FACULTY OF HEALTH AND MEDICAL SCIENCES
UNIVERSITY OF COPENHAGEN
PhD thesis
Hans Okkels Birk
Freedom of choice of provider as a
governance tool in the public sector
Case: freedom of choice of hospital in Denmark
This thesis has been submitted to the Graduate School of The Faculty of Health and Medical
Sciences, University of Copenhagen.
University representatives: Allan Krasnik and Karsten Vrangbæk
Submitted: 27/02/2015
Institutnavn:
Institut for Folkesundhedsvidenskab
Name of department:
Department of Public Health
Author:
Hans Okkels Birk
Title / Subtitle:
Freedom of choice of provider as a governance tool in the public
sector. Case: freedom of choice of hospital
University representatives: Allan Krasnik and Karsten Vrangbæk
Assessment committee
Professor Mogens Grønvold (Chairperson)
Department of Public Health
University of Copenhagen
Associate Research Professor Camilla Palmhøj Nielsen
Department of Public Health
Aarhus University
Professor Runo Axelsson
Department of Sociology and Social Work
Aalborg University
Submitted:
27 February 2015
Defense date
29 June 2015
2
Table of contents
Table of contents ...................................................................................................................................................... 3
Preface ....................................................................................................................................................................... 5
List of papers ............................................................................................................................................................ 5
1 Introduction ........................................................................................................................................................... 6
1.1 Public intervention as a tool to remedy market failure ..................................................................................... 7
1.2 Characteristics of the market for health care .................................................................................................... 8
1.3 Content and introduction of “choice” ................................................................................................................ 9
1.4 Introduction of freedom of choice of hospital in Denmark ............................................................................. 11
1.5 Implications of freedom of choice – topics for research ................................................................................ 11
1.6 Aims of the present study ..................................................................................................................................12
1.7 Theoretical approach .........................................................................................................................................12
1.8 Specific objectives and delimitations ...............................................................................................................14
1.9 Sub-studies .........................................................................................................................................................16
2 Materials and methods......................................................................................................................................... 17
2.1 The Danish health care system ......................................................................................................................... 17
2.2 The study areas .................................................................................................................................................. 17
2.2.1 Study area A: Roskilde County (study I and II) ........................................................................................... 18
2.2.2 Study area B: the counties of Roskilde, Storstrøm and Vestsjælland (study III). ..................................... 18
2.3 Study I: inpatients’ choice behaviour .............................................................................................................. 18
2.4 Study II: outpatients’ choice behaviour ...........................................................................................................19
2.5 Study III: general practitioners’ choice behaviour on behalf of patients ...................................................... 20
3 Results ...................................................................................................................................................................21
3.1 Results, Study I: inpatients’ choice behaviour .................................................................................................21
3.2 Results, Study II: outpatients’ choice behaviour .............................................................................................21
3.3 Results: study III: GPs’ choice behaviour on behalf of patients .................................................................... 22
4 Discussion ............................................................................................................................................................ 23
4.1 Do patients – and/or their agents - choose the hospital? .............................................................................. 23
3
4.2 Which sources of information determine choice of hospital? ....................................................................... 25
4.2.1 Patients’ sources of information ................................................................................................................... 25
4.2.2 GPs’ sources of information ......................................................................................................................... 28
4.3 Influence of sociodemographic characteristics on choice behaviour? .......................................................... 29
4.4 Is performance the primary factor behind choice of hospital?...................................................................... 30
4.5 A model of patients’ and their agents’ choice of hospital ................................................................................31
4.6 Strengths and weaknesses of the study ........................................................................................................... 32
5 Conclusions and perspectives ............................................................................................................................. 35
5.1 Conclusion ......................................................................................................................................................... 35
5.2 Implications ...................................................................................................................................................... 36
5.3 Future research................................................................................................................................................. 37
Summary ................................................................................................................................................................. 38
Danish summary .................................................................................................................................................... 39
Acknowledgments .................................................................................................................................................. 40
References ................................................................................................................................................................41
Appendices ............................................................................................................................................................... 51
Appendix 1: Free choice of hospital in Denmark – the framework ...................................................................... 51
Appendix 2-4: Original papers ...............................................................................................................................61
Appendix 1: Free choice of hospital in Denmark – changing institutional limits over time
Appendix 2 (Paper I): Birk HO, Henriksen LO. Why do not all hip- and knee-patients facing
long waiting times accept re-referral to hospitals with short waiting time? Questionnaire study.
Health Policy 2006; 77: 318-25.
Appendix 3 (Paper II): Birk HO, Gut R, Henriksen LO. Patients’ experience of choosing an
out-patient clinic in one county in Denmark: results of a patient survey. BMC Health Serv Res
2011; 11: 262.
Appendix 4 (Paper III): Birk HO, Henriksen LO. Which factors decided general practitioners’
choice of hospital on behalf of their patients in an area with free choice of public hospital? A
questionnaire study. BMC Health Serv Res 2012, 12: 126.
4
Preface
The results presented in this PhD-thesis stem from studies carried out during my employment at
the Department of Health, Roskilde County, and the department of Quality & Development, Region
Zealand, Denmark. Roskilde County initiated the study and the county and Region Zealand
provided most of the funding for the study, which was also facilitated by grants from The
Programme for Promotion of Regional Cooperation on Medical Research in Eastern Denmark; the
Health Research Foundations in the Counties of Eastern Denmark, and Section for Health Services
Research, Department of Public Health, University of Copenhagen.
The study was performed in accordance with the Helsinki Declaration. The study was not notifiable
to the Danish research ethics committee system: according to section eight in the Danish Act of 28
May 2003 on a Biomedical Ethics Committee System and the Processing of Biomedical Research
Projects questionnaire studies not involving biological material were not notifiable to the Danish
research ethics committee system.
List of papers
The present PhD-thesis is based on the following three papers, which will be referred to by their
Roman numerals:
Paper I: Birk HO, Henriksen LO. Why do not all hip- and knee-patients facing long waiting
times accept re-referral to hospitals with short waiting time? Question
naire study. Health Policy 2006; 77: 318-25.
Paper II: Birk HO, Gut R, Henriksen LO. Patients’ experience of choosing an out-patient clinic in
one county in Denmark: results of a patient survey. BMC Health Serv Res 2011; 11: 262.
Paper III: Birk HO, Henriksen LO. Which factors decided general practitioners’ choice of
hospital on behalf of their patients in an area with free choice of public hospital? A questionnaire
study. BMC Health Serv Res 2012, 12: 126.
Reprints of the papers were provided with permission from the editors (papers II and III were
published in open-access journals).
5
1 Introduction
Access to health care constitutes a key part of modern welfare states’ services [1]. Public and private
health care systems in developed countries face growing challenges arising from: new medical
technologies; wider indications for treatment due to evidence from medical research; a growing
share of citizens with chronic diseases; increasing prevalence of obesity resulting in more patients
with chronic life style-related conditions; greater expectations from the population concerning
quality, service and the range of health care services provided; and a combination of the elderly
population’s growing share of the whole population, and the work force’s decreasing share of the
population.
These general trends put pressure on the health care systems to strengthen their efficiency and
responsiveness, while the public sector traditionally tends to focus on processes and cost control
[2]. On this background the Anglo-Saxon and Nordic countries, especially, have introduced
management tools, which resemble those, which are used in the private sector, under the umbrella
term of ‘New Public Management’ to strengthen the public sector’s responsiveness and efficiency
[3]. The introduction and use of such tools must take into account, that there are major differences
between the conditions facing providers in the private and in the public sector [4,5]. In the private
sector providers traditionally face a single performance measure (profit), providers may stay away
from unprofitable markets, and customers pay for the consumed services – providing the providers
with an income and with a restriction on the demand. On the contrary in the public sector
providers face multiple and competing criteria of success, provision of some services is mandatory,
and services may be provided as a right for the consumer and free at the point of delivery without a
price mechanism to connect demand with producers’ costs and consumers’ utility. Furthermore
managers’ and employees’ motivation differ by sector [6]. Due to such fundamental institutional
differences between the two sectors, management/governance tools may have different effects in
each sector, warranting studies of the actual effects of governance tools.
Quasi markets, including free choice of provider – i.e. a hospital, constitute one of the NPM-tools
which play a still stronger role in current governance, although the available evidence on patients’
actual choice behaviour is ambiguous and has had little impact on design of public management
[7]. Likewise few studies of the choice behaviour of patients’ agents – the GPs - are available.
In 1999 a Danish county bought hospital capacity outside the county in order to reduce patients’
waiting time on the assumption that they had strong preferences for short waiting times for
treatment and were willing to travel to other hospitals to shorten their waiting time. However, it
turned out, that much fewer patients than expected by politicians and civil servants chose
treatment at the hospital with short waiting time, thereby questioning key assumptions about
patients’ preferences and thereby in effect also questioning preconditions for utilizing quasi
markets as a governance tool: if patients are unlikely to collect information about various providers
and choose the hospital offering the best performance, quasi markets in health care may not put
pressure on providers to improve their performance. Therefore it is important to test the
assumptions behind utilization of quasi markets as a governance tool.
6
1.1 Public intervention as a tool to remedy market failure
The central theorem of neoclassical welfare economics is that under certain strict conditions1, an
equilibrium achieved by competitive markets will satisfy the same conditions as a Pareto optimum,
meaning that it will not be possible to make somebody better off without making somebody worse
off [9]. The theorem indicates that a system of competitive markets is the optimal tool for
allocation of scarce resources in society [10], where consumers, in possession of initial resources,
choose consumption bundles to maximize utility at a given set of prices, while providers choose
production bundles so as to maximize their profits at the same set of prices [9]. However, actual
markets may not satisfy the strict conditions in this competitive market model (the standard
model), resulting in some degree of market failure [8;11]. For example agents may not have all the
relevant information (e.g. consumers may not be fully aware of the total utility derived from one or
more goods)[10], individual consumers and providers do not take the negative or positive effects of
their choices on the rest of society (externalities) into account, when they maximise their own
utility [12], and monopolies or oligopolies may eliminate competition and thereby the incentive to
efficient production. A typical tool to remedy market failure is the development of planned
economy systems (nonmarket systems)2 in which central authorities decide how much should be
produced of certain goods, and for whom. However, government intervention to remedy market
failure may fail [14] and result in intrinsic nonmarket failure/government failure [11;15;16] due to
the structure of incentives, lack of information and/or the organization of government action [17],
which lead to:
Internalities: Producers who lack signals from a market may develop and follow their own ‘private’
goals and standards, ignoring consumers’ preferences and even ‘public’ goals specified by
politicians; e.g. a manager may maximise the budget of their organization rather than quality and
service [18].
Redundant costs: Nonmarket organisations operating without a profit motive may be less likely to
maximise productivity than organisations operating on a market and following a profit motive,
resulting in an inefficient production.
Derived externalities: Initiatives to correct market failure may have unintended side effects.
Public choice theory may be viewed as a theory of nonmarket failure [14]. Professor in social policy
Julian Le Grand has defined four basic models of management of public (nonmarket) service
provision – clearly favouring the fourth model [19;20]:
The Trust Model: political and administrative decision makers trust professionals and managers to
deliver services of high quality without interference, on the assumption that the professionals know
what is in the consumers’ best interests, and that their main motivation is the welfare of those they
are serving and not their own self-interest.
The Command and Control Model: central management sets targets and follow up on providers’
performance – professionals may be hostile to the targets and focus on achieving targets rather
than providing the relevant services (‘gaming’).
1
All goods are private; no externalities; information is free and accurate, and each buyer or seller is a price taker [8].
“A command economy or planned economy is an economy where … a central authority … assigns quantitative
production goals and allots raw materials to productive enterprises … the central planners work out the assortment of
goods to be produced and the quotas for each enterprise. Consumers may influence the planners’ decisions indirectly if
the planners take into consideration the surpluses and shortages that have developed in the market. The only direct
choice made by consumers, however, is among the commodities already produced” [13].
2
7
The Voice Model: consumers express their (dis)satisfaction directly to the providers; this is a kind
of bottom-up management with incentives of uncertain strength on providers to respond.
Providers may focus on voice from vocal and articulate groups
The Quasi Market Model: each consumer chooses one of several competing public or private
providers, and the services are paid by the government - a quasi market thereby constitutes a proxy
for a traditional market mechanism3.
The four models are not mutually exclusive but may co-exist. Development over time may reflect
changes in the mix between models rather than substitution of one model for another [21]. One
common trait in public sector governance reforms in the latest two decades within the broad
concept of New Public Management (NPM) to remedy government failure is a development from
collective systems towards an individual-based democracy model [13;22], where individual citizens
are viewed as autonomous consumers (queens) rather than passive clients (pawns)[23] and are
expected by politicians and central management to take responsibility for allocation of resources to
providers4 by utilising consumers’ rights [24] like freedom of choice of hospital to choose a
provider of a public service more or less freely among several competing public or private
providers. In combination with financial incentives to attract consumers, choice may result in a
quasi-market, strengthening providers’ incentives to improve responsiveness [3;25], as the public
sector moves from government towards governance (see chapter 1.9).
1.2 Characteristics of the market for health care
Citizens’ demand for health care services is characterized by structural market failure due to
imperfect and asymmetric information about the need for and effect of services, low predictability
of need/demand, socioeconomic inequalities, and high costs of services [26]. Therefore the
production of health care usually is not decided in a competitive market. In developed societies
agents like general practitioners (GPs) act on behalf of patients (principals) to overcome
information asymmetry, and health care is usually financed or subsidized by a third part, thereby
partly protecting patients from the financial effects of their individual consumption of health care,
unlike consumers in a standard model market [27]. This may result in excess consumption [26]
and moral hazard5 among consumers, and adverse selection6 among providers and in the financing
third part [28]. Due to supplier induced demand [29] monopolistic providers may experience
weaker incentives to improve responsiveness, quality or productivity than in a competitive market.
Provision of services may be the responsibility of private or public organisations, and services may
be produced by public as well as private providers irrespectively of the ownership of the
organisation responsible for provision. Health care systems in the Nordic countries are public and
universal: tax-financed public organizations are responsible for provision, while actual services are
provided by public and/or private providers working under a contract with a public body. Thereby
health care in the Nordic countries shares basic features with the National Health Service models
(‘Beveridge-models’) in the United Kingdom, New Zealand and Southern Europe [30]. Most health
care systems based on the Beveridge-model constitute planned economy structures in the public
sector, where the supply side – integrating the supplier of health care and the financing third part allocates resources to the sector and sets targets and priorities.
3
Such initiatives have been associated with privatizations but may also be presented as alternatives to privatization.
Or, rather, be used unconsciously for this purpose.
5 A situation where an agent tends to be accept a risk because the potential associated costs will not be felt by the agent
but by others, e.g. by the financing third part.
6 Situations where the third part avoids consumers which are especially likely to consume health care services.
4
8
1.3 Content and introduction of “choice”
Until a couple of decades ago, patients in public health care systems faced few opportunities for
choice among providers. Typically GPs would refer a patient to a hospital prescribed by the public
authority responsible for health care or the closest hospital, irrespectively of the patient’s wishes,
service and clinical quality [31]. For example the right to choose a hospital in other uptake areas
was restricted in all of the Nordic countries at the start of the 1990s [30]. During the latest couple
of decades the perception of patients’ role has gradually changed from that of a passive recipient7
of health care services prescribed by public authorities and providers to a consumer, who utilizes
some degree of free choice of provider to pick and choose among different providers [33]. However
a consumer may be perceived to act in very different ways, ranging from a rational, empowered,
informed chooser/customer to a consumer who consciously delegates choices of care to an agent
due to information asymmetry between the consumer and potential providers [34]. The meaning
of choice varies between countries [35], as free choice of provider covers a wide range of
institutional arrangements, as it works at different levels [36]: e.g. the provider may be chosen
among potential suppliers by the governing body responsible for provision of health care, a GP, a
fund holding GP or the patient [37] In health care patients or agents acting on their behalf may
choose or be involved in the choice of treatment (examination, intervention etc.)(“what”), provider
of health care (“where”), individual health professional (“by whom”), appointment time/date
(“when”), and/or method of communication (“how”)[20]. Information asymmetry is a major
challenge to patients’ opportunities for choosing a specific treatment, and providers’ work
schedules and lack of published information about individual clinicians and their results limit the
opportunities for patients to choose among individual health professionals. Choice has been
introduced in other service areas provided by the public sector, e.g. education [38;39], social
services [40;41], elderly care [42] and public employment services [43], but health care constitutes
one part of the public sector, where the introduction of governance–tools has been especially
important in countries applying the Beveridge-model, like England, Denmark and Sweden [44;45],
and to a lesser degree in countries where health care is organized in accordance with the Bismarckmodel, like France, Germany and the Netherlands [46;47]. The introduction of choice in public
health care reflects two fundamental arguments [48]: an ideological viewpoint: an opportunity to
choose a supplier is an objective in itself, as it strengthens personal freedom [49-52]8. Individuals
consistently express strong preferences for having a choice and may derive process utility from
choosing the hospital by themselves [53], although the value per se of choice has been questioned:
even if patients are aware of their needs, the properties of alternatives, and of their opportunities to
choose, they often report that they are overwhelmed by the choice, lack the resources needed to
make it, and experience unintended negative consequences, including cognitive overload and fear
that they may regret their choice and shoulder responsibility for the choice [54-60]. On the other
hand, patients (principals) may delegate the choice to an agent, and in a perfect principal-agentrelation the agent makes the same choice as the principal would have made, if the principal had
had the same information as the agent [29]. An instrumental viewpoint (the intrinsic value of
choice): providers and employees in the public sector in general may not be viewed as altruistic
(“knights’) but as agents acting out of self-interest (“knaves’)[23]. Therefore politicians (and
administrators) may attempt to improve responsiveness and performance by developing self-
7
As reflected in the adjective ‘patient’: “able to wait for a long time or accept annoying behaviour or difficulties
without becoming angry.” [32]
8 Amartya Sen underscores the long tradition behind this idea by quoting Aristotle in the introduction to [52].
9
correcting mechanisms, which resemble the market mechanism working in competitive markets by
introducing competition among (numerous) suppliers [15;61;62]. Ideally a combination of choice
and activity-based financing will [63]:
1) Communicate consumers’ preferences to providers more efficiently than command
economies by turning passive consumers - “pawns’ - into active consumers - “queens’
[23;64].
2) Provide the suppliers with financial incentives to utilise this information and satisfy the
consumers’ demand, because “the money follows the patient”[19;61;65-68], thereby
challenging traditional planning and cost control mechanisms in publicly funded health
care systems [69].
