and jointly accounting for
about 75% of the total statin market. From 2007 onwards, we have
witnessed a clear shift in prescribing habits of physicians as
atorvastatin became the leading molecule, reaching a market
share of 63% of the entire C10AA market by 2010. The decline in
295
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 290–299
2,500,000
2,000,000
1,500,000
10 mg
20 mg
40 mg
1,000,000
80 mg
500,000
0
2003
2004
2005
2006
2007
2008
2009
2010
atorvastatin - packs-per-30-tablets equivalents
2003
10 mg
316,913
20 mg
41,473
40 mg
52,610
80 mg
0
Total mg
183,089,700
year-on-year change - total mg
2004
53,171
22,926
284,359
0
370,937,700
103%
2005
2006
2007
2008
2009
2010
102,943
217,556
275,965
393,193
456,090
460,146
165,500
499,472
730,063
1,331,798
1,773,735
1,935,304
449,340
637,291
773,674
1,194,641
1,588,795
1,755,123
13,167
32,139
62,320
80,226
97,807
97,099
700,991,600 1,206,832,700 1,598,804,100 2,543,148,300 3,342,358,800 3,638,411,400
89%
72%
32%
59%
31%
9%
Fig. 5 – Number of units (in per-30-tablets equivalents) and total milligrams of atorvastatin dispensed in Hungarian
pharmacies between 2003 and 2010 [2].
simvastatin is notable even if we add the uptake of combination
of simvastatin and ezetimibe, which lies outside the C10AA
ATC group.
Table 1 outlines the public expenditure for statins along with
out-of-pocket and social welfare co-payments for the period 2006
to 2010. As in the total prescription drug market, the share of outof-pocket co-payment increased from 13% in 2006 to 20% in 2007,
while in the absolute amount, it increased by 29%. The hike was
mostly due to higher co-payments and the unwillingness of
patients to switch to cheaper substitution: study of GfK Hungaria
noted that “only one tenth of population chooses their medication based on price” and while general practitioners offered
cheaper alternatives to 61% of their patients, only 30% of the
patients actually agreed to the substitution [21].
years among the top 10 leading molecules. In 2010, atorvastatin’s
share of the total prescription market was 4.3%. Figure 4 shows
the public expenditure for atorvastatin along with out-of-pocket
and social welfare co-payments for the period 2003 to 2010. A
counterfactual curve was constructed on the basis of average
year-on-year added expenditure between 2004 and 2006: we
estimate that in 2007 and 2008 alone, the PEA measures saved
9.2 billion HUF (30.6 million EUR). Out-of-pocket co-payment
increased steadily since 2004, with the share of co-payment
representing 12% to 14% of the total atorvastatin market. Similarly as in the total and statin market, the co-payment share
increased to 18% in 2007 from 13% in 2006.
So far, we have not related contracting of public expenditure
to the trend in the number of atorvastatin units and one could
intuitively argue that the lower expenditure simply resulted from
patients failing to fulfill their prescriptions or patients covering
the difference by out-of-pocket co-payment. Yet, Figure 5 shows
that the total milligrams of atorvastatin issued at pharmacies
grew faster (þ32%) in 2007 versus 2006 than either public
Atorvastatin (ATC5 group C10AA05)
Among statins, the most important single molecule is atorvastatin, which has been in the Hungarian prescription market for
Table 2 – Average out-of-pocket co-payment per milligram and public expenditure per milligram (with 12%
rebate included) in Hungary for all strengths (10 mg, 20 mg, 40 mg, and 80 mg) of atorvastatin between 2003 and
2010 [2].
Co-payment per milligram (HUF)
Year-on-year change (%)
Public expenditure per milligram (HUF), 12% rebate included
Year-on-year change (%)
HUF, Hungarian forint.
2003
2004
2005
2006
2007
2008
2009
2010
5.0
1.6
67
9.9
20
1.2
27
10.3
7
1.2
3
8.5
13
1.1
7
4.1
36
0.9
19
3.3
16
1.0
10
2.9
12
1.1
6
2.8
4
12.4
296
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 290–299
Prices (in HUN) - atorvastatin 40 mg x 30
8000
7000
6000
Off-patent originator
5000
Generic #1
Generic #2
4000
Generic #3
Generic #4
3000
Generic #5
2000
Generic #6
1000
0
Jan-07
Jan-08
Jan-09
Jan-10
Jan-11
Fig. 6 – Prices (in HUF) of atorvastatin 40 mg (30 tablets) for an off-patent originator and various generic bioequivalent
alternatives between January 2007 and January 2011 [2]. HUF, Hungarian forint.
Table 3 – ATC5 market share of off-patent atorvastatin originator in units per strength and in public
expenditure [2].
2003
2004
2005
2006
2007
2008
41
24
92
81
21
16
64
69
0.01
0.002
27
50
0.001
14
50
53
16
Off-patent originator ATC5 market share (%) in units
10 mg
100
100
20 mg
100
99
40 mg
100
100
80 mg
Off-patent originator ATC5 market share (%) in public expenditure
All strengths
100
100
88
9
2009
2010
9
42
7
37
6
5
Table 4 – Number of drug firms marketing various atorvastatin strengths in the Hungarian prescription drug
market between 2003 and 2010 [2].
10
20
40
80
mg
mg
mg
mg
2003
2004
2005
2006
2007
2008
2009
2010
1
1
1
0
2
2
2
0
4
4
5
3
7
8
6
3
8
8
9
4
9
10
11
4
10
11
12
4
13
13
16
4
expenditure (15%) or out-of-pocket co-payment (þ23%). The
trend continued in 2008, with total milligrams increasing by 59%,
public expenditure by 33%, and co-payment by 29%. In other words,
this means that both the NHIF-OEP and patients were paying less
per milligram than before the implementation of the PEA, as is
corroborated by Table 2. Table 2 reveals two important additional
trends, though: since 2009, out-of-pocket co-payment per milligram started increasing again, while the decrease in public expenditure per milligram (with included 12% rebate) markedly slowed
down. Both trends clearly show the unwillingness of the off-patent
branded firm and generic firms to further reduce prices.
Figure 6 substantiates this claim by explicitly showing an
example of the price dynamics for atorvastatin 40 mg during the
period following the implementation of the PEA. The generic
paradox [24,25] could not apply here because the off-patent
branded drug would be delisted if its price was increased,
although the branded drug does retain some premium over
generics as has been noted elsewhere [11,26]. The premium of
off-patent originator appears a logical choice in the Hungarian
market because of patients’ relative price insensitivity. Lingering
presence of the off-patent originator and price convergence of
generics, however, strongly suggest that both the NHIF-OEP and
patients may be financing rents to which neither the branded firm
nor generic firms are entitled. But can we estimate the price that
would be close to the marginal cost of production of atorvastatin
without having direct access to this proprietary information?
Table 3 shows the ATC5 market share of the off-patent
atorvastatin originator, which encountered the first generic
entries toward the end of 2004 and rapidly lost the share in 10mg and 20-mg strengths. As indicated by Table 4, the number of
entries grew larger with each year and at least formally formed
the conditions for the perfectly competitive market, which
should drive the prices toward the marginal cost of production [27,28]. In Table 2, we can deduce that the average price
297
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 290–299
Fig. 7 – Total milligrams of long-acting atypical
antipsychotics dispensed in the Hungarian prescription
drug market between 2005 and 2010 [2].
Fig. 8 – Number of units of bortezomib dispensed in the
Hungarian prescription drug market between 2004 and 2010
[2].
(equaling public price plus co-payment) per milligram of the then
branded atorvastatin with monopoly position in the market was
between 11.5 HUF (in 2004) and 17.4 HUF (in 2003). Small
molecules launched in 1990s had—as a general rule of thumb—
cost of goods sold (COGS) typically at 10% (or even lower) of
their monopoly price. From this rather crude, but unlikely inaccurate, estimate, the price between 1.15 HUF per mg and 1.74 HUF
per mg should be comfortable enough and should even retain
some considerable rent for generic firms. Table 2 reveals that in
2010 the average price per mg was 4.7 HUF, about three times
higher than the estimate above. On the basis of this estimate, we
may conclude that the NHIF-OEP could easily save two thirds of
13.2 billion HUF (i.e., 8.8 billion HUF—29 million EUR) in its annual
expenditure for atorvastatin and patients could save two thirds of
their annual out-of-pocket co-payments of 2.9 billion HUF (i.e., 1.9
billion HUF—6.3 million EUR).
controlled nature of growth, the long-acting atypical antipsychotic achieved in Hungary diffusion that was globally surpassed
only by Spain; in April 2011, its market share in the total
antipsychotic market (N05A) was 23.7% (vs. 26.7% in Spain,
17.1% in France, 13.4% in Italy, 10.9% in Germany, and 6.8% in
the United Kingdom) [35].
Co-payment surged in 2007, and the absolute amount has
been with an average monthly dose of 75 mg around 525 HUF (1.7
EUR), which is likely not excessive.
Long-acting injectable risperidone (ATC5 group N05AX08)
To examine dynamic efficiency, we first analyzed the long-acting
injectable risperidone (administered once every 2 weeks), which
represents a line extension of the oral risperidone, thus addressing the issue of adherence of patients suffering from schizophrenia [29–32] and associated higher hospitalization rates and costs
[33,34]. The MAH/branded firm entered in 2005 into a pricevolume agreement with the NHIF-OEP, which was renewed in
2008. The controlled nature of growth is obvious from the
examination of total milligrams dispensed (Fig. 7) and public
expenditure including 12% rebate (Table 5). Because the NHIF-OEP
did not resort to international reference pricing once the branded
drug had entered the market, the price per milligram remained
unchanged until the 12% rebate was mandated. In spite of the
Oncology drugs (ATC5 groups L01XX32 and L01XE01)
Oncology drugs are good proxies for examining dynamic efficiency; here, we will consider bortezomib (ATC5 group L01XX32).
As with the long-acting atypical antipsychotic, its MAHs/branded
firm entered into price-volume agreements with the NHIF-OEP in
2005 (Figure 8). Bortezomib’s consumption per capita was in 2010
in Hungary on par with that in Italy and the most developed CEE
jurisdiction Slovenia, and although lower than in Spain, France,
and Germany, it was markedly higher than in the United Kingdom [36].
Since 2007, co-payment amounted to 300 HUF per prescribed
3.5-mg bortezomib vial (Table 6). As with the long-acting atypical
antipsychotic, the absence of international reference pricing
implies that public expenditure per milligram changed only in
2007 with the introduction of the 12% rebate.
Discussion
To our knowledge, this is the first study to systematically
evaluate the impact of the PEA in the period between 2007 and
Table 5 – Public expenditure along with out-of-pocket co-payment for the long-acting atypical antipsychotic
dispensed in the Hungarian prescription drug market between 2005 and 2010 [2].
2005
2006
2007
2008
2009
2010
Public expenditure
Social welfare “co-pay”
Co-payment
Total market
Public expenditure minus 12% rebate
1,093
0
0
1,093
3,168
0
0
3,168
3,633
1
24
3,658
3,197
4,067
3
26
4,096
3,579
4,546
4
27
4,578
4,001
4,950
5
29
4,984
4,356
Public expenditure per milligram (HUF), rebate included
1,133
1,128
984
983
983
983
0.3
0.0
7.2
7.1
6.7
6.4
Co-payment per milligram (HUF)
Note. Values are in million HUF unless otherwise indicated.
HUF, Hungarian forint.
298
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 290–299
Table 6 – Public expenditure along with out-of-pocket copayment for bortezomib dispensed in the Hungarian
prescription drug market between 2005 and 2010 [2].
Public expenditure
Social welfare “co-pay”
Co-payment
Total market
Public expenditure minus 12% tax
Public expenditure per milligram (HUF), rebate included
Co-payment per milligram (HUF)
2005
2006
2007
2008
2009
2010
366
0
0
366
1,551
0
0
1,551
1,729
0.0
1.6
1,731
1,521
1,934
0
2
1,936
1,702
2,303
0
2
2,305
2,027
2,562
0
2
2,565
2,255
88,227
87,792
77,186
77,182
77,182
77,016
20.0
0.0
79.4
84.0
84.8
85.1
Note. Values are in million HUF unless otherwise indicated.
HUF, Hungarian forint.
2010 and the results of our study could have potential implications for prescription drug policy in Hungary and other jurisdictions within CEE and EU-15 regions.
There is no point denying that the NHIF-OEP managed to
control the prescription drug costs by implementing the PEA in
January 2007. The PEA caused a substantial decrease in prices of
off-patent drugs and imposed a 12% rebate on pharmaceutical
firms. While the share of out-of-pocket co-payments markedly
increased and the reimbursement was lowered, the concurrent
price decreases often meant that the co-payment per milligram
of a given dispensed drug was actually lower than that before the
act, thereby benefiting the patient. We also have no reason to
believe that the cost-containment policies were excluding innovation: examples of long-acting atypical antipsychotic and bortezomib suggest that the Hungarian prescription market under the
PEA has enabled diffusion of new technologies, on per-capita
basis comparable to those of G-5 countries.
The NHIF-OEP set up before the PEA an excellent information
technology system for tracking the number of prescription drugs
dispensed and associated public expenditure and co-payments;
while most of the jurisdictions in the EU have in place electronic
prescribing systems, the analysis provided to the public by the
NHIF-OEP is exemplary in the CEE region and maybe even wider.
With such accolades, it is almost impossible to resist the
temptation in adopting the NHIF-OEP “model” all over Europe.
The PEA, however, has some serious shortcomings, which are
promoting inefficiencies. An obvious inefficiency is that the PEA
is still financing excessive rents for off-patent drugs, which
remain a global challenge for payers [22]. The Hungarian offpatent market reinforces in particular the general notion that
regulating generic prices may lead to price convergence and
inefficiencies [7,37] and that generic competition is not sufficiently fierce in a regulated environment.
It may well be that the NHIF-OEP also needed to make in
design of its policies “compromises” with the local pharmaceutical industry (e.g., Gedeon-Richter), which are significant local
employers and contributors to local taxes. As an example, NHIFOEP’s continued efforts of international-nonproprietary-name
prescribing were blocked by the dominant generic player
Gedeon-Richter.
A 12% rebate (currently 20% rebate and for drugs older than 5
years 30% rebate) imposed on the pharmaceutical industry
deserves a comment. This was indeed a bold move that could
lead to a serious unintended consequence of branded firms
leaving the Hungarian market en masse as their profits and
particularly profit versus net-trade-sales ratios plummeted way
below the expectations of their shareholders. The fact that
branded firms stayed behind indirectly speaks for the relatively
low production costs of prescription drugs, which is also the
reason why protection of intellectual property rights is needed
[38,39]. In this way, the NHIF-OEP not only collected financial
rebates but also optimized social welfare via the optimization of
payments for drugs. Overall, NHIF-OEP’s 12% or even 20% rebate
could be interpreted as a hostile act against mostly branded and
foreign generic firms (as local generic firms are exempt from it
because of their local R&D investment), which indeed led to
downsizing of branded firms’ sales forces and reduction in their
promotional capabilities. The advantage of the 12% rebate is in its
simplicity of implementation and in avoiding legal hassle when
ex-factory prices potentially decrease below the COGS—particularly of some older products with usually low expenditure but
that are critical for health care (e.g., fentanyl ampoules). Another
argument might be that the 12% rebate is just a starting tool that
could be always accompanied by a reduction in ex-factory prices
either via international reference pricing or the so-called risksharing mimicking the UK bortezomib scheme [40,41].
The PEA was launched into the milieu of co-payments while
many jurisdictions in the CEE region still largely exempt patients
from co-payments. When introducing co-payments de novo, the
adverse effect on adherence could be severe. Also, co-payments
in Hungary are collected diligently, while it may well be that
administration costs of applying co-payments may exceed in
some instance the amount of co-payment. Further studies would
be needed to address this point.
Our study has some obvious limitations. We have examined
a limited number of products and one might well find the
product or a group of products that would deviate from our
conclusions. In fact, we are planning to extend our analysis to
the wider range of ATC groups, including antihypertensives and
proton pump inhibitors and a wider number of innovative drugs.
We also plan to examine the impact of the reforms that were
introduced after July 1, 2011, and which we have referred to in
the text.
One may argue that deriving the marginal cost of production
from the assumption of 10% COGS is invalid and may at best
apply to a limited number of drugs. A good counterargument is a
real-life case from Romania in which in March 2005 the first
generic after the patent expiry entered the antipsychotic market
with 12% to 15% (depending on strength) of the price of the offpatent originator [42]. Other generics quickly followed by decreasing the price even below the 10% mark of the original monopoly
price. Another more recent example also from the Romanian
pharmaceutical market is an entry of a generic version of an
oncological drug at 7% of the originator’s price [43].
Population health in Hungary is among the worst in Europe,
and it may well be that the PEA might adversely contribute to the
mid-term and long-term health prognosis. Our analysis cannot
assess this impact. In the same way, we cannot address whether
lower expenditure on drugs increased the use of other health care
services (e.g., on the secondary or tertiary levels) via the so-called
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 290–299
squeezed balloon effect because we have not considered expenditures outside of the prescription drug market.
Finally, which are the tools of the NHIF-OEP that could be
shared with other jurisdictions in the CEE region and beyond?
The overall philosophy of focusing on volume rather than price is
a good guiding beacon and price-volume agreements appear to
work and also provide enough room for innovation to strive. In
other particulars, the tools to be shared without modifications are
less obvious due to shortcomings/inefficiencies discussed above.
Making data on prescription drug expenditures and associated
co-payments publicly available is an item that should be definitely followed by the surrounding jurisdictions.
Source of financial support: These findings are the result of
work supported by the author. The views expressed in this article
are those of the author, and no official endorsement by the
University of Ljubljana is intended or should be inferred.
R EF E R EN CE S
[1] Pharma Marketletter. Pharma-Economic Act now passed by the
Hungarian Parliament. December 4, 2006. Available from: http://www.
highbeam.com/doc/1G1-155673153.html. [Accessed May 31, 2013].
[2] National Health Insurance Fund – Orszagos Egeszebiztositasi Pentzar.
2011. Available from: http://www.oep.hu/portal/page?_pageid=
35,21341107&_dad=portal&_schema=PORTAL and http://www.oep.hu/
portal/page?_pageid=35,21341860&_dad=portal&_schema=PORTAL
(in Hungarian). [Accessed May 31, 2013].
[3] Harris CM, Scrivener G. Fundholders’ prescribing costs: the first five
years. BMJ 1996;313:1531–4.
[4] Gosden T, Torgerson DJ. The effect of fundholding on precribing and
referral costs: a review of the evidence. Health Policy 1997;40:103–14.
[5] Walley T, Mossialos E. Financial incentives and prescribing. In:
Mossialos E, Mrazek M, Walley T,eds., Regulating Pharmaceuticals in
Europe: Striving for Efficiency, Equity, and Quality. Maidenhead: Open
University Press, 2004.
[6] Busse R, Howerth C. Cost containment in Germany: twenty years
experience. In: Mossialos E, Le Grand J,eds., Health Care and Cost
Containment in the European Union. Aldershot: Ashgate, 1999.
[7] Danzon PM, Liu H. Reference Pricing and Physician Drug Budgets: The
German Experience in Controlling Pharmaceutical Expenditures.
Working paper. Philadelphia, PA: Wharton School, University of
Pennsylvania, 1996.
[8] Mrazek M, Mossialos E. Regulating pharmaceutical prices in the
European Union. In: Mossialos E, Mrazek M, Walley T,eds., Regulating
Pharmaceuticals in Europe: Striving for Efficiency, Equity, and Quality.
Maidenhead: Open University Press, 2004.
[9] Pharma Marketletter. Hungary’s new drug act “needs revision” (Pharma
Economic Act). February 26, 2007. Available from: http://www.
highbeam.com/doc/1G1-160184516.html. [Accessed May 31, 2013].
[10] Santerre RE, Neun PE. Health Economics: Theories, Insights, and Industry
Studies(5th ed.). Mason, OH: South-Western College Learning, 2010.
[11] Hren R. Overview of economic principles of competition in prescription
drug markets. Farmakoekonomika a liekova politika 2011;7:3–11.
[12] Gaal P. Health Care in Transition – Hungary. Copenhagen: European
Observatory on Health Care Systems, 2004.
[13] McGuire A, Raikou M, Kanavos P. Pricing pharmaceuticals: value-based
pricing in what sense. Eurohealth 2008;14:3–6.
[14] Mrazek M, de Jonchere K, Petrova G, Mossialos E. The pharmaceutical
sector and regulation in the countries of Central and Eastern Europe.
In: Mossialos E, Mrazek M, Walley T,eds., Regulating Pharmaceuticals
in Europe: Striving for Efficiency, Equity, and Quality. Maidenhead:
Open University Press, 2004.
[15] IHS Global Insight. IHS Global Insight Report: Hungary (Healthcare and
Pharma). Budapest, Hungary: IHS Global Insight, Inc., February 23, 2010.
[16] OECD health data. 2006. Available from: http://www.oecd-ilibrary.org/
social-issues-migration-health/data/oecd-health-statistics_health-data-en.
[Accessed May 31, 2013].
299
[17] OECD health data. 2005. Available from: http://www.adbi.org/
3rdpartycdrom/2005/06/13/1342.public.health.data/. [Accessed May 31, 2013].
[18] United Nations, Department of Economic and Social Affairs, Population
Division. United Nations Population Division 2005-2010 data. New York,
NY: United Nations, 2013.