If the two mechanisms are successful and lead to an increase in quality, service and efficiency,
patients will derive consequentialist utility from choice [53].
Ideally, patients’ and providers’ individual (egoistical) utility maximisation through rational
choices will result in an optimal resource allocation in society at the national [70] and even the
international level9[72], when choice and activity-based financing creates financial incentives for
providers to attract consumers, and thereby incentives to adjust their services’ properties to
consumers’ preferences [73], which are assumed to equal society’s objectives, as patients’ choices
move resources from inferior providers to superior providers and level out waiting times as a sideeffect [74].
Information asymmetry resulting in agent-principal-relationships may influence the interpretation
of the content of choice: if a patient is referred to and treated at the hospital closest to his or her
home, the referral process may be interpreted as a conscious, active choice of the nearby hospital,
or as a failure by the GP or the patient to utilise choice, passively going to the hospital closest to
the patient’s home. The interpretation depends on the observers’ views on choice: if choice is
viewed as a right introduced to increase citizens’ utility by strengthening patients’ freedom to
choose, the actual choices are of little concern, if they reflect patients’ preferences rather than
barriers to choice of more distant hospitals. However, if choice is introduced as a governance tool
to level out waiting times and put pressure on providers to increase responsiveness, quality and
productivity, the tool’s success depends on patients’ and their agents’ willingness to compare
providers’ performance and select the hospital with the best performance, no matter where it is
located. Otherwise the providers may ignore the risk of patients and agents utilizing choice.
Likewise, the criteria for success in evaluations of choice depend on the objectives of choice and the
views on actual referral/choice patterns:
If choice is viewed as a right, the criterion of success is that patients are treated at the hospitals,
where they want to go for treatment, but the actual utilization of distant hospitals (or hospitals
nearby) is not a relevant criterion of success: a high utilization of choice may be interpreted as a
success: it indicates that choice and the health care service “works’, as patients utilize their rights.
However, choice of distant hospitals may also be viewed as an indication of failure in parts of the
health care system, the choice indicates that providers close by do not live up to patients’ and GPs’
expectations, forcing them to go to distant providers. Likewise, a low utilisation of choice may be
interpreted as a success – an indication of a high degree of satisfaction with hospitals, as patients
9
E.g. through introduction of free choice of health care provider in the EC and/or other areas [71].
10
and agents abstain from utilising choice. Or it may be viewed as an indication of failure: patients
and agents are forced by barriers to choice to abstain from utilisation of choice.
Nor is the actual utilization of hospitals a relevant criterion of success, if choice is viewed as a
governance tool. Its impact may be evaluated by measuring providers’ performance rather than
patients’ and agents’ utilization of choice. Major variation in providers’ performance - including
waiting times, clinical quality and patient satisfaction – indicates that utilisation of choice is
insufficient to level out waiting times and move resources to the most efficient hospitals, resulting
in failure for the governance tool and the health care system as a whole. Such failure may reflect
barriers to choice or it may reflect that patients’ and agents’ preferences and choice behaviour
make them less inclined to utilise choice than assumed in quasi market-models.
1.4 Introduction of freedom of choice of hospital in Denmark
Choice policies were introduced in Sweden (1991) and Denmark (1993) earlier than in the UK
(2002) and the Netherlands (2006), which on the other hand had introduced quasi-markets with
little or no choice at patient level [69]. When choice was introduced in Denmark and Sweden, the
emphasis was predominantly on the intrinsic value of choice, and the reform was associated with
patients’ rights and patient empowerment, although choice was also viewed as an instrumental
policy to reduce waiting times, improve allocation of resources, and strengthen providers’
responsiveness, like the motives behind introduction of choice in the UK and the Netherlands later
on [66]. Most Danish patients did not choose hospitals in other counties/regions, thereby
indirectly influencing the subsequent development of choice, as national politicians have
interpreted the low utilization of choice [45] as an indication of institutional barriers. Suspecting
that public providers undermine utilization of choice, e.g. by not informing patients about their
rights, centre-right and centre-left governments have extended patients’ rights in several steps
(Appendix 1) by offering choice to more patient groups, by providing patients and agents with
more information on choice, by including more providers of health care in choice, by
providing more sources of information about quality and service, and by strengthening
incentives for hospitals to attract patients (higher activity based financing).
1.5 Implications of freedom of choice – topics for research
As indicated above, introduction of choice gives rise to a number of topics for research in patients’,
agents’ and providers’ actual behaviour, and the effect of choice on performance and equity [75],
e.g. how patients and/or agents actually choose a hospital, how patients’ and/or agents utilize
information for decision-making, and the net effect of choice on providers. Two necessary
preconditions for a positive effect of choice on performance are that valid and reliable data on
providers’ general performance is available, and that patients and/or agents utilise the data10. In
health care few consumers can make a choice based on their own experiences with different
providers of a specific service11, and the inherent strong information asymmetries and uncertain
effects of care on individual patients make it difficult to evaluate the quality of care ex ante or even
ex post for individual patients12, increasing the risk that care may be judged by proxy measures
10
If providers assume that performance data influence choice, this may be a sufficient condition for influencing
performance, as management may respond forcefully to bad performance [20;76;77]. For an opposing view, see [78].
11 I.e. a patient can only get his or her appendix removed once.
12 Most patients do not have the information needed to assess their chances for being cured or to judge whether an
apparently negative outcome is due to low quality care.
11
rather than actual quality13. Politicians’ and civil servants’ persistent confidence in choice as a
governance tool, as reflected in the stepwise extension of choice, makes it still more important to
improve our understanding of patients’ and agents’ preferences, the actual responsibility for
choice, how decision makers choose, which data influence their choices14[72], whether choice
behaviour is consistent with the assumptions behind the introduction of quasi-markets, and
whether these markets achieve the intended objectives.
A Danish case was especially relevant for this purpose, because free choice of hospital had been in
place for several years and included all hospitals15, standardized surveys of waiting times and
patients’ experiences were available, and referring GPs faced no financial incentives to choose a
particular hospital on behalf of their patients. The cases and study areas in the present study were
especially relevant, because more information on waiting times was available to GPs than in the
country as a whole, and patients’ transportation costs were subsidized to a higher degree, thereby
reducing a negative impact of transportation costs on choice of hospital. When the study was
initiated, biannual nationwide surveys already recorded patients’ awareness of choice and the
factors determining their choice [80], and so the present study aimed at investigating patients’
utilisation of various sources of information for choice in more detail than the nationwide
aggregated surveys.
1.6 Aims of the present study
The objective of the study was to investigate choice as a governance tool in a public health care
system. More specifically the aims of the study were to investigate, whether patients’ and GPs’
choice behaviour was consistent with key assumptions behind use of quasi markets in health care
as a governance tool (see 1.8).
1.7 Theoretical approach
In accordance with the aims of the study, quasi-market theory was selected as the theoretical
framework. Quasi-markets are established by the public sector when it opens its service production
to other providers – public or private, for-profit or non-profit – by offering consumers or their
agents a choice of provider [81]. The term ‘Quasi-market’ has no exact definition but may be
characterised as an intermediate form between hierarchy and a market with perfect competition
[82;83]. Quasi-markets are ‘markets’, because they replace monopolistic public providers with
competitive and independent providers, and they are ‘quasi’, by differing from conventional
markets in key areas on the supply side as well as the demand side [82;84]:
o Providers do not necessarily maximise their profits – a key assumption about providers’
behaviour on a competitive market [84]: some providers’ behaviour may be altruistic [85],
public organizations may aim to maximise their budget [86], or their behaviour may reflect
a multitude of criteria for success [87].
o Consumers’ purchasing power may take the form of an earmarked budget or voucher [84],
presenting the consumer with a right to a specified service, which cannot be traded for a
13
In Denmark people critical of utilizing patients’ evaluations of care may parody patients’ evaluations by attributing
judgments like “the nurses were very kind” to patients, thereby indicating that patients’ judgments tend to draw the
emphasis away from “real” quality measures towards proxy measures of quality and/or service.
14 In 2006, the European Health Policy Group decided to explore choice under a theme of access, choice and equity
[79].
15 Only an insignificant private hospital sector was not included.
12
different kind of service, money flowing from the public purchaser to the provider without
involving the consumer. In health care the combination of a referral from a GP and health
care provided free at the point of delivery may be viewed as similar to a voucher.
o For some kinds of services, the consumers do not choose the provider by themselves [84].
Not only is the ‘voucher’ provided from the outside, but the actual choice may also be made
by an agent.
The theoretical framework presented briefly in section 1.2 was developed by Julian le Grand and
his group [20;23] on the background of British reforms of the welfare state from the 1980s and
onwards in a move away from public provision of services by monopolistic vertically integrated and
hierarchical public organisations [88], but the phenomenon, the term, and the critique of quasimarkets are decades older. Apart from Le Grand’s understanding of the term, it is used to
characterize other set-ups which differ from competitive markets, including situations where a
public body buys services from private or public providers and makes them available to its
consumers but does not offer consumers a choice of provider, as competition may be introduced
without choice and vice versa [89]. One such case was the introduction of an ‘internal market’’ in
the British National Health Service in the 1990s16, which built on a principal-agent-relation where
patients presented their needs to clinicians, who purchased the care they considered relevant on
behalf of the patients [34]. British reforms introduced after 2000 put more emphasis on turning
patients into informed consumers [91] with a choice of provider. The current development was
partly motivated by a suspicion that excessive vertical integration in hierarchies and monopolies
lead to inefficient allocation of resources [92-94], while quasi-markets are supposed to provide
providers with incentives to be more responsive to consumers’ demands by allocating resources to
the providers chosen by consumers, and use resources more economically, improving efficiency
[84;95]. This view reflects standard neoclassical economic theory, which assumes that the quality
of services is sustained and extended through innovation in competitive environments, as
providers’ survival depends on their capacity to develop and produce services, which consumers
want [96]. More specifically quasi markets are expected to work by putting pressure on providers
through one or both of two mechanisms [81]:
o Separation of provision of services and production of services by way of vouchers/freedom
of choice of provider – or in Le Grand’s terms: ‘choice’ [23].
o The Tiebout model [97]: consumers are provided with an opportunity to ‘vote with their
feet’ – or in Hirschmann’s and Le Grand’s terminology: ’exit’ [23;98]. Even if only a
resourceful minority of patients exercise choice the threat of exit of this group may lead to
improvements in hospital performance (see section 1.6) [20;23;98].
Little evidence is available on the workings of the first British quasi-markets in health care
developed in the 1980s and 1990s [99;100], but like markets and planned economies quasimarkets may fail, probably due to: First, failure in market formation (monopoly or oligopoly) due
to legal barriers to entry to the “market”, a “corner” in a resource (e.g. MDs), or declining marginal
costs and thereby economies of scale favouring big producers. Second: failure by preference error
due to lack of information, providers’ manipulation of consumers’ preferences, externalities not
reflected in prices, or choices between bundles of services instead of single services. Third: failure
16
Based on the working paper ”Working for Patients”[90].
13
by preference substitution, where consumers’ choices are based on other service characteristics
than assumed by the providers.
Like market failure and government failure, quasi-market failure may not be obvious: even if
consumers utilize vouchers and/or vote with their feet, consumer sovereignty may be compromised
because of too few providers or preferences biased by lack of information, manipulation or
externalities [81].
The traditional neo-classical framework for analysis of markets is not useful for analyses of quasimarkets, because quasi-markets violate two core assumptions in traditional neo-classical
economics: consumers pay for their own consumption, and transaction costs are zero. On the
contrary, in quasi markets consumption is paid, partly or in full, by the public sector, and
transaction costs are high due to uncertainty, bounded rationality and imperfect information [101].
Therefore, it is necessary to develop a specific framework for this purpose, and Le Grand [20] and
Dixon [102]17 have developed detailed models of quasi-markets and of choice as a governance tool
(Table 1). Viewing patients as fully informed and autonomous consumers [103] the models are
consistent with each other and build on assumptions which are also consistent with the model of
‘economic man’, homo economicus. The two models may be merged into a combined Le GrandDixon model for use of choice as a governance tool, including 15 assumptions about a market’s
institutional characteristics, actors’ information, actors’ preferences, and actors’ actual behaviour.18
1.8 Specific objectives and delimitations
The specific objective of the present study was to investigate whether agents’ choice behaviour on
the demand side after the introduction of choice in a hitherto supplier-driven health care system
supported key assumptions behind a model of patient choice as a governance tool. Specifically the
objective was to investigate who chose the hospital and the perspectives behind the choice by
answering the following study questions derived from six of the assumptions behind the combined
Le Grand-Dixon model (Table 1):
o Do patients – and/or the agents acting on their behalf (the GPs) – utilize their right to
choose the hospital? (assumptions 9, 13 and 15)
o Which sources of information determine choice of hospital by or on behalf of patients?
(assumption 3, 10 and 15)
o Do patients’ and GPs’ sociodemographic characteristics influence their choice behaviour?
o Which preferences for health care services’ properties are reflected in patients’ selection of
data sources? More specifically: is quality the primary factor determining the choice of
hospital (assumption 11)?
Most studies investigating patients’ utilisation of choice concern citizens’ choice of GP and
inpatients’ choice of hospital, but more and more patients in need of specialised treatment are
treated on an outpatient basis [104], and therefore the present study included a study of
outpatients as well as a study of inpatients’ choice behaviour. Due to the GPs’ role in agentprincipal-relationships a study of GPs’ choice-behaviour was performed.
17
Dixon refers to another source, but her presentation is much more condensed and specific than the original source.
When quasi-market theory was developed, much emphasis was put on comparisons of transaction costs in market vs.
nonmarket solutions [101]. Nowadays this subject plays a minor role and was not addressed in the present study.
18
14
Table 1: A combined Le Grand-Dixon model – assumptions behind use of choice as a
governance model.
Assumptions
The market
1 Opportunities for
choice
Le Grand
Dixon
Combined model
Consumers/agents may
choose the provider
Consumers are offered a
choice of provider
2 Externalities from
utilization of choice
Individual consumption
of health care generate
no externalities
Useful and accessible
information enable
consumers/agents to
make informed choices
Providers are paid by the
state
Subsidies supports
utilization of choice
1 Consumers or agents
have the opportunity to
choose the provider
2 Individual consumption of health care
generate no externalities
3 Consumers/agents
have access to relevant
and appropriate information on quality
4 Providers are paid by
the state
5 Subsidies support
utilization of choice
3 Access to
information
4 Responsibility for
payment
5 Transportation
costs
Supply: providers
6 Access to the
market
7 Availability
8 Incentives
Demand: consumers
9 Preferences
Providers are able to
enter the market and to
expand/reduce activity
Alternative providers are
available
Providers are motivated
to attract consumers –
without driving out
altruistic behavior or
encouraging creamskimming
Consumers want to
choose the provider
10 Awareness of
choice
11 Behaviour
Demand: agents
12 Agency and the
market
13 Agents’ beliefs
14 Resources
15 Behaviour
Consumers/agents have
access to relevant and
appropriate information
on quality
6 Providers are able to
enter the market and to
expand/reduce activity
7 Alternative providers
are available
8 Providers are
motivated to attract
consumers – without
driving out altruistic
behavior or encouraging
cream-skimming
Consumers want to
choose and think that
choice is important
Consumers are aware of
freedom of choice and
are able to interpret data
on quality
Quality is consumers’
primary discriminator in
choice of provider
9 Consumers want to
choose and think that
choice is important
10 Consumers are aware
of freedom of choice and
are able to interpret data
on quality
11 Quality is consumers’
primary discriminator in
choice of provider
Freedom of choice is
important to consumers
Agents have time and
resources to help
consumers make
informed choices
Agents offer choice and
information to consumers, involving them in
decision making
12 Agency does not
challenge the market’s
function
13 Freedom of choice is
important to consumers
14 Agents have time and
resources to help
consumers make
informed choices
15 Agents offer choice
and information to
consumers, involving
them in decision making
Agency does not
challenge the market’s
function
15
1.9 Sub-studies
Three sub-studies were performed to answer the study questions (Table 2):
Study 1 was a questionnaire study to investigate patients’ reasons for their choice behaviour. The
study was performed on the basis of a registry study to take advantage of the opportunity offered by
this design to document the choice behaviour of all patients in the study group. Studies 2 and 3
were performed as questionnaire studies to investigate major decision maker populations’ stated
reasons for their choice behaviour.
Study I: A questionnaire study of which factors, including information sources, decide choice of
hospital among patients facing a choice between a short waiting time and long distance to hospital
vs. a long waiting time and short distance to hospital [104].
Study II: A questionnaire study of outpatients’ utilisation of data sources when they utilise choice
[105].
Study III: A questionnaire study of the factors deciding GPs’ choice behaviour on behalf of
patients and GPs’ utilization of data sources on hospital performance (quality and service) [106].
Table 2. Study questions and sub-studies.
Study question
Do patients utilise choice of hospital?
How is the trade-off between short
waiting-time and short distance to
hospital?
Which data sources on hospital performance are utilised by patients in choice?
Is patients’ awareness and utilisation of
choice related to their background?
Is patients’ utilisation of information in
choice associated with their background?
Which data sources on hospital
performance are utilised by GPs in choice?
Is GPs’ utilization of data sources in
choice related to their background?
Are six specific assumptions behind the Le
Grand-Dixon model for choice as a
governance-tool supported by the
evidence?
Study I
X
X
Study II
X
X
X
X
X
Study III
X
X
X
X
X
X
X
X
16
2 Materials and methods
2.1 The Danish health care system
The public Danish health care system is a universal, tax financed, partly decentralized
Beveridge/NHS-system. Five public bodies, ‘regions’, each of which is headed by a council elected
by the population in the region, are responsible for provision of inpatient and outpatient health
care provided free at the point of delivery by hospitals, GPs and specialists. The regions own and
manage public hospitals and enter into agreements with GPs and specialists who are self-employed
and responsible for their own facilities and never carry out their work in a hospital.. Until a reform
coming into force on January 1 2007 the regions’ tasks were performed by similar but smaller
public administrative bodies, ‘counties’ which were abolished in the reform [107]. Examination and
treatment by general practitioners (GPs) and at public hospitals is financed by the patient’s home
region. Each Dane chooses a local GP, who is responsible for basic examinations and treatments.
GPs may refer elective patients to admittance or outpatient services at any public hospital for
specialised services. In the study period the private Danish hospital sector owned less than 1 % of
the Danish hospital beds, but private hospitals’ share of patients varied greatly between specialties
from a large share of orthopaedic patients to no or very few patients with diagnoses relevant to
departments of internal medicine.
2.2 The study areas
Studies I-III were performed in two study areas. Both of the study areas presented an opportunity
to investigate the study questions in settings without institutional barriers to utilization of free
choice of hospital:
o Each of the two study areas constituted a single uptake area (no ‘lock-in’-effects).
o No physicians’ visiting rights: GPs faced no incentives to refer patients to specific hospitals.
o Transport costs were subsidized to a greater degree than in the rest of the country.
o No user charges for hospital treatment
o GPs used the same procedure for referral to all the hospitals within the study areas: the GPs
did not face incentives to choose a hospital in the county their practice was situated in.
o Standardised information about expected waiting times etc. was available.
The study areas were partly selected on the assumption that patients’ choice of hospital would
reflect their preferences, due to the reduced institutional barriers to choice.
The three counties jointly published waiting time-prognoses for common surgical procedures. At
first these prognoses were sent to each GP by letter, later they were published on a website
managed by the Danish counties and accessible to the public and a national website maintained by
the Ministry for the Interior and Health. National data on clinical quality at hospital or department
level were not available, but the ministry published data on individual departments’ volume for
common surgical interventions as a proxy for quality; departments held regular information
meetings for the GPs about the interventions provided at the department and the department’s
procedures, and the counties and the ministry jointly published standardised data on patients’
17
experiences with hospitals. Some departments and medical societies published data on individual
departments’ performance as part of quality development or clinical research19.
2.2.1 Study area A: Roskilde County (study I and II)
Roskilde County was a mixed urban/rural county in eastern Denmark with 230,368 inhabitants as
of July 1, 1999 (area: 891.5 km2, population density: 258 inhabitants/km2), concentrated in one
large town, Roskilde, and a suburban area along the county’s Baltic coast, including the town of
Køge. In 1999, 41 % of the county’s economically active workforce worked in the Danish capital of
Copenhagen, east of the study area. A single public hospital in the study area (Køge) hosted an
orthopaedic department performing hip and knee replacement surgery. Travel distances within the
study area were small by international standards: no point in the county was farther than 66 km by
road from any of the hospitals included in study I. Roskilde and the Baltic coast were closely
connected to Copenhagen and with each other by highways and public transport.
2.2.2 Study area B: the counties of Roskilde, Storstrøm and Vestsjælland (study III).
The three mixed urban/rural counties of Roskilde, Storstrøm and Vestsjælland (801,452
inhabitants on January 1 2004 in total) constituted a single uptake area providing hospital
treatment at 13 public hospitals evenly distributed within their area. No point in the study area was
more than 30 km from the nearest public hospital in a bee line. Each specialty represented in the
study area was available at two hospitals or more, except for dermatology and plastic surgery,
which were only available at one hospital each. Patients in need of highly specialised treatment
were referred to hospitals outside the region. Each public hospital was obliged to accept any
referral from any GP in any of the three counties. Patients entitled by law to free travel to the
hospital closest to their home were offered free travel to any public hospital in the study area to
strengthen patients’ opportunities to utilise freedom of choice.
2.3 Study I: inpatients’ choice behaviour
Study question: Which factors decided choice of hospital in elective joint arthroplasty patients
facing a trade-off between short waiting time and short distance to hospital?
Setting: Roskilde County (study area A).
Source group: Patients on a waiting list for hip or knee replacement at hospital A. The patients did
not know when they would be operated on at Hospital A, except that the waiting time from the
examination to hospitalisation was 4–5 months or more (we did not know how many patients had
been notified of their date of surgery). The source group was offered treatment at the more distant
hospitals B or C with a shorter waiting time.
Study group: The study group included all the patients who accepted re-referral (“accepters”) or
preferred to remain on the waiting list (“decliners”).
Questionnaire: The members of the study group received a questionnaire developed by the
authors. The questionnaires were validated by interviews with five members of the study group by
use of an interview guide developed by Unit of Patient Evaluation, Denmark, for validation of
questionnaires regarding patients’ experiences. Data on all patients’ age, gender, diagnosis, the
county’s offer of hospital and the patients’ decisions were obtained from an administrative
database used for billing purposes.
19 For an example see reports from The Danish Society for Vascular Surgery - one of the first Danish societies to
publish data on individual departments’ results regularly – currently only in Danish [108].
18
Statistical analysis: Data were fed into a database (EPIINFO Version 3.2.2—April 14, 2004).
Respondents’/non-respondents’ and accepters’/decliners’ gender, diagnosis and decisions were
compared by use of a chi-square-test. We used the t-test to compare mean age in respondents/nonrespondents and in patients who regretted/did not regret their decision. We compared the
presence of a car in responding accepters’ and decliners’ household by use of a chi-square-test.
Respondents’ likelihood of regret was compared by use of a chi-square-test (decision, gender,
diagnosis and hospital) and a t-test (age). Accepters and decliners were compared by gender,
diagnosis and hospital by use of multivariate unconditional logistic regression.
2.4 Study II: outpatients’ choice behaviour
Study question: Were outpatients aware of their opportunities for free choice of hospital and which
sources of information were important to their choice?
Study area: The former Roskilde County (study area A).
Study group: Outpatients referred to examination, treatment (including surgery) or follow-up at
one or more of 11 somatic outpatient clinics in Roskilde County in two months of 2002. The outpatient clinics included the following specialties: Internal medicine (2), general surgery (2),
orthopaedic surgery (1), rheumatology (1), neurology (1), ophthalmology (1), paediatrics (1),
gynaecology and obstetrics (1), and otorhinolaryngology (Ear, Nose, Throat (EAT), or Head and
neck surgery) (1). In order to obtain data from all of the specialties we randomly allocated 400
patients from each outpatient clinic to the study group. For clinics visited by less than 400 patients
in the two months all patients were included in the study group. Patients were only included in the
study group once for each outpatient clinic they attended.
Questionnaire: The study group received a standardised questionnaire developed for use in a
biannual survey of inpatients’ experiences with Danish public hospitals. The original version of the
questionnaire was validated for readability and understanding by interviews with 80 inpatients and
was used for two nationwide studies of Danish inpatients’ experiences. A revised questionnaire
aimed at outpatients was validated by interviews with 12 patients from five outpatient clinics using
a standardised interview guide.
Data: The patients were asked about their awareness and utilisation of free choice of hospital.
Respondents who were aware of and utilised free choice of outpatient clinic were asked, why they
chose to be examined or treated in this outpatient clinic? No questions concerning patients’ use of
published data on clinics’ quality or service level were included. Data on each patient in the study
group included gender, age, specialty, education and employment.
Statistical analysis: Data were entered into a database (SAS). We weighted the responses from
each specialty in accordance with the specialty’s share of the number of outpatients which attended
the clinics during the study period and met the inclusion criteria. Respondents’ and nonrespondents’ gender, age and specialty were compared by a univariate chi-square-test.
Respondents’ awareness and utilisation of free choice of hospital was analysed by gender,
education and employment by a univariate chi-square test. Respondents aware of and utilising free
choice of hospital were compared by gender, age (0-60 vs. 61+ years), referring doctor and
education (none/short, medium and long) and specialty (surgical vs. medical specialty but not by
single specialties) by use of a logistic regression analysis which did not control for other factors.
Respondents’ reasons for choice of hospital were analysed by specialty (surgical vs. medical
specialty), gender, education (none/short, medium and long), employment (in employment vs.
other) and age (0-60 vs. 61+ years) using a univariate chi-square test.
19
2.5 Study III: general practitioners’ choice behaviour on behalf of patients
Study questions: Did the GPs consider themselves or the patients responsible for choice of
hospital? Which factors decided GPs’ choice of hospital on behalf of their patients? Which formal
and informal sources of information were utilised by GPs to obtain information about various
hospital departments’ standards of quality and service?
Study area: the counties of Roskilde, Storstrøm and Vestsjælland (study area B).
Study group: all of the 474 GPs practising in the study area.
Questionnaire: The questionnaire was developed after a review of the literature and face-to-face
discussions with GPs from two of the three counties. The questionnaire was validated by interviews
with three GPs. The final questionnaire included questions on the GP and on the latest three
somatic patients referred to hospital (department or outpatient clinic) by the GP:
For each of the patients: gender; age; specialty; responsibility for choice of hospital in the GP’s
opinion, and which of 16 factors were important for the choice of hospital in the GP’s opinion?
Which of 15 sources of information on quality at hospital departments did the GP in general
consider most relevant? How often did the GP use six specified sources of information on expected
waiting time at hospital departments (the GPs could choose between four categories)?
Data: The following data was recorded: the number of patients where the hospital was chosen by
the GP, the patient or the patient’s relatives. For GPs who had chosen the hospital on behalf of one
patient or more, we recorded the reasons for the choice on behalf of the latest patient referred to
hospital. Each GP was only included once in the study of reasons for choice.
Statistical analysis: Data was recorded in a database (EPIINFO Version 3.2.2. April 14, 2004).
Respondents were compared with the study population by univariate analyses of gender, county
(chi²) and number of years since graduation (t-test). This analysis was repeated for GPs who had
chosen the hospital on behalf of one or more patients.
For GPs who had chosen the hospital on behalf of one or more patients the GP’s reasons for choice
on behalf of the most recent patient were compared by univariate analysis for gender (chi²) and
years since graduation (t-test), and by logistic multiple regression analysis with the GPs’ gender
and years since graduation as the independent variables. We tested for correlation between the
number of factors for choice and the GPs’ gender and years since graduation by use of a multiple
regression analysis. The GPs’ use of information sources on quality and expected waiting time at
various hospitals were compared by univariate analyses for association with gender (chi²) or years
since graduation (t-test) and by multiple logistic regression analysis with gender and years since
graduation as the independent variables.
20
3 Results
3.1 Results, Study I: inpatients’ choice behaviour
The source group consisted of 184 patients. Forty patients (22 %) did not respond to the invitation
or asked to be deleted from the waiting list. The 86 accepters and 58 decliners made up the study
group (n = 144). One hundred and twenty-five (87 %) filled in and returned the questionnaire.
Univariate and multivariate analysis of the whole study group showed that choice of local hospital
(decliners) versus choice of distant hospital (accepters) was not associated with gender, diagnosis
or hospital. Short distance, short transport time and prior experience with Hospital A were the
most important factors behind decliners’ choice. Some participants had already been informed of
their date of surgery at Hospital A, when they received the offer of re-referral, and 50 % of the
decliners stated that this information had been important for their choice. Short waiting time was
the most important factor behind accepters’ choice. Nine patients attributed their choice to a
negative impression of Hospital A, of which six patients had been hospitalised at Hospital A.
3.2 Results, Study II: outpatients’ choice behaviour
The study group included 4,232 patients; 2,272 (54 %) filled in and returned the questionnaire. The
respondents did not differ significantly from the study group with regard to specialty, gender and
age. Forty-one percent of the respondents reported that they were aware of their right to choose the
hospital. Patients referred to surgical specialties were significantly more likely than patients
referred to medical specialties to report that they were aware of their right to choose. Reported
awareness was the greatest among patients referred to ophthalmology (54 %) and the lowest
among patients referred to neurology or rheumatology (34 %) and parents to paediatric patients
(33 %). Patients’ reported awareness also differed significantly by gender, education and
employment; female patients, patients with longer education and salaried employees in the public
sector being especially likely to report that they were aware of their right to choose the hospital by
themselves. Fifty-three percent of respondents, who reported that they were aware of their right to
choose, reported that they chose the hospital. In univariate and logistic regression analysis female
patients were significantly more likely than men to report that they chose the hospital. The share of
parents which utilised free choice of hospital on behalf of their children was markedly lower than
other patient groups. Reported utilisation of free choice was also low in patients referred to outpatient clinics in neurology, while utilisation was high in patients referred to ophthalmology (where
awareness was high as well). Utilisation in the other specialties did not differ markedly. Patients,
who were referred to a surgical specialty, were especially likely to report that they chose the
hospital (44 vs. 37 %), but this tendency was not statistically significant. Patients who had an
education of long duration, and patients who were self-employed or salaried employees in the
public sector were especially likely to choose the hospital. The statistically significant univariate
association between education and utilisation of choice disappeared in logistic regression, unlike
the association with awareness of choice. Patients 20-39 years old were especially likely to choose
the hospital, but age was not statistically significant in univariate or logistic regression. Distance to
hospital was the factor, which the greatest number of patients reported to be important for their
choice (44 %). GPs’ recommendation and waiting time were the second most important factors
behind patients’ choice in the present study (24 %). Twenty-two percent of outpatients reported,
that their own experiences influenced their choice of hospital, while seven percent were influenced
by their friends’ and five percent by their family’s experiences.
21
3.3 Results: study III: GPs’ choice behaviour on behalf of patients
Two hundred forty of 474 GPs (51 %) returned a filled-in questionnaire. Male and female GPs’
response rates were 50 % and 54 %, respectively. Respondents did not differ significantly from the
study population with regard to county, gender or number of years since graduation. Among the
240 respondents 183 (76%) reported that in their view they chose the hospital on behalf of the
latest patient referred to hospital, 35 (15%) reported that the patient made the choice, two (1%)
reported that the patients’ relatives made the choice, and 20 (8%) did not state, who chose the
hospital or ticked off several categories. Several of these GPs commented that they chose the
hospital in cooperation with the patient or that the patient agreed with the GP. Ninety-two percent
of the respondents (220 of 240) reported that they had chosen the hospital on behalf of one or
more of the latest three patients referred to hospital. These 220 respondents did not differ
significantly from the study population with regard to county, gender or number of years since
graduation. Four GPs did not tick off any reasons for choice of hospital and were excluded from
this part of the study. Eighty-seven of the 216 GPs (40 %) reported that a single factor decided their
choice for the patient, short distance to the hospital being the decisive factor for 75 of the 87 GPs.
Ninety-five GPs reported that 2-5 factors were very important for their choice, 25 quoted 6-9
factors, and nine GPs quoted ten or more factors. Short distances to hospital, the department’s
serious consideration of referrals from the GP, excellent cooperation between the GP and the
department and comments from previous patients referred to the department were the most
common factors behind GPs’ choice of hospital on behalf of patients. The importance of each factor
behind choice of hospital was not associated with the GP’s gender. Multiple regression analysis
showed no significant association between the number of reasons for choice and the GP’s gender
and the number of years since the GP’s graduation. In univariate analysis of each factor and the
number of years since graduation, GPs who based their choice on their personal experiences with
the department as employees were significantly younger than the other GPs. The most frequently
used sources of information on quality at a hospital department were reports from patients
referred to the department or the hospital by the GP previously, and other GPs’ comments on the
department. In univariate analysis female GPs were significantly more likely than male GPs to
consider official information from departments or hospitals an important source of information,
and GPs who quoted their experiences as trainees at a department or a hospital or previously
referred patients’ comments on a department as important sources of information on quality were
significantly younger than those GPs who did not. Multivariate analyses confirmed the statistically
significant associations between age and utilisation of information from previously referred
patients, and between gender and use of official information. Multivariate analysis found no
association between GPs’ age or gender and their quoting experiences from employment at a
department or a hospital as sources of information on quality. The hospitals’ letters of confirmation
of referrals were the GPs’ most important source of information on waiting times. Information
available from websites was used less often than information on paper. Multiple, logistic regression
analyses showed that male and younger GPs were especially likely to use prognoses on the internet.
22
4 Discussion
The introduction of choice as a patient right builds on the assumption, that patients derive
consequentialist utility from improved hospital performance and/or process utility [53] from
making the choice by themselves, or consequentialist utility by being treated at a better hospital
than otherwise – the present study focused on the assumptions about choice behaviour behind
utilization of patient rights (free choice of hospital) as a governance tool. The assumptions may be
illustrated in a model, where autonomous patients are aware of their opportunities for choice, want
to choose the provider, have preferences for high performance hospitals, and have access to and
collect information about various providers’ performance from information sources, including their
GP; or delegate their choice to a GP who acts in full accordance with the patient’s preference.
Figure 1: Model behind utilization of choice as a governance tool.
Patient
Patient
Specific choice of hospital
GP
Published
performance data
4.1 Do patients – and/or their agents - choose the hospital?
In study I and II we examined specific patient groups’ awareness and utilization of free choice of
hospital. In Study II outpatients’ awareness was lower than Danish inpatients’ awareness, which
was higher and growing in other studies (81 % in 2000 [109] and 87 % in 2006 [110]20). We found
no studies on this subject, despite the trend from inpatient- towards outpatient-care. Awareness
differed by specialty [111], indicating that a governance effect of choice may also differ by specialty.
International comparisons were difficult due to differences regarding the content of rights,
differences between study groups regarding specialty and patients’ mobility, and time-lags, as
choice was introduced much later in e.g. the English NHS than in Denmark, but by 2008 approx.
46 % of elective English patients recalled being offered a choice of provider at the point of referral
[112].
Approximately 49 % of Danish inpatients who were aware of choice reported that they exercised
this right, and the share is growing [109;110;113], although Danish patients’ utilization of new
rights quickly reached a stable level [114]. International comparisons were difficult due to time-lags
and different patient rights, study groups and research questions, but a review found that few
patients choose their healthcare provider [115], while studies found widely differing shares in
individual countries (31 % [74] and 61 % [116] in the Netherlands or 14-41 % in three US studies
20
Whereupon the questions about inpatients’ awareness and utilization of choice in the biannual studies were deleted
from the national surveys of patients’ experiences.
23
[117-119]). Patients’ emphasis on choice differs by age, education and income level [120;121],
indicating that the governance effect of choice is smaller in specialties with a great share of older
patients and patients with short education and low income. Le Grand challenges the assumption
that laymen face significant challenges in choices in health care, pointing out that a patient only
needs to obtain information about the relevant disease and does not need to obtain a degree in
medicine in general: “…the patient has to know only about those [diseases] potentially or actually
affecting himself or herself. Medical handbooks can aid self-diagnosis; and, once their illness is
diagnosed, especially in these days of the Internet, patients can – and indeed often do – ‘train’
themselves in their own disease.”[23] This reasoning demonstrates great confidence in patients’
ability and willingness to collect, analyze and utilize information, but results from studies of
patients’ practical decision-making processes indicate that rational decision-making is challenging
for a patient [122]. However, Le Grand also explicitly points to patients’ opportunities to take
advantage of agency-relationships, stating that: “…[patients] are exercising their royal powers by
delegating authority to someone else. The voluntary pawn is not really a pawn…”[23], thereby
dismissing critique pointing to irrationality in decision-making and challenges associated with
obtaining and processing medical information and uncertainty. According to Study I and II the GP
was the most important single factor behind choice of hospital, confirming consistent findings that
patients rely on their GP to choose for them or choose the nearest provider21[74;115;124], thereby
underscoring the need to investigate GPs’ preferences and choice behavior.