[19] IHS Global Insight. Hungarian pharma law change will affect industry
adversely: Egis warns of R&D freeze. 2011. Available from: http://www.
ihs.com/products/global-insight/industry-economic-report.aspx?
ID=1065930004. [Accessed May 31, 2013].
[20] National Pharmaceutical Institute – Orszagos Gyogyszerszeti Intezet.
2011. Available from: http://www.ogyi.hu/nyitooldal/ (in Hungarian).
[Accessed May 31, 2013].
[21] Business Monitor International. Hungary Pharmaceuticals & Healthcare
Report Q1 2009. London: Business Monitor International, 2009.
[22] Kanavos P. Generic policies: rhetoric vs. reality. Euro Observer
2008;10:1–6.
[23] OECD. OECD Policy Roundtables: Generic Pharmaceuticals. OECD, 2010.
Available from: http://www.oecd.org/competition/abuse/46138891.pdf
[Accessed May 31, 2013].
[24] Grabowski HG, Vernon JM. Brand loyalty, entry, and price competition
in pharmaceuticals after the 1984 Drug Act. J Law Econ 1992;35:331–50.
[25] Frank R, Salkever D. Generic entry and the pricing of pharmaceuticals. J
Econ Manag Strat 1997;6:75–90.
[26] Vandoros S. Generic policies and the ‘generics paradox’. Euro Observer
2008;10:7–8.
[27] Folland S, Goodman AC, Stano M. The Economics of Health and Health
Care(6th ed.). London: Pearson, 2010.
[28] Frank RH. Microeconomics and Behavior(8th ed.). New York: McGrawHill, 2010.
[29] Wahlbeck K, Tuunainen A, Ahokas A, Leucht S. Dropout rates in
randomised antipsychotic drug trials. Psychopharmacology (Berl)
2001;155:230–3.
[30] Valenstein M, Ganoczy D, McCarthy JF, et al. Antipsychotic adherence
over time among patients receiving treatment for schizophrenia: a
retrospective review. J Clin Psychiatry 2006;67:1542–50.
[31] Lieberman JA, Stroup TS, McEvoy JP, et al. Effectiveness of
antipsychotic drugs in patients with chronic schizophrenia. N Engl J
Med 2005;353:1209–23.
[32] Keith SJ, Kane JM. Partial compliance and patient consequences in
schizophrenia: our patients can do better. J Clin Psychiatry
2003;64:1308–15.
[33] Gilmer TP, Dolder CR, Lacro JP. Adherence to treatment with
antipsychotic medication and health care costs among Medicaid
beneficiaries with schizophrenia. Am J Psychiatry 2004;161:692–9.
[34] Kozma C, Grogg A. Medication compliance and hospitalization in
schizophrenia. Paper presented at Sixth Annual Meeting of the College of
Psychiatric and Neurologic Pharmacists, May 1–4, 2003, Charleston, SC.
[35] IMS. Long-acting injectable risperidone data in European Union.
Danbury, CT: IMS Health, 2011.
[36] IMS. Bortezomib data in European Union. Danbury, CT: IMS Health,
2010.
[37] Danzon PM, Ketcham JD. Reference pricing of pharmaceuticals
for Medicare: evidence from Germany, the Netherlands and
New Zealand. In: Cutler DM, Garber AM (eds.), Frontiers in Health Policy
Research Vol. 7. National Bureau of Economic Research and
MIT Press, 2004.
[38] Beier FK. The significance of the patent system for the technical,
economic and social progress. Intern Rev Indus Prop Copyright
1980;11:563-71.
[39] Landes WM, Posner RA. The Economic Structure of Intellectual
Property Law. London: Harvard University Press, 2003.
[40] Towse. Value based pricing, research and development, and patient
access schemes: will the United Kingdom get it right or wrong? Br J Clin
Pharmacol 2010;70:360–6.
[41] Jack A. NHS drug adviser questions drug policy. Financial Times,
August 24, 2010.
[42] Canamed. Romanian national price registry of medicines. 2005.
Available from: http://www.ms-preturi.ro/msf-dgf26/download.htm,
historical data available from the Romanian Ministry of Health upon
request. [Accessed May 31, 2013].
[43] Canamed. Romanian national price registry of medicines. 2011.
Available from: http://www.ms-preturi.ro/msf-dgf26/download.htm,
historical data available from the Romanian Ministry of Health upon
request. [Accessed May 31, 2013].
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 300–305
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Performance Assessment of Ga District Mutual Health Insurance Scheme,
Greater Accra Region, Ghana
Eric Nsiah-Boateng, MPH1,*, Moses Aikins, PhD2
1
Claims Department, National Health Insurance Scheme, Ashiedu Keteke Submetro Office, Accra, Ghana; 2Health Policy, Planning and Management Department,
School of Public Health, University of Ghana, Accra, Ghana
AB STR A CT
Objective: This study assessed performance of the Ga District Mutual
Health Insurance Scheme over the period 2007-2009. Methods: The
desk review method was used to collect secondary data on membership coverage, revenue, expenditure, and claims settlement patterns
of the scheme. A household survey was also conducted in the Madina
Township by using a self-administered semi-structured questionnaire
to determine community coverage of the scheme. Results: The study
showed membership coverage of 21.8% and community coverage of
22.2%. The main reasons why respondents had not registered with the
scheme are that contributions are high and it does not offer the
services needed. Financially, the scheme depended largely on subsidies and reinsurance from the National Health Insurance Authority
for 89.8% of its revenue. Approximately 92% of the total revenue was
spent on medical claims, and 99% of provider claims were settled
beyond the stipulated 4-week period. Conclusions: There is an
increasing trend in medical claims expenditure and lengthy delay in
claims settlements, with most of them being paid beyond the
mandatory 4-week period. Introduction of cost-containment measures including co-payment and capitation payment mechanism
would be necessary to reduce the escalating cost of medical claims.
Adherence to the 4-week stipulated period for payment of medical
claims would be important to ensure that health care providers are
financially resourced to deliver continuous health services to insured
members. Furthermore, resourcing the scheme would be useful for
speedy vetting of claims and also, community education on the
National Health Insurance Scheme to improve membership coverage
and revenue from the informal sector.
Introduction
though implementation in terms of access to benefits began in
November 2005 [3–5]. Its policy objective is that “within the next
five years, every resident of Ghana shall belong to a health
insurance scheme that adequately covers him or her against
the need to pay out-of-pocket at point of service use in order to
obtain access to a defined package of acceptable quality health
services” [6]. The NHIS was designed as a mandatory health
insurance system, with risk pooling across district schemes,
funded from members’ contributions and a levy on the valueadded tax charged on selected goods and services [3–6].
As a key social sector initiative to support the Ghana Poverty
Reduction Strategy II policy objective of ensuring sustainable
financial arrangements that protect the poor, the NHIS’s performance and long-term sustainability are significant. The performance assessment of the scheme is also key to Ghana’s attainment
of the Millennium Development Goals 1, 3, 4, and 5.
Since its full implementation in 2005, the NHIS has been
facing major structural and administrative challenges, including
significant delays in issuing membership cards; lack of a uniform contribution system across schemes, which has implications for portability and equity within the national scheme; and
Many low- and middle-income countries are challenged with
how to finance their health care systems to achieve universal
coverage of health services. In 2005, the member states of the
World Health Organization adopted a resolution encouraging
countries to develop health financing systems aimed at providing
universal coverage [1]. This was defined as securing access for all
to appropriate promotive, preventive, curative, and rehabilitative
services at an affordable cost.
In the 1990s, a number of mutual health organizations were
established in Ghana, with funding and technical support from
external partners. Most of these mutual health organizations,
however, primarily focused on providing financial protection
against the potentially catastrophic costs of a limited range of
inpatient services for the disadvantaged people in society [2]. The
National Health Insurance Scheme (NHIS) was introduced in 2004
to build on these organizations and provide comprehensive
health services to all citizens in Ghana [3].
The National Health Insurance Act, Act 650, was passed into
law in Ghana in 2003 through the Legislative Instrument (LI 1809),
Keywords: claims settlements, Ghana, membership coverage, National
Health Insurance Scheme..
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
* Address correspondence to: Eric Nsiah-Boateng, MPH, Claims Department, National Health Insurance Scheme, Ashiedu Keteke
Submetro Office, Derby Avenue, Accra, Ghana.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.005
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 300–305
considerable delays in provider claims reimbursement [3,7,8]. An
independent health sector review report shows that at the end of
2008, the health facilities had outstanding claims worth GH¢49
million [7]. A number of initiatives and activities by researchers
and development partners aimed at tackling performance challenges facing the District Mutual Health Insurance Schemes
(DMHISs) have taken place. Most of these research and initiatives,
however, focused on inventories of DMHISs and access, utilization, and quality of care and were largely uncoordinated. Moreover, key findings and associated recommendations were left
unimplemented [9]. Therefore, little is known about performance
elements of membership coverage, revenue mobilization, expenditure, and medical claims settlements of the DMHISs. The study
aimed at filling this gap by providing performance assessment of
Ga DMHIS.
According to the 2010 NHIS Annual Report, there are 145
DMHISs nationwide, with 10 in the Greater Accra region [10]. The
Ga DMHIS is the biggest DMHIS in the Greater Accra region in
terms of catchment area. The Ga district has a large number of
suburban and rural communities, making it suitable for this
study. Madina is the largest cosmopolitan settlement in the
district, which also made it appropriate for the household survey.
This article reports performance assessment of the scheme for
the 2007-2009 period and recommendations for improving its
operations.
Conceptual Framework
The conceptual framework for the study was adopted from the
World Health Organization proposed framework for health systems performance assessment and modified to reflect the International Labour Organization’s core performance indicators for
assessing social health insurance schemes [11,12]. According to
the framework, high membership coverage, high revenue base,
low expenditure, and prompt settlement of provider claims
enhance performance ratios such as coverage rate, renewal rate,
expense ratio, and claims ratio, which, in turn, results in high
performance and an improved health status of the target population (Fig. 1).
301
Data Collection Method
Desk review
Documents on membership, operational reports, audited reports,
financial statements, and claims payment books of the scheme
were reviewed. The registration files were reviewed in terms of
the number of people registered, number of membership cards
issued, and number of renewals for each year under review. The
audited accounts for 2007-2008 and unaudited accounts for 2009
were examined for total contribution collected, subsidies received
from the National Health Insurance Authority (NHIA), donor
support, and other internally generated funds. Information on
administrative expenditure and medical bills was also collected.
The claims submission registers and claims payments for 2009
were reviewed to determine the number of days between submission of claims and reimbursement.
Face-to-face interview
A community household survey was conducted in the Madina
Township to determine the community coverage rate. A multistage sampling method was used to select the study subjects. In
all, 376 household heads were sampled on the basis of an
estimated prevalence rate of 43% membership coverage, a confidence level of 95%, and 5% margin of error. The questionnaire
covered background characteristics and membership in the NHIS.
The household membership section looked at knowledge on the
NHIS, membership status, and reasons for not enrolling into the
scheme.
Data Analysis
The performance indicators for the study were analyzed as
follows.
Coverage rate
Membership files for the period 2007-2009 were reviewed to
determine the total number of valid card-bearing members in
each year. The coverage rate was determined by dividing the total
number of valid card-bearing members in each year by the
estimated district population in the same year.
Methods
Community coverage rate
Study Area
This was estimated by dividing the total number of participants
with valid membership cards as of March 2010 by the total
number of participants interviewed.
The study was conducted at Ga DMHIS and Madina Township, all
in the Ga district of the Greater Accra region. The Ga DMHIS has a
staff strength of nine and 74 contracted health care providers.
The Ga district lies in the northern part of the Greater Accra
region and is bounded in the north by Akuapim South district, in
the east by Tema Municipal, and in the south by Accra Metropolis. It has three subdistricts, namely, Ga South, Ga East, and Ga
West, with 594 communities comprising mixed settlements:
urban, periurban, and rural areas. The district has an estimated
population of 891,609 and a growth rate of 4.4%. There are 58
health facilities in the district comprising public, private, and
Christian Health Association of Ghana facilities. The main economic activities in the district are public service, trading, farming,
and craftsmanship.
Study Design
The study was a cross-sectional survey of households and a
retrospective analysis of membership, revenue, and expenditure
records of the Ga DMHIS for the period 2007-2009. The study
population consisted of membership data of Ga DMHIS and
selected heads of surveyed households in the Madina Township.
Annual revenue
This was estimated by adding the total amount of money
collected from contributors, received from the NHIA and other
donor agencies, and investment returns in each year of the
period under review (2007–2009).
Annual expenditure
This was estimated by adding administrative expenses and
medical claims expenses in each year of the period 2007-2009.
The administrative expenses comprised salaries, membership
cards’ processing cost, and other operational costs, while the
medical claims expenses constituted payment of outpatient,
inpatient, specialized medical services and essential medicines
in the NHIS minimum benefit package.
Expense ratio
This was estimated by dividing the total administrative expenses
incurred in each year by the total amount of contributions
collected in the same year for the period 2007-2009.
302
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 300–305
High membership
coverage
High
renewal rate
and growth
rate
-Registered
members with
valid ID Cards
High revenue
-Premium
-NHIA subsidy
Low
expense
ratio and
claims ratio
-Other income
High
performance
Low expenditure
Improved
health
status of the
population
-Administrative
-Medical bills
(claims)
Prompt provider
claims settlement
Fig. 1 – Conceptual framework for assessment of Ga DMHIS. DMHIS, District Mutual Health Insurance Scheme; NHIA, National
Health Insurance Authority.
Claims ratio
This was estimated by dividing the total medical claims expenses
incurred in each year by the total amount of contributions
collected in the same year for the period 2007-2009.
Combined ratio
This was calculated by adding the expense ratio and the
claims ratio.
Promptness of claims settlements
All provider claims submitted within the October-December 2009
period were reviewed to determine the number of days it took to
settle them. The settlements days for each provider claim were
grouped according to a defined schedule of 0 to 28 days, and more
than 28 days. Based on the stipulated period of 4 weeks for claims
settlements, all claims paid within 0 to 28 days were considered
as prompt payment and those paid beyond 28 days as delayed
payment.
Ethical consideration
Ethical clearance for the study was obtained from Ghana Health
Service Ethical Review Committee on Research Involving Human
Subjects. Approval was sought from the Ga DMHIS to use its data,
and permission was sought from the Municipal Health Services to
conduct the household survey in the Madina Township.
Participants were informed of the objective of the study and
that they were free to participate and to leave at any point.
A total of 365 household heads out of the 376 sampled voluntarily
provided signed informed consent and were interviewed.
Table 1 – Membership by category (2007–2009).
Year
District
population
Formal
sector
Informal
sector
(18–69 y)
SSNIT
pensioners
Aged
(≥70 y)
Younger
than
18 y
Indigent
Pregnant
women
Total
coverage
(%)
2007
2008
2009
749,022
782,715
817,924
4,957
9,794
15,943
19,035
42,318
55,561
237
1,171
3,484
1,990
4,027
7,909
25,598
47,240
63,512
831
2,920
2,920
NA
11,997
28,594
7.0
15.3
21.8
NA, not applicable (the free maternal care policy was introduced in 2008); SSNIT, Social Security and National Insurance Trust.
303
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 300–305
Table 2 – Reasons for not registering with the
Ga DMHIS.
Reason (n ¼ 284)
Revenue and expenditure
In all, the expenditure of the scheme for the period under study
exceeded the revenue. This occurred in 2007 and 2008 but was
reversed in 2009 by a significant increase in NHIA support (Fig. 2).
In 2009, the scheme generated 60.9% of the total revenue and
expended 41.8%. The total revenue generated and the expenditure incurred for the period under study were GH¢12.8 million
and GH¢13.7 million, respectively.
%
Contribution is expensive
Not sick now
Hospital is too far
Treat elsewhere
Does not offer services needed
Belong to other district scheme (NHIS)
Total
35.3
10.5
1.0
9.5
33.2
10.5
100
Annual revenue
While proportions of NHIA support in the form of subsidy for the
exempt group and reinsurance for claims payment increased
significantly from 82.9% in 2008 to 93.9% in 2009, contributions
collected and other income generated showed a decreasing trend
(Fig. 3).
DMHIS, District Mutual Health Insurance Scheme; NHIS, National
Health Insurance Scheme.
Limitations of the study
Annual expenditure
First, there were no separate data for active and nonactive
members between the 2007 and 2008 period, making it difficult
to determine the true membership coverage of card-bearing
members in that period. Second, the financial audit report for
2009 was not available; hence, revenue and expenditure data
were retrieved from account ledger books and financial statements, which might not give the true picture of financial flows
in 2009.
The main expenditure areas of the scheme were administrative
and medical claims. The total expenditure for the period under
review doubled from GH¢2.2 million in 2007 to GH¢5.8 million in
2008, and dropped slightly in 2009 (i.e., GH¢5.7 million) (Table 3).
The proportion of administrative expenses showed a downward
trend from 18.2% in 2007 to 3.6% in 2009, while that of medical
claims expenses went up from 81.8% in 2007 to 96.4% in 2009.
Claims settlements
Results
About 99% of 38,737 medical claims reviewed for the period
October-December 2009 were paid beyond the stipulated period
of 28 days for claims settlements.
Coverage Rate
Expense, claims, and combined ratios
In general, the membership coverage of the scheme increased
from 7% in 2007 to 21.8% in 2009 (Table 1). The number of
registered members, however, dropped from 8.3 percentage
points in the first year (2007–2008) to 6.5 in the second year
(2008–2009). In all, the informal sector constituted one-third of
the registered members; there were no registrations of pregnant
women in 2007 and indigent in 2009.
The expense ratio showed a downward trend from 2.7 in 2007 to
0.7 in 2009, while the claims ratio increased from 12.2 in 2007 to
18.1 in 2009 (Table 4). The combined ratio, which is the sum of the
expense ratio and the claims ratio, increased from 15.0 in 2007 to
18.8 in 2009.
Discussion
The membership of the scheme is categorized into formal sector
workers, informal sector workers, pensioners, the aged (70 years
and older), children younger than 18 years, indigent, and pregnant
women. The membership coverage increased over the period
under study, driven mainly by the exempt groups: children
younger than 18 years, the aged, indigent, and pregnant women;
only one-third of the informal sector group who pays contributions is registered. The increasing trend in annual membership
Community coverage rate
Out of the total number of 365 household heads who participated
in the survey, 81 (22.2%) were registered members with valid ID
cards. The main reasons why some of the respondents had not
registered with the scheme were “contribution is expensive” and
“does not offer services needed” (Table 2).
70
60.9
Percentage (%)
60
50
41.8
42
40
26.9
30
20
16.3
12.2
10
0
2007
Revenue (GHc
2008
Years
2009
Expenditure (GHc
Fig. 2 – Revenue and expenditure status (2007–2009).
304
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 300–305
100%
90%
80%
Percentage
70%
60%
50%
84.9
82.9
93.9
2007
2008
2009
40%
30%
20%
10%
0%
Year
Contributions collected
NHIA support
Other income
Fig. 3 – Revenue distribution of the scheme (2007–2009). NHIA, National Health Insurance Authority.
coverage is good for membership growth and development of the
scheme as shown in previous studies [11,13,14]. It indicates that
registered members have accepted the NHIS program and are
ready to pool their resources to seek a measure of protection from
the risks that they face [13]. The small proportion of registered
informal group who pays contributions, however, could affect the
revenue base and presents long-term sustainability problems [3].
In spite of the increasing trend in annual membership coverage,
the number of registered members declined from 8.3 percentage
points in the first year (2007–2008) to 6.5 in the second year (2008–
2009), a trend found in other studies [3,8,15]. The cause of this
decline in membership registration could be attributed to reasons
found in the household survey.
The household survey shows a membership coverage of 22%,
which gives credence to that obtained from the desk review at
the scheme office. Most of the respondents considered the
annual contribution of the scheme, which ranges from GH¢
10.00 to GH¢24.00, as expensive, and hence they did not register
for the scheme. Others believe that the scheme does not offer the
services that they need or are simply not sick; hence, they do not
see the need to register. These reasons show that respondents do
not understand the importance of the NHIS; therefore, continuous community education would be useful to sensitize people on
the role that health insurance plays in stabilizing their situation.
In relation to revenue, the main sources are support from the
NHIA in the form of subsidy and reinsurance, contributions from
the informal sector, and internally generated funds such as
membership cards’ processing fees and interest on investments.
The total revenue of the scheme increased significantly over the
study period, which is good for growth and long-term sustainability as evidenced in a study by Guy [4]. In spite of the annual
increase in revenue, NHIA support was the predominant contributing factor, which constitutes about 90% of total revenue of the
Table 3 – Expenditure status of the scheme (2007–
2009).
Year
2007
2008
2009
Administrative
expense (%)
18.2
8.9
3.6
Medical
claims
expense (%)
81.8
91.1
96.4
Total
expenditure
(GH¢)
2,232,163.34
5,757,550.21
5,732,104.33
scheme. In absolute terms, the contributions collected and other
income generated increased in the first year (2007–2008) and
declined in the second year (2008–2009), which might be due to a
similar trend observed in the membership registration, particularly informal sector registration. The decline in contributions
from the informal sector, if it continues in subsequent years,
would reduce the revenue base and cause further days in the
payment of provider claims. In the long term, it could threaten
the financial viability of the scheme if support from the NHIA
goes down. These trends in revenue were also found in the study
by Witter and Garshong [3].