We examined GPs’ choice of hospital on behalf of patients, distinguishing between patients’ and
GPs’ choice behaviour. This distinction may be challenged, as patients and GPs may share the
choice: Study III, like several other studies, started out on the assumption that in each referral the
main responsibility for choice of hospital could be attributed to a single person: the GP or the
patient [125]. However, Study III and another Danish study [113] indicated that in 15 % or more of
the referrals the patient and the GP found that they had made the choice. The divergent findings
may be due to shared decision making, where both parts reported that they made the choice. Two
US studies found that 62 % of randomly selected US citizens outside hospital and 42 % of Medicare
patients referred to major high-risk surgery [118] reported that the choice was shared with the GP
[120], the last time they were hospitalised.
One study divided GPs into the three categories choice enthusiasts, choice sceptics and choice
paternalists [126]. Study III did not enable us to apply this distinction, but several GPs reported
that they regarded provision of information to patients about freedom of choice and providers’
performance as a task for other health care personnel or the patients themselves. The literature
points to several reasons for GPs to choose the hospital on behalf of the patient even if they are not
asked to make the choice: they may question the value of choice to patients [102;126], or they may
not consider information to patients about choice to be a part of their job[102;127]. They may
consider this task to be too time-consuming [102;128]; distrust data published by the providers
[126;128;129]; find that use of performance data interferes with their professional role [128], think
that choice reduces continuity of care [128], want to prevent being blamed by patients for
presenting faulty data [57], or simply not be aware of where to find performance data [128]. GPs’
choice behaviour varies by GP [126] and by the patients’ diagnoses [130], English GPs being more
likely to offer choice to patients in need of a routine intervention, and patients who are relatively
healthy [102]. Swedish GPs in principle approved of choice, but specific questions about their
actual referrals revealed a preference for choosing the hospital on behalf of patients [127]. The GPs
21
See for example [123].
24
attributed their choices on behalf of patients to medical reasons, and they did not regularly provide
patients with information about their rights and alternatives available [127]. English GPs described
a significant increase in the work burden, when patients were provided with information about
choice, although each GP only referred two patients a week to hospital and each of these two
patient visits were only extended by 36 seconds [131]. This apparent imbalance between actual time
spent on providing advice and the perceived burden may reflect that GPs must stay up-to-date on
various providers’ performance, or that they find time spent on advising patients about providers
less rewarding professionally than time spent on diagnostic procedures and information about the
patient’s symptoms, disease and prognosis. These results challenge the assumption that quasi
markets constitute major drivers of performance improvement, as GPs’ choice behaviour only
partly reflects performance, thereby underscoring the need to investigate the reasons determining
GPs’ choice of hospital on behalf of patients [102;126;127;132].
4.2 Which sources of information determine choice of hospital?
The present study included patients and GPs but did not include healthy citizens. However, studies
of healthy citizens’ preferences in hypothetical choices of hospital indicate that they take a keen
interest in data on providers’ performance [124;133-137]. This is especially so, if there is only a
short distance to the closest alternative hospital [138]. Major shares of citizens and patients who do
not face a choice of hospital report, that if they had faced a choice of hospital, their choice would be
based on performance data [134;139;140], and that they think that they will put more emphasis on
finding and utilizing performance data at their next hospitalization than they did before their latest
hospitalization [139]. These findings support the assumptions behind quasi markets as a
governance tool.
4.2.1 Patients’ sources of information
In Study I and II patients did not report that they utilized published performance data from
specific sources, supporting conclusions from other studies, that data has little impact on choice
[141], even when they are aware of data. Likewise, patients show an interest in hypothetical
performance data [74;137], but little interest in available comparative data, and publishing of
outcome data is not followed by patient movements towards the hospitals with the best outcomes
[142]. Utilization of performance data differs by sociodemographic groups, young patients, well-off
groups and patients with higher education being more likely to utilize such information [142-6].
In 2005 it became mandatory for Danish hospitals to inform patients, that they have the
opportunity to call hospitals and ask for data on their volume, although it is not possible to make
general conclusions about a relationship between clinical volume and outcome [147]. The present
study did not include questions about the impact of volume, but other studies indicated that
patients are less likely than clinicians to subscribe to a positive association between volume and
clinical performance [148], few patients reporting that hospital or surgeon volume influences their
choice [118;123]. Studies from the US and France have found that utilization of low volumehospitals is associated with rural residency (probably confounded by long distance to alternative
hospitals), low income, short education and attributing great weight to short distance to hospital
[149;150]. The French study unexpectedly found that some patients, who chose the surgeon by
themselves, chose a low volume-surgeon [150], and similarly a third of coronary bypass graft
patients in a US study bypassed a high-volume hospital to get to a low-volume hospital for surgery
[151]. These findings confirm conclusions from studies finding mixed evidence for the effect of
publicly released performance data on patients’ selection of provider [152], and may contribute to
understanding why studies of US hospital volumes after the publication of data on hospital quality
25
only found modest, transient effects limited to certain geographic areas and sociodemographic
groups [144]. Even when data is utilized, patients may misinterpret performance indicators and
make choices which do not reflect their preferences [153].
Patients’ limited interest in performance data may be due to several reasons [154]. Patients and
their preferences are heterogeneous: on the one hand, some patients find that they face too many
alternatives [151] or the information is too detailed resulting in cognitive overload [155], while
other patients find that the data is too aggregated and puts too much emphasis on indirect
measures of quality [156]. Patients may not be aware of the data [142]. Utilization of performance
data is only relevant to elective patients, and may necessitate postponement of treatment to search
for and compare data [157]. The interpretation of data may be a challenge, as many patients are
insufficiently informed to utilize data for choice [155;158-163] despite Le Grand’s confidence in
patients’ abilities; one study of simple one-dimensional data on mortality found a small but
significant positive effect on choice of hospital [164]. One Dutch study found that patients utilized
performance data in different and selective ways, some respondents selecting information which
confirmed their own views [165]. Apparently very specific data is more likely to strengthen patients’
interest [145]: in a US study of mandatory publishing of data on quality at fertility clinics, the
introduction of report cards was followed by an increase in market share for clinics with relatively
high birth rates. The authors’ interpretation was that the effect of performance data will be higher
when patient groups find outcome data easy to interpret, patients are young, well-educated, and
wealthy, or patients are less influenced by their GP or alternative information sources [166]. The
importance of the presentation of data is underscored by findings that patients become confused, if
they must interpret whether an indicator makes a high or a low rating preferable [153]. Patients
may also lack confidence in the data [167], or their decision making may be influenced by fatalism
[168] or anticipated regret [55;56], or they may not be aware of the number of alternatives [142].
The present study did not enable us to evaluate the influence of each of these reasons for non-use
of available performance data. Patients’ small interest in performance data may also be interpreted
as an indication that high clinical quality is taken for granted by patients or is viewed as a
professional responsibility, which patients do not need to take into consideration [169]. If this is a
major factor behind non-use of data, efforts to improve the presentation of data may not influence
utilization and the governance effect of choice.
In study I and II the patients’ own experiences with hospital departments or hospitals was one of
the most important factors behind patients’ choice, confirming results from studies which
consistently demonstrate that patients’ personal experiences constitutes a very important source of
information about alternative providers and that patients rely on their previous experiences when
they choose a provider [115]. For example a US study found that past experience was the most
important source of information to patients [170], and 52 % of Dutch patients stated that their
latest choice of hospital was influenced by their previous experiences with hospitals. A slightly
higher share (56 %) assumed that their previous experiences would also influence future choices
[116]. Canadian patients put more emphasis on previous experiences with individual surgeons than
on shorter waiting time [171], but Danish patients face a choice of department and not of clinician.
In Study I relatives’ and friends’ experiences with hospitals available played a major role for
patients’ choice of hospital, thereby resembling Dutch findings, in which 24 % of former patients
considered relatives’ and friends’ experiences important to their choice [116], and US studies of
clinically healthy citizens’ hypothetical choice, which indicate that patients’ own experiences and
other people’s reports about their experiences is more important or just as important for the choice
26
of hospital as performance data and the GP’s advice [117;159;162;171]. Other Dutch patients
attributed at least as much weight to former patients’ experiences as to performance data [141].
Hospitals’ or individual clinicians’ reputation is a common but nebulous and often ill-defined
variable in studies of choice, and several variables in studies I and II may be compared to “hospital
reputation” in other studies. A US study of elderly patients attending major surgery showed that 79
and 80 % of patients stated that hospital and surgeon reputation was “extremely” or “very”
important to their choice of hospital or surgeon, and that the GP was the most important source of
information on hospital reputation. Sixty-four percent attributed their impression of the hospital’s
reputation to the GP, while family and friends were the second most important source [124].
In Study I patients were offered a specific combination of shortened waiting time and more distant
destination, while patients in study II could utilize waiting time forecasts, if forecasts were
available for their specific diagnosis. The utilization of choice in study I was considered surprisingly
low in Denmark but high by international standards [172], as it reached the same level as the
London Choice Project. Apparently patients in Study I who utilised the offer were averse to
uncertainty about the length of the waiting time. Two British studies support the findings in Study
I on the importance of uncertain but long waiting time: elective patients facing very long waiting
times and uncertainty about its length were willing to travel much further than in Study I in order
to reduce the waiting time and its uncertainty [173;174]. A majority (57-67%) of English patients
offered a shorter waiting time at an alternative hospital took advantage of the offer [112], and in
London cataract patients’ mobility was so high, that mean waiting times at different hospitals
converged after the introduction of choice [175]. In a Danish study 31 % of patients who had chosen
the hospital by themselves reported, that the expected waiting time influenced their choice [109],
and Dutch patients put at least as much weight on waiting time data as on data on clinical
performance [141]. In Study II patients referred to surgical specialties were more sensitive to
waiting time than patients referred to medical specialties, probably because patients referred to
surgical specialties are more likely to be elective than patients referred to medical specialties. Other
studies have found associations between specialty or disease and the impact of waiting time on
choice of hospital, a short waiting time being especially important to patients referred to
departments of ophthalmology, orthopaedics or otorhinolaryngology [111]. These results may
reflect that 1) these three specialties concern abilities which are of essential importance to patients’
mobility and daily functioning, namely the patients’ ability to see, move and hear, and 2) waiting
times are usually longer in surgical than in medical specialties. The importance of the latter point is
underscored by a Norwegian study in which patients on a waiting list were more likely to consider
choosing the hospital by themselves, than patients who did not have to wait [176].
In study I the media played a minor role for patients’ evaluation and choice of hospital. This may
be due to the few media reports about local hospitals in the mass media in the study period, but
even when one of the distant alternatives was criticized in the mass media in the study period for
its standards of hygiene, patients’ refusal of re-referral to the hospital was rarely attributed to
media reports. A Dutch study also found that the media influenced few patients’ (3 %) choice of
hospital, although a slightly higher (in absolute terms) share (6 %) would like to include media
reports in future choices [116]. A US study found that reports of single sensational events at a
hospital are more important for patients’ choice of a hospital than data on general mortality [164].
A Cochrane review found that the mass media are important for the decision to utilize specific
health services [177], but the findings indicate that media reports only have minor influence on the
specific choice of provider.
27
In Study I and II a large share of the patients, who reported that they chose the hospital by
themselves (20-35 %), also reported that the GP’s recommendations were important to their choice
[109;113], confirming results from other studies of the importance of the GP to patients’ choice: in
a study from the Netherlands, where the GP acts like a gatekeeper like in Denmark, 67 % of
patients reported that the GP advised them on choice of hospital at their latest hospitalization, and
81 % reported that they thought that advice from the GP would be important or extremely
important to their future choices of hospital or specialist [116].
4.2.2 GPs’ sources of information
Despite the GPs’ major formal and documented role in choice of hospital, only few studies of GPs’
choice behaviour have been performed, some of which were performed at a time where less
information on performance was available to GPs than today.
In study III little department- and diagnoses-specific data on clinical quality was available to the
GPs, including departments’ reports on their own performance and data from some specialty
societies which published department-specific data on their performance, enabling GPs to compare
providers’ historical performance with national standards or other providers. However, even when
data on clinical quality was available, GPs – especially male GPs - showed little interest in the data.
GPs face patients with a broad range of symptoms or diseases, and constant up-dating on results
from various databases was described as a major challenge, some GPs commenting that patient
visits were of such short duration, that they could not allocate time to searching for performance
data on the relevant intervention, thereby indicating that more detailed data may not have resulted
in a greater utilization of data. These results confirm results from countries, where performance
data has been available for several years, but where GPs still show little interest in the data [178],
indicating that non-use was not just due to lack of experience.
In Study III GPs put little emphasis on standardized waiting time forecasts. Some of the reasons
may be that forecasts were not available for all diagnoses, and that waiting times were shorter than
in for example the UK. Even when formalised and prospective information was available, GPs were
more likely to rely on their memory of waiting time in past referrals or data on paper, because
accessing information on the web took more time. Some GPs stated that they intended to use
information from the web in the future. However, this reported intention may be compared with
the results of British studies from the 1990s: one British study from the early 1990s found that GPs’
previous experiences with hospitals and reports from patients played a major role for GPs’ choices,
“waiting time for first consultant appointment” being quoted as the most important reported factor
behind GPs’ choice of hospital [179]. Meanwhile British GPs presented with hypothetical scenarios
of visiting patients considered that waiting time would be the most important factor for the choice
of hospital [180;181]. However, when monthly forecasts on hospital waiting times were introduced,
the forecasts were not followed by major movements of patients from one area to another, although
some GPs stopped referring patients to distant hospitals with equally long waiting times as the
local hospitals [182].
In study III patients’ descriptions of their past experiences with various hospitals was important for
GPs’ current choices of hospital, a result conformed in a later Dutch study [128], and the GPs
generalised the reports about one hospital department to other departments at the same hospital,
thereby using patients’ reported experiences at one department as a proxy for service and/or
quality at the hospital in general. Likewise an English study found, that GPs’ previous experience
with hospitals and reports from patients played a major role for GPs’ choices, GPs reporting that
28
“confidence in consultant” was the second most important, and “quality and speed of treatment
information” the third most important factor behind choice of hospital [179].
Information from colleagues about departments and hospitals was also important for GPs’ choice
on behalf of patients. This result was consistent with other studies of GPs’ utilisation of sources of
information, which refer to GPs as having “a sort of ‘mental filing cabinet’ of “informal
information” or ”soft intelligence” [126]. The strong influence on current choices of hospital of
patients’ and GPs’ past experience on current choice of hospital is likely to introduce inertia in GP’s
choice of hospital and recommendations to patients.
The media played a minor role for GPs’ evaluation and choice of hospital in Study III. This may be
due to the few media reports about local hospitals in the mass media in the study period or GPs
may be reluctant to attribute their evaluation of hospitals to the mass media.
The results of the present study challenge the assumptions behind utilization of choice – and
thereby quasi markets – as governance tools, as patients’ and GPs’ choices of hospital were not
based on recent data on hospital performance but on reports about proxy measures on past
performance.
4.3 Influence of sociodemographic characteristics on choice behaviour?
We found some relationships between patients and GPs’ sociodemographic characteristics which
were not included in sections 4.1 and 4.2 but may still be relevant to choice behaviour:
In Study II female outpatients were significantly more likely than men to choose the clinic closest
to their home; male outpatients were significantly more likely to attribute their choice to waiting
time than female outpatients, and male outpatients were more likely than female patients to make
their choice based on their personal experiences with clinics.
Age was not a statistically significant factor behind patients’ reported utilization of choice in Study
II. Other studies reached contradictory conclusions: younger patients may need assistance to think
through what is important to them, making older patients with their greater experience with
providers and with decision making more active choosers despite their reduced mobility, although
accessing and utilising data on the internet may constitute a challenge [183]. However, older
patients tend to prefer less autonomy and seek less information on their own than young patients
[121].
In Study II male outpatients were especially likely to make their choice based on their personal
experiences with clinics.
In Study II patients’ choice behaviour varied by specialty: family’s and friends’ experiences was
especially important to outpatients referred to surgical specialties, and outpatients referred to
surgical specialties were significantly more likely to make their choice based on waiting time than
outpatients referred to medical specialties. Two interrelated reasons for paediatric patients’ low
utilisation could be a small number of paediatric departments (few alternatives) and quite long
distances between these departments: parents may have to pick up their child at a kindergarten or
school before they go to the clinic, leading to an even longer transport time for the parents, making
them especially sensitive to distance. We did not disaggregate from specialty level to diagnoses or
the severity of the specific disease. Concern about severe diseases may make patients more likely to
utilize choice, and inpatients are more likely than outpatients to suffer from serious disease [102].
In Study II outpatients not in employment were especially likely to make their choice based on
their personal experience with clinics. People in lower social positions tend to be conservative in
their choice behaviour and stick to habits rather than make new choices [184], while more highly
29
educated patients and patients with higher income make an active choice more often [115]. In
South Korea pregnant women’s awareness of published Caesarean section rates by hospital was
positively correlated with family income [185]. In a French interview study pregnant women with a
higher educational level were especially likely to choose a maternity unit with special technical
attributes [186], and such demands may contribute to explaining why these patients were
especially likely to choose specialized hospital departments.
In the US the GP’s role in choice of hospital is inversely related to the patient’s socioeconomic class,
probably due to institutional reasons, as poorer patients are more likely to use walk-in clinics or
emergency rooms [187].
The results indicate that in the study period many patients and GPs did not consider utilization of
choice to be important, thereby reducing choice’s impact as a governance tool. However, patients’
and GPs’ preferences and choicer behaviour may change over time.
4.4 Is performance the primary factor behind choice of hospital?
Introduction of choice as a governance tool to promote quality and service in health care depends
on the assumption that providers’ performance is the most important factor behind patients’ choice
of provider. However, studies from several continents with different health care systems have
shown that short distance is of great importance to the actual choice of hospital [188;189], thereby
questioning a key assumption behind choice, although a more recent (based on data from 2008-10)
Dutch study found that published performance data also have a significant effect on patients’
choice of hospital [190]. In Study I and II short distance to the closest hospital was inpatients’
primary reason for choice. A US study from a metropolitan area confirmed the importance of
distance to choice of hospital, although patients were more sensitive to travel time than to travel
distance. By rural standards numerous alternatives were available, but metropolitan patients were
still strongly biased in favour of nearby hospitals [191]. In a US registry study 60 % of patients
chose the hospital which was closest to their home. The share did not differ between the cities and
rural areas [192], suggesting that there is no threshold below which distance becomes unimportant.
Distance is a relative rather than an absolute factor when patients or GPs choose a hospital: the
farther patients live from the closest hospital, the more likely they are to choose or be referred to an
even more distant hospital [130;193]. Female patients [193], older patients [192;194-6], patients
with a short or no education [74], and disadvantaged patients in general [186] are less likely to
bypass a large local hospital and travel to a distant hospital. Patients’ educational level and/or
income may constitute confounders for distance to the closest hospital, when people with higher
education and income are more likely to live in suburbs, while hospitals tend to be located in city
centres, where people with lower educational levels and income and shorter distance to hospital
live [197]. Several studies have found a trade-off between distance to hospital and other properties
of health care services like waiting time, but the patients most willing to travel (male patients less
than 67 years old, with higher education, referred late in the study period of 2001-3), were only
willing to choose a hospital associated with an extra transport time of one hour, if the waiting time
there was 32 weeks shorter [198], while an English study found, that patients required a reduction
of 2.3 months to accept an extra transport time of one hour [199].
In Study III GPs reported, that short distance to hospital was the most important factor behind
their choice of hospital on behalf of patients, while waiting time at the local hospital made
Norwegian GPs more likely to refer patients to a distant hospital [130].
The present study confirmed results from other studies that distance is the major factor deciding
choice of hospital on choice of hospital leading to an inverse relationship between distance to
30
alternative hospitals and likelihood of choosing a hospital [116;123;198;200-5]. Distance is
independent of quality, service and efficiency, thereby challenging use of quasi markets and choice
as a governance tool aimed at improving quality and service.
4.5 A model of patients’ and their agents’ choice of hospital
Patients – and GPs - and their preferences are heterogeneous, with individual sociodemographic
backgrounds and experiences with the health care system which influence their choice of hospital,
but patients and their agents are most likely to choose the hospital closest to their home. The
findings in the present study are illustrated in a model (Figure 2) which differs from the original
model (Figure 1) by including more informal sources of information and introducing feedback
mechanisms, where information from several sources on past experiences influence patients’ and
GPs’ present choices, giving rise to new experiences which are shared with other decision-makers
and thereby influence future choices. The model emphasizes the influence of patients’ sociodemographic background on their awareness of and utilization of choice, and on the importance the
patient attributes to health care services’ characteristics and to information sources, thereby
influencing the choice of provider.
Figure 2: Model of factors behind patients’ and agents’ choice of hospital.
The patient’s
relatives and
friends
Sociodemographic
background
The patient’s experiences
with the hospital
Patient
Patient
Specific choice of hospital
GP
The GP’s experiences with
cooperation with the
hospital
Published
performance data
Colleagues
Although each patient’s experiences with individual hospital departments influence their future
choice of hospital, the influence is limited to the small number of hospital departments, where the
patient has been hospitalized in the past. Patients compensate for these limitations by utilizing
information from the GP, their friends and their family; by generalising their experiences with one
department to the whole hospital and/or by delegating the choice to the GP.
The GPs strongly influence patients’ choice when they choose the hospital on their behalf and/or
provide information about providers, based on (informal) information about previous hospitalisations from their own sources of information and their interaction with hospitals, which reflect the
31
GPs’ past experience with hospitals, rather than the hospitals’ current performance. The GPs’
choice behaviour is influenced to a minor degree by their age and gender.
Patients share their experiences at hospital departments and hospitals with the GPs, who share this
information with their colleagues. Published performance data may add to GPs’ information base,
but play a minor role as recalling informal information takes less time.
The major influence of past experiences with hospital departments on patients’ choices implies that
many sources of information influence choice of hospital and a tendency to inertia in referral
patterns, as information about proxy performance measures is provided through feedbackmechanisms with time-lags. The emphasis on informal information about hospitals implies a
minor governance effect of choice, if providers are aware, that an excellent performance may not
attract patients, while a bad performance may not keep patients away from the provider.
4.6 Strengths and weaknesses of the study
The two study areas constituted relevant subjects for the study. The health care system was an
example of the public systems whose performance quasi markets in general and choice in
particular were developed to improve. Patients and GPs chose the hospital independently of
financial considerations, and patients had a greater freedom of choice than in a NHS-system like in
England, although institutional differences between health care systems influence patients’ and
GPs’ opportunities for choice, complicating international comparisons and generalizations of
findings from studies performed in a single country. The content of choice was almost unchanged
during the study period despite changes in national legislation. Hospitals were obliged to accept
patients from other parts of the study area – thereby creating a “single market”. The counties paid
for transport costs at more distant hospitals within the region, if patients were entitled to
reimbursement to the hospital they would usually be referred to; still transport constituted a
barrier to utilization of choice, if patients wanted to be treated at a hospital in another region. The
study was performed a few years after the introduction of choice, and a study performed in study
area B found that utilisation of new opportunities for choice reached a stable level a few years after
its introduction [114]. The present study did not include the small Danish private hospital sector or
the small number of patients travelling abroad for treatment, and it was performed before the
introduction of extended freedom of choice of hospital, which probably is especially relevant to
relatively mobile and young patients. Therefore patients may be more likely to utilise choice now,
where patients have more experience and more providers are available, and the conclusions from
the present study may not necessarily be applicable to extended choice, which is especially relevant
to certain specialties and where providers’ behaviour may differ from that of public hospitals and
their employees.
The population in Study 1 was unselected but homogeneous with regard to diagnosis and age,
reducing the opportunities to generalise the conclusions to other patient groups. A high response
rate reduced the risk of selection bias. Patients were not informed about their specific expected
waiting time, presenting all the patients with the same choice based on standardised information
about their opportunities of choice, except that some patients knew when they would be operated
on at A. Patients were not asked about their social background, and data on health status was not
collected, even though patients’ views on acceptable waiting times for surgery are influenced by
their symptoms [206;207]. There was a risk of recall bias, as the participants received the questionnaire more than three years after making their choice. Respondents may have simplified the
choices they made. Many respondents elaborated on their answers, but we cannot say whether
these comments reflected that patients recalled their choices or a rationalization.
32
Study II included a major, representative study group and utilized questions validated in
comparable studies. However, a medium response rate increased the risk of selection bias. Some
patient groups may have been especially likely to answer, e.g. patients with strong views on choice
of hospital. Female patients were especially likely to be aware of and utilize choice than men, and
they were also more likely to respond. Therefore the study may have exaggerated 1) the importance
of any single factor for choice, 2) patients’ likelihood to utilize choice, and 3) factors which were
especially important to female patients. The study included many statistical tests and some
apparently statistically significant findings may reflect mass significance rather than causality.
Recall bias may constitute a problem, and chronic patients attending a regular check-up may have
chosen the clinic several years before the study. Respondents may have provided a simplified and
idealized description of the decision making. The study did not provide evidence on outpatients’
utilisation of published performance data. The study was performed nine years after the introduction of freedom of choice of public hospital in Denmark and before introduction of extended
freedom of choice. Danish patients have become increasingly used to choice, and Study II may
reflect patients’ behaviour in public health care systems several years after its introduction.
In Study III the response rate of 52 % apparently was normal for studies of GPs. Respondents did
not differ from non-respondents with regard to age, gender or county, but GPs with a stronger than
average interest in questions concerning choice of hospital may have been especially likely to
participate. Therefore the study may exaggerate the number of factors which influence choice and
of the importance of each individual factor. Usually studies should be performed prospectively to
reduce bias. However, we chose a retrospective design in order not to influence the GPs’ choice
behaviour but thereby increasing the risk of recall bias. Patients and GPs may have shared the
choice of hospital. Therefore the study may have underestimated patients’ influence on choice.
The present study did not concern quasi-markets, where a public or private purchaser chooses a
provider on behalf of patients without offering patients a choice among providers, or only a choice
among very few providers. Nor did the study concern citizens’ or patients’ choice of a financing
third part, a specific clinician at a hospital or a GP; patients’ thresholds for attending their GP for
examination and possible referral to hospital; the GPs’ thresholds for referring patients to hospital;
the effect of co-payments on choice of hospital; quantification of the effect of introduction of choice
of hospital on equity or on hospitals’ quality, production or productivity [36]; changes over time in
politicians’ reasons for extending choice22, or management of private providers – the results in the
present study should only be generalized to these topics with great care.
Assumptions 1-6 in the Le Grand-Dixon model were evaluated in the study, while assumptions 7-10
were not evaluated because an evaluation of one or more of these assumptions would necessitate
more substudies. Assumptions 11-15 were not evaluated in the present study because their degree
of fulfilment was decided by the Danish care system’s structure.
Assumption 1: Consumers and agents have access to relevant and appropriate information on
quality.
Assumption 2: Consumers want to choose and think that choice is important.
Assumption 3: Consumers are aware of freedom of choice and are able to interpret data on
quality.
Assumption 4: Quality is consumers’ primary discriminator in choice of provider.
22
A brief description of the content of freedom of choice in Denmark is provided in section 1.3 and in Appendix 1.
33
Assumption 5: Freedom of choice is important to consumers.
Assumption 6: Agents offer choice to consumers, convey information to consumers, involving
patients in decision making.
Assumption 7: Individual consumption does not generate externalities. The topic of externalities
from health care is a major one, which could not be covered within the limits of the present study.
Assumption 8: Providers are motivated to attract users to the services they provide – without
driving out altruistic behaviour or encouraging cream-skimming. Le Grand argues that the
introduction of quasi-markets does not reduce “knights” to “knaves”[20;23]. Cream-skimming is
associated with the combination of choice and activity-based financing and not by choice per se.
Departments were forbidden from declining referrals from within the study areas. Letters-ofreferral contained little and unspecific information, which could be used to sift patients. Targets
were defined at department- and hospital-level.
Assumption 9: Agency does not challenge the market’s function. This very broad assumption
was partly tested in the study, but it was hardly possible to cover every possible challenge from
agency to the market in this study.
Assumption 10: Agents have time and resources to support consumers to make an informed
choice. GPs were obliged to assist patients in making a choice and the counties/regions employed
patient advisors who provided advice on patients’ rights and opportunities for choice.
Assumption 11: Consumers or their agents have the opportunity to choose the provider. The
assumption was met in the study areas: by law patients could choose the provider freely among
providers within or outside the study areas.
Assumption 12: Providers are paid by the public sector. The assumption was fulfilled, as the
study was performed in a universal health care system without user charges for hospital care.
Assumption 13: Assistance supports consumers’ utilization of choice. The influence of subsidized
transport was not included in the present study. Danish patients’ rights depended on a detailed set
of regulation, involving patients’ age, income and mobility. Therefore Denmark was not a useful
area to study whether subsidizing relatively poor patients’ transportation costs promotes
competition and reduces inequality as proposed by Le Grand [20]. In the study areas the rules were
more supportive of choice of a distant hospital within the study area than the rules in the rest of the
country.
Assumption 14: Providers are able to enter the market and to expand/reduce activity. The
assumption was not fulfilled in the study area: interventions were distributed between the hospitals
by political and administrative decision-makers, reducing the impact of competition. It was not
possible for private providers to enter the market. Closures of public hospitals met with public
critique of politicians, thereby reducing politicians’ degrees of freedom.
Assumption 15: A number of alternative providers are available. The assumption was fulfilled:
all treatments performed in the study areas were available at several hospitals within the study area
or outside the study area in Denmark.
34
5 Conclusions and perspectives
5.1 Conclusion
Utilisation of choice as a quasi market governance tool in health care builds on an assumption, that
actors on the demand side are aware of choice, want to choose the provider by themselves based on
the alternative providers’ quality and service and thereby put pressure on providers to improve
their performance in order to attract customers. This assumption is consistent with experiences
from other sectors [208] and with healthy citizens’ views, as they tend to assume that if they were
patients, they would utilize data on performance and ‘shop around’ to select the best possible
provider by way of a rational decision making process [124;133;134;142;209]. However, the choice
behaviour which healthy people assume that they would practice in a choice situation differs from
patients’ actual choice behaviour [155], and the present study confirmed findings from other
studies that challenge key assumptions and preconditions behind the introduction of quasi markets
to improve performance in the public sector. Several years after the introduction of the quasi
market, actors only to a limited extent acted like autonomous customers in conventional markets
for goods and services, even though GPs and a major share of patients were aware of the
opportunity to choose the hospital and utilized the opportunity.
The GP was the most important source of information for patients choosing a hospital and made
the choice together with or on behalf of the rest of the patients.
Patients’ and GPs’ choice behaviour sent signals with a delay to providers to improve their
performance, as patients and their agents’ decision making processes were dominated by feedbackmechanisms, where information about past experiences influenced patients’ and GPs’ present
choices, giving rise to new experiences which in time were shared with other decision-makers and
thereby influenced future choices. Short distance to hospital was a major factor behind patients’
and GPs’ choice of hospital, and a major share of patients and GPs attributed great weight to
informal and unsystematically collected information about their own and other decision-makers’
past experiences with alternative providers, utilizing proxy information about performance even
when standardized data on quality and service was available, and relying on agent-principalrelationships to overcome challenges in interpreting even simple data.
The findings confirmed results from studies of quasi markets where performance data have been
available for a longer time. The limited utilisation of published data was attributed to several
reasons including GPs being suspicious of the data’s validity and reliability, technological
challenges, and GPs’ views on what agency implied. GPs were reluctant to spend more time with
patients on investigating their preferences and searching for data. Consequently, GPs tended to
choose the hospital based on rules-of-thump decision rules utilizing past experiences.
Outpatients’ sociodemographic characteristics significantly influenced their choice behaviour.
Consequentially the governance effect of choice may be stronger in surgical specialties than in
internal medicine. GPs’ utilisation of information sources was influenced by their age and gender,
as younger GPs were more likely to utilise information from patients referred previously, and
female GPs were more likely to utilise data published from official data sources.
Other studies of the present study’s subjects were published in scientific journals and the results
have been presented in reviews aimed at a more general public, e.g. in Denmark [210]. This
evidence has only had minor impact on policy. When evidence indicated that only a small
proportion of patients chose distant hospitals, the data was not followed by critical discussions of
35
the assumptions behind quasi markets in general and choice in particular but were interpreted as
an indication of barriers to utilization of choice.
5.2 Implications
Patients may derive utility from utilizing free choice of provider, but the present study raised
questions about its effects as a governance tool, which warrants further research into the effects of
quasi markets.
The major influence of past experiences on actual choices implies a tendency to inertia in referral
patterns, as information about providers’ performance is disseminated through feedbackmechanisms with time-lags. If GPs and patients utilize proxy data about past performance data for
decision-making in choice, the data and the choices may not reflect clinical quality but contribute
to path-dependency in choice behaviour rather than improved performance. Differences in patient
groups’ emphasis on choice indicates that the governance effect of choice is small in specialties
with a greater share of older patients and patients with short education and low income [120;121].
Even if patients and their agents utilize choice, a quasi market may not achieve all of its objectives,
as quasi markets are viewed as universal tools capable of achieving several objectives at one and
the same time, although the underlying mechanisms build on conflicting assumptions about
patient preferences: choice is expected to level out waiting times at hospitals on the assumption
that patients will go to the hospital with the shortest possible waiting time – thereby presupposing
that patients view each kind of examination and treatment as a homogeneous good, where
providers only differ by waiting time. At the same time patients are assumed to put pressure on
providers to secure high quality by staying away from hospitals with less than optimal performance
– on the assumption that patients view health care services as heterogeneous goods.
Choice is assumed by some authors to promote efficiency in health care although the mechanism
behind this effect is not quite clear and probably reflects an analogy to competitive markets where
customers pay directly for their consumption. The effect depends on the interplay between choice
and activity based financing and is sensitive to the resource allocation mechanism. It is not clear
how patients will respond, if they become aware that a certain provider operates with low costs.
Patients may view costs as a proxy for quality and suspect that less expensive hospitals compromise
patient safety.
An international tendency towards specialisation and centralisation of hospitals interacts with
choice and citizens’/patients’ preferences. The effects of extending patients freedom of choice may
be counteracted by centralisation, if patients have to move still farther for treatment by alternative
providers. It remains to be seen whether centralization results in greater mobility, making patients
more likely to choose other distant hospitals, or makes patients less likely to choose other even
more distant hospitals in accordance with results from other studies [130;193].
If the providers attempt to forestall patients’ and agents’ expected choice behaviour, quasi markets
may indirectly achieve part of the objectives. Therefore strengthening the demand side’s influence
through introduction of choice may achieve the intended effects on providers’ performance, if
department and hospital managers share the assumptions regarding patients’ choice behaviour and
are not aware of their actual choice behaviour. However, if managers are or become aware of the
divergences between model assumptions and actual patient behaviour, they may ignore the
opportunities for recruiting patients, as the management may conclude that patient mobility is
small.
Le Grand describes quasi markets as a supplement to rather than a substitute for three traditional
management models. A quasi market like choice of hospital constitutes a supplement to a
36
multitude of management tools, for example: representative democracy, management by
professionals, global budgeting and/or activity-based financing, quality measurements and/or
accreditation models to improve quality, surveys of users’ experiences etc. Therefore, even though
some of the assumptions in the model were not confirmed, other management concepts may
compensate for weaknesses inherent in quasi markets, thereby enabling them to contribute to
improving the multidimensional performance of public institutions.
The impact of quasi markets may change over time: if younger patients’ greater likelihood of
utilizing choice in the present study represents a cohort effect rather than an age effect, the effects
of quasi markets will increase in the long run.
5.3 Future research
The choice of hospital and the division of tasks between patient and GP is a complex process, and
we know little about which characteristics of health care services GPs and patients attribute
importance to. Apparently, the decision process varies greatly by patient and by GP, warranting
further studies of the interaction between GP and patient in choice of hospital, preferably by a
combination of direct observation and interviews with the GP as well as the patient rather than
questionnaire studies. Research into outpatients’ choice behaviour should distinguish between
referrals of outpatients to a clinic and outpatients attending a clinic for a check-up for an ongoing
medical condition or after discharge.
Because conclusions from studies of single quasi markets should only be generalized to quasi
markets in general with great care, detailed studies of the functioning of individual quasi markets
as management tools are warranted to take into account the characteristics of each quasi market.
37
Summary
One reason for public provision (but not necessarily public production) of health care services is
that the market for health care services is characterized by market failure. However, partly due to
incentives facing institutions and individual employees in the public sector, public provision of
health care and other services may be characterized by government failure. Attempts to reduce the
impact of government failure include introduction of market-like mechanisms like free choice of
provider, i.e. free choice of hospital, to provide an incentive for providers to improve quality,
service, responsiveness to consumers, and productivity.
The introduction of free choice of hospital in Denmark constitutes one such case of a market-like
mechanism aimed at providing patients with an opportunity to achieve ‘process utility’ by choosing
a provider, while at the same time utilizing choice to provide producers of health care services with
an incentive to attract patients by offering services of high quality and service, including short
waiting times.