The main expenditure areas of the scheme are administrative
expenses and medical claims expenses. The annual expenditure
went up in the first year of the study period and dropped slightly
in the second year. The total expenditure, however, exceeded
revenue, with medical claims expenses as the major contributing
factor. This trend in the scheme’s expenditure was also seen in
the estimation of the expense and medical claims ratios. The
combined ratio increased consistently over the study period, with
medical claims ratio as the driving force. The increasing trend in
medical expenses could pose a serious financial challenge to the
scheme. In the long term, it may result in diminished social
protection and value to the insured members. The change in
claims payment mechanism from fee-for-service to Diagnostic
Related Groupings and an upward review of the medicine price
list in 2008 might partly account for the high medical claims
expenditure observed in 2008 [3]. According to the 2010 independent health sector review, claims for medicines totaled 60% of
all claims in 2009 [7]. These factors contributed to a growth in
distressed schemes and failure to pay outstanding provider
claims in 2008 [3]. As emphasized earlier, it would be important
for the scheme to attract more members and to retain them over
long periods during which they consume no or few services [16].
The increasing rate of medical bills expenditure as explained by
the management is due to 1) comprehensive benefit package with
no co-payment mechanism to control cost; 2) ineffective gatekeeper system, which contributes to increase in health care
utilization and cost at secondary and tertiary facilities; and 3)
client and health care provider abuse.
The claims settlements pattern of the scheme shows lengthy
delays in the payment of provider claims. According to the NHIS
Legislative Instrument (LI 1809), the stipulated period for vetting
and payment of claims is 4 weeks after receipt of the claims from a
health care provider [5]. About 99% of sampled claims submitted
within October-December 2009, however, were settled beyond the
mandatory period. Because the main source of income for health
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 300–305
Table 4 – Expense, claims, and combined ratios
(2007–2009).
Year
2007
2008
2009
Expense
ratio
2.7
1.4
0.7
Claims
ratio
12.2
14.0
18.1
Combined
ratio
14.9
15.4
18.8
care providers is the internally generated fund that comes from
claims reimbursed by the NHIA, this situation would affect their
revenue base and may result in poor delivery of health service to
insured members [7]. Delays in claims settlements remains a
problem in most of the DMHISs, and this has raised acute
problems of bankruptcy of certain providers and lack of trust by
suppliers [7]. It has been found that paying claims promptly is an
important aspect of service and good value [17]. It is also an
essential element of the financial incentives of health care
providers and as a result, a key factor affecting provider behavior
[18]. Early settlement of provider claims will enable health care
providers to render continuous service to insured members. More
often, health care providers require funds immediately after the
provision of service; hence, significant delays in claims settlements may force them to take other measures that defeat the
purpose of the scheme [17]. For instance, they may indulge in
unhealthy practices including extortion of money from insured
members and refusing health care services to insured members.
These provider-abuse practices may affect the renewal rate
because dissatisfied insured members see no value in the scheme
and as such are less likely to renew their memberships [13].
Since its introduction in 2004, the NHIS has partly accounted
for reduction in availability of Government of Ghana resources for
operations and public health activities of health care providers
because service claims reimbursed by the NHIS cover some of the
facility-based operational costs [7]. The scheme has also been a
contributory factor to delays in the release of funds from the
central government, which affects the implementation of planned
activities of health care providers. According to the Ministry of
Health, by December 2009, the ministry received only 20% of
expected transfers from the National Health Insurance Fund,
mainly because of delays in inflows in the National Health
Insurance Fund [7]. As a result, health care providers largely
depend on claims reimbursement funds to cope with the significant increase in health cost and workload. Further delays in claims
payment by the DMHISs will therefore cause cash problems for
health care providers. The factors contributing to delays in claims
settlements as explained by the management of the scheme were
delays in the transfer of subsidies and reinsurance from the NHIA,
large volumes of medical claims, inadequate number of personnel,
and ineffective claims processing software. The review of the
medical claims showed that the scheme receives an average
volume of 30,000 claims per month, which makes it practically
challenging to vet and settle payments within the mandatory 28
days. These factors were also found to be contributing to delays
associated with provider claims settlements in 2009 [7].
Conclusions
There are increasing trends in membership coverage and
revenue that are largely driven by the exempt groups and
subsidies from the NHIA, respectively. The medical claims
expenditure is increasing with significant delays in settlement.
Introduction of cost-containment measures including copayments and capitation payment mechanism would be necessary to reduce the escalating cost of medical claims.
305
Adherence to the stipulated 4-week period for claims payment
would also be important to ensure that health care providers
are financially resourced to deliver continuous health services
to insured members. Furthermore, the scheme should be
adequately resourced to ensure speedy vetting of medical
claims and also to facilitate community education in the
district to improve membership coverage and revenue from
the informal sector.
Acknowledgments
We acknowledge contributions of the management and staff of
Ga DMHIS, field workers, and field supervisors. We thank the Ga
East District Director of Health Services for permitting us to
conduct the household survey in the Madina Township and the
staff of Madina Polyclinic for their support in the recruitment and
training of field workers.
Source of financial support: This study was self-financed by
the authors. The views expressed in this article are those of the
authors.
R EF E R EN C ES
[1] World Health Organization. Sustainable Health Financing, Universal
Coverage and Social Health Insurance [A58/33]. Geneva, Switzerland:
World Health Organization, 2005.
[2] Atim C, Grey S, Apoya P, et al. A Survey of Health Financing Schemes in
Ghana. Bethesda, MD: Partners for Health Reformplus Project, Abt
Associates, Inc., 2001.
[3] Witter S, Garshong B. Something old or something new? Social health
insurance in Ghana (Article). BMC Int Health Hum Rights 2009;9:20.
[4] Republic of Ghana. National Health Insurance Scheme Act, Act 650.
Accra, Ghana: Ghana Publishing Corporation, 2003.
[5] Republic of Ghana. National Health Insurance Scheme, Legislative
Instrument, LI 1809. Accra, Ghana: Ghana Publishing Corporation, 2004.
[6] Ministry of Health. National Health Insurance Policy Framework for
Ghana, Revised Version. Accra, Ghana: Ministry of Health, 2004.
[7] Ministry of Health. Independent Review: Health Sector Programme of
Work 2009, Ghana. Accra, Ghana: Ministry of Health, April 2010:8–19.
[8] Ministry of Health. Independent Review: Health Sector Programme of
Work 2007, Ghana. Report. Accra, Ghana: Ministry of Health, April
2008:54–8.
[9] National Health Insurance Authority (NHIA) and Development Partners
(DP). Ghana National Insurance Scheme: joint development mission.
Aide Memoire 2007. Available at: http://www.snvworld.org/sites/www.
snvworld.org/files/publications/5.ghana_national_health_insurance_
scheme_joint_dp_mission_aide_memoire.pdf. [Accessed August 30,
2013].
[10] National Health Insurance Scheme. Annual Report 2010. Accra, Ghana:
National Health Insurance Authority, 2011.
[11] Wealth Health Organization. World Health Report 2000. Health
Systems: Improving Performance. Health Financing Policy Issue Paper.
Geneva, Switzerland: Wealth Health Organization, 2000.
[12] International Labour Organization. Performance Indicators of Statutory
Social Insurance Schemes: Global Extension of Social Security. Working
Paper 2009.
[13] Wipf J, Garand D. Performance Indicators for Microinsurance: A
Handbook for Microinsurance Practitioners. Luxembourg: ADA asbl,
2008.
[14] Guy C. Social health insurance in developing countries: a continuing
challenge. Int Soc Secur Rev 2002;55:2.
[15] Yevutsey SK, Akins M. Financial viability of District Mutual Health
Insurance Schemes of Lawra and Sissala East Districts, Upper West
Region, Ghana. Ghana Med J 2010;44(4):130–7.
[16] Dror D. Health insurance for the poor through community schemes—is
it viable? Defining an agenda for poverty reduction. Proc First Asia
Pacific Forum Poverty 2007;2:193–208.
[17] CGAP Working Group on Microinsurance. Performance Indicators for
Microinsurance Practitioners. Luxembourg: ADA asbl, 2006; Workshop
report. October 16–17, 2006.
[18] Kwon S. Thirty years of national health insurance in South Korea:
lessons for achieving universal health care coverage. Health Policy
Planning 2009;24:63–71.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 306–311
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
The Process of Privatization of Health Care Provision in Poland
Krzysztof Kaczmarek, PhD1, Hannah Flynn, MPH2, Edyta Letka-Paralusz, MA3, Krzysztof Krajewski-Siuda, MD, PhD4,5,
Christian A. Gericke, MD, PhD6,7,8,
1
Department of Health Policy, Medical University of Silesia, Katowice, Poland; 2PenCLAHRC, National Institute for Health Research, Plymouth University Schools
of Medicine and Dentistry, Plymouth, UK; 3Department of Public Health, Medical University of Silesia, Katowice, Poland; 4Sobieski Institute, Warsaw, Poland;
5
University of Information Technology and Management, Rzeszow, Poland; 6The Wesley Research Institute, Brisbane, Australia; 7University of Queensland School
of Population Health, Brisbane, Australia; 8Queensland University of Technology School of Public Health, Brisbane, Australia
AB STR A CT
Objectives: In January 1999, a new institutional structure for Poland's
health care system was laid out, instigated by the dramatic change in
both the political and economic system. Following the dissolution of
state socialism, private financing of health care services was encouraged to fill an important role in meeting rising consumer demand and
to encourage a more efficient use of resources through competition
and private initiative. However, from the outset of the intended
transformations, systemic limitations to the privatization process
hindered progression, resulting in varying rates of privatization
amongst the distinct health care sectors. The aim of this paper is to
describe the privatization process and to analyze its pace and differences in strategic approach in all major health care sectors. Methods:
Policy analysis of legislation, government directives, and published
national and international scientific literature on Polish health
reforms between 1999 and 2012 was conducted. Results: The analysis
demonstrates a clear disparity in privatization rates in different
sectors. The pharmaceutical industry is fully privatized in 2012, and
the ambulatory and dental sectors both systematically increased their
private market shares to around 70% of all services provided. However,
despite a steady increase in the number of private hospitals in Poland
since 1999, their overall role in the health care system is comparatively
limited. Conclusions: Unclear legal regulations have resulted in a gray
area between public and private health care, where informal payments
impede the intended function of the system. If left unchanged, official
health care in Poland is likely to become an increasingly residual
service for the worst-off population segments that are unable to afford
the legal private sector or the informal payments which guarantee a
higher quality service in the public sector.
Introduction
Initial Drivers for Health Care Reform
On January 1, 1999, a new institutional structure for Poland’s
health care system was founded, instigated by a dramatic change
in both the political system and the economic system [1]. In the
years preceding such change, a state-funded and centralized
health care system had operated where the public sector had
dominated in terms of both funding and service provision. The
collapse of state socialism in 1989 because of increased opposition and a failing economy, however, had severe consequences
on the state’s ability to provide health care coverage [2]. This
resulted in a growing imbalance between the needs expressed by
the population and the system’s ability to meet them, exacerbated by the ever-increasing cost of health care service provision.
In an attempt to address this, Poland transformed the health care
system and encouraged competition and private initiative [3,4].
From the outset of the intended transformations, however,
systemic limitations to the privatization process have hindered
progression. This has resulted in varying rates of privatization
among the distinct health care sectors and an ambiguous
relationship between public and private health care provision.
During state socialism, Poland, like many other Soviet bloc
nations, adopted the Semashko model for health care [5]. Statefunded through taxation and heavily centralized, this particular
system was designed with the intention of guaranteeing egalitarian health care coverage for the entire population. After the
dissolution of the Soviet Union, however, Poland along with
many other Central European countries suffered severe economic
difficulties that significantly affected health care provision [6].
Because of cuts in government expenditure and a shortage of
providers, public health care facilities became overcrowded and
had long waiting lists, scarce medical supplies, and out-of-date
technologies [7]. Receptive to this, Poland began to allow limited
private providers to manage demand for public health services
[8]. The principal idea envisaged was to establish a new set of
institutions and market-type mechanisms that would ensure a
more efficient use of productive assets by creating stronger
incentives arising from ownership, thereby increasing productivity and efficiency [3,9,10]. This signaled an initial step toward
privatization, defined as follows: “the transfer of ownership and
Keywords: health care provision, health care reform, health policy,
Poland, privatization.
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Christian Gericke, The Wesley Research Institute, PO Box 499, Toowong QLD 4066, Australia.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.001
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 306–311
control of government or state assets, firms and operations to
private investors” [11].
After this, the public sector gradually began to devolve further
until budgeting of health care services was replaced with an
insurance-based system of financing. Undoubtedly, this radically
changed the population’s right to health services, as access was
instantly linked to registration with a mandatory health insurance and payment of contributions [5]. As insurance funds were
initially regional and given autonomy, conditions were set for
private sector service provision, which then intensified as official
out-of-pocket payments for health services were started [12].
Alongside hospitals, clinics, and health centers, foundations or
voluntary associations were established, which accepted payments for performing better quality or difficult-to-access services.
This divided health care provision into both public and private,
with a gray sphere of informal payments emerging between the
two [2] that continued a long-standing history of informal payments in the socialist health system.
In the 2007 Stefan Batory Foundation’s Corruption Barometer,
78 (9%) of 870 respondents declared that they had made informal
payments in the last year, 52% of which were for informal
payments in health care [13].
In the larger Social Diagnosis panel of 3000 Polish households,
1.8% of households declared informal payments in 2007, 1.3% in
2009, and 1.7% in 2011 [14]. In 2011, the average informal payment
for health services was estimated at 1244 Polish Zloty (300 euros)
per year and household. Furthermore, 18.1% of households
declared that they refrained from purchasing necessary medicines, 17.3% could not afford dental treatment, and 13.9% could
not afford medical treatment [14].
307
care facilities from the central administration units to the local
self-governments, enabling them at the same time (under some
conditions) to transform those facilities into private entities.
These laws include
● the law of March 8, 1990, on local self-government [19];
● the law of November 24, 1995, on a change in the range of
responsibilities of some cities on the municipal areas of public
services [20];
● the law of June 5, 1998, on regional self-government [21]; and
● the law of June 5, 1998, on district self-government [22].
None of these legal acts, however, has directly and systematically regulated the issue of privatization of health care facilities. This has resulted in a process that is complicated, legally
unclear, and vulnerable to abuses, particularly in the case of
hospitals that are the most controversial in terms of their
privatization. During the last decade, successive governments
have tried on three occasions to establish such a law but none of
these efforts has been successful, each time being blocked during
the legislative process, or even earlier, at the stage of preparation.
In its first attempt, the Ministry of Health tried to implement
obligatory transformation of all health care organizations into
commercial law companies, entitled “Law on Commercialisation
and Privatisation of Independent Public Health Care Facilities
(2001).” Nevertheless, because of unfavorable political conditions
(forthcoming elections, a breakdown of the governing coalition,
and a strong political disintegration), the project was withdrawn
and replaced with a less radical approach.
Progress of Privatization in Poland
Legal Basis for Privatization
Between 1989 and 2001, approximately 20 new laws relating to
health care provision were adopted in Poland, which facilitated
the development of the private sector. In particular, the law of
July 13, 1990, which related to the privatization of state enterprises [15], and after its abolition the law of August 30, 1996,
which related to the commercialization and privatization of state
enterprises [16], were exceptionally influential in instigating the
privatization process. Although these acts did not directly refer to
health care services, they drew a general framework for the
process of privatization in Poland after the fall of communism.
The most important and far-reaching legislative acts to affect
health care were those that shaped the contracting environment.
The Health Care Organisation Act passed in 1991 introduced
contracting in place of administrative relationships, allowing
private surgeries and organizations to sign contracts for the
provision of services to people entitled to care financed from
public resources [17]. In doing so, categories of entities authorized
to provide health services (including those that are established by
nonpublic entities or individuals) were defined, as well as the
technical requirements that such entities must fulfill.
This was followed by perhaps the most influential act—The
General Health Insurance Act 1999, which introduced a social
health insurance system in Poland of 16 regional sickness funds
and 1 sickness fund for employees of military services [18]. This
caused a vast increase in the number of private organizations
holding public contracts because the regional sickness funds
were allowed to contract services with private health care
institutions as long as they met the required conditions and
offered cheaper service costs [8]. This was the first time private
providers were able to act within the public system of financing
health services.
In addition to these, a package of laws regulating the competences of local self-government units have since been passed,
which have gradually transferred the ownership duties of health
An analysis of the scale of privatization in the Polish health care
system shows significant disparity between the different health
care sectors. Changes in the pharmaceutical sector and in
ambulatory, dental, and hospital care differ in terms of pace,
strategic approach, and public resistance. To understand
these fundamental differences, each sector will be discussed
separately.
Pharmaceutical sector.
The commercialization of health services began with the privatization of the pharmaceutical industry. This was based on the
Freedom of Economic Activity Act (1988), which came into
fruition at the very beginning of the postcommunist transformation period [23]. Around the same time, the number of private
pharmacies accounted for approximately 43.9% of the total
number. Following a program implemented in 1994 devoted to
the privatization of pharmacies, however, almost all pharmaceutical outlets belonging to the Treasury have subsequently
been privatized [24]. This dynamic transformation in pharmacy
ownership between 1990 and 2006 is illustrated in Fig. 1.
Since the introduction of co-payments for dental care,
patients have started to purchase services offered by private
practices and clinics more willingly, even when required to cover
the total cost of the treatment. In doing so, they are able to
receive a faster and perceived better quality treatment. Because
of this high acceptability, the private dental sector developed
quickly in the early 1990s. After the Law on Social Health
Insurance came into force in 1999, private dental practices
started to offer treatment contracted within the Social Health
Insurance system. As a result, the number of facilities offering
services that are available only for out-of-pocket payments has
started to decrease gradually since 1999 [25]. Currently, more
than 80% of the active dentists work in the private sector and
approximately 85% of the services are provided by nonpublic
providers [26].
308
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 306–311
Fig. 1 – Dynamics changes in pharmacy ownership between 1990 and 2006.
Ambulatory care.
Until 1990, hospitals were responsible for the delivery of specialist outpatient services as well as for laboratory and imaging
diagnostics. The separation of hospitals and ambulatory care was
the first step toward the transformation of service provision. This
subsequently offered a great opportunity to invest in the ambulatory care infrastructure when previously inpatient services that
are in general more expensive than outpatient services had
received financial priority.
In the mid-1990s, the process of privatization was directed
toward outpatient care. In the first phase of the process, nonpublic
facilities were established mainly by individuals (doctors, nurses,
and other medical practitioners) and companies, who became
responsible and accountable for providing health care services [27].
Later, the local self-government units joined the process by
transforming the facilities they owned when the package of laws
reforming the system of public administration came into force.
These were, however, based on legal regulations that enabled only
large cities to do so in 1995 and all other local government units
from 1999 onward. This coincided, however, with the introduction
of the Social Health Insurance Law in 1999, which enabled nonpublic health care providers to enter into contracts with public
insurers [28]. This allowed self-government units to privatize their
ambulatory care facilities in two distinct ways:
1. The first involves holistic transformation of the public entity
through its liquidation. In this case, the duty to provide
services is transferred to a private entity together with ownership of the technical infrastructure. Formally, the unit that
opts for such a procedure must adopt a resolution to liquidate
the facility as well as to define the procedure of its transfer to
the nonpublic entity [29].
2. The second method involves separation of the service provision from the structures of the facility and a transfer of the
provision function to a private entity. Contrary to the first
method, in this case the self-government unit does not liquidate the facility but adopts a resolution on its restructuration.
Formally, the infrastructure ownership remains public [28].
In view of the above-mentioned legal changes, ambulatory care
privatization has been accelerated since the mid-1990s. This is
illustrated in Fig. 2, which highlights the increase in the number
of private ambulatory care providers since 1990.
As stated previously, the past two decades have seen an
immense increase in the number of private ambulatory care
facilities in Poland. In 1990, 4.5% of the ambulatory care
facilities were privately operated; by 2008 this had increased
to 77.8% [26]. The increase in privately owned facilities was
particularly fast during the period between 1999 and 2002
before stabilizing to an approximate increase of 3% each year.
During this time, public facilities decreased by approximately
47.5% [26].
Currently, the private sector is dominating outpatient care,
which is illustrated by the number of services provided. In 2008, a
total of 290,553,000 services were provided; of these, 202,785,000
were provided by private facilities, meaning that services provided by public facilities account for only 30% of all services
provided [30].
Hospital care.
The postwar history of private hospitals in Poland is relatively
linear. Those that resumed activity after 1945 were closed a few
years later because the government continued to transform
both the economic and the health system into the Semashko
model. The networks of private providers, however, started to
reestablish themselves after the fall of communism in the
1990s. The first private hospitals were established between
1993 and 1994 and began their operations as single departments
established mostly alongside outpatient health centers, and
only later developed further more advanced services. The
process of the development of the private sector in inpatient
care is much slower than in the other sectors previously
described. The main reason behind this largely relates to the
extent of the higher financial risk and investment required, and
the widespread existence of political beliefs opposing hospital
privatization.
Perhaps most significantly, there is often a public and political
unwillingness toward hospital privatization that is of particular
concern for postcommunist countries. During the communist
period, health care was generally considered to be a “social service”
that should not be determined by economic measures of efficiency
309
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 306–311
Fig. 2 – Structure of the ownership of ambulatory care facilities in Poland (1990–2008).