This introduction of free choice of hospital as a New Public Management tool builds on an
assumption that patients’ choice behaviour fits the assumptions in economic theory about rational
consumers, including that patients seek out information about the characteristics of alternative
providers and choose the optimal combination of high quality and service (including short waiting
time). The patients may delegate the choice to an agent (the GP) acting on their behalf, ideally
making exactly that choice which the patient would have made, if she or he had had the same
education and information as the GP.
The primary contributions of the present study are that:
- Confirming results from other studies (short) distance to hospital was the primary factor
determining patients’ choice of hospital.
- Short distance to hospital was also the primary factor deciding GPs’ choice of hospital on
behalf of patients.
- GPs’ most important information sources in choice was not published performance data
(available in the study period) but reports from patients previously referred to hospital
departments and comments from other GPs - presumably based on reports from patients
they had referred to hospital.
- GPs as well as patients generalized their experience from one department at a hospital to
other departments at the hospital.
- Outpatients were less likely to be aware of and utilize freedom of choice than in-patients.
The present study implies that patients’ and GPs’ choices are made based on data which represent
proxy measures for quality and service and have been collected with major time lags.
The present study thereby challenged important preconditions for utilization in the study period of
quasi markets to promote better performance in organizations operating in nonmarket conditions.
38
Danish summary
Én årsag til offentlig tilvejebringelse (men ikke nødvendigvis produktion) af sundhedsydelser er, at
markedet for sundhedsydelser er karakteriseret ved markedsfejl. Imidlertid kan offentlig
tilvejebringelse af sundhedsydelser indebære såkaldt government failure, bl.a. pga. de incitamenter som enheder og medarbejdere i den offentlige sektor arbejder under. Initiativer mhp. at reducere betydningen af government failure omfatter bl.a. indførelse af markedslignende mekanismer
som frit-valgs-ordninger, herunder frit sygehusvalg, for at indføre incitamenter for udbyderne til at
forbedre kvalitet, service, lydhørhed over for brugerne samt produktiviteten.
Indførelsen af frit sygehusvalg i Danmark udgør sådan et eksempel på en markedslignende
mekanisme med henblik på at give patienter mulighed for at opnå ”procesnytte” ved at vælge en
udbyder, samtidig med at frit sygehusvalg benyttes til at give leverandørerne af sygehusydelser et
incitament til at tiltrække patienter ved at tilbyde ydelser med et højt kvalitets- og serviceniveau,
herunder korte ventetider.
Indførelsen af frit sygehusvalg som et New Public Management-redskab bygger på en forudsætning
om, at patienters adfærd i valgsituationer opfylder forudsætningerne i økonomisk teori om
rationelle forbrugere, herunder forudsætningerne om at de vil opsøge information om alternative
leverandører og vælge den optimale kombination af høj kvalitet og et højt serviceniveau (herunder
kort ventetid). Patienterne kan vælge at overlade valget til en agent (egen læge), der kan handle på
patientens vegne og – ideelt set - vil træffe det valg, som patienten selv ville have truffet, hvis denne
havde haft samme uddannelse og informationsgrundlag i øvrigt som lægen.
De væsentligste bidrag fra nærværende studie er at:
- Studiet bekræftede fund fra andre studier af, at (kort) afstand til sygehus er den primære
faktor bag patienters valg af sygehus.
- Kort afstand til sygehus var også den væsentligste faktor bag egen læges valg af sygehus på
vegne af patienter.
- De praktiserende lægers væsentligste informationskilder ved valg af sygehus var ikke
offentliggjorte data om sygehusafdelingers kvalitet og service (i det omfang de forelå) men
tidligere indlagte patienters beskrivelser af deres erfaringer med sygehusafdelinger samt
kollegers oplysninger, der formentlig byggede på oplysninger fra de patienter, de selv havde
henvist til sygehus.
- Både læger og patienter generaliserede erfaringer fra én afdeling til andre afdelinger på
sygehuset.
- Ambulante patienter var mindre tilbøjelige til at være opmærksomme på og benytte det frie
sygehusvalg end indlagte patienter.
Nærværende studie indikerer, at patienters og egen læges valg træffes på grundlag af data, der
afspejler proxy-mål for kvalitet og service og er indsamlet lang tid før tidspunktet for valget.
Studiets resultater udfordrede dermed væsentlige forudsætninger i studieperioden for brug af
kvasimarkeder som styringsredskab med henblik på at fremme kvalitet og service i organisationer,
der ikke fungerer på markedsvilkår.
39
Acknowledgments
I am greatly indebted to a large number of people who assisted me in my research. I am especially
indebted to professor, MPH, PhD Allan Krasnik, Section for Health Services Research, Department
of Public Health, University of Copenhagen; general manager Lars Onsberg Henriksen, MD,
Region Zealand; and professor, PhD Karsten Vrangbæk, Department of Political Science and
Department of Public Health, University of Copenhagen, who provided very valuable assistance
before and throughout the study period.
The project was made possible by enthusiastic support from politicians, management and
colleagues at Roskilde County and Region Zealand. I am grateful to the management and
administration at Roskilde Hospital for assistance in data collection and inspirational discussions.
Part of the study was carried out in cooperation with the former counties of Vestsjælland and
Storstrøm, and I would like to express my gratitude for the assistance provided by the departments
of health in the two counties which merged with Roskilde County to form Region Zealand.
Throughout the study period I was attached to Section for Health Services Research at the
Department of Public Health, University of Copenhagen, where my colleagues were very
supportive. I am especially grateful to professor, PhD Klaus Høyer, postdoc, PhD Andreas
Rudkjøbing, associate professor, PhD Jørgen Holm Petersen, Section of Biostatistics, Department
of Public Health, University of Copenhagen, and to chief advisor Rikke Gut, Center for Patient
Experience and Evaluation, the Capital Region of Denmark. It has been a great pleasure and very
inspirational to teach at the Department of Public Health and to supervise its very intelligent and
hardworking students including outstanding public health students like Sarah Wadmann
Lauritsen, Camilla Lund-Cramer, Ane Lind Møldrup, Liv Høst Dørflinger, Rie L.R. Johansen, Anne
Vinggaard Christensen and Anne Hjøllund Christiansen, who provided very relevant inputs to the
study.
During the most intensive part of the study I worked, in my spare time, as the warden at the
university student residency “Nordisk Kollegium” in Copenhagen, which constituted a very
stimulating environment for my work. I am greatly indebted to the residency’s staff and alumni for
their support, patience and kindness.
Finally I would like to thank the three persons who spurred my interest in health services research:
the late professor in health economics Gavin Mooney, who supervised me when I wrote my
master’s thesis; former CEO for Health Kjeld Kjeldsen, Ringkjøbing County, and general manager,
MD Lars Onsberg Henriksen, Region Zealand.
40
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
Moran M. Understanding the welfare state: the case of health care. The British Journal of
Politics & International Relations 2000; 2: 135-60.
Olsen JP. The ups and downs of bureaucratic organization. Annual Review of Political
Science 2008; 11: 13-37.
Hood C. A public management for all seasons? Public Adm 1991; 69: 3-19.
Pesch U. The publicness of public administration. Administration & Society 2008; 40:
170-93.
Rainey HG, Bozeman B. Comparing public and private organizations: empirical research
and the power of the a priori. Journal of Public Administration Research and Theory
2000; 10: 447-70.
van der Wal Z, de Graaf G, Lasthuizen K. What’s valued most? Similarities and
differences between the organizational values of the public and private sector. Public
Administration 2008; 86: 465-82..
Fotaki M. Are all consumers the same? Choice in health, social care and education in
England and elsewhere. Public Money Manage 2009; 29: 87-94.
Bator FM. The anatomy of market failure. Q J Econ 1958; 72: 351-79.
Arrow KJ. General economic equilibrium: purpose, analytic techniques, collective choice.
Am Econ Rev 1974; 64: 253-72.
Stiglitz J. The private uses of public interests: incentives and institutions. J Econ Perspec
1998; 12: 3-22.
Pigou AC. The economics of welfare. London: MacMillan and Company, 1920. Quoted in:
Keech WR, Munger MC, Simon C. Market failure and government failure. Paper
submitted to Public Choice World Congress 2012, Miami. Public version 1.0 - 2-27-12.
Arrow KJ. The organization of economic activity: issues pertinent to the choice of market
versus non-market allocation. In: Analysis and Evaluation of Public Expenditures: The
PPP System. Washington DC: Joint Economic Committee of Congress, 1969.
Encyclopædia Britannica Online (accessed January 4 2013).
Stigler GJ. The economist and the state. Am Econ Rev 1965; 55: 1-18.
Wolf C. A theory of nonmarket failure: framework for implementation analysis. J Law
Econ 1979; 22: 107-39.
Le Grand J. The theory of government failure. Br J Pol Sci 1991; 21: 423-42 (a critical
discussion of (8).
Keech WR, Munger MC, Simon C. The anatomy of government failure. Duke PPE
Working Paper 13.0216. Paper submitted to the 2013 meetings of the Public Choice
Society, New Orleans, Louisiana, 2013.
Niskanen W. Bureaucracy and representative government. Chicago: Aldine-Atherton,
1971.
Le Grand J. Choice and competition in publicly funded health care. Health Econ Policy
Law 2009; 4: 479-88.
Le Grand J. The other invisible hand. Delivering public services through choice and
competition. Oxford: Princeton University Press, 2007.
Exworthy M, Powell M, Mohan J. Markets, bureaucracy and public management: The
NHS: quasi-market, quasi-hierarchy and quasi-network? Public Money Manage 1999;
19: 15-22.
Winblad U, Ringard Å. Meeting rising public expectations: the changing roles of patients
and citizens. In: Magnussen J, Vrangbæk K, Saltman RB (eds.). Nordic health care
systems. Recent reforms and current policy challenges. Maidenhead: Open University
Press, 2009.
Le Grand J. Motivation, agency, and public policy. Of Knights & Knaves, Pawns &
Queens. Oxford: Oxford University Press, 2003.
41
24. Vrangbæk K, Østergren K. Patient empowerment and the introduction of hospital choice
in Denmark and Norway. Health Econ Policy Law 2006; 1: 371-94.
25. Appleby J, Smith P, Ranade W, Little V, Robinson R. Competition and the NHS:
monitoring the market. In: Tilley I (ed.). Managing the market. Liverpool: Paul Chapman
Publishing Ltd., 1993.
26. Arrow KJ. Uncertainty and the welfare economics of health care. Am Econ Rev 1963: 53:
941-73.
27. Mills A, Ranson MK. The design of health systems. In: Merson MH, Black R, Mills A
(eds.). International public health – diseases, programs, systems and policies.
Gaithersburg, Maryland: Aspen Publishers, 2001: 515-56.
28. Le Grand J, Propper C, Robinson R. The economics of social problems. 3rd edn. London:
Palgrave, Macmillan, 1992. Quoted in Greener I, Powell M. The other Le Grand?
Evaluating the ‘Other Invisble Hand’ in Welfare Services in England. Soc Policy Admin
2009; 43: 557-70.
29. Labelle R, Stoddart G, Rice T. A reexamination of the meaning and importance of
supplier-induced demand. J Health Econ 1994; 13: 347-68.
30. Hagen TP, Vrangbæk K. The changing political governance structures of Nordic health
care systems. In: Magnussen J, Vrangbæk K, Saltman RB (eds). Nordic Health care
systems: recent reforms and current policy changes. Maidenhead: Open University
Press, 2009.
31. Vrangbæk K, Robertson R, Winblad U, van de Bovenkamp H, Dixon A. Choice policies in
Northern European health systems. Health Econ Policy Law 2012; 7: 47-71.
32. Hornby AS. Oxford Advanced Learner’s dictionary of current English. Sixth edition.
Oxford: Oxford University Press, 2000.
33. Newman J, Vidler E. Discriminating customers, responsible patients, empowered users:
consumerism and the modernization of health care. Jnl Soc Pol 2006; 35: 193-2009.
34. Greener I. Towards a history of choice in UK health policy. Sociol Health Ill 2009; 31:
309-24.
35. Conner-Spady BL, Marshall DA, Bohm E, Dunbar MJ, Loucks L, Hennigar AW, Frank C,
Noseworthy TW. Patient factors in referral choice for total joint replacement surgery.
Med Care 2014; 52: 300-6.
36. Propper C, Wilson D, Burgess S. Extending choice in English health care: the
implications of the economic evidence. Jnl Soc Pol 2006; 34: 537-57.
37. Clarke J, Newman J, Wetstmarland L. The antagonisms of choice: New Labour and the
reform of public services. Social Policy &Society 2007; 7: 245-53.
38. Whitty G, Power S. Marketization and privatization in mass education systems. Int J
Educ Dev 2000; 20: 93-107.
39. Hoxby CM. School choice and school competition: Evidence from the United States.
Swedish Economic Policy Review 2003; 10: 13-67.
40. Blomqvist P. The choice revolution: privatization of Swedish welfare services in the
1990s. Soc Policy Admin 2004; 38 (2): 139-55.
41. Fotaki M, Boyd A. From plan to market: a comparison of health and old age care policies
in the UK and Sweden. Pub Money Manage 2005; 4: 237-43.
42. Swedish Association of Local Authorities and Regions. Inventering av valfrihet i
äldreomsorg Inventory of choice of elderly care. Stockholm: Swedish Association of Local
Authorities and Regions; 2009. Quoted in: Ahgren B. Competition and integration in
Swedish health care. Health Policy 2010; 96: 91-7.
43. Bredgaard T, Larsen F. Quasi markets in employment policy. Do they deliver on
promises? Social Policy and Society 2008; 7: 341-52.
44. Lewis R. More patient choice in England’s National Health service. Int J Health Serv
2005; 35: 479-83.
45. Vrangbæk K, Østergren K, Birk HO, Winblad U. Patient reactions to hospital choice in
Norway, Denmark and Sweden. Health Econ Policy Law 2007; 2: 125-52.
42
46. Pr Z, Cases C, Lisac M, Vrangbæk K, Winblad U, Bevan G. Are health problems systemic?
Politics of access and choice under Beveridge and Bismarck systems. Health Econ Policy
Law 2010; 5: 269-93.
47. Lisac M, Reimers L, Henke KD, Schlette S. Access and choice – competition under the
roof of solidarity in German health care: an analysis of health policy reforms since 2004.
Health Econ Policy Law 2010; 5: 31-52.
48. Victoor A, Friele RD, Delnoij DMJ, Rademakers JJDJM. Free choice of healthcare
providers in the Netherlands is both a goal in itself and a precondition: modelling the
policy assumptions underlying the promotion of patient choice through documentary
analysis and interviews. BMC Health Serv Res 2012; 12: 441.
49. Morris J. Independent living: the role of the disability movement in the development of
government policy. In: Glendinning C, Kemp PA (eds). Cash and Care: Policy Challenges
in the Welfare State. Bristol: Policy Press, 2006: 235-48.
50. Dowding K, John P. The value of choice in public policy. Pub Adm 2009; 87: 219-33.
51. Wilmot S. A fair range of choice: justifying maximum patient choice in the British
National Health Service. Health Care Anal 2007; 15: 59-72.
52. Sen A. Freedom of choice: concept and content. Eur Econ Rev 1988; 32: 269-94.
53. Gerard K, Mooney G. QALY league tables: handle with care. Health Econ 1993; 2: 59-64.
54. Reibling N, Wendt. Gatekeeping and provider choice in OECD healthcare systems. Curr
Sociol 2012; 60: 489-505.
55. Bell D. Regret in decision making under uncertainty. Oper Res 1982; 30: 961-81.
56. Loomes G, Sugden R. Regret theory: an alternative theory of choice under uncertainty
Econ J 1982; 86: 805-24
57. Barnett J, Ogden J, Daniells E. The value of choice: a qualitative study. Br J Gen Pract
2008; 58: 609-13.
58. Schwartz B. The paradox of choice – why more is less. New York: Harper Perennial,
2004
59. Dowding K. Choice: its increase and value. Brit J Polit Sci 1992; 22: 301-14.
60. Peters E, Klein W, Kaufman A, Meilleur L, Dixon A. More is not always better: intuitions
about effective public policy can lead to unintended consequences. Social Issues and
Policy Review 2013; 7: 114-48.
61. Porter ME, Teisberg EO. Redefining healthcare. Creating value-based competition on
results. Cambridge: Harvard Business School Press, 2006.
62. Perri G. Giving consumers of British public services more choice: what can be learned
from recent history. Jnl Soc Pol 2003; 32: 239-70.
63. Tai-Seale M. Voting with their feet: patient exit and intergroup differences in propensity
for switching usual source of care. Journal of Health Politics, Policy and Law 2004
29(3):491-514.
64. Hayek FA. The use of knowledge in society. Am Econ Rev 1945; 35: 519-30.
65. Smith A. The wealth of nations. Book 4-5, p. 32. New York: Penguin Books, 1997.
66. Vrangbæk K. The interplay between central and sub-central levels: the development of a
systematic standard based programme for governing medical performance in Denmark.
Health Econ Policy Law 2009; 4: 305-27.
67. Maynard A. Competition and quality: rhetoric and reality. Int J Qual Health Care 1998;
10: 379-84.
68. Appleby J, Harrison A, Devlin N. What is the real cost of more patient choice. London:
King’s Fund, 2003.
69. Vrangbæk K, Robertson R, Winblad U, van de Bovenkamp H, Dixon A. Choice policies in
Northern European health systems. Health Econ Policy Law 2012; 7: 47-71.
70. Friedman M. The social responsibility of business is to increase its profits. The New York
Times Magazine 1970; 122 (September 13): 32-33. Reprinted in: Zimmerli WC,
Holzinger M, Richter K (eds). Corporate Ethics and Corporate Governance. Wolfsburg:
Springer, 2007, 173-8.
43
71. Martinsen DS, Vrangbæk K. The Europeanization of health care governance:
implementing the market imperatives of Europe. Pub Adm 2008; 86: 169-84.
72. Bernstein AB, Gauthier AK. Choices in health care: what are they and what are they
worth? Med Care Res Rev 1999; 56: 5-23.
73. Besley T, Ghatak M. Incentives, choice, and accountability in the provision of public
services. Oxford Rev Econ Pol 2003; 19: 235-49.
74. Lako CJ, Rosenau P. Demand-driven care and hospital choice. Dutch health policy
toward demand-driven care: results from a survey into hospital choice. Health Care Anal
2009; 17: 20-35.
75. Kreisz FP. Gericke C. User choice in European health systems: towards a systematic
framework for analysis. Health Econ Policy Law 2010; 5: 13-30.
76. Ettinger WH, Hylka SM, Phillips RA, Harrison Jr. LH, Cayr JA, Sussman AJ. When
things go wrong: the impact of being a statistical outlier in publicly reported coronary
artery bypass graft surgery mortality data. Am J Med Qual 2008; 23: 90-5.
77. Mukamel DB, Weimer DL, Mushlin AI. Interpreting market share changes as evidence
for effectiveness of quality report cards. Med Care 2007; 45: 1227-32
78. Bevan G, Helderman JK, Wilsford D. Changing choices in health care: implications for
equity, efficiency and cost. Editorial. Health Econ Policy Law 2010; 5: 251-67.
79. http://www.lse.ac.uk/LSEHealthAndSocialCare/research/LSEHealth/ResearchNetwork
s/EHPGSEPTEMBER2006/Programme.doc (p. 5; accessed January 29 2015)
80. http://www.patientoplevelser.dk/center-patient-experience-and-evaluation
(presentation in English, accessed October 15 2014)
81. Lowery D. Consumer sovereignty and quasi-market failure. J Publ Adm Res Theor 1998;
8: 137-72.
82. Kähkönen L. Quasi-markets, competition and market failures in local government
services. Kommunal ekonomi och politik 2004; 8: 31-47.
83. Kähkonen L. Costs and efficiency of quasi-markets in practice. Local Gov Stud 2005; 31:
85-97.
84. Le Grand J. Quasi-markets and social policy. Econ J 1991; 101: 1256-67.
85. Titmuss R. The Gift Relationship: from Human Blood to Social Policy. London: Allen &
Unwin, 1970.
86. Niskanen WA. The peculiar economics of bureaucracy. Am Econ Rev 1968; 58: 293-305.
87. Boyne GA, Meier KJ, O’Toole LJ, Walker RM. Where next? Research directions on
performance in public organizations. JPART 2005; 15: 633-9.
88. Brugnoli A, Vittadini G. Subsidiarity: positive anthropology and social organization.
Foundations for a new conception of state and market and key elements of the
experience in Lombardy. Milano: Guerini e Associati SpA, 2009.
89. Hunter DJ. The case against choice and competition. Health Econ Policy Law 2009; 4:
489-501.
90. Department of Health. Working for patients. London: Her Majesty’s Stationary Office,
1989. (Cm 555)
91. Clarke J, Smith N, Vidler E. The indeterminacy of choice: political, policy and
organisational implications. Social Policy and Society 2006; 5: 327-36.
92. Enthoven AC. Reflections on the management of the National Health Service. London:
Nuffield Provincial Hospital Trust, 1985.
93. Jones PR, Cullis JG. Decision making in quasi-markets: A pedagogic analysis. J Health
Econ 1996; 15: 187-208.
94. Forster R, Gabe J. Voice or choice? Patient and public involvement in the National
Health Service in England under New Labour. Int J Health Serv 2008; 38: 333-56.
95. Jones PR, Cullis JG. Decision making in quasi-markets: A pedagogic analysis. J Health
Econ 1996; 15: 187-208.
96. Glennerster H. Competition and quality in health care: the UK experience. Int J Qual
Health Care 1998; 10: 403-10.
44
97. Tiebout CM. A Pure Theory of Local Expenditures. J Polit Econ 1965; 64 (5): 416–24.
98. Hirschman AO. Exit, voice and loyalty. Response to decline in firms, organizations and
states. Cambridge: Harvard University Press, 1970.
99. Le Grand J. Competition, cooperation or control. Tales from the British National Health
service. Health Aff 1999; 18: 27-39.
100.Fotaki M, Roland R, Boyd A, McDonald R, Scheaff R, Smith L. What benefits will choice
bring to patients? Literature review and assessment of implications. J Health Serv Res
Policy 2008; 13: 178-84.
101. Bartlett W. Quasi-markets and contracts: A markets and hierarchies perspective on NHS
reform. Public Money Manage 1991; 11 (3): 53-61.
102. Dixon A, Robertson R, Appleby J, Burge J, Devlin N, Magee H. Patient choice. How
patients choose and how providers respond. London: The King’s Fund, 2010.
103. Keaney M. Are patients really consumers? Int J Soc Econ 1999; 26: 695-706.
104. Birk HO, Gut R, Henriksen LO. Patients’ experience of choosing an out-patient clinic in
one county in Denmark: results of a patient survey. BMC Health Serv Res 2011; 11: 262.
105. Birk HO, Henriksen LO. Why do not all hip- and knee-patients facing long waiting times
accept re-referral to hospitals with short waiting time? Questionnaire study. Health
Policy 2006; 77: 318-25.
106. Birk HO, Henriksen LO. Which factors decided general practitioners’ choice of hospital
on behalf of their patients in an area with free choice of public hospital? A questionnaire
study. BMC Health Serv Res 2012, 12: 126.
107. Olejaz M, Juul Nielsen A, Rudkjøbing A, Okkels Birk H, Krasnik A, Hernández-Quevedo
C. Denmark: Health system review. Health Systems in Transition, 2012, 14(2):1 – 192.
http://www.euro.who.int/__data/assets/pdf_file/0004/160519/e96442.pdf (accessed
January 29 2015).
108.The Danish Vascular Registry: http://www.karbase.dk/English/english.htm (accessed
January 29 2015).
109. Unit of Patient Evaluation. Patienters vurdering af landets sygehuse 2000. Spørgeskemaundersøgelse blandt 34.000 patienter [in Danish: Patients’ experiences with hospitals
in Denmark 2000. Questionnaire study including 34,000 patients]. Glostrup: Enheden
for brugerundersøgelser i Københavns Amts sundhedsvæsen [Unit of Patient Evaluation,
Copenhagen County], 2001.
110. Unit of Patient Evaluation. Patienters oplevelser på landets sygehuse 2006. Spørgeskemaundersøgelse blandt 26.045 indlagte patienter [in Danish: Patients’ experiences
with hospitals 2006. Questionnaire study including 26,045 inpatients]. København:
Enheden for Brugerundersøgelser [Unit of Patient Evaluation], 2007. Summary in
English available at:
http://patientoplevelser.dk/files/dokumenter/artikel/survey_2006_english.pdf
(accessed January 29 2015).
111. Unit of Patient Evaluation. Patienters vurdering af sygehusafdelinger i Roskilde Amt.
Spørgeskemaundersøgelse blandt 3.400 patienter [in Danish: Patients’ evaluation of
hospital departments in Roskilde County 2002. Questionnaire study including 3,400
patients]. Glostrup: Enheden for brugerundersøgelser i Københavns Amts
sundhedsvæsen [Unit of Patient Evaluation], 2003.
112. Jones L, Mays N. Systematic review of the impact of patient choice of provider in the
English NHS. Working paper. London: London School of Hygiene and Tropical
Medicine. Department of Public Health and policy. Health Services Research Unit, 2009.
113. Unit of Patient Evaluation. Patienters vurdering af landets sygehuse 2002. Spørgeskemaundersøgelse blandt 32.000 patienter [in Danish. Patients’ evaluation of hospitals in
Denmark 2002. Questionnaire study including 32,000 patients]. Glostrup: Enheden for
brugerundersøgelser i Københavns Amts sundhedsvæsen [Unit of Patient Evaluation],
2003.
45
114. Birk HO, Henriksen LO. Brugen af det frie sygehusvalg inden for tre amter 1991-1999 [in
Danish: Utilization of freedom of choice of hospital within three counties, 1991-1999].
Ugeskr Laeger 2003; 165: 3613-6.
115. Victoor A, Delnoij DMJ, Fiele RD et al. Determinants of patient choice of healthcare
providers: a scoping review. BMC Health Serv Res 2012; 12: 272.
116. Berendsen AJ, de Jong GM, Schuling J, Bosveld HEP, de Waal MWM, Mitchell GK, van
der Meer K, Meyboom-de-Jong B. Patients’ need for choice and information across the
interface between primary and secondary care: A survey. Patient Educ Couns 2010; 79:
100-5.
117. Gooding SK. The effect of consumer perceptions of quality and sacrifice on hospital
choice: a suburban vs. urban competitive scenario. J Hosp Mark 1996; 11: 81-94.
118. Wilson CT, Woloshin FD, Schwartz LM. Choosing where to have major surgery. Who
makes the decision? Arch Surg 2007; 142: 242-6.
119. Kurz RZ, Wolinsky FD. Who picks the hospital: practitioner or patient. Hosp Health Serv
Adm 1985; 30: 95-106.
120. Coulter A, le Maistre N, Henderson L. Patients’ experience of choosing where to undergo
surgical treatment. Evalution of the London Choice Scheme. Oxford: Picker Institute,
2005.
121. Reed AE, Mikels JA, Simon KI. Older adults prefer less choice than young adults. Psychol
Aging 2008; 23: 671-5.
122. Tversky A, Kahnemann D. Judgment under uncertainty: Heuristics and biases. Science
1974; 185: 1124-31.
123. Unit of Patient Evaluation. Patienters oplevelser på landets sygehuse 2004.
Spørgeskemaundersøgelse blandt 26.300 indlagte patienter [in Danish: Patients’
experiences with hospitals 2004. Questionnaire study including 26,300 inpatients].
Glostrup: Enheden for Brugerundersøgelser i Københavns Amt. [Unit of Patient
Evaluation, Copenhagen County], 2005. Summary in English available at:
http://www.patientoplevelser.dk/sites/patientoplevelser.dk/files/survey_2004.pdf
(accessed January 29 2015).
124. Schwartz LM, Woloshin S, Birkmeyer JD. How do elderly patients decide where to go for
major surgery? Telephone interview survey. BMJ 2005; 331: 821-7.
125. Sloane G, Tidwell P, Horsfield M. Identification of the decision maker for a patient’s
hospital choice: who decides which hospital? J Hosp Mark 1999; 13: 57-77.
126. Rosen R, Florin D, Hutt R. An anatomy of GP referral decisions. A qualitative study of
GPs’ views on their role in supporting patient choice. London: The King’s Fund, 2007.
127. Wiinblad U. Do physicians care about patient choice? Soc Sci Med 2008; 67: 1502-11.
128. Ketelaar NABM, Faber MJ, Elwyn G, Westert GP, Braspenning JC. Comparative
performance information plays no role in the referral behaviour of GPs. BMC family
Practice 2014; 15: 146.
129. Doering N, Maarse H. The use of publicly available quality information when choosing a
hospital or health-care provider: the role of the GP. Health Expectations 2014 [Epub
ahead of print].
130. Ringard Å. Why do general practitioners abandon the local hospital? An analysis of
referral decisions related to elective treatment. Scand J Public Health 2010; 38: 597-604.
131. Taylor R, Pringle M, Coupland C. Implications of offering “Patient Choice” for routine
adult surgical referrals. Report submitted to the Department of Health. Nottingham:
University of Nottingham and Dr. Foster, 2004.
132. Dixon A, Robertson R, Bal R. The experience of implementing choice at point of referral:
a comparison of the Netherlands and England. Health Econ Policy Law 2010; 5: 295-317.
133. Kang HY, Kim SJ, Cho W, Lee S. Consumer use of publicly released hospital performance
information: assessment of the National Hospital Evaluation Program in Korea. Health
Policy 2009; 89: 174-83.
46
134. Sofaer S, Crofton C, Goldstein E, Hoy E, Crabb J. What do consumers want to know
about the quality of care in hospitals? Health Serv Res 2005; 40: 2018-36.
135. Lubalin JS, Harris-Kojetin LD. What do consumers want and need to know in making
health care choices? Med Care Res Rev 1999; 56: 67-102.
136. Det umuliges kunst – omprioriteringer i velfærdsstaten belyst ved sygehusvæsenet [in
Danish: The art of the impossible – re-prioritising in the welfare state. Case: the
hospitals]. Århus: PLS Consult, 1999.
137. Faber M, Bosch M, Wollersheim H, Leatherman S, Grol R. Public reporting in health
care: how do consumers use quality-of-care information? Med Care 2009; 47: 1-8.
138. Victoor A, Rademakers J, Reitsma-van Rooijen M, de Jong J, Delnoij D, Friele R. The
effect of the proximity of patients’ nearest alternative hospital on their intention to
search for information on hospital quality. J Health Serv Res Policy 2014; 19: 4-11.
139. Dijs-Elsinga J, Otten W, Versluijs MM, Smeets HJ, Kievit J, Vree R, van der Made WJ,
Marang-van de Mheen PJ. Choosing a hospital for surgery: the importance of
information on quality of care. Med Decis Making 2010; 30: 544-55.
140. Marang-van de Mheen PJ, Dijs-Elsinga J, Otten W, Versluijs M, Smeets HJ, Vree R, van
der Made WJ, Kievit J. The relative importance of quality of care information when
choosing a hospital for surgical treatment: a hospital choice experiment. Med Decis
Making 2011; 31: 816-27.
141. De Groot IB, Otten W, Smeets HJ, Marang-van de Mheen PJ. Is the impact of hospital
performance data greater in patients who have compared hospitals? BMC Health Serv
Res 2011; 11: 214.
142. Schneider EC, Epstein AM. Use of public performance reports: a survey of patients
undergoing cardiac surgery. JAMA 1998; 279: 1638-42.
143. Howard DH, Kaplan B. Do report cards influence hospital choice? The case of kidney
transplantation. Inquiry 2006; 43: 150-9.
144. Romano PS, Zhou H. Do well-publicized risk-adjusted outcomes reports affect hospital
volume? Med Care 2004; 42: 367-77.
145. Schauffler HH, Mordavsky JK. Consumer reports in health care: do they make a
difference? Annu Rev Public Health 2001; 22: 69-89.
146. Davies HTO, Washington AE, Bindman AB. Health care report cards: implications for
vulnerable patient groups and the organizations providing them care. Journal of Health
Care Politics, Policy and Law 2002; 27: 379-99.
147. Teisberg P, Hansen FH, Hotvedt R, Ingebrigtsen T, Kvalvik AG, Lund E, Myhre HO,
Skjeldestand FE, Vatten L, Noderharug I. Pasientvolum og behandlingskvalitet.
Metodevurdering basert på egen og internasjonal litteratugransking. [in Norwegian:
Hospital volume and quality of health outcome] Rapport nr. 2/2001. Oslo: Senter for
medisinsk metodevurdering (SMM), 2001.
148. Blendon RJ, DesRoches CM, Brodie M, Benson JM, Rosen AB, Schneider E, Altman DE,
Zapert K, Herrmann MJ, Steffenson AE. N Engl J Med 2002; 347: 1933-40.
149. Losina E, Wright EA, Kessler CL, Barrett JA, Fossel AH, Creel AH, Mahomed NN, Baron
JA, Katz JN. Neighborhoods matter. Use of hospitals with worse outcomes following
total knee replacement by patients from vulnerable populations. Arch Intern Med 2007;
167: 182-7.
150. Bouce G, Migeot V, Mathoulin-Pélissier S, Salamon R, Ingrand P. Breast cancer surgery:
do all patients want to go to high-volume hospitals? Surgery 2008; 143: 699-705.
151. Kronebusch K. Quality information and fragmented markets: patient responses to
hospital volume thresholds. J Health Polit Policy Law 2009; 34: 777-827.
152. Fung CH, Lim YW, Mattke S, Damberg C, Shekelle PG. Systematic review: the evidence
that publishing patient care performance data improves quality of care. Ann Intern Med
2008; 148: 111-23.
47
153. Jewett JJ, Hibbard JH. Comprehension of quality care indicators: differences among
privately insured, publicly insured and uninsured. Health Care Financ Rev 1996; 18: 7594.
154. Mannion R, Goddard M. Public disclosure of comparative clinical performance data:
lessons from the Scottish experience. J Eval Clin Pract 2003; 9: 277-86.
155. Øvretveit J. Informed choice? Health service quality and outcome information for
patients. Health Policy 1996; 37: 75-90.
156. Broder MS, Payne-Simon L, Brook RH. Measures of surgical quality: what will patients
know by 2005? J Eval Clin Pract 2005; 11: 209-17.
157. Hibbard J, Sofaer S, Jewett J. Condition-specific performance information: assessing
salience, comprehension and approaches for communicating quality. Health Care Financ
Rev 1996; 18: 95-109.
158. Stoop A, Vrangbæk K, Berg M. Theory and practice of waiting time data as a
performance indicator in health care. A case study from the Netherlands. Health Policy
2005; 73: 41-51.
159. Marshall MN, Shekelle PG Leatherman S, Brook RH, The public release of performance
data: what do we expect to gain? A review of the evidence. JAMA 2000; 283: 1866-74.
160. Entwistle VA, Sheldon TA, Sowden A, Watt IS. Evidence-informed patient choice. Int J
Tech Assess Health Care 1998; 14: 212-25, quoted in: Fotaki M, Roland R, Boyd A,
McDonald R, Scheaff R, Smith L. What benefits will choice bring to patients? Literature
review and assessment of implications. J Health Serv Res Policy 2008; 13: 178-84.
161. Vaiana M, McGlynn E. What cognitive science tells us about the design of reports for
consumers. Med Care Res Rev 2002; 59: 3-35.
162. Robinson S, Brodie M. Understanding the quality challenge for health consumers: the
Kaiser/AHCPR survey. Jt Comm J Qual Improv 1997; 23: 239-44.
163. Schneider EC, Lieberman T. Publicly disclosed information about the quality of health
care: response of the US public. Qual Health Care 2001; 10: 96-103.
164. Mennemeyer ST, Morrissey MA, Howard LZ. Death and reputation: how consumers
acted upon HCFA mortality. Inquiry 1997; 34: 117-28.
165. Moser A, KorstjensI, van der Weijden T, Tange H. Themes affecting health-care
consumers’ choice of a hospital for elective surgery when receiving web-based
comparative consumer information. Patient Educ Couns 2010; 78: 365-71.
166. Bundrof MK, Chun N, Goda GS, Kessler DP. Do markets respond to quality information?
The case of fertility clinics. J Health Econ 2009; 28: 718-27.
167. Bentley JM, Nash DB. How Pennsylvania hospitals have responded to publicly released
reports on coronary artery bypass graft surgery. Jt Comm J Qual Improv 1998; 24: 40-9.
168. Hibbard JH, Jewett JJ. Will quality report cards help consumers? Health Aff 1997; 16:
218-28.
169. Vuori H, Aaku T, Aine E, Erkko R, Johansson R. Doctor-patient relationship in the light
of patients’ experiences. Soc Sci Med 1972; 6: 723-30.