[31]. Consequently, strong public opposition remains one of the key
barriers to the privatization of health care, driven by beliefs that
which makes them so-called nonpublic self-government-owned
hospitals. The difference between such an organization and a
traditional one is that these hospitals are acting as companies
under commercial law and not as the “independent public health
care facilities” as it was before they were restructured. Table 1
summarizes the dynamics of the development of the nonpublic
hospital sector in Poland over the past two decades.
Of the 120 nonpublic hospitals currently operating in Poland,
77 are those that have been restructured by local self-government
units. This constitutes 64.2% of the number of private hospitals
and 10.4% of the total number of private hospitals in Poland [26].
In recent years, the general share of private hospitals equates to
approximately 20% of the total, and when taking the number of
beds into consideration, it is less than 6%. In 2007, nonpublic
hospitals signed contracts with the social health insurance funds
for an amount of more than 600 million Polish Zloty (145 million
euros), which equates to approximately 3.2% of the total public
resources spent on hospital care [26].
● hospital services are of a specific nature and they should not
be subject to a profit motive;
● private hospitals are acting mainly for the profit of their
owners, which is in conflict with the mission of health care
facilities;
● privatization of hospitals equals charging patients for the
services; and
● nonpublic hospitals are unwilling to provide highly complex
medical procedures because of their unprofitability, which
may cause a limitation of access to such services.
Furthermore, the lack of a comprehensive legal basis for
privatization constitutes an obstacle for the development of this
sector. Despite these limitations, the process is still progressing,
largely through the local self-governments, for which the transformation of hospital facilities they own is becoming increasingly
popular [32].
Since 1999, self-government units have transformed 130
hospital units into commercial companies; this includes 77
hospitals and 53 single hospital departments [33]. In many cases,
the situation is becoming paradoxical because self-government
units still remain formal owners of the privatized hospitals,
Cost-Effectiveness of Privatization
Because of issues regarding the availability of data, a comparison
of the cost-effectiveness of public and private health care poses a
significant problem. Private entities generally do not release
relevant information, and available sources are generally limited
Table 1 – Structure of ownership of hospital care facilities in Poland (1999–2007).
No. of private
hospitals
No. of beds
Total hospitals
Total beds
% of private
hospitals
% of private
beds
1999
2000
2001
2002
2003
2004
2005
2006
2007
21
30
45
61
72
147
170
153
170
446
715
198,688
2.9
1,574
716
190,952
4.2
2,476
736
188,234
6.1
4,221
739
188,038
8.3
5,171
732
186,043
9.1
7,649
790
183,280
18.6
8,215
781
179,493
21.8
9,318
742
176,673
20.6
10,204
748
175,023
22.7
0.2
0.8
1.3
2.2
2.8
4.2
4.6
5.3
5.8
310
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 306–311
to the presentation of a few selected individuals. Existing data,
however, suggest an average growth of within 20% to 30% in
private sector revenues annually. The overall size of the sector
was estimated at about 1915 million Polish Zloty (462 million
euros). This information confirms an analysis by the National
Association of Private Employer Health Care (2009), who conducted a study involving nine private companies providing
medical services. The analysis indicated that in 2007 and 2008,
the annual growth of the sector revenue amounted to 31.1% and
33.1%, respectively, and the value of income in 2008 was estimated to be 930.4 million Polish Zloty (225 million euros). On
average, survey participants within the health profession indicated that their income in 2006 increased by approximately 32.9%
in 2007 and 35.3% in 2008. In comparison with the public sector,
employers’ contributions to medical care for their employees
were the dominant source of revenue for private providers. These
accounted for nearly 44% of the revenue [34].
comparatively limited. This is largely due to complicated and
unclear legal regulations, resulting in a gray area between public
and private health care where informal payments impede the
intended function of the system. If this is left unchanged, the
official public health care sector in Poland is likely to become an
increasingly residual service for the worst-off patients, who are
unable to afford the legal private sector or the informal payments
that guarantee a higher quality service in the public sector.
Of prime importance, however, as with many postcommunist
countries, one of the key barriers to health care privatization is
the traditional belief that health care should be a “public service,”
not determined by economic measures of efficiency. With the
heritage of the communist period hard to overcome, strong
public opposition will remain one of the key barriers to the
privatization of health care.
Source of financial support: This study was supported by the
National Institute for Health Research Collaboration for Applied
Health Research and Care for the South West Peninsula (to H.F.).
Quality of Care—Private versus Public
The private and public sectors in Poland also differ in terms of the
perceived quality of care reflected in patients’ opinions. Several
opinion polls dedicated to this issue share the same conclusion—
in general, private services are considered to be of a higher
quality, with the most significant differences stated to be the
physician’s manner toward patients. In a survey conducted by
the marketing research and opinion poll company Partner in
Business Strategies in 2007 [35], respondents compared public
and private services with respect to atmosphere, doctor’s commitment, and respect to patient privacy. For these three criteria,
private providers were considered significantly better than the
public sector. With respect to atmosphere, 54% of the respondents declared that private providers are better, with only 10%
declaring the opposite. Results for doctors’ commitment and
respect for patient’s privacy were also in favor of the private
sector (51% and 47% for private providers and 9% and 8% for
public providers, respectively). The general quality of services
was also recognized as being better in the private sector, and of
all respondents, 45% declared the private sector as being superior,
with only 10% having the opposite opinion. One of the few areas
in which the study showed no significant advantage to the
private sector was with respect to the experience of the doctors.
Although 19% of the respondents thought that more experienced
and better qualified doctors worked in private health care, 17% of
the respondents considered the quality of doctors better in the
public sector. It is worth noting at this point that almost twothirds of the respondents had no opinion on this matter, which
may suggest that differences in knowledge and experience are
difficult to judge for patients.
Conclusions
A free health care market does not exist in Poland, nor is it
planned to be one. Since the dissolution of state socialism,
private financing of health care services has increased substantially and has filled an important role in meeting the increasing
consumer demand in some areas of health care and, to some
extent, encouraging a more efficient use of scarce resources
through competition and private initiative for public health
services. This step toward privatization is mostly evident within
the pharmaceutical industry, which is now fully privatized. In
addition, the ambulatory sector has systematically increased its
private market share to approximately 70% of all services provided and is continually moving toward full privatization. Despite
the regular increase in the number of private hospitals in
Poland, however, their role in the health care system remains
R EF E R EN C ES
[1] McMenamin I, Timonen V. Poland’s health reform: politics, markets
and informal payments. J Soc Policy 2002;31:103–18.
[2] Mishtal J. Neoliberal reforms and privatisation of reproductive health
services in post-socialist Poland. Reprod Health Matters 2010;18:56–66.
[3] Ostrowski T, Wdowiak L. Prywatyzacja opieki zdrowotnej w Polsce
[Characteristics of the privatization of health care in the province].
Zdrowie Publiczne 2002;112:86–92.
[4] Jaworska-Łuczak B. Prywatyzacja w ochronie zdrowia w Polsce-analiza
stanu obecnego i prognoza na przyszłość [The privatization of health
care in Poland—analysis of the current state and outlook for the
future]. Zdrowie Publiczne 2002;112:93–103.
[5] Kuszewski K, Gericke C. Health Systems in Transition: Poland (Vol 7).
Copenhagen, Denmark: World Health Organization Regional Office for
Europe on behalf of the European Observatory on Health Systems and
Policies, 2005:1–106.
[6] Maarse H. The privatization of health care in Europe: an eight-country
analysis. J Health Polit Policy Law 2006;31:981–1014.
[7] McLaughlin D, Smith D. Doctors go west in Polish brain drain. The
Observer May 15, 2005.
[8] Brandt T, Schulten T. Liberalisation, Privatisation and Regulation in the
Polish Healthcare Sector/Hospitals: Privatisation of Public Services and
the Impact on Quality, Employment and Productivity (PIQUE).
Düsseldorf, Germany: Wirtschafts- und Sozialwissenschaftliches
Institut, 2006.
[9] Kruk W, Wdowiak L. Prywatyzacja w opiece zdrowotnej [Analysis of the
current state and outlook for the future]. Medycyna Ogólna
2006;12:171–85.
[10] Tymowska K. Prywatyzacja opieki zdrowotnej w teorii i praktyce
[Prywatyzacja opieki zdrowotnej w teorii i praktyce]. Prawo Medycyna
2000;5:138–46.
[11] Organisation for Economic and Co-operation Development (OECD).
Glossary of Industrial Organisation Economics and Competition Law. Paris,
France: Organisation for Economic and Co-operation Development, 1993.
[12] Chawla M, Berman P, Kawiorska D. Financing health services in
Poland: new evidence on private expenditures. Health Econ 1998;7:337–46.
[13] Kopiak A. Barometr korupcji 2007. Raport z badań [2007 Corruption
Barometer. Research Report]. Warsaw: Stefan Batory Foundation, 2007.
Available from: http://www.batory.org.pl/doc/barometr-korupcji-2007.
pdf [Accessed May 8, 2013].
[14] Białowolski P, Czapiński J, Grabowska I, et al. Warunki życia
gospodarstw domowych. Diagnoza Społeczna 2011. Warunki i Jakość
Życia Polaków [Social diagnosis 2011. Objective and subjective quality
of life in Poland]. Contemp Econ 2011;5:45–159.
[15] Polish Parliament. Ustawa z dnia 13 lipca 1990 r. o prywatyzacji
przedsiębiorstw państwowych. Dz. U. 51, item 298 [The Act of 13 July
1990 on the privatization of state enterprises. No. 51, item 298]. SEJM of
the Republic of Poland, 1990.
[16] Polish Parliament. Ustawa z dnia 30 sierpnia 1996 r. o komercjalizacji i
prywatyzacji przedsiębiorstw państwowych. Dz.U. 118, item 561 [The
Act of 30 August 1996 on commercialisation and privatization. No. 118,
item 561]. SEJM of the Republic of Poland, 1996.
[17] Polish Parliament. Ustawa z dnia 30 sierpnia 1991 r. o zakładach opieki
zdrowotnej. Dz. U. 91, item 408 [The Act of 30 August 1991 on health
care. No. 91, item 408]. SEJM of the Republic of Poland, 1991.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 306–311
[18] Watson P. Fighting for life: health care and democracy in capitalist. Crit
Soc Policy 2011;31:53–76.
[19] Polish Parliament. Ustawa z dnia 8 marca 1990 r. o samorządzie
terytorialnym. Dz. U. 16, item 95 [The Act of 8 March 1990 on the
local government. No. 16, item 95]. SEJM of the Republic of Poland,
1990.
[20] Polish Parliament. Ustawa z dnia 24 listopada 1995 o zmianie zakresu
działania niektórych miast oraz o miejskich strefach usług
publicznych. Dz. U. 141, item 692 [The Act of 24 November 1995 to
change the scope of some of the cities and urban areas of public
services. No. 141, item 692]. SEJM of the Republic of Poland, 1995.
[21] Polish Parliament. Ustawa z dnia 5 czerwca 1998 r. o samorządzie
województwa. Dz. U. 91, item 576 [The Act of 5 June 1998 on the
provincial self-government. No. 91, item 576]. SEJM of the Republic of
Poland, 1998.
[22] Polish Parliament. Ustawa z dnia 5 czerwca 1998 r. o samorządzie
powiatowym. Dz. U. 91, item 578 [The Act of 5 June 1998 on the county
government. No. 91, item 578]. SEJM of the Republic of Poland, 1998.
[23] Polish Parliament. Ustawa z dnia 23 grudnia 1988 o działalności
gospodarczej. Dz. U. 41, item 324 [The Act of 23 December 1988 on
economic activity. No. 41, item 324]. SEJM of the Republic of Poland, 1988.
[24] Trzecia W, ed. Zespół do przygotowania raportu: Finansowanie
ochrony zdrowia w Polsce—Zielona Księga [Financing Health Care in
Poland—Green Paper]. 2005.
[25] Izabella-Michalewicz M. Sektor prywatny w systemie ochrony zdrowia
w Polsce [Private Sector in the Health Care System in Poland]. Warsaw,
Poland: Wydział Analiz Ekonomicznych i Społecznych, Kancelarii
Sejmu, 2004.
[26] Central Statistical Office. Mały rocznik statystyczny [Polish Statistical
Yearbook]. Warsaw, Poland: Central Statistical Office, 2009.
311
[27] Kachniarz M. Komercjalizacja samodzielnego publicznego zakładu
opieki zdrowotnej. Kluczowe warunki osiągnięcia sukcesu
[Commercialization of Independent Public Health Care. Key Conditions
for Success]. (1st ed.). Krakow, Poland: ABC - Wolters Kluwer Polska,
2008.
[28] Król ZJ. Prywatyzacja ambulatoryjnej opieki zdrowotnej [Privatization
of ambulatory health care]. Zdrowie Zarządzanie II 2000:29–32.
[29] Kowalska K. Managed care and a process of integration in health care
sector: a case study from Poland. Health Policy 2007;84:308–20.
[30] Golinowska S, Kozierkiewicz A. Quality in and Equality of Access to
Healthcare Services: country report for Poland. Brussels, Belgium:
European Commission, 2008.
[31] Profaska J. Prywatyzacja samodzielnego zakładu opieki zdrowotnej
jako decydujący element przekształceń własnościowych w ochronie
zdrowia [Privatization of health care themselves as a decisive element
of privatization in health care]. Przewodnik Menedżera Zdrowia
2001;1:15–9.
[32] Wojciechowski A. Prywatyzacja szpitali. [The privatisation of
hospitals]. Przegląd Urologiczny 2001;7:83–94.
[33] Ministry of Health. Informacja o przekształceniach własnościowych w
sektorze ochrony zdrowia, przeprowadzonych decyzją jednostek
samorządu terytorialnego w latach 1999–2009 (I półrocze). [Information
about Ownership Transformations in the Health Sector, Carried Out the
Decision of Local Government Units in the Years 1999–2009]. Warsaw,
Poland: Ministerstwo Zdrowia RP, 2009.
[34] Mrozowicki A. Poland: Industrial Relations in the Health Care Sector.
Wroclaw, Poland: Institute of Public Affairs, University of Wroclaw,
2011.
[35] Partner in Business Strategies. Służba zdrowia — publiczna a prywatna
[Healthcare—Public and Private]. Warsaw, Poland: Partner in Business
Strategies, 2007.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 312–318
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Transforming Public Servants’ Health Care Organization in Greece
through the Implementation of an Electronic Referral Project
Kyriakos Souliotis, PhD1,, Vasiliki Mantzana, PhD2, Manto Papageorgiou, MSc3
1
Faculty of Social Sciences, University of Peloponnese, Corinth, Greece; 2Department of Digital Systems, University of Piraeus, Piraeus, Greece; 3National School of
Public Health, Athens, Greece
AB STR A CT
Objective: The Greek Public Servants’ Health Care Organization aiming to organize, monitor, and enhance the health care services
provided to 1,500,000 public servants decided to respond to the
national alert of the economic crisis through the reduction of costs
caused by diagnostic tests (€300,000,000 claims for 2008), to improve
working conditions of contracted physicians and laboratories, and to
enhance services provided to insured members. In September 2010,
the Greek Public Servants’ Health Care Organization initiated a pilot
project that electronically records the prescription process of the
diagnostic tests, which is Web-based, is open source, and was
provided for free to the contracted physicians and diagnostic centers.
Methods: In this article, we present some interesting findings resulting from the implementation of the pilot electronic referral project by
examining a 9-month period. Results: Fifty-eight percent of the
physicians had the necessary equipment for the operation of the
system, more than 3600 physicians used it, 17,495 public servants had
been served through the system, and 178,456 paraclinical examinations had been prescribed with a cost of €1,394,980. In addition, the
analysis revealed that the implementation of an electronic referral
system could provide significant benefits, such as a faster referral
process, valid and coherent information, minimization of the risk of
misinterpreting the electronic referral due to illegibility of handwriting, and improvement in quality of services. Conclusions: The Greek
electronic referral system was one of the first attempts toward
creating the basis of a society of transparency and cost control. The
lessons learnt from this article should not be ignored in the process of
redesigning and improving the electronic referral system for Greece.
Keywords: E-health, electronic referral project, Greece, insurance
funds, physician.
Introduction
countless number of hours have been invested in the development of electronic health, more needs to be done.
In Greece, the adoption of e-health systems is happening
at an exceptionally slow pace. According to records of the
Health Ministry, the actions included in the second Community
Structural Fund relative to the sector of information technology
were of small scale and limited to the planning level. For this
reason, in the third Community Structural Fund, the main focus
was on the integration of information systems in health care
organizations. Handwritten prescriptions and referral letters had
been the main method for physicians to communicate decisions
on diagnostic tests and for diagnostic centers to conduct the tests
while being a source of information for the payer at the
same time.
In an effort to put an end to directed prescribing, to elevate
the medical expense in great heights, and to crack down the
phenomenon of overprescribing, Greek social insurance funds
decided to implement and adopt an electronic referral system.
Currently, the electronic referral project in Greece is at the
transitional stage between paper and Web. The transition from
the traditional process to the new electronic era offers unique
opportunities and challenges.
Health care systems consist of large and complex processes that
affect and get affected by multiple interacting actors, such as
physicians, nurses, patients, citizens, medical suppliers, and
health insurance providers, with different backgrounds, knowledge, organizational beliefs, interests, and culture. Health care
systems all over the world, however, face challenges, such as
1) increasing demand for health and social services due to an
aging population; 2) higher expectations by citizens who request
provision of high-quality health care services and reduction of
inequalities; 3) management of large amounts of health data that
should be available on request safely at any place and should be
processed for administrative purposes; 4) need to provide the
best possible health care services in conditions of budgetary
constraints; and 5) reduction of medical errors that lead to loss
of life or causing irreparable injury.
It has been reported that the implementation and use of
e-health systems, such as e-prescribing, electronic referral
systems, and electronic health care records, could support governments and organizations to confront the aforementioned
challenges. Although hundreds of billions of dollars and
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Kyriakos Souliotis, Faculty of Social Sciences, University of Peloponnese, Damaskinou & Kolokotroni Street,
Corinth 20100, Greece.
E-mail: [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.003
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 312–318
The introduction of electronic referral systems has been
proved to have a positive effect on prescribing and dispensing
processes, in the sense that an electronic referral system can
improve safety, quality, efficiency, and cost-effectiveness [1].
The purpose of this article was to present some interesting
findings and outcomes that the pilot electronic referral project
induced to Greek Public Servants’ Health Care Organization
(GPSHCO) by examining a 9-month period (September 1, 2010,
to June 1, 2011). In doing so, the authors present a review of the
literature on e-health systems as well as the methodology used
to conduct this research. Afterward, they analyze GPSHCO’s effort
to enhance health care services provided to 1.5 million insured
public servants and reduce costs that are provoked by diagnostic
tests prescribed to GPSHCO members (€300,000,000 claims for
2008) by launching the pilot electronic referral project on May 1,
2010. Moreover, they present the outcomes that this project
induced to the organization and draw conclusions.
Literature Review
Information systems (ISs) play an increasingly crucial role in the
health care sector by providing an infrastructure to integrate
people, processes, and technologies [2,3]. Health care information
systems (HISs) are defined as “computerized systems designed to
facilitate the management and operation of all technical (biomedical) and administrative data for the entire health care
system, for a number of its functional units, for a single health
care institution, or even for an institutional department or unit”
[4]. The use of computers in the health care sector can be traced
to the 1960s. The first attempts to adopt HISs were made during
the early 1970s [5]. In the past, however, HIS initiatives were
limited to the automation of business processes related to
administration and health care tools as well as techniques
related to various medical procedures such as diagnostic, therapeutic, and surgical procedures. During the 1980s, innovative
patterns in database designs and applications related to HISs led
to developments in planning and administration of the health
care data. In parallel, HISs also introduced low-cost financial
systems for hospitals fewer than 200 beds in size [6]. It should be
noted that the early computerized systems were limited to big
hospitals and government projects (military).
As the information technology industry flourished, the HIS
technology was populated with various network applications.
The net period with the Internet, intranet, and extranet affected
the communication of data in hospitals, especially in the 1990s.
In the middle of the same decade, the interface engine emerged
as a product to support the integration of applications, as best-ofbreed applications became harder to manage [7].
E-health systems such as electronic prescribing [8], electronic
referral systems [9], personal health records [10,11], asynchronous health care communication systems [12], and picture
archiving communication systems [8] have been applied in
health care to improve the capabilities of physicians and clinical
staff and provided increased services to patients, caregivers, and
citizens in general. E-health is an emerging field in the intersection of medical informatics, public health, and business,
referring to health services and information delivered or
enhanced through the Internet and related technologies.
World Health Organization defined e-health as being “the
cost-effective and secure use of information and communication
technologies in support of health and health-related fields,
including health-care services, health surveillance, health literature, and health education, knowledge and research.” Healthcare Information and Management Systems Society’s E-Health
SIG defined e-health as “the application of Internet and other
related technologies in the health care industry to improve the
313
access, efficiency, effectiveness, and quality of clinical and
business processes utilized by health care organizations, practitioners, patients, and consumers to improve the health status of
patients” [13]. In a broader sense, the term characterizes not only
technical development but also a way of thinking, an attitude,
and a commitment for networked, global thinking to improve
health care locally, regionally, and worldwide by using information and communication technology [14].
In a study conducted by Delloite/Ipsos in 2011 for the European Commission regarding e-health systems in Europe, it was
stressed that integrated systems for electronic referral are among
the e-health systems that need the greater attention. In the same
study, it was reported that an integrated system able to send or
receive electronic referral letters is currently available in 34% of
the 900 acute hospitals in the 30 countries surveyed [15].
E-referral projects among various countries differ a lot. It is
not likely that there is one solution that proves to be beneficial for
all countries because the solution depends on legislation, organization of the health care system, and cultural differences.
Electronic referral systems seem to have a large potential for
economic savings as a whole, but it takes a longer time than
expected to realize this potential.
Research Methodology
The authors have developed an empirical research methodology
to study electronic referral projects. This methodology is based
on three development stages: 1) research design development,
2) case study data collection, and 3) case study data analysis. The
research design adopted was the first independent part of the
empirical research methodology. The starting point was to review
the literature to develop an understanding of the research area
under investigation. From the literature review, several research
issues were highlighted for a more focused study. For the purpose
of this article, a single case study strategy was used to explore
and understand the topic under research. A qualitative case
study strategy can offer a “holistic” view of the processes
involved as well as a realization of the topic under research
[16]. To collect data, we used various techniques, such as questionnaires, documentation, and observation.
Case Study
In 1983, the Greek National Health System (ESY) was established
under the Law 1397/1983 [17]. The Ministry of Health and Welfare
Fig. 1 – Health care provision infrastructure; data from the
World Health Organization [18].
314
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 312–318
is responsible for the health care provision as well as for the
national health policy and strategy development. The ESY consists of the following three subsystems (Fig. 1), which operate
almost independently [19].
The National Health System
The ESY comprises public hospitals, health centers, and the
National Centre of Emergency Care. It provides inpatient care
and emergency prehospital care on a universal basis. It aims to
provide free and comprehensive health care coverage. The ESY
constitutes 123 general and specialized public hospitals (36,621
beds) and 9 psychiatric clinics (3,500 beds). Moreover, 32 of the
123 hospitals provide tertiary and highly specialized care [17].
Health care services are also provided by public hospitals, which
include 13 military hospitals financed by the Ministry of Defense,
5 hospitals of the Social Security Institution, and 2 university
hospitals operating under the authority of the National and
Kapodistrian University of Athens. Emergency prehospital care
is provided by the National Centre of Emergency Care, which is a
National Health System agency. Health centers also provide
emergency services, short-term hospitalization, and follow-up
of recovering patients, dental treatment, family planning services, vaccination services, and health education. In addition,
health care services are provided extensively by private health
care organizations (26%). Nowadays, 234 private hospitals
and clinics operate in Greece, with a total capacity of 15,397 beds
[17]. It appears that the ESY is a mixed system of public-private
funding and provision of health care services.
Social Insurance Funds
Health care services in Greece are offered through hospitals,
private clinics, and physicians as long as one has been registered
with a social insurance fund. Every individual (employer,
employee, pensioner, dependent family member, and child)
living in Greece has a unique personal Social Security Number,
named AMKA, to benefit from social security and to obtain a
health booklet. There are several social security institutions in
Greece, such as the GPSHCO and the Social Security Institution.
Registration with such institutions requires part-time or full-time
employment. Social security benefits are available only to
those who are registered and contributing to the social security
system. Social security contributions are paid by both the
employer and the employee or by the individual, in case one is
self-employed.
Refused to
join
15.4%
Will join in the
immediate
future
16.9%
Have joined
the pilot
project
66.6%
Fig. 3 – Acceptance of the electronic referral system for
diagnostic tests.
Private Health Sector
The private health sector has numerous diagnostic centers,
private clinics, laboratories, and so forth.
This case study was conducted at the GPSHCO whose aim is to
organize, monitor, control, and enhance the health care services
provided to public servants in terms of quality and funding.
The structure of the GPSHCO includes the Central Office and
57 regional health care services for the insured public servants
(YPAD). The GPSHCO has approximately 1,500,000 insured
members (public servants and family members). It offers health
care services to public servants. More precisely, the GPSHCO has
contracted partners (10,000 pharmacists, 13,000 physicians, 3,000
diagnostic centers and private clinics, and 700 physiotherapists)
that provide health care services to the insured members.
The remuneration system for social security funds is based on
Greek law. The GPSHCO receives revenues from the contributions
of the insurers that are paid on a monthly basis. In particular, the
GPSHCO receives 2.55% of the monthly wage of each insured
member paid by the employee plus an amount equal to 5.1% of
the monthly wage of each insured member paid by the employer,
that is, the state. These contributions are collected via the
Ministry of Finance that subsidizes the GPSHCO to pay its
obligations. The aforementioned contracted partners are paid
via the regional offices of the GPSHCO.
The GPSHCO has a fully integrated IS, which was implemented in early 2003 and consists of the following subsystems:
Fig. 2 – Digital behavior of the medical community.
Other reasons
for not joining
1.1%
protocol and price management of hospital stays of 57 YPAD;
protocol and economic management of GPSHCO’s Central
Office;
insured members’ registry for the Central Office and YPAD;
warehouse for the Central Office and YPAD;
management of insured members’ hospitalization in private
clinics;
payroll of GPSHCO’s employees;
Management information system—Business intelligence
based on budget figures for the Central Office and YPAD;
Safety management system (create users’ accounts, assign
roles, etc.); and
Disaster recovery system that covers all GPSHCO’s systems.
Although the GPSHCO has a quite advanced IS, it is not
effective enough in identifying abuse of the system by the
insured members. In particular, the insured member, presenting
with his or her health booklet, can visit a physician contracted
with the GPSHCO. The latter can refer the insured member for
diagnostic tests using the health booklet, and the insured
member can then visit a physician contracted with the GPSHCO
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 312–318
315
Fig. 4 – Number of physicians who used the system from October 1, 2010, to July 7, 2011.
Fig. 5 – Number of electronic referrals issued from October 1, 2010, to July 7, 2011.
Fig. 6 – Cost of electronic referrals (€) from October 1, 2010, to July 7, 2011.
or diagnostic center to execute the referral and conduct the tests.
The physician or the diagnostic center, at the end of each month,
submits claims to the GPSHCO. The GPSHCO registers to the IS
the claims made by each physician or the diagnostic center on an
invoice basis without registering any information regarding the
diagnosis or the type and cost of tests ordered. As a result, it is
316
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 312–318
Fig. 7 – Mean cost of electronic referrals (€) per patient from October 1, 2010, to July 7, 2011.
impossible for GPSHCO’s management to monitor and control the
increasingly rising health care expenditure. More specifically, in
2003, the cost of diagnostic tests and visits was €155,494,142.8
whereas in 2006, it was €266,391,245.4 and in 2009, €297,051,270.7.
At this point, it should be mentioned that during this period, the
number of insured members and their health status have not
changed.
In an effort to 1) control and reduce expenses that are provoked
by diagnostic tests and visits to physicians by GPSHCO’s insured
members, 2) improve the working conditions of contracted physicians and laboratories, and 3) enhance the services provided to
insured public servants of the organization, GPSHCO’s Committee
Board initiated a pilot project on September 1, 2010, that electronically records the prescription process of the diagnostic tests. The
platform is a Web-based open source and was provided for free to
the contracted physicians and diagnostic centers, which are its
main users. The contracted physicians use the platform to refer the
patient (insured member) for diagnostic tests by using a coded list
of diagnostic tests and diagnoses according to International Classification of Diseases, Tenth Revision, and diagnostic centers execute the
diagnostic tests and charge the GPSHCO.
The project has been assisted by a call center that promotes
the implementation of the system, communicates and supports
physicians, and provides information as well as technical support
Table 1 – Comparing the cost of diagnostic tests
between 2009 and 2010.
Month
Cost of diagnostic tests per year (€)
2009
2010
January
February
March
April
May
June
July
August
September
October
November
December
8,984,396.21
7,319,735.93
8,195,129.24
8,079,929.95
8,234,257.64
7,510,277.69
8,824,800.53
9,048,589.53
5,850,758.07
8,047,418.45
7,562,862.22
8,083,606.29
7,294,334.95
5,846,991.50
6,421,289.59
6,495,760.75
6,189,177.16
5,586,522.53
5,754,905.08
4,230,721.10
4,577,497.58
4,882,519.44
4,169,340.00
4,167,922.29
for the whole project. Furthermore, a Web site providing information on the application and answering frequently asked
questions posed by the stakeholders concerning the management of electronic referrals has been developed.
Discussion
The electronic referral project started on a voluntary pilot basis
in September 2010, and after 9 months (June 2011) its use
became mandatory for all physicians contracted with the
GPSHCO. During the first month of the pilot project implementation, a survey was conducted to study the digital behavior of
the medical community. According to the results of this survey,
44.5% of the physicians mentioned that they have all the
necessary equipment for the operation of the system, 22.5%
stated that they have part of the necessary equipment (usually
just a PC and a printer), and 29.1% said that they do not have
any of the necessary equipment (Fig. 2). Moreover, it should be
noted that 66.6% of them accepted to use the electronic referral
system (more than 3600 physicians) while 16.9% of the contracted physicians said they would start using it in the immediate future (Fig. 3).
In the following paragraphs, we present data regarding the
use of the system on a monthly basis for the pilot period under
study (September 2010 to June 2011).
People react differently toward new ideas, practices, or objects
as a result of variations in innovativeness. The diffusion of
innovation theory operationalizes innovative individuals as early
adopters of innovations in contrast with majority and late
adopters [20]. The electronic referral project started on a voluntary pilot basis in September 2010, and after 9 months (June 2011)
its use became mandatory for all physicians contracted with the
GPSHCO. At the beginning of the project, the number of physicians was relatively low, which started increasing in the middle
of the pilot project and decreased in the last 3 months of the
project (Fig. 4). When the GPSHCO’s Committee Board realized
this, and as during the pilot phase the system was continuously
improving, they made the use of the system mandatory for all the
contracted physicians.
Although the number of contracted physicians decreased in
the last 3 months of the pilot project, the number of insured
members asking for an electronic referral increased constantly
because the system provided significant advantages to them.
317
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 312–318
Table 2 – Summarizing data regarding electronic referral system from January 2011 to June 2011.
Month
January 2011
February 2011
March 2011
April 2011
May 2011
June 2011
No. of physicians
using the system
No. of insured
members visiting
physician
Visits
No. of electronic
referrals
Cost of electronic
referrals (€)
Mean cost of
electronic
referrals (€)
1,701
2,859
3,559
2,880
2,482
2,204
54,571
94,088
130,945
109,972
95,296
88,154
55,249
94,787
136,720
122,047
131,647
129,354
13,996
19,806
43,205
37,216
37,804
41,200
807,272.49
1,072,562.93
2,039,216.66
1,755,829.23
1,875,444.88
1,979,538.59
57.68
54.15
47.20
47.18
49.61
48.05
More specifically, the advantages of the implementation of the
electronic referral system included faster referral process, valid
and complete (coherent) information, and minimization of the
risk of misinterpreting the electronic referral due to illegibility of
handwriting. Moreover, such an implementation can increase the
quality of services provided to insured members. It should be
noted that in the 9-month pilot period, 88,154 insured public
servants had been served through the system and 41,200 electronic referrals had been issued (Fig. 5).
Moreover, 178,456 diagnostic tests had been prescribed, with a
total cost of €1,979,538.59 (mean cost of €48.05) (Figs. 6 and 7). The
mean cost of the electronic referrals per patient became approximately €47 compared with the mean cost on January 2010, when
the project was initiated, which was approximately €58 (Table 1).
Moreover, it should be mentioned that the cost for diagnostic
tests decreased by approximately 32% in 2010 (January to December) compared with 2009 (January to December), not only because of
the use of the electronic system but also because of administrative
decisions that the committee decided to implement (Table 1).
The vast majority of physicians using the system seem to have
normal prescribing behavior. Approximately 60% of them induce
less demand for diagnostic tests than the average, 16% of the
physicians induce greater demand than the average because of
their medical (Table 2) specialization, and 40% of the physicians
cause demand below the average, regardless of their specialty.
Conclusions
The Greek electronic referral system was one of the first attempts
toward creating the basis of a society of transparency and cost
control. Similar to other projects, the launch of the electronic
referral system aims at supporting Greece in tackling the economic crisis, and thus it is important to note that the memorandum of economic and financial policies published mentioned
that “[the] Government ensures that the e-prescribing system for
diagnostic tests currently piloted by GPSHCO is extended to all
social security funds” [21].
Moreover, the empirical evidence indicates the lack of technological education among physicians and their denial to invest in
new technologies. Inevitably, physicians have to deal with many
changes in their work environment as the progress of medical
science and technology is rapid. It is imperative for them to
improve their skills and requalify themselves at all times. In
addition, the analysis revealed that the implementation of an
electronic referral system could provide significant benefits, such
as faster referral process, valid and complete (coherent) information, and minimization of the risk of misinterpreting the electronic
referral due to illegibility of handwriting, and can also increase the
quality of services provided to insured members. Moreover, such
an implementation can decrease the mean cost of the electronic
referrals per patient, which became approximately €47 compared
with the mean cost on January 2010 when the project was initiated,
which was approximately €58. In addition, it can decrease the cost
for diagnostic tests. In this case study, the cost has decreased by
approximately 32% in 2010 (January to December) compared with
that in 2009 (January to December), not only because of the use of
the electronic system but also because of administrative decisions
that the committee decided to implement.
The lessons learned from the international experience should
not be ignored in the process of redesigning and improving the
electronic referral system for Greece. Electronic referral in Greece
is a reality against all predictions. Electronic referral can and
should be the beginning for the widespread of e-health, given
that the electronic referral administration affects public health
and economy. No claim for generalization is made, and the
lessons learned are a result of the description provided and do
not seek to be prescriptive. These lessons might be helpful to
those willing to implement electronic referral projects as well as
to researchers and practitioners.
Source of financial support: The authors have no other
financial relationships to disclose.
R EF E R EN C ES
[1] Franklin BD, O’Grady K, Donyai P, et al. The impact of a closed-loop
electronic prescribing and administration system on prescribing errors,
administration errors and staff time: a before-and-after study. Qual Saf
Health Care 2007;16:279–84.
[2] Wanless D, Charlesworth A, Walker I, et al. In: H. Treasury, ed., Securing
Our Future Health: Taking a Long-Term View (Vol. 2004). London: HM
Treasury, 2002.
[3] Ragupathi W. Healthcare Information Systems. Commun ACM
1997;28:81–2.
[4] Rodrigues RJ, Gattini G, Aalmeida G. Setting up Healthcare Services
Information Systems: A Guide for Requirement Analysis, Application
Specification, and Procurement. Washington, DC: Essential Drugs and
Technology Program, Division of Health Systems and Services
Development, PAHO/WHO, 1999.
[5] Hodge JG, Gostin LO, Jacobson PD. Legal issues concerning electronic
health information: privacy, quality, and liability. JAMA
1999;282:1466–71.
[6] Straggers N, Thompson CB, Snyder-Halpern R. History and trends in
clinical information systems in the United States. J Nurs Sch
2001;33:75–81.
[7] Beaver K. Healthcare Information Systems (2nd ed.). Boca Raton, FL.
Auerbach Publications, 2002.
[8] Menachemi N, Burke DE, Ayers D. Factors affecting the adoption of
telemedicine—a multiple adopter perspective. J Med Syst
2004;28:617–32.
[9] Nicholson C, Jackson CL, Wright B, et al. Online referral and OPD
booking from the GP desktop. Aust Health Rev 2006;30:397–404.
[10] Lafky DB, Tulu B, Horan TA. A user-driven approach to personal health
records. Commun Assoc Inf Syst 2006;17:1028–9.
[11] Lafky DB, Horan TA. Prospective personal health record use among
different user groups: results of a multi-wave study. Paper presented at
the Proceedings of the 41st Hawaii International Conference on System
Sciences, Waikoloa, Big Island, HI, 2008.
[12] Wilson EV. Asynchronous health care communication. Commun ACM
2003;46:79–84.
[13] Griskewicz M. HIMSS SIG develops proposed e-health definition
(Vol. 13). Chicago, IL: Healthcare Information and Management
Systems Society, 2002: 12.
318
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 312–318
[14] Eysenbach G. What is e-health? J Med Internet Res 2001;33:E20.
[15] Deloitte/Ipsos. eHealth Benchmarking (Phase III): Final Report. Brussels:
Deloitte/Ipsos, 2011.
[16] Zmud RW, Olson MH, Hauser R. Field experiment in MIS research.
Boston, MA: Harvard Business School Research Colloquium, 1989:97–
111.
[17] Hellenic Association of Pharmaceutical Companies. The
Pharmaceutical Market in Greece: Facts and Figures. Athens, Greece:
Hellenic Association of Pharmaceutical Companies, 2003.
[18] World Health Organization. WHO Highlights on Health in Greece.
Copenhagen, Denmark: World Health Organization, 2004.
[19] Koutsouris D, Aggelidis P, Berler A, Tagaris A. Integration of Healthcare
Information Systems. Athens, Greece: ebusiness Forum, Z3 Research
Team, 2005.
[20] Rogers EM. Diffusion of Innovations (4th ed.). New York: The Free Press,
1995.
[21] Hellenic Ministry of Finance. Memorandum of understanding
between the European Commision acting on behalf of the Euroarea member states, and the Hellenic Republic, 2010, p. 43.
Available from: http://ec.europa.eu/economy_finance/eu_
borrower/mou/2012-03-01-greece-mou_en.pdf. [Accessed
November 22, 2010].
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 319–327
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
CONCEPTUAL PAPER
Recommendations for Reporting Pharmacoeconomic Evaluations in Egypt
Gihan H. Elsisi, MSc1,, Zoltán Kaló, MSc, MD, PhD2, Randa Eldessouki, MSc, MD3,4, Mahmoud D. Elmahdawy, PharmD5,6,
Amr Saad, MSc, PhD7, Samah Ragab, MPA8, Amr M. Elshalakani, MD, MBA9, Sherif Abaza, MBA10
1
Pharmacoeconomic Unit, Central Administration for Pharmaceutical Affairs, Cairo, Egypt; 2Health Economics Research Centre, Eötvös Loránd University,
Budapest, Hungary; 3Scientific and Health Policy Initiatives, International Society for Pharmacoeconomics and Outcomes Research, NJ, USA; 4Faculty of Medicine,
Fayoum University, Fayoum, Egypt; 5Hospital Pharmacy Administration, Central Administration for Pharmaceutical Affairs, Cairo, Egypt; 6Misr International
University, School of Pharmacy, Cairo, Egypt; 7Pharmacovigilance Center, Central Administration for Pharmaceutical Affairs, Cairo, Egypt; 8Head Technical Office,
Central Administration for Pharmaceutical Affairs, Cairo, Egypt; 9Health Economics Unit, Ministry of Health, Cairo, Egypt; 10Market Access, Roche, Cairo, Egypt
AB STR A CT
Objective: Introduction of economic evaluations for pharmaceuticals
or other health technologies can help the optimization of outcomes
from resource allocations. This article aims to provide recommendations for researchers in presenting pharmacoeconomic evaluations in
Egypt with special focus on pricing and/or reimbursement applications of pharmaceuticals. Methods: The Minister of Health approved
the initiative of establishing a focus group of decision makers that
included academic and industry experts with experience in health
economics, pharmacovigilance, and clinical pharmacy. The focus
group has reviewed 17 economic evaluation guidelines available on
the Web site of the International Society for Pharmacoeconomics and
Outcomes Research for reporting health economic evaluations. To
develop core assumptions before preparing a draft report, focus group
meetings were held on a regular basis starting June 2012. The
recommendations were developed by using the Quasi-Delphi method,
taking into account current practices and capacities for conducting
pharmacoeconomic evaluations in Egypt. Conclusions: Worldwide,
health care decision makers are challenged to set priorities in an
environment in which the demand for health care services outweighs
the allocated resources. Effective pharmaceutical pricing and
reimbursement systems, based on health technology assessment
(HTA) that encompasses economic evaluations, are essential to an
efficient sustainable health care system. The Egyptian Ministry of
Health and Population was encouraged to establish a pharmacoeconomic unit, as an initial step, for the support of pricing and reimbursement decisions. We anticipate that standardization of reporting
would lead to a progressive improvement in the quality of submissions over time and provide the Egyptian health care system with
health economic evidence often unavailable in the past. Therefore,
recommendations for pharmacoeconomic evaluations provide an
essential tool for the support of a transparent and uniform process
in the evaluation of the clinical benefit and costs of drugs that do not
rely on the use of low acquisition cost as the primary basis for
selection. These recommendations will help inform health care
decisions in improving health care systems and achieving better
health for the Egyptian population.
Keywords: economic evaluation, Egypt, recommendations, reporting.
Introduction
pharmaceuticals and health technologies are critical for efficient
allocation of the limited resources.
To better allocate resources and with the growing awareness
of the importance of health technology assessment (HTA), the
Ministry of Health and Population (MOHP) established a pharmacoeconomic unit to support and inform pricing and reimbursement decisions [3]. No economic evaluations guidelines or
standards, however, have been set up yet.
This article provides recommendations based on reviewing
other countries’ national guidelines for economic evaluation as
well as experts’ opinions. Other factors influencing the feasibility
of conducting such studies in Egypt, including the complexity of
the health sector, the availability of data on health care outcomes
Egypt’s general budget devotes limited amounts to the health
sector. In the period 2008 to 2009, Egypt spent LE 61.4 billion
(Egyptian pounds) on health, which represents 5.9% of the
country’s gross domestic product. Out of the total health care
expenditure, pharmaceutical expenses constitute a large portion,
34%. [1]. In addition, over the past 16 years, the share of out-ofpocket spending in total health spending has increased dramatically from 51% to 72% [2]. These numbers suggest the increasing
need for optimizing the limited resources available. With the
growing public demand for improving health care services and
reducing the out-of-pocket expenses, economic evaluations of
Copyright & 2013, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article.
Address correspondence to: Gihan H. Elsisi, Central Administration for Pharmaceutical Affairs, 21 Abd Elaziz Alsoud Street, PO 11451,
Elmanial, Cairo, Egypt.
E-mail: [email protected]; [email protected].
2212-1099/$36.00 – see front matter Copyright & 2013, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.06.014
320
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 319–327
and the costs data, and current capacities for conducting pharmacoeconomic evaluations, were put into consideration in developing these recommendations.
Objective
This article aims to provide recommendations for researchers to
present pharmacoeconomic evaluations in Egypt with special
focus on pricing and/or reimbursement applications of pharmaceuticals. Policymakers are encouraged to consider these recommendations in developing the national guidelines for the
economic evaluation of pharmaceuticals.
Methods
As a self-initiated activity by government personnel, with the
approval of the Ministry of Health at the time, a focus group was
formed. The aim of the focus group was to develop a set of
recommendations and standards for economic evaluation studies used in applying for reimbursement and coverage to 1)
promote the concept of combining efficacy, safety, effectiveness,
and economic evaluation in the decision-making process; 2)
provide instructions for drug manufacturers: how to supply
information directly to health care decision makers to support
the use of their products; and 3) emphasize that simple assessment of acquisition cost is not a sufficient approach for the
control of overall health care expenditures.
To develop the recommendations, two steps were undertaken. The first step was to review the available national
economic guidelines. It included a review of 17 recently published
national economic evaluation guidelines for conducting and
reporting of economic evaluations (Table 1) that included an
English version available on the Web site of the International
Society for Pharmacoeconomics and Outcomes Research [4].
The second step was to solicit inputs and feedback from key
leaders and stakeholders through focus groups. For a comprehensive representation of key stakeholders in health care, focus
groups included decision makers experienced in health economics, pharmacovigilance, and clinical pharmacy, health providers
as well as researchers and experts selected from both industry
and academia, as shown in Table 2.
A consensus approach developed by using the Quasi-Delphi
method consisted of an iterative series of meetings and interrogations. Anonymous responses were synthesized into a series
of statements. Then, the synthesized statements were submitted
to the focus group members for comment until convergence or
stasis of opinion was identified in the third round.
Starting June 2012, focus group meetings were held on a
regular basis to develop core assumptions before preparing a
draft report. The discussions were recorded in written minutes.
The recommendations were developed by consensus approach,
taking into account current practices and capacities for conducting pharmacoeconomic evaluations in Egypt.
Developing Recommendations for Reporting
Pharmacoeconomic Evaluations
Disease and Product Background
Economic evaluations should provide information about the epidemiology of the disease and treatment pathways according to
most recent treatment guidelines. Data on the product should
include pharmacological class, proposed dosing regimen, route of
administration, and results of clinical studies performed to date [5].
Study Design
The study question should address the needs of the decision
makers by clearly establishing the context of the study. It should
provide details of the study perspective, the proposed product
and its comparator(s), the target population, and the effect on
specific subgroups where appropriate. Secondary questions that
relate to the primary study question should be clearly stated [6].
Perspective should be relevant to the research question and
adapted to benefits gained by the health care system. The
perspective adopted should maximize the health gain for the
population while representing the most efficient use of the finite
resources available to the Ministry of Health [7]. It should include
direct medical costs as well as additional costs, savings, or other
benefits when data are available.
The proposed product should be used primarily in the
approved indications with detailed information about its
Table 1 – Focus group members’ information.
Member of Focus
Group
Degree
Title
Organization
Government
Employee
Head of Pharmacoeconomic
Unit
Director, Scientific and Health
Policy Initiatives/Lecturer
Central Administration for
Pharmaceutical Affairs, Cairo, Egypt
International Society for
Pharmacoeconomics and Outcomes
Research, NJ, USA/Faculty of
Medicine, Fayoum University, Egypt
Central Administration for
Pharmaceutical Affairs/Misr
International University, Cairo,
Egypt
Central Administration for
Pharmaceutical Affairs, Cairo, Egypt
Central Administration for
Pharmaceutical Affairs, Cairo, Egypt
Ministry Of Health, Cairo, Egypt
Yes
Yes
Hoffmann-La Roche Ltd. Cairo, Egypt
No
Gihan H. Elsisi
MSc
Randa Eldessouki
MSc, MD
Mahmoud D. Elmahdawy
PharmD
Manager of Hospital Pharmacy
Administration/Part Time
Lecturer of clinical pharmacy
Amr Saad
MSc, PhD
Samah Ragab
MPA
Amr M. Elshalakani
MBBch,
MSc, MBA
MBA
Head of Pharmacovigilance
Center
Director of the Technical
Support Office
Head of Health Economics Unit
Sherif Abaza
Market Access & Governmental
Affairs Manager
No
Yes
Yes
Yes
321
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 319–327
Table 2 – The national health economic guidelines reviewed by the focus group members.
Title of the document
Source, Country
Published
Year
Guidelines for the Economic Evaluation of Health Technologies: Canada
Canadian Agency for Drugs and Technologies in
Health (CADTH), Canada
Taiwan Society for Pharmacoeconomics and
Outcomes Research, Taiwan
The Pharmaceutical Management Agency
(PHARMAC), New Zealand
Scottish Medicines Consortium, Scotland
2006
Guidelines of Methodological Standards for Pharmacoeconomic
Evaluations in Taiwan
Prescription for Pharmacoeconomic Analysis: Methods for Cost-utility
Analysis
Guidance to Manufacturers for Completion of New Product Assessment
Form
Guide to the Methods of Technology Appraisal
Guidelines for Preparing Submissions to the Pharmaceutical Benefits
Advisory Committee
Guidelines for Pharmacoeconomic Evaluations in Belgium
Health Technology Assessment Guideline
The Academy of Managed Care Pharmacy Format for Formulary
Submissions
General Methods for the Assessment of the Relation of Benefits to Costs
Guidelines for conducting Health Technology Assessment
Decree of the Ministry of Social Affairs and Health on applications and
price notifications made to the Pharmaceuticals Pricing Board–
Appendix: Guidelines for Preparing a Health Economic Evaluation
Procedure for Clinical and Economic Evaluation of Drug Lists That Are
Submitted for Reimbursement Coverage from Public Health Care
Budget.
The Guidelines for Pharmacoeconomic Evaluations of Medicines and
Scheduled Substances
Guidelines for the Submission of a Request to Include a Pharmaceutical
Product in the National List of Health Services
Guidelines for the Economic Evaluation of Health Technologies in
Ireland
Guidelines on How to conduct Pharmacoeconomic Analyses
technical characteristics (to differentiate it from its comparators),
regulatory status, and the specific application.
The selection of the comparator has to be justified. Comparators should be policy relevant; therefore, widely used and reimbursed health care technology for a given patient group and
indication is the preferred option. If no such technologies are
reimbursed in the tender list at the time the assessment is
conducted, the investigated product can be compared with the
most frequently used technologies to treat the same patient
groups. If a new product is used as first-line, second-line, or
third-line therapy, it should be compared with first-, second-, or
third-line therapies, respectively.
The targeted population should include both those who are
insured by the Egyptian health system and those who are uninsured. Parameters to define the population include baseline demographic characteristics, disease characteristics, treatment setting,
the context of past treatment, and any confounders adjusted [5].
Specific subgroups should be identified for those for whom
clinical effectiveness and cost-effectiveness may be expected to
differ from those of the overall population. Stratified analysis used
to quantify the differences in cost-effectiveness that may exist in
different subgroups is recommended because it may contribute
important information to the final advice. The evidence supporting the clinical plausibility of the subgroup effect should be fully
documented, including details of statistical analysis [8].
National Institute for Health and Clinical
Excellence (NICE), England and Wales
Pharmaceutical Benefits Advisory Committee
(PBAC), Australia
Belgian Health Care Knowledge Center (KCE),
Belgium
Journal of the Medical Association of Thailand,
Thailand
Academy Of Managed Care Pharmacy, United
States
German national institute for quality and
efficiency in health care (IQWiG), Germany
Poland Agency for Health Technology
Assessment, Poland
Ministry of Social Affairs and Health, Finland
2006
2007
2007
2008
2008
2008
2008
2009
2009
2009
2009
ISPOR Russia HTA Regional chapter, Russian
State Medical University, Russian Federation
2010
National Department of Health, South Africa
2010
Pharmaceutical Administration, Israel
2010
Health Information and Quality Authority, The
National Centre for Pharmacoeconomics,
Ireland
Norwegian Medicines Agency, Norway
2010
2012
Appropriate Pharmacoeconomic Method
The choice of method of analysis depends on the research
question and must be justified. If the compared health technologies result in equal health gain, cost minimization analysis is
the preferred analytical approach.
If at least one of the compared health technologies is better
than the other, and the clinical benefit can be aggregated and
interpreted as naturalistic clinical outcomes, cost-effectiveness
analysis (CEA) is the preferred method. CEA, where an intermediate marker is chosen, must have a validated, wellestablished link with an important hard end point (e.g., patient
survival, heart attack, and bone fracture) [9]. Because the measure of primary clinical outcome may differ in different therapeutic areas, CEA cannot be used to compare or rank the costeffectiveness of a broad set of products.
If the quality of life of patients is an important clinical
outcome in the treatment course of patients, cost-utility analysis
is the preferred analytical approach. In cost-utility analysis, the
health gain is expressed in a combined single measure of lifeyears and health-related quality of life (HRQOL), for example, in
quality-adjusted life-years (QALYs) [10]. Ignoring quality-of-life
differences among products would provide less than complete
data to decision makers to address the health care dilemma of
where to allocate resources [11]. Adherence to the reference case
322
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 319–327
approach for estimating QALYs for inclusion in economic evaluations would facilitate comparability [12].
Time Horizon
In choosing the time horizon, it should be ensured that the chosen
outcome and the resource consumption of the treatment alternatives are observable in this period to reflect the course of the disease
and the effects of the interventions. The same time horizon should
be applied to both costs and outcomes [9]. A decision to use a shorter
time frame should be justified. When extrapolating data beyond the
duration of the study, assumptions regarding future treatment
effects and disease progression should be clearly outlined. Censoring
might be used to account for the incomplete information [13].
Choice of Outcome Measure
The choice of outcome parameters depends both on the indication and on the research question. Primary outcome measures are
the first choice whenever possible. When an intermediate end
point is used, it must have a high degree of predictability of the
final end point.
HRQOL is an appropriate outcome indicator for the evaluation of
health status. HRQOL can be measured by using generic questionnaires, disease-specific questionnaires, or preference-based measures. If HRQOL is to be included in the study design, this variable
must be measured by validated instruments. The direct use of the
EuroQol five-dimensional questionnaire, six-dimensional health
state short form (derived from short-form 36 health survey), or
similar generic measures is recommended, because they are easy to
use and interpret and are based on preferences of the general
public. If the use of disease-specific HRQOL instruments increases
the sensitivity of measurement, mapping of disease-specific HRQOL
results with the EuroQol five-dimensional questionnaire or similar
generic measures can be useful to translate the findings into QALYs.
Information on the changes in the health state should be
reported directly by the patient or the caregiver. A valuation of
these changes in the health state should then be reported for the
general population. The outcome parameter chosen must be
sensitive, valid, and consistent [14].
Synthesis of Clinical and Economic Evidence
Evidence synthesis has to be based on objective, systematic, and
reproducible search criteria. Estimation of health gain must be
based on scientific literature review and/or results of primary
data collection, and the best available evidence should be
considered. Meta-analysis based on large randomized controlled
trials is the highest hierarchy of evidence with the heterogeneity
of data accounted for. If compared drug therapies differ in
adherence or persistence of patients, then these factors should
be incorporated in calculating the relative effectiveness. In case
of orphan drugs where randomized controlled clinical studies
have not been conducted, the results of uncontrolled clinical
studies can be accepted, including studies with small sample
size. All product safety data need to be included whether from
clinical studies or from national and foreign pharmacovigilance
centers and patient registries with attention given to those that
differ substantively among the products being compared [15].
Economic evidence should be synthesized from systematic
review of the local data sources and the best available evidence.
Costs Determination
Resource use data should be obtained mainly from primary data
collection (e.g. health care providers or non-interventional studies) from Egypt; if not available, secondary data sources such as
local administration, accounting data, or patient chart review data
can be used. Official sources of unit cost data for products
(e.g., tender lists) are preferable. In the absence of a published
tender list price, the price submitted by a manufacturer for a
product may be used. The quality, validity, relevance, and generalizability of local data should be clearly described. Both
estimated consumption of resources and their unit prices must
reflect real-world settings in Egypt because relative and absolute
price levels differ among countries [16].
Resource use and costs should be identified, measured in their
natural units and values [17]. The primary perspective for these
studies is the overall health care services. Therefore, the resources that should be considered are direct medical costs, which
include drugs, medical devices, medical services including procedures, laboratory, or diagnostic tests, hospital services and
emergency department visits, and primary care visits. Other
direct nonmedical and indirect costs paid by patients, including
lost productivity costs, might be included only in the sensitivity
analysis. If indirect costs are included in the analysis, the rationality of the costs and how they are estimated should be
explained. Current and future costs arising as a consequence of
a product, and occurring during the specified time frame of the
study, should also be included. Mean values should be used.
Different costs or costs of the same resources that are used in
different quantities should be included in the analysis [18].
Out of the two general approaches to determine costs, microcosting and macro-costing, macrocosting is preferred [19]. The
source of cost data must be reported in detail. Data should be the
most recently available, with the cost year specified. Retrospective
input costs should be inflated to the most recent calendar year by
using the Consumer Price Index for health [20]. The drug cost used
should reflect the formulation and pack size that gives the lowest
cost. For drugs available in the outpatient pharmacies, the full
public price should be used for calculating costs. For hospital
products, the wholesale price should be used for cost calculations.
Future costs should be calculated at constant current costs; therefore, results are not subject to uncertainty in future inflation rates.
Modeling
Economic modeling based on prospectively collected data is the
preferred method by decision makers in an increasing number of
countries to aggregate the expected costs and health effects for
all options relating to appropriate population and subpopulations, based on the full range of existing evidence [21]. The major
aim of applying modeling techniques is to aggregate short- and
long-term outcomes in the most appropriate time horizon.
The results of economic modeling studies presented should take
into account the following requirements: 1) the model should be
described in detail and should correspond to real practice of patient
management; 2) the model should be as simple as possible, and
easily understood; and 3) to facilitate assessment of the outputs of a
model, full documentation of the structure, data elements, and
validation of the model should be addressed in a clear manner,
with justification provided for the options chosen and presented
through diagrams (e.g., decision trees and Markov models) [22].
In addition, the model should be adapted to exclude clinical
events not expected to differ among the comparator products
[20]. For state transition models, such as Markov models, the
cycle length should be sufficiently short to ensure that multiple
changes in disease, treatment decisions, or costs do not occur
within a single cycle. Heterogeneity in the population should be
accounted by disaggregating the population into clinically plausible subgroups that require different structural assumptions.
The internal validity of the model should be tested before using
to ensure that the model is robust. The external validity should
be tested by comparison of the results with those generated by
other models and explaining differences if they exist.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 319–327
Discounting
Discounting should be made according to the time horizon. Any
costs or outcomes occurring beyond 1 year should be discounted
by using standard methods [19]. For comparability of results
across evaluations, it is important that a common discount rate
is used. Because constant prices and outcomes are used in the
economic evaluation, there is no need to take into account
inflation in the discount rate. A real discount rate of 3.5% per
year should be used for both costs and health gains. The discount
rate should be varied from 2% to 6% in the sensitivity analysis.
Uncertainty
Data for a health economic analysis are derived from various
sources, and this may be incomplete and affected by uncertainties. In a sensitivity analysis, critical component(s) in the calculation should be varied through a relevant range or from the
worst case to the best case, and the results recalculated [13].
Probabilistic sensitivity analysis (PSA) is an appropriate
method for exploring uncertainty around the true mean values
of cost and efficacy inputs in decision-analytic modeling. In PSA,
however, probability distributions are applied by using specified
plausible ranges for the key parameters rather than the use of
varied point estimates for each parameter. Its results are difficult
to interpret for decision makers, while the stochastic approach,
such as deterministic sensitivity analysis (DSA), examines how
parameter variables (included as point estimates) affect the
model output [23]. We propose, given the difficulty in interpreting
the PSA, that DSA should be required, while PSA remains
optional.
To avoid potential bias and uncertainty that arise from the
modeling process, assumptions about the model structure should
be clearly stated and justified and their impact on costeffectiveness explored though a series of plausible scenario
analyses so that whether the study results will be changed can
be observed. All choices and the ranges of the parameters, and
the method used in sensitivity analysis, should be clearly
explained.
Present Study Results
Total costs and health outcomes must be reported separately,
and the aggregated result be explained. All parameters used in
the estimation of clinical effectiveness and cost-effectiveness
should be itemized in a tabular form with data sources transparently. Negative results should be reported. Incremental costeffectiveness ratio has to be calculated, unless one of the
compared health technologies dominates the other one. In
addition, the potential impact of the introduction of the new
treatment on the society needs to be assessed [24].
Where more than two products are being compared, the
results should be presented in the order of increasing costs and
the incremental cost-effectiveness ratio calculated by comparing
each product with the one above it, excluding those products that
are dominated. Equity issues, affordability, and resource constraints should be considered in judging the cost-effectiveness of
a product for reimbursement [20].
Tornado diagrams are useful tools to display DSA. If PSAs are
performed, the probability that the intervention is cost-effective
at a range of threshold values should be reported and the data
should be displayed graphically to facilitate the uncertainty
interpretation [9].
Equity and Generalizability Issues
To meet the needs of the decision makers, an attempt should be
made to include equity considerations in the study report. The
323
equity assumption of the basic case in economic evaluations
means that all patients should have a fair participation opportunity and obtain the expected treatment outcomes.
To determine equity in economic evaluation, we propose that
all lives, life-years, or QALYs should be valued equally, regardless
of the age, gender, or socioeconomic status of individuals in the
population [12]. The equity assumption should be included in
every model and analytical method of economic evaluations and
must be clearly stated.
Analysts must consider two specific areas of concern regarding the generalizability of clinical and economic data in the
assessment of technologies. The first area of concern is the
extent to which the clinical efficacy data are representative of
the likely effectiveness and similarly the extent to which economic data are representative of the costs and resource utilization [8]. The second area of concern is the generalizability of the
economic and clinical data across different patient ages and
genders as well as regional differences in health care practice
within Egypt. These areas of concern should be identified and
discussed, and the likely effect on the results and conclusions of
the report should be highlighted [25].
Discussion
There is an increasing need for justification of resource allocations and policy decisions, especially with the scarcity of public
resources. Fig. 1 shows that among all middle-income countries
in the region, Egypt invests a smaller proportion of its gross
domestic product on health care [1]. Investments in economic
evaluation studies and development of pharmacoecomic guidelines and expertise will help in allocating these limited resources
in the most efficient way to improve health care services.
The current reimbursement decisions in Egypt are based on
the lowest price after clinical review and approval of efficacy and
safety of the medication by the Procurement Technical Committee. The Procurement Technical Committee reviews all MOHP
hospitals and primary care units’ needs of medications and
applications submitted by drug manufacturers. It then decides
whether this medication is to be listed or not, according to
pharmacokinetics, pharmacodynamics, safety, and efficacy.
Then, the applications go to the Committee for Financial Offers
at the MOHP to review the financial issues and decide which drug
manufacturer or wholesaler, the one that presents the lowest
price for each active ingredient (medication), is to get reimbursement. Drug manufacturers or wholesalers who submitted identical price levels for the same active ingredient are given
reimbursement by an equal process. Fig. 2 presents the decision
makers and influencers in the Egyptian pricing and reimbursement decision-making processes [26].
There is a growing need to incorporate high-quality economic
evaluation studies into the reimbursement decision-making
process to adequately evaluate clinical and economic benefits
of medications in addition to the assessment of their acquisition
costs. These evaluations will improve decision making with
prioritizing our resources, which results in reducing our huge
expenditure on pharmaceuticals and save these resources to be
allocated to other cost-effective health technologies. In Egypt,
there is no limited budget that should be allocated for drug
coverage only but is allocated to the whole health sector.
The submission of an economic evaluation is currently recommended in Egypt. And an economic evaluation guideline to
standardize the process and provide a transparent and uniform
approach was approved by the Ministry of Health. There is a big
chance that these recommendations will be implemented in
Egypt. Policymakers are encouraged to consider these
324
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 319–327
10.00%
9.00%
8.00%
7.00%
6.00%
5.00%
4.00%
3.00%
2.00%
1.00%
0.00%
Overall Health Care Spending as Percentage of Gross Domesc Product
Fig. 1 – Egypt in comparison with other middle-income countries in overall health care spending as percentage of gross
domestic product. Data from USAID [1].
recommendations in developing the national guidelines for the
economic evaluation of pharmaceuticals.
The Canadian guidelines reported that by providing standards
for conducting and reporting of economic evaluations, the current limitations of evaluations can be addressed and lead to
better study [9]. It is important to note that the standardization of
reporting and other policies in the United States shared in the
bulk of the estimated $2 trillion savings [27]. We anticipate that
the standardization of reporting would lead to a progressive
improvement in the quality of submissions over time and provide
the Egyptian health care system with data often unavailable in
the past.
In developing those recommendations, we chose to build on
the learning experience from other countries and modify and
adapt the knowledge acquired to fit the Egyptian setting. In doing
so, duplication of efforts and use of resources much needed
elsewhere are avoided. As a rule, certain elements of HTA reports
are transferable, but adjustment to local data is absolutely
necessary [16]. Copying recommendations based on international
HTA without local adjustment may do more harm than good.
Putting this into consideration, our recommendations were tailored to the current settings and environment in Egypt while using
the current guidelines as an initial benchmark. Our starting point
is built on many years of experiences and expertise worldwide.
Fig. 2 – The Egyptian pricing and reimbursement decision-making processes [26]. Bold boxes, decision-making bodies; boxes,
decision influencer bodies; bold arrows, required step in the decision-making process; arrows, may or may not affect
decision.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 319–327
325
Table 3 – Key elements of the recommendations for reporting pharmacoeconomic evaluations in Egypt.
Key elements
The Egyptian recommendations
Differences/
similarities across
the national
guidelines
reviewed
Rationale for inclusion in the Egyptian
setting
It is a common agreed-upon element that
captures all the benefits when data are
available representing the most efficient
use of the finite resources.
According to the Egyptian Ministry of Health
regulations, the use of the product in
unapproved indications is forbidden.
Because of policy-related problems such as
drug supply shortage, we have to use the
available technologies.
Perspective
It should be relevant to the research
question and adapted to benefits gained
by the health care system.
Common
Indication
It should be used in the approved
indications.
Common
Choice of
comparator
Comparators should be policy relevant. The
widely used and reimbursed health care
technology for a given patient group is the
preferred option.
Both those who are insured and uninsured
by the Egyptian health care system.
Different
Subgroup
analysis
Only for those for whom clinical
effectiveness and cost-effectiveness may
be expected to differ from that of the
overall population.
Common
Preferred
analytical
technique
Any of CMA, CEA, and CUA considered.
Different
Time horizon
It should be ensured that the chosen
outcome and the resource consumption of
the treatment alternatives are observable
in this period.
Primary outcome measures are the first
choice. CEA, where the intermediate
marker is chosen, must have a validated,
well-established link with an important
hard end point. In CUA, outcomes are
measured in QALYs gained.
The direct use of the EQ-5D questionnaire,
SF-6D, or similar generic measures is
recommended.
Common
Target
population
Choice of
outcome
measure
Preferred
method to
derive utility
Synthesis of
clinical and
economic
evidence
Evidence synthesis has to be based on
objective, systematic, and reproducible
search criteria. The results of metaanalysis are preferable with the
heterogeneity of data accounted for.
Economic evidence should be synthesized
from systematic review of the local data
sources and the best available evidence.
Different
Because of the existing widespread
uninsured population that is covered by
other forms of health coverage, there is a
need to assess the effectiveness among
different categories of access to health
care.
When a distinct group differs from the
overall population, a subgroup analysis is
essential to reflect the actual clinical
benefit and provide an accurate estimate
of the cost-effectiveness of the therapy
across all population groups to better
inform decision on reimbursement.
It depends on the research question. When
the clinical benefit is interpreted as
naturalistic clinical outcomes, CEA is the
preferred method while CUA is the second
option because the concept of using
QALYs might not be well understood by
the majority of decision makers.
To accurately reflect the course of the
disease and the total effects of the
interventions.
Common
It depends both on the indication and on the
research question.
Common
They are easy to use and interpret and are
relevant to the Egyptian public
educational level and preferences. After a
period of time allowing knowledge
building and according to the learning
curve, more sophisticated instruments
might be considered.
Meta-analysis based on large randomized
controlled trials is the highest hierarchy of
clinical evidence and is recommended for
clinical benefit evidence. However, economic benefit evidence is nontransferable
among the countries and should be
obtained from local data sources and the
best available evidence.
Common
326
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 319–327
Table 3 – continued
Key elements
The Egyptian recommendations
Differences/
similarities across
the national
guidelines
reviewed
Rationale for inclusion in the Egyptian
setting
In most cases, data and information on
indirect costs are lacking in Egypt;
therefore, direct costs estimation is
recommended.
Both estimated consumption of resources
and their unit prices must reflect realworld settings in Egypt because relative
and absolute price levels differ among
countries.
Costs to be
included
Direct medical costs as well as additional
costs, savings, or other benefits when data
are available.
Different
Sources of
costs
Primary data collection; if unavailable,
secondary data sources can be used such
as local administration, accounting data,
and patient chart review. Official sources
of unit cost data for products (e.g., tender
lists) are preferable.
Modeling options include decision trees and
Markov models. The model should be
described in detail and should correspond
to real practice of patient management.
Common
Discounting
costs and
outcomes
A discount rate of 3.5% per year should be
used for costs and outcomes.
Common
Uncertainty
Critical component(s) in the calculation
should be varied through a relevant range
or from the worst case to the best case.
DSA should be required, while PSA
remains optional.
All lives, life-years, or QALYs should be
valued equally, regardless of the age,
gender, or socioeconomic status of
individuals in the population.
The generalizability and the extent to which
the clinical efficacy data and the
economic data are representative should
be identified and discussed.
Total costs and health outcomes must be
reported separately, and the aggregated
result be explained. ICER has to be
calculated. The probability that the
intervention is cost-effective at a range of
threshold values should be reported and
displayed graphically.
Different
Modeling
Equity issues
Generalizability
Presenting
results
Common
These models are easy to use, interpret, and
aggregate the expected costs and shortand long-term outcomes captured in the
most appropriate time horizon relating to
the Egyptian population and subpopulations.
Because constant prices and outcomes are
used in the economic evaluation, there is
no need to take into account inflation in
the discount rate.
With the current level of knowledge, results
of PSA are difficult to interpret by
personnel reviewing the studies for
coverage decisions.
Common
All patients should have a fair participation
opportunity and obtain the expected
treatment outcomes.
Common
Because of the presence of wide regional
differences in health care practice among
urban and rural areas, generalizability of
the studies should be discussed in detail.
Detailed information should be provided to
facilitate the interpretation of results,
thus allowing for a more transparent and
uniform process for the final coverage
decision.
Common
CEA, cost-effectiveness analysis; CMA, cost minimization analysis; CUA, cost-utility analysis; DSA, deterministic sensitivity analysis; EQ-5D,
EuroQol five-dimensional; SF-6D, six-dimensional health state short form (derived from short-form 36 health survey); ICER, incremental costeffectiveness ratio; PSA, probabilistic sensitivity analysis; QALYs, quality-adjusted life-years.
Key elements common across all national guidelines
reviewed were included in the recommendations. Other key
elements differed between the various guidelines such as the
choice of comparator, preferred analytical technique, target
population, costs to be included, and uncertainty. Through a
consensus approach between all focus group members
for these elements, we recommended the best fit to Egyptian
settings that are applicable to the current Egyptian environment. A summary of the key elements of the recommendations
and the rationale for their inclusion within the Egyptian setting are presented in Table 3 highlighting the elements
that were common across all guidelines and the ones that
varied.
HTA implementation in Egypt, however, is significantly challenged by the diversity and heterogeneity of the health care
system, limited tradition for national treatment guidelines, and
limited availability of epidemiological, health outcomes, and cost
data. Data for economic evaluations are low quality, region and
provider specific, unavailable in electronic records, and, in most
cases, not updated. So, the common opinion is “HTA cannot be
implemented in Egypt.” In fact, there are no perfect data for
health care research; we have to assess realistically how wrong
they have to be not to be considered useful. Having some data is
better than having no data at all; conducting pharmacoeconomic
evaluations and outcomes research in Egypt would also greatly
improve the quality of current data.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 319–327
Therefore, it will be important to evaluate the effect of the
implementation of these recommendations on reporting in
future economic evaluations in a manner similar to Consolidated
Health Economic Evaluation Reporting Standards (CHEERS):
ISPOR Task force report [28]. As methods for the conduct of
economic evaluation continue to evolve, it will also be important
to revisit our recommendations. These recommendations were
presented to the Assistant Minister of Health, and initial steps
required to start building the capacity of the pharmacoeconomic
unit are underway. A young generation of government personnel
is enthusiastic to enter this field; recent graduates from the first
health economic diploma program in the Middle East are keen to
facilitate the implementation of HTA in Egypt.
[8]
[9]
[10]
[11]
[12]
[13]
Conclusions
Worldwide, health care decision makers are challenged to set
priorities in an environment in which the demand for health care
services outweighs the allocated resources [29]. Effective pharmaceutical pricing and reimbursement systems, based on HTA that
encompasses economic evaluations, are essential to an efficient
sustainable health care system [30]. The MOHP was encouraged to
establish a pharmacoeconomic unit, as an initial step, for the
support of pricing and reimbursement decisions. We anticipate
that the standardization of reporting would lead to a progressive
improvement in the quality of submissions over time and provide
the Egyptian health care system with health economic evidence
often unavailable in the past. Therefore, recommendations for
pharmacoeconomic evaluations provide an essential tool for the
support of a transparent and uniform process in the evaluation of
the clinical benefit and costs of drugs that do not rely on the use of
low acquisition cost as the primary basis for selection. Eventually,
these recommendations will help inform the health care decisions
in improving health care systems and achieving better health for
the Egyptian population.
Source of financial support: The authors have no other
financial relationships to disclose.
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
R EF E R EN CE S
[23]
[24]
[1] USAID. Summary key findings national health accounts 2008/2009. 2009.
Available from: http://egypt.usaid.gov/en/procurement/Documents/
keyfindings_nha2008_09EN.pdf. [Accessed May 31, 2012].
[2] Egypt’s Central Agency for Public Mobilization and Statistics in
collaboration with Egypt’s Ministry of Health and Population.
Household Health Care Utilization and Expenditure Survey 2010. Cairo,
Egypt: Ministry of Health and Population, 2010.
[3] International Society for Pharmacoeconomics and Outcomes Research.
Pros and cons of pricing and reimbursement systems in Egypt.
Available from: http://www.ispor.org/meetings/WashingtonDC0512/
releasedpresentations/Updated-Final-Egypt-ISPOR-presentation_06032
012.pdf. [Accessed August 15, 2013].
[4] International Society for Pharmacoeconomics and Outcomes Research.
Pharmacoeconomic guidelines around the world. Available from:
http://www.ispor.org/PEguidelines/index.asp. [Accessed August 15,
2013].
[5] Spooner JJ, Gandhi PK, Connelly SB. AMCP Format dossier requests:
manufacturer response and formulary implications for one large health
plan. J Manag Care Pharm 2007;13:37–43.
[6] International Society for Pharmacoeconomics and Outcomes Research.
Pharmacoeconomic guidelines around the world, Austria. Available
from: http://www.ispor.org/PEguidelines/source/Guidelines_Austria.
pdf. [Accessed August 15, 2013].
[7] Drummond MF, Schwartz JS, Jonsson B, et al. Key principles for the
improved conduct of health technology assessments for resource
[25]
[26]
[27]
[28]
[29]
[30]
327
allocation decisions. Int J Technol Assess Health Care 2008;24:
244–58.
Fry RN, Avey SG, Sullivan SD. The academy of managed care pharmacy
format for formulary submissions: an evolving standard—a foundation
for managed care pharmacy task force report. Value Health
2003;6:505–21.
International Society for Pharmacoeconomics and Outcomes Research.
Pharmacoeconomic guidelines around the world, Canada. Available
from: http://www.ispor.org/PEguidelines/source/HTAGuide
linesfortheEconomicEvaluationofHealthTechnologies-Canada.pdf.
[Accessed August 15, 2013].
Drummond M. Introduction to pharmacoeconomics. Eur J Hosp Pharm
Pract 2008;14:17–9.
Nord E, Daniels N, Kamlet M. QALYs: some challenges. Value Health
2008;12:S10–5.
Drummond M, Brixner D, Gold M, et al. Toward a consensus on the
QALY. Value Health 2009;12(Suppl.):S31–5.
Berger ML, Bingefors K, Hedblom EC, et al. Health Care Cost, Quality,
and Outcomes: ISPOR Book of Terms. (1st ed.). Lawrenceville,
New Jersey: ISPOR, 2003.
Elliot R, Payne K. Essentials of Economic Evaluation in Healthcare. (1st
ed.). London: Pharmaceutical Press, 2005.
International Society for Pharmacoeconomics and Outcomes Research.
Pharmacoeconomic guidelines around the world, Russia. Available
from: http://www.ispor.org/PEguidelines/source/Russia_PE_
Recommendations_english_fnal_13_03.pdf. [Accessed August 15, 2013].
Drummond M, Barbieri M, Cook J, et al. Transferability of economic
evaluations across jurisdictions: ISPOR Good Research Practices Task
Force Report. Value Health 2009;12:409–18.
Hay JW, Smeeding J, Carroll NV, et al. Good research practices for
measuring drug costs in cost effectiveness analyses: issues and
recommendations: the ISPOR Drug Cost Task Force Report—part I.
Value Health 2009;13:3–7.
International Society for Pharmacoeconomics and Outcomes Research.
Pharmacoeconomic guidelines around the world, Taiwan. Available
from: http://www.ispor.org/PEguidelines/source/2006_PEG_EN_2009.
pdf. [Accessed August 15, 2013].
Rascati KL. Essentials of Pharmacoeconomics. Philadelphia: Lippincott
Williams and Wilkins, 2009.
International Society for Pharmacoeconomics and Outcomes Research.
Pharmacoeconomic guidelines around the world, Ireland. Available
from: http://www.ispor.org/PEguidelines/source/Ireland_Eco
nomic_Guidelines_2010.pdf. [Accessed August 15, 2013].
Bodrogi J, Kaló Z. Principles of pharmacoeconomics and their impact on
strategic imperatives of pharmaceutical research and development. Br J
Pharmacol 2010;159:1367–73.
Weinstein MC, Brien BO, Hornberger J, et al. Principles of good practice
for decision analytic modeling in health-care evaluation: report of the
ISPOR Task Force on Good Research Practices—modeling studies. Value
Health 2003;6:9–17.
Gold MR, Siegel JE, Russell LB, et al. Cost-effectiveness in Health and
Medicine. New York: Oxford University Press, 1996.
Ngorsuraches S. Defining types of economic evaluation. J Med Assoc
Thailand 2008;91(Suppl.):S21.
Drummond M, Brown R, Fendrick AM, et al. Use of pharmacoeconomics
information: report of the ISPOR Task Force on use of
pharmacoeconomic/health economic information in health-care
decision making. Value Health 2003;6:407–16.
International Society for Pharmacoeconomics and Outcomes Research.
ISPOR global healthcare systems road map for pharmaceuticals in
Egypt. Available from: http://www.ispor.org/HTARoadMaps/EgyptPH.
asp. [Accessed August 15, 2013].
Commonwealth Fund. Confronting costs: stabilizing U.S. health
spending while moving toward a high performance health care system.
Available from: http://www.commonwealth
fund.org/Publications/Fund-Reports/2013/Jan/Confronting-Costs.aspx?
page=all. [Accessed August 15, 2013].
Husereau D, Drummond M, Petrou S, et al. Consolidated Health
Economic Evaluation Reporting Standards (CHEERS)—explanation and
elaboration: a report of the ISPOR Health Economic Evaluation
Publication Guidelines Good Reporting Practices Task Force. Value
Health 2013;16:23–50.
Gibson JL, Martin DK, Singer PA. Priority setting for new technologies in
medicine: a transdisciplinary study. BMC Health Serv Res 2002;2:
article14.
Eldessouki R, Smith MD. Health care system information sharing: a
step toward better health globally. Value Health Regional 2012;1:
118–29.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 328
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Note from the Editors
A journal article is sometimes an unfinished piece, which is
finished not after the peer review process [1], but after its original
publication and the letters sent to the journal by readers and the
wider audience. An example of this is the case of the article by
Azevedo et al. [2] published in Volume 1, Issue 2, of Value in Health
Regional Issues (VIHRI) (focusing on Latin America). After receiving
a Letter to the Editors by Dr. Jose Elias Rizk Aziz regarding this
article, the editorial team reviewed the article and sent the letter,
as well as other comments, to the authors, who promptly
responded with a response Letter to the Editors, as well as with
an erratum of the original article. We think that these letters and
the new version of the article, published in this VIHRI issue,
contribute to the iterative process of improving publication
through the interaction of authors, with the readers and editors.
Best regards,
Federico Augustovski, MD, MSc, PhD
Bon-Ming Yan, PhD
Dan Greenberg, PhD
Value in Health Regional Issues Co-Editors-in-Chief
2212-1099/$36.00 – see front matter Copyright & 2013,
International Society for Pharmacoeconomics and Outcomes
Research (ISPOR). Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.05.006
R EF E R EN C ES
[1] Smith R. Peer review: a flawed process at the heart of science and
journals. J R Soc Med 2006;99:178–82.
[2] Azevedo VF, Sandorff E, Siemak B, Halbert RJ. Potential regulatory and
commercial environment for biosimilars in Latin America. Value Health
Regional 2012;1:228–34.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 329–330
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
ERRATUM
In the article, “Potential Regulatory and Commercial Environment for Biosimilars in Latin America,” by Valderilio Feijó Azevedo, Erik
Sandorff, Brian Siemak, Ronald J. Halbert, which appeared in Value in Health Regional Issues, Volume 1, Issue 2 (December 2012), the
following items are to be addressed and clarified:
1) On page 231, Table 3, the correct table should be:
Table 3 – Attributes potentially affecting the development of biosimilar policies in five Latin American
countries.
Attribute
Brazil
Argentina
Chile
Mexico
Venezuela
Market size (population)
Large
(205 million)
High
Midsized
(42 million)
High
Small
(17 million)
Medium
Large
(115 million)
High
Midsized
(28 million)
Medium
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
No
No
Yes
NA
Yes
Yes
Yes
Yes
Yes
No
Yes
No
2010
2005/2011
≥2012
2009 (in
development)
None
Interaction with international
thought leaders
Local production capabilities
Safety issues with biosimilars
in the past
Expected differences in
regulatory
requirements based on the
complexity of the molecule
Regulatory pathway specific for
biosimilars
Year of biosimilar regulatory
pathway
NA, not available/applicable; WHO, World Health Organization.
WHO guidelines were finalized in 2009.
2) Also on page 231, the Potential Issues section should read:
Potential issues
Although there was no regulatory pathway until recently, biosimilars have been available in Mexico for many years. Because of
the increasing number of biosimilars coming into the market, the new biosimilar pathway was designed to increase access to
biosimilars while maintaining quality, efficacy, and safety. As the regulatory pathway is still fluid, there are a few potential issues
that could affect future regulation and subsequent utilization. Thus far only relatively simple biosimilars have been approved for
use in Mexico. As the regulatory pathway will most likely depend on the complexity of the biosimilar, more complex biosimilars
may be subject to the same clinical trial requirements as the originator biologic, as described in other biosimilar regulatory
guidance [5].
3) On page 232, the Potential Issues section should read:
Chile has local production capacity for biologics. As the current regulatory guidelines for biosimilars are in development, the
future regulatory landscape remains open. Chile’s eventual biosimilar pathway, however, will most likely follow European
Medicines Agency or WHO guidelines. Our findings indicate that there may be standardized guidelines for all biosimilars with
specific requirements depending on the complexity of the molecule. Safety issues could also influence the future regulatory
environment. Although potential safety issues might slow down the progress toward developing biosimilar regulatory guidelines,
our primary research indicates that the Chilean government seems determined to press forward with a distinct regulatory
pathway for biosimilars.
330
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 329–330
4) On page 233, the second and third paragraph in the Discussion section should read:
Of these two pathways, the comparability pathway is fairly conventional in its similarity to the requirements brought forward by
the WHO Similar Biotherapeutic Products guidelines. The individual development pathway, however, includes reduced
requirements, opening the door to lower complexity biosimilar products. Such a pathway deviates from the vision for biosimilars
defined by the WHO and is reflective of a hypothesis that has emerged from this research that a conceptual bias may exist within
Brazil regarding the appropriate threshold for regulatory requirements for follow-on biologic products.
Safety problems highlight the need for distinct pathways to regulate review, approval, and pharmacovigilance processes for
biosimilars, and argue for greater transparency of government actions to incentivize the domestic production of biologics.
Observed safety issues, however, have apparently not had a major impact on governmental actions to increase access to
biosimilars in the region to date.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 331–332
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
LETTERS TO THE EDITOR
Response to “Potential Regulatory and Commercial Environment for
Biosimilars in Latin America” by Azevedo et al.
I refer to the article titled “Potential regulatory and commercial
environment for biosimilars in Latin America” written by Azevedo et al., published in Value in Health Regional Issues 1 (2012):228–
34, which contains incorrect information about biosimilars’
regulation and market environment in Mexico.
It must be emphasized that until now there have been no
reported cases of anaphylactic shock related to Kikuzubam in
Mexico or elsewhere and that there are no safety concerns with
the product. The authors based their assertion on a newspaper
article that consisted of an interview with a competitor’s executive at its facility in Palo Alto, CA [1]. The mentioned rash and
anaphylactic shock cases never existed and were never reported.
This is why Probiomed would like to explain to your distinguished readers that there are no Kikuzubam’s safety data
concerns published by any scientific journal or official communication from the Mexican health authorities supporting those
claims against the product.
In addition, we identified that the Mexican current situation
described by Sandorff et al. has several inaccuracies regarding the
regulatory framework for biotechnology drugs. The authors omit
to mention that Mexican health authorities issued a decree
amending and adding various provisions of the regulation of
health products in October 2011 [2].
Although Sandorff et al. mentioned that phase III comparative
trials may not be required for biosimilars and the main factor
influencing the decision will be the product type, the regulation
cited above describes how biotechnology products require pharmacokinetics, pharmacodynamics, safety, and efficacy clinical
trials. Kikuzubam complied with all these requirements.
The safety and scientific requirements endorsed in the Emergency Mexican Official Standard NOM-EM-001-SSA1-2012 for
biotechnology drugs and biopharmaceuticals published in September 2012 include Good Manufacturing Practices compliance,
technical and scientific safety, efficacy and quality proof, labeling,
and requirements for biocomparability studies and active pharmacovigilance [3].
Mexican health authorities’ initiatives confirm that the new
biosimilar registration pathway has been designed to align the
Mexican regulatory framework with International Conference of
Harmonization requirements, instead of just pretending to
increase access to biosimilars as Sandorff et al. claimed.
Finally, it is a known fact that biotechnology generic drugs (as
they were called before the terms “biosimilars” or “biocomparables” were coined) have been in the Mexican health system for
over 20 years.
Probiomed has developed such products since 1996; the
portfolio includes recombinant proteins, cytokines, hormones,
recombinant vaccines, monoclonal antibodies, and fusion proteins. In approximately 15 years, Probiomed has supplied more
than 76 million doses of biotechnology products to health
institutions, health care professionals, and patients in Mexico
and several countries, without receiving any safety or efficacy
complains.
More than 15 million patients have been exposed to Probiomed’s biotechnology generic drugs as detailed in Table 1.
We appreciate and thank you for giving us the opportunity to
make these clarifications.
Jorge Revilla Beltri, MD
Medical Director, Probiomed, Mexico D.F., Mexico
2212-1099/$36.00 – see front matter Copyright & 2013,
International Society for Pharmacoeconomics and Outcomes
Research (ISPOR). Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.05.007
R EF E R EN C ES
[1] Milenio. Medicamento biotecnológico del ISSSTE desata pleito legal.
Available from: http://jalisco.milenio.com/cdb/doc/impreso/9043982.
[Accessed April 24, 2013].
[2] DECRETO por el que se reforman y adicionan diversas disposiciones del
Reglamento de Insumos para la Salud. Available from: http://dof.gob.
mx/nota_detalle.php?codigo=5214882&fecha=19/10/2011. [Accessed
April 24, 2013].
[3] Norma Oficial Mexicana Emergente NOM-EM-001-SSA1-2012,
Medicamentos biotecnológicos y sus biofármacos. Buenas prácticas de
fabricación. Características técnicas y científicas que deben cumplir
éstos para demostrar su seguridad, eficacia y calidad. Etiquetado.
Requisitos para realizar los estudios de biocomparabilidad y
farmacovigilancia. Available from: http://dof.gob.mx/nota_detalle.php?
codigo=5269530&fecha=20/09/2012. [Accessed April 24, 2013].
332
Table 1 – Patients treated with biosimilars manufactured by Probiomed.
Probiomed biosimilar
name
Probivac pediatric
Probivac adult
Filatil
Protophin
Glinux
Uribeta
Emaxem
Jumtab
Kikuzubam
Infinitam
Molgramostim
Alfa 2a IFN
Alfa 2b IFN
Eritropoyetin
(rHu-EPO)
HBv Vaccine
HBv Vaccine
Filgrastim
Somatropin
Insulin
Beta 1b IFN 8 MUI
Beta 1a IFN 12
MUI
Beta 1a IFN 6 MUI
Rituximab
Etanercept
Original brand/
laboratory
Approval
year
Leucomax/Sch-P
Roferón/Roche
Intrón-A/Sch-P
Recormón/Roche
Units distributed per year
Total units
Treated
patients†
2010*
2011
2012
2013
1996
1996
1997
1998
157,633
169,492
977,341
12,764,557
3,652
–
170,035
2,198,892
227
–
183,316
2,317,365
3,654
–
49,745
4,887,978
165,166
169,492
1,380,437
22,168,792
23,595
1,412
11,504
213,161
Engerix-B/Glaxo
Engerix-B/Glaxo
Neupogen/Roche
Humatrope/Lilly
Humulin/Lilly
Betaferon/Bayer
Rebif/Merck-Se
2000
2000
2001
2001
2001
2002
2004
4,532,144
17,030,081
436,274
130,294
3,709,686
360,720
188,442
3,875,000
4,969,950
79,072
47,424
224,231
106,083
13,922
3,927,069
1,296,456
84,133
44,016
55,330
121,742
14,544
7,500,000
2,500,000
317,639
18,000
23,400
154,669
16,391
19,834,213
25,796,487
917,118
239,734
4,012,647
743,214
233,299
6,611,404
8,598,829
131,017
7,882
131,923
4,764
1,496
Avonex/Bayer
Mabthera/Roche
Enbrel/W-Pfizer
2006
2010
2012
111,244
–
–
20,883
12,685
–
25,852
40,593
–
13,575
63,461
248,000
171,554
116,739
248,000
Total biosimilar
units
Total patients
treated
3,299
7,708
2,583
76,166,892
15,750,577
Note. For more than 15 years, Probiomed has offered more than 76 million doses of biosimilar products for the treatment of more than 15 million patients in Mexico and in the global market.
Probiomed has never received a report of immunogenicity or adverse events on its biosimilars, different from those reported by the original product. All Probiomed products are under strict
pharmacovigilance programs.
From year of registration up to 2010.
†
The number of patients was calculated according to current posological schemes.
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 331–332
Gramal
Proquiferón
Urifrón
Bioyetin
API
VALUE IN HEALTH REGIONAL ISSUES 2 (2013) 333
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Response to Letter from Dr. Jorge Revilla Beltri dated 22nd March, 2013
The letter from Dr. Revilla Beltri, a Medical Director at Probiomed,
raises two points.
First, that we cited a newspaper article interview with a
source who appears to have been an industry lawyer. On this
basis, we withdraw the paragraph in question.
Second, Dr. Revilla Beltri states that there are inaccuracies in
the published article with regard to the Mexican regulatory
framework surrounding biosimilars. We stand by our assertion
in the article that a new regulatory pathway had only been
recently created in Mexico to reflect the increasing number of
biosimilars coming into that market and at the time the article
was submitted for publication, that regulatory pathway was still
evolving. In fact, the October 2011 Decreto referenced in
Dr. Revilla Beltri’s letter specifies that biosimilar regulation in
Mexico should be addressed by a subcommittee on a caseby-case basis but did not specify the regulatory structure that
would surround such subcommittees. The second regulation
quoted by Dr. Revilla Beltri was enacted after our article was
accepted for publication.
Erik Sandorff, MA, MBA
PriceSpective, Blue Bell, PA, USA
Copyright & 2013, International Society for Pharmacoeconomics
and Outcomes Research (ISPOR). Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.vhri.2013.05.005
GUIDE FOR AUTHORS
Value in Health Regional Issues is a peer-reviewed publication of the
International Society for Pharmacoeconomics and Outcomes Research,
focusing on Asia, Latin American, Central & Eastern Europe, Western Asia,
and Africa. Its mission is to provide a forum for the advancement and dissemination of knowledge and research in pharmacoeconomics and the
healthrelated outcomes of disease and treatment processes, and therefore solicits original contributions in health care policty analysis, outcomes
research (clinical, economic, and patient-reported) empirical studies,
methodological studies, and manuscripts on health care resources in the
region. Health care decision-maker commentaries from health outcomes
researchers and policy makers are welcome. Commentaries are expected
to include discussion on how researchers can better respond to the needs
of those making clinical and financial decisions in health care.
Value in Health Regional Issues does not consider papers reporting
data series or data sets that do not include appropriate statistical confidence intervals and/or other measures of statistical imprecision. Value
in Health Regional Issues also does not consider papers reporting modeling results that do not include sensitivity analysis of key and influential
model parameters.
Authors for whom English is a second language that would like to
submit a manuscript in English may choose to have their manuscript
professionally edited before submisstion or during the review process.
Authors wishing to employ a professional English-language editing
service (English manuscript submissions only) should make contact
and arrange payment with the editing service of their choice. For more
details regarding the recommended services, see http://support.
elsevier.com/.
Value in Health Regional Issues articles may be published in a language
of the region other than English (i.e. Chinese, Korean, Japanese, Thai,
Indonesian, etc. for Asia; Spanish, Portuguese, etc. for Latin America;
and Russian, Polish, Arabic, etc. for Central & Eastern Europe, Western
Asia and African region). All manuscripts accepted for publication
in Value in Health Regional Issues must have an abstract and
keywords in English as well as the language of the manuscript (if
not English). Languages for a specific issue are decided by the Value in
Health Regional Issues Editorial Board.
The criteria for a manuscript to be considered for Value in Health Regional
Issues are as follows:
First author must reside in the region to which a specific regional
issue is dedicated; and
The empirical study manuscript submitted to Value in Health
Regional Issues must include subjects from population(s) in the
region to which a specific regional issue is dedicated.
For further details concerning the types of articles and eligible regions
suitable for Value in Health Regional Issues, see: http://www.ispor.org/
publications/VIHRI/VIHRImain.asp.
•
•
I. ETHICS IN PUBLISHING
For information on Ethics in Publishing and Ethical guidelines for journal
publication see http://www.elsevier.com/publishingethics and http://
www.elsevier.com/ethicalguidelines.
II. CONFLICT OF INTEREST
All authors must disclose any financial and personal relationships with
other people or organisations that could inappropriately influence (bias)
their work. Examples of potential conflicts of interest include employment, consultancies, stock ownership, honoraria, paid expert testimony,
patent applications/registrations, and grants or other funding. See also
http://www.elsevier.com/conflictsofinterest.
III. SUBMISSION DECLARATION
Submission of an article implies that the work described has not been
published previously (except in the form of an abstract or as part of a
published lecture or academic thesis), that it is not under consideration
for publication elsewhere, that its publication is approved by all authors
and tacitly or explicitly by the responsible authorities where the work
was carried out, and that, if accepted, it will not be published elsewhere
including electronically in the same form, in English or in any other language, without the written consent of the copyright-holder.
IV. RETAINED AUTHOR RIGHTS
As an author you (or your employer or institution) retain certain rights;
for details you are referred to: http://www.elsevier.com/authorsrights.
V. FUNDING BODY AGREEMENTS AND POLICIES
Elsevier has established agreements and developed policies to allow
authors whose articles appear in journals published by Elsevier, to comply
with potential manuscript archiving requirements as specified as conditions of their grant awards. To learn more about existing agreements and
policies please visit http://www.elsevier.com/fundingbodies.
VI. MANUSCRIPT SUBMISSION AND SPECIFICATIONS
Submissions received eight months prior to the publication date will be
considered by the upcoming issue. Publication timeline for a specific
regional issue is below:
Value in Health Regional Issues for Asia is published in May.
Value in Health Regional Issues for Central & Eastern Europe, Western
Asia and Africa is published in August.
Value in Health Regional Issues for Latin America is published in
December.
To submit a manuscript to Value in Health Regional Issues, please
go to: http://www.ispor.org/publications/VIHRI/submission_instruction.
asp. For assistance, authors may contact the Value in Health Regional
Issues editorial office at: [email protected].
If submissions are larger than 500 KB, they should be compressed
using PKZIP or WINZIP.
Authors will be required to assign copyright of their papers. Copyright
assignment is a condition of publication and papers will not be passed
to the publisher for production unless copyright has been assigned. An
appropriate copyright assignment form can be found at the following
address: http://www.ispor.org/publications/VIHRI/submission_instruction.asp. A faxed copy of this completed and signed form is acceptable;
fax to 609-586-4982 or email to: [email protected].
If excerpts from other copyrighted works are included, the author(s)
must obtain written permission from the copyright owners and credit the
source(s) in the article. Elsevier has preprinted forms for use by authors
in these cases: please consult: http://www.elsevier.com/permissions.
•
•
•
Each Submission should contain separate documents as follows:
i. COVER LETTER. The cover letter should include: 1) title of the manuscript; 2) name of the document file(s) containing the manuscript and the
software (and version) used; 3) name and all contact information for the
corresponding author and a statement as to whether the data, models,
or methodology used in the research are proprietary; 4) names of all
sponsors of the research and a statement of all direct or indirect financial
relationships the authors have with the sponsors; and 5) if applicable, a
statement that the publication of study results was not contingent on the
sponsor’s approval or censorship of the manuscript.
For a sample cover letter, see:
http://www.ispor.org/publications/VIHRI/SampleCoverLetter.doc.
ii. TITLE PAGE. The title page should contain the following: 1) title; 2)
full names (first and surname) of all authors including academic degrees
and affiliation(s); 3) name, mailing and email addresses, telephone and fax
numbers of corresponding author (with whom all correspondence will take
place unless other arrangements are made); 4) all sources of financial or
other support for the manuscript (if no funding was received, this should be
noted on the title page); 5) at least four key words for indexing purposes;
and 6) a running title of not more than 45 characters including spaces.
For a sample title page, see:
http://www.ispor.org/publications/VIHRI/SampleTitlePg.doc.
iii. MANUSCRIPTS. Manuscripts must be typed in either Microsoft
Word (Version 5.0 or later) or WordPerfect (version 5.1 or later). Manuscripts should be in 8.5x11-inch page format, double-spaced with 1-inch
margins on all sides and size 10 font (Arial or Times New Roman fonts
are preferred). Minimal formatting should be used, i.e., no justification, italics, bold, indenting, etc. There should be no hard returns at
the end of lines. Double-spacing after each element is requested (e.g.,
headings, titles, paragraphs, legends). The ‘Uniform Requirements for
Manuscripts Submitted to Biomedical Journals’ should be consulted for
specific style issues not addressed here (www.acponline.org, Ann Intern
Med 1997;126:36-47). The word count should be less than 4500.
a. ABSTRACT. An abstract of 250 words or less is required, summarizing
the work reported in the manuscript. Original research manuscripts should
use a structured format for the abstract, i.e., Objectives, Methods, Results, and Conclusions. A non-English research manuscript requires an
abstract in English as well as in the language of the manuscript.
b. TEXT. The body of the manuscript should be divided into sections that facilitate reading and comprehension of the material. This
should normally include sections with the major headings: Introduction, Methods, Results, Conclusions, Acknowledgments (if needed),
and References. There should be no footnotes. Figures (inclusive
of figure legends) and Tables must be submitted each as separate
documents.
GUIDE FOR AUTHORS – continued
c. REFERENCES. References should be listed in a separate section
and numbered consecutively with Arabic numerals in the order in which
they are cited in the text. Citing unpublished or non-peer-reviewed work
such as abstracts and presented papers is discouraged. Personal
communications may be indicated in the text as long as written acknowledgment from the authors of the communications accompanies
the manuscript. Reference style should follow that of Index Medicus.
Spell out single-word journals and abbreviate all others according to
the style of Index Medicus. If there are more than four authors, use only
the names of the first three, followed by et al.
The three most common types of references are illustrated
below for example.
Journal article: Surname and initials of author(s), title of article,
name of journal, year, volume number, first and last page.
Arocho R, McMillan CA. Discriminant and criterion evaluation of the
U.S.-Spanish version of the SF-36 Health Survey in a Cuban-American
population with benign hyperplasia. Med Care 1998;36:766–72.
Book: Surname and initials of author(s)/editor(s), title and subtitle,
volume, edition (other than first), city, publisher, year.
Johnston J. Econometric Methods (3rd ed.). New York: McGraw-Hill,
1984.
Chapter in Book: Surname and initials of author(s), title of chapter,
author(s)/editor(s) of book, title of book, volume, edition (other than
first), city, publisher, year.
Luce BR, Manning WG, Siegel JE, et al. Estimating costs in costeffectiveness analysis. In: Gold MR, Siegel JE, Russell LB, et al.,
eds., Cost-effectiveness in Health and Medicine. New York: Oxford
University Press, 1996.
Website: Title of article on web. www.document. Available from:
http://www... [Accessed Month day, year].
International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Good Outcomes Research Practices Index. Available from: http://www.ispor.org/workpaper/practices_index-.asp.
[Acessed January 1, 2011].
For a sample manuscript, see:
http://www.ispor.org/publications/VIHRI/SampleManuscript.pdf.
iv. TABLES. Tables should be clearly labeled, neatly typed, and easy
to understand without reference to the text. Each should be double
spaced and presented on a separate page. Statistical estimates
should indicate parameter estimates and, as appropriate, t ratios or
standard error, statistical significance, sample size, and other relevant
information. All abbreviations must be explained below each table.
Each table should be numbered and have a self-explanatory title.
For a sample table, see:
http://www.ispor.org/publications/VIHRI/SampleTbls.doc.
v. FIGURES. Figures should each be submitted as a separate image
file, not embedded in the manuscript document or in a slide presentation.
Cite figures consecutively, as they appear in the text, with Arabic numbers (Figure 1, Figure 2, Figure 3A, etc.). If, together with your accepted
article, you submit usable color figures then the Journal will ensure, at
no additional charge, that these figures will appear in color on the Web
(e.g., ScienceDirect and other sites) regardless of whether or not these
illustrations are reproduced in color in the printed version. There will be a
charge for color reproduction in print; you will receive information regarding the costs from Elsevier after receipt of your accepted article. Please
indicate your preference for color in print or on the Web only. Each figure
must be assigned a brief title (as few words as possible, and reserving abbreviations for the legend) as well as a legend. The corresponding legend
should be typed double-spaced on a separate page. All symbols, arrows,
and abbreviations must be explained in the legend. Please submit files
with a resolution of at least 300 DPI. Line artwork should contain a resolution of least 1000 DPI. Elsevier recommends submitting figures in the
following formats: TIFF, JPG, EPS, and PDF. Please be sure to delete any
identifying patient information such as name, social security number, etc.
Photographs in which a person’s face is recognizable must be accompa-
nied by a letter of release from that person explicitly granting permission
for publication in the Journal. For any previously published material, written permission for both print and electronic reprint rights must be obtained from the copyright holder. For further explanation and examples of
artwork preparation, see Elsevier’s Author Artwork Instructions at www.
elsevier.com/artwork.
For a sample figure, see:
http://www.ispor.org/publications/VIHRI/SampleFigure.doc.
vi. SUPPLEMENTARY MATERIAL OR SUPPLEMENTAL DATA. You
may submit appendixes that describe either methods or results in more
detail if these are needed for clarity of understanding by either peer
reviewers or readers. If appropriate, materials suitable for Web publication but not print publication (eg, audio or video files, see below) can
also be submitted. If you do so, indicate the particular reasons for the
appendix and whether you are submitting it for possible Web publication
or simply for peer review purposes. Value in Health Regional Issues accepts audio and video files as ancillaries to the main article. Audio files
should be in .mp3 format; the recommended upper limit for the size of
a single file is 10 Mb. Video files should be submitted in .mpg or .mp4
format, the recommended upper limit for the size of a single file is 10
Mb. Any alternative format supplied may be subject to conversion (if
technically possible) prior to online publication.
vii. SURVEY INSTRUMENT. For papers analyzing preferences, Value
in Health Regional Issues requires the submission of a copy of the survey instrument used to generate the preference data. This is to help
in the review process and the survey instrument need not appear in a
final publication. If the authors wish the questionnaire to be published
with the paper, it should be submitted as part of the paper. If they do
not wish it to be published, it should be submitted as Supporting Information and then will be sent to the reviewers as a reviewer’s appendix.
VII. DATA, MODELS, AND METHODOLOGY
All authors must agree to make their data available at the Editor’s
request for examination and re-analysis by referees or other persons
designated by the Editor. All models and methodologies must be presented in sufficient detail to be fully comprehensible to readers.
VIII. AUTHOR ANONYMITY
It is the policy of Value in Health Regional Issues that peer review of
submitted manuscripts is double blinded, i.e., the reviewers do not know
the names of the authors of manuscripts and the authors do not know
the names of the reviewers. Blinded reviews are common practice at
many important scientific and medical journals.
IX. THE REVIEW PROCESS
All manuscripts deemed appropriate for Value in Health Regional Issues after
initial screening will be reviewed by at least two peer reviewers. The objective
of the journal is to complete peer review and reach editorial decision within
ten to twelve weeks of submission, at which time the corresponding author
will receive written notification, including anonymous reviewer commentary.
X. AUTHOR TRACKING SERVICES
Authors may track accepted manuscripts (English only) at:
http://www.elsevier.com/trackarticle and set up e-mail alerts to inform
them when an article’s status has changed. Contact details for questions
arising after acceptance of an manuscript, especially those relating to
proofs, will be provided by the publisher. For manuscripts not submitted
in English, authors may query the Value in Health Regional Issues Editorial
office at: [email protected], [email protected], or [email protected].
XI. PROOFS
Proofs are to be sent electronically to Authors to review for printer’s
errors. Substantive changes or additions to the edited manuscript
cannot be allowed at this stage. Corrected proofs should be returned
to the publisher within 2 days of receipt.
XII. OFFPRINTS
The corresponding author, at no cost, will be provided with a PDF file of
the article via e-mail. For an extra charge, paper offprints can be ordered
via the offprint order form which is sent once the manuscript is accepted
for publication. The PDF file is a watermarked version of the published
article and includes a cover sheet with the journal cover image and a
disclaimer outlining the terms and conditions of use.