170. Gooding SK. The relative importance of information sources in consumers’ choice of
hospitals. J Ambul Care Mark 1995; 6: 99-108.
171. Wolinsky FD, Kurz RS. How the Public Chooses and Views Hospitals. Hosp Health Serv
Adm 1984; 29: 58-67.
172. Conner-Spady B, Sanmartin C, Johnston G, McGurran J, Kehler M, Noseworthy T.
Willingness of patients to change surgeons for a shorter waiting time for joint
arthroplasty. CMAJ 2008; 179: 327-32.
173. Howell GP, Richardson D, Forester A, Sibson J, Ryan JM, Morgans BT. Long distance
travel for routine elective surgery: questionnaire survey of patients’ attitudes. BMJ 1990;
300: 1171-1173.
174. Nofal F, Moran MM. Long distance travel by children for tonsillectomy: experience of the
ORL department at Princess Alexandra Hospital (PAH), Royal Air Force, Wroughton,
Swindon. J Laryngol Otol 1990; 104: 417-418.
48
175. Dawson D, Jacobs R, Martin S, Smith P. Is patient choice an effective mechanism to
reduce waiting times. Appl Health Econ Health Policy 2004; 3: 195-203.
176. Godager G, Iversen T. Hvem bruker retten til fritt sygehusvalg? Resultater fra Samordnet
levekårsundersøkelse 2002. Økonomisk Forum 2004; 58 (4/5): 49-56. [In Norwegian:
Who utilises the right of freedom of choice of hospital? Results from the integrated study
of living conditions 2002].
177. Grilli R, RamsayC, Minozzi S. Massmedia interventions: effects on health services
utilisation. Cochrane Database of Systematic Reviews 2002, Issue 1. Art. No.:
CD000389.
178. Schneider EC, Epstein AM. Influence of cardiac-surgery performance reports on referral
practices and access to care. N Engl J Med 1996; 335: 251-6.
179. Whynes DK, Reed G. Fundholders’ referral patterns and perceptions of service quality in
hospital provision of elective general surgery. Br J Gen Pract 1994; 44: 557-60.
180.McArdle PJ, Whitnall M. The referral practice of general medical practitioners to the surgical specialties: implications for the future. Br J Oral Maxillofac Surg 1996; 34: 394-9.
181. Kennedy F, McConnell B. General practitioner referral patterns. J Public Health Med
1993; 15: 83-7.
182. French JA, Stevenson CH, Eglinton J, Bailey JE. Effect of information about waiting lists
on referral patterns of general practitioners. Br J Gen Pract 1990; 40: 186-9.
183. Boyce T, Dixon A, Fasolo B, Reutskaja E. Choosing a high-quality hospital. The role of
nudges, scorecard design and information. London: The King’s Fund: 2010.
184. Lindbladh E, Lyttkens CH. Habit versus choice: the process of decision-making in
health-related behaviour. Soc Sci Med 2002; 55: 451-65.
185. Khang YH, Yun SC, Jo MW, Lee MS, Lee SI. Public release of institutional Caesarean
section rates in South Korea: which women were aware of the information? Health Policy
2008; 86: 10-6.
186. Combier E, Zeitlin J, de Courcel N, Vasseur S, Lalouf A, Amat-Roze JM, de Pouvourville
G. Choosing where to deliver: decision criteria among women with low-risk pregnancies
in France. Soc Sci Med 2004; 58: 2279-2289.
187. Cohen MA, Lee HL. The determinants of spatial distribution of hospital utilization in a
region. Med Care 1985; 23: 27-38.
188. Odell A. A study of patient referrals. Publ Hlth Lond 1983; 97: 109-14.
189. Mahon A, Whitehouse C, Wilkin D, Nocon A. Factors that influence general
practitioners’ choice of hospital when referring patients for elective surgery. Br J Gen
Pract 1993; 43: 272-6.
190. Beukers PDC, Kemp RGM, Varkevisser M. Patient hospital choice for hip replacement:
empirical evidence from the Netherlands. Eur J Health Econ 2014: 15: 927-36.
191. McGuirk MA, Porell FW. Spatial patterns of hospital utilization: the impact of distance
and time. Inquiry 1984; 21: 84-95.
192. Adams EK, Wright GE. Hospital choice of Medicare beneficiaries in a rural market: why
not the closest? J Rural Health 1991a; 7: 134-152.
193. Bronstein JM, Morrisey MA. Bypassing rural hospitals for obstetrics care. J Health Polit
Policy Law 1991; 16: 87-118.
194. Tai WT, Porell FW, Adams EK. Hospital choice of rural Medicare beneficiaries: patient,
hospital attributes, and the patient-physician relationship. Health Serv Res 2004; 39:
1903-22.
195. Adams EK, Houchens R, Wright GE, Robbins J. Predicting Hospital Choice for Rural
Medicare Beneficiaries: The Role of Severity of Illness. Health Serv Res 1991; 26: 583612.
196. Buczko W. Nonuse of local hospitals by rural Medicare beneficiaries. J Health Hum Serv
Adm 1997; 19: 319-40.
197. Bashshur RL, Shannon GW, Metzner CA. Some ecological differentials in the use of
medical services. Health Serv Res 1971; 6: 61-75.
49
198. Monstad K, Engesæter LB, Espehaug B. Patients’ preferences for choice of hospital. In:
Monstad K. Essays on the Economics of health and fertility. PhD-thesis. Bergen: the
Norwegian School of Economics and Business Administration, 2007.
199. Burge P, Devlin N, Appleby J, Rohr C, Grant J. Do patients always prefer quicker
treatment?: a discrete choice analysis of patients’ stated preferences in the London
Choice Project. Appl Health Econ Health Policy 2004; 3: 183-94.
200. Goddard M, Hobden C. Patient choice: a review. Report to the Department of Health.
York: University of York, Centre for Health Economics, 2003.
201. Burns LR, Wholey DR. The impact of physician characteristics in conditional choice
models for hospital care. J Health Econ 1992; 11: 43-62.
202. Chang R-KR, Joyce JJ, Castillo J, Ceja J, Quan P, Klitzner TS. Parental preference
regarding hospitals for children undergoing surgery: a trade-off between travel distance
and potential outcome improvement. Can J Cardiol 2004; 20: 877-82.
203. Al-Doghaither AH, Abdelrhman BM, Wahid Saeed AA, Magzoub MEMA. Factors
influencing patient choice of hospitals in Riyadh, Saudi Arabia. J R Soc Promot Health
2003; 123: 105-9.
204. Oliveira M. A flow demand model to predict hospital utilisation. LSE Health and
Social Care Discussion Paper Number 5. London: London School of Economics and
Political Science, 2002.
205. Hansen TB. Hvilke faktorer har betydning for valg af sygehus ved tilskadekomst? En
undersøgelse af forholdene i Ringkøbing Amt [in Danish. What factors affect choice of
hospital in cases of trauma? A study of conditions in the county of Ringkobing]. Ugeskr
Laeger 1994; 156: 652-655.
206. Lofvendahl S, Eckerlund I, Hansagi H, Malmqvist B, Resch S, Hanning M.Waiting for
orthopaedic surgery: factors associated with waiting times and patients’ opinion. Int J
Qual Health Care 2005; 17: 133–40.
207. Derrett S, Paul C, Morris JM. Waiting for elective surgery: effects on health-related
quality of life. Int J Qual Health Care 1999; 11: 47–57.
208. Jin GZ, Leslie P. The effect of information on product quality. Evidence from restaurant
hygiene grade cards. Q J Econ 2003; 118: 409-51.
209. Cheng SH, Song HY. Physician performance information and consumer choice: a
survey of subjects with the freedom to choose between doctors. Qual Saf Health Care
2004; 13: 98-101.
210. Christensen M. Offentliggørelse af kvalitetsdata [in Danish: publishing of data on
clinical quality]. DSI report 2003.01. Copenhagen: DSI The Danish Institute for Health
Services Research, 2003.
50
Appendices
Appendix 1: Free choice of hospital in Denmark – the framework
In the early 1990s the Danish public sector was a supply-driven sector characterized by:
o Production of services with properties (quantity, content, quality, location etc.) defined by
politicians, expert administrators and professionals.
o Influence by clients [1] through “voice” and regular election of politicians in the
representative democracy [2].
o Politically set constraints - physical and financial (global budgets).
During two decades of reform the sector has changed towards a supply- and demand-driven sector
characterized by:
o Production of services with properties (content, quality etc.) set by responsive
administrators and professionals.
o Actual consumption (quantity) and service properties (location, content, service level etc.)
decided through some degree of choice by consumers [1], thereby influencing provision of
services and their properties directly through exit as well as traditional representative
government and voice [2].
o Reallocation of resources to providers chosen by consumers by means of activity based
financing. However, activity based financing works within the limits of global budgets,
thereby in effect maintaining global budgeting as a government tool.
o Performance data are published with several purposes in mind – e.g. to facilitate
consumers’ choice of provider [3-5]
A broad majority in the Danish parliament which introduced choice by law by January 1 1993 had
several objectives in mind, some of them reflecting a wish to introduce a market in health care,
others focusing on patients’ rights. In the parliamentary debates it was suggested, that choice could
contribute to evening out waiting times across Denmark, as patients would choose hospitals with
short waiting times; that choice could develop into a quasi market [6], and that choice would
favour efficient hospital departments. The assumed mechanism behind the assumption about
increasing efficiency was not explained by the proponents, but later researchers [7] have pointed to
empirical data from the UK, which indicate that competition led to reduced costs and/or increased
quality, when prices were fixed [8]23.
1 Choice before introduction of choice by law
Before 1991 each of the 16 Danish counties typically divided their area into a number of hospital
uptake areas, although they were free to let patients choose a hospital within the home county.
Some counties entered into bilateral agreements allowing patients living in certain areas close to
county borders to choose a hospital in a neighbouring county. This option was relevant to few
patients, and hospitals’ financial incentives to accept these patients were small. This limited access
to choice of hospital reflected a strong emphasis on controlling public health care costs through a
supply-driven health care system.
On September 1 1991, the three counties of Roskilde, Storstrøm and Vestsjælland on the island of
Zealand introduced freedom of choice of somatic hospital for elective care within the three
23
For opposing results, see [9].
51
counties, in effect creating a single common uptake area. Likewise, three counties in the northern
part of Jutland created a single uptake area.
On October 1 1992 the Danish counties introduced freedom of choice of public hospital throughout
Denmark, probably in an attempt to forestall legislation on the subject and control the future
design of freedom of choice [10]. However, in the autumn of 1992 a broad majority of the Danish
parliament decided to introduce freedom of choice effective from January 1 1993, allowing patients
freedom of choice among public hospitals at the lowest sufficient level of specialisation, if the
receiving hospital was willing to accept the patient. The Danish parliament’s decision was a
response to:
o Criticism of long waiting times for hospital treatment
o Criticism of counties’ unwillingness to refer patients to hospitals in other counties, even if
the hospital closest to their home was situated in another county, or if patients had had
unpleasant experiences with the local hospital
o Preferences for competition in the public sector to put pressure on non-performing
hospitals [6;11]
A few non-profit hospitals – self-governed institutions or hospitals owned by four patient
organisations24 - were covered by the law, stipulating that each non-profit hospital had the right to
receive patients up to a certain limit on hospital turnover.
Over time the potential influence of choice has been extended by a number of initiatives which may
be divided into six categories:
o More patient groups have been included
o More providers of health care have been included
o More sources of information on hospital quality have been provided
o The association between hospital departments’ budgets and their activity has been
strengthened by activity based financing, leading to strengthened financial incentives to
accept patients utilizing choice
o Patients have been provided with more information on choice
o More information to patients on the providers they may choose among
In parallel with the extensions of patients’ formal rights to choose, Danish health care policy has
followed a path of centralization, reducing the actual number of providers, which patients may
choose among. For decades the counties/regions have centralized hospital capacity to achieve
economies of scale with regard to costs and quality, leading to a steady fall in the number of public
hospitals [3]. By 2007 this path was supplemented by new national regulation to increase hospital
volume for individual interventions in order to improve clinical quality: regions and private
hospitals who want to perform specialized interventions must apply to the National Health and
Medicines Authority for permission to perform the interventions, which the National Health and
Medicines Authority has decided to centralize at a few hospitals in each region or to 1-3 hospitals in
the country as a whole, thereby reducing the number of alternatives for patients and creating
barriers to entry to the “market” for specialized care. In order to improve patients’ opportunities
for making informed choices of hospital department the National Health and Medicines Authority
published proxy data on quality and service at hospital departments in the shape of a “star rating
system”25, and the National Health and Medicines Authority still publishes data on clinical quality
24
Associations concerned with examination for and treatment of epilepsy, brain damage, multiple sclerosis or
rheumatic diseases; and one private center for rehabilitation.
25 Abolished later on.
52
for specific interventions at various hospitals [12], and forecasts of waiting time at hospital level for
common interventions [13]. The regions and the Ministry of Health publish data from biannual
surveys of patients’ experience with hospitals [14].
To strengthen the understanding of the extensions and facilitation of choice a chronological list of
these initiatives is provided below. The introduction of choice was associated with some anxiety
from the counties and the Ministry of Finance that choice would weaken cost control. However,
this turned out not to be the case – or the efforts to reduce the financial risks associated with choice
were successful.
Some of the extensions reflect that the utilization of choice turned out to be to smaller than
expected, which was interpreted as a symptom of unwanted barriers to choice: the original
legislation on choice only provided patients with a right to treatment at a hospital in another
county if the hospital was willing to accept the patient, and hospitals accepting patients from other
counties only received a small financial compensation (set at the national level) for treating
patients from other regions utilizing choice26. Officially the reasoning behind the nominally small
financial incentive to accept patients was, that hospitals would utilise spare capacity for these
patients, and therefore the payment should reflect the marginal (and supposedly small) costs
rather than the (greater) average costs. Another underlying rationale for the decision to limit the
financial incentive was probably fear that choice could undermine the cost control in the Danish
public sector which was achieved during the 1980s through global budgeting with fixed limits to
each hospital’s budget – as one of the insights from health economics is, that there is basically no
upper limit to the demand for tax-financed health care provided free at the point of delivery.
The rules on counties’/regions’ right to refuse to receive patients referred from other counties/regions are vague; at no time since the introduction of choice has it been clarified when hospitals
are allowed to refuse referrals of patients from other counties/regions.
2 Introduction of choice by law by January 1 1993
On January 1 1993 freedom of choice was introduced by law, allowing patients freedom of
choice among public hospitals at the lowest sufficient level of specialisation, if the receiving
hospital was willing to accept the patient [15].
1997
By July 1 1997 highly specialized hospital departments were allowed to receive patients referred to
hospital with symptoms/diseases which could be diagnosed/treated at departments performing
basic hospital tasks [16], thereby increasing the number of providers involved in freedom
of choice for more patient groups, if the specialized hospital departments were willing to
accept patients referred to basic hospital treatment. The original law on choice only provided
patients with a right to choose a hospital department at the lowest effective level of specialization.
However, some patients with ailments which could be treated at specialized hospitals departments
complained to the Ministry of the Interior and Health that they would like to be treated at highly
specialized departments (“tertiary” departments in WHO’s terminology), on the assumption that
tertiary departments were especially good not only at highly specialized interventions but at
standard interventions as well. However, opening up highly specialized departments for patients
with minor ailments ran counter to national efforts to strengthen specialization at tertiary hospital
26 The financial compensation included a bed-day rate supplemented with specific supplementary rates for drugs,
prostheses etc.
53
departments, and could increase health care costs, as highly specialized departments’
reimbursement rates were higher than standard departments’ rates. Therefore the highly
specialized departments’ financial incentive to accept such patients was limited to the (relatively
small) payment which the standard department would receive, thereby reducing the incentive to
accept patients who could be cared for at lower levels of specialization. The patient’s home county
could demand that patients were examined at a local hospital before the patient was referred to the
highly standardized department to establish whether the patient was in need of treatment of highly
specialized treatment and thereby how much the home county should pay27.
By July 1 1997 it became mandatory for hospital departments to inform patients about choice, if
waiting time exceeded waiting times at other departments considerably, thereby providing patients
with more information on choice from a new source [16].
1998
By 1997 several counties employed patient advisors, but by January 1 1998 it became mandatory
for the counties to employ one or more patient advisors whose task was to provide patients with
information on their rights, thereby providing patients with access to another source of
information on choice [17].
1999
By 1998 choice apparently had not led to major patient movements, or to deficits in the health care
sector as feared. Meanwhile interest in strengthening market-like governance mechanisms and
other NPM-tools continued to grow. By January 1 1999 the Danish parliament authorized the
secretary of state for health to order the counties to inform patients about the number of
treatments and waiting times at various hospitals, thereby introducing another source of
information and more information on the providers, which patients may choose
among [18].
2000
In 1998 the Danish parliament decided to introduce payment of 100 % DRG-rates effective by
January 2000 for patients utilizing choice, and made activity based financing of hospitals
mandatory: the allocation of at least 10 % of each hospital’s budget should be decided by the size of
the clinical production28. These initiatives strengthened the hospitals’ and/or the counties’
financial incentives to accept patients from other counties.
By July 1 2000 three private hospices were included in choice increasing the number of
providers involved in freedom of choice and the patient groups covered by choice [19].
2002
By 1 July 2002 “extended freedom of choice” was introduced by the national government, allowing
somatic patients to choose a private hospital, which the counties had entered into agreement with,
27
This rule enabled the county to 1) present itself to the patient (and hopefully - from the department’s and county’s
viewpoint – to convince the patient that there was no need for referral to the highly specialized department), and 2)
examine the patient and establish whether the department found that the county should pay the relatively low standard
reimbursement rate og the higher rate for patients in need of highly specialized treatment.
28 In activity-based financing no distinction was made between patients from the county and choice patients from other
counties. Therefore there was no financial incentive for hospitals and departments to distinguish between patients from
within the county and patients from other counties.
54
if the region could not provide examination/treatment within two months after the referral at one
of the county’s own hospitals or one of the hospitals which the county usually cooperated with. The
initiative increased the number of providers involved in freedom of choice. The
government made it clear, that it expected the counties to enter into agreement with so many
private clinics/hospitals that the new set of rules would provide a real increase in the number of
alternatives for patients. However, there were several restrictions on utilisation of “extended choice
of hospital”. Extended choice did not apply:
o If the GP’s diagnosis was insufficient to establish whether the intervention could be
performed at a private hospital
o If the intervention at a public hospital was delayed by the patient
o If there was a medical indication for delaying the intervention
o If waiting time at the private clinic/hospital was longer than at the public hospital
o To the following interventions: transplantations, sterilisation, IVF, re-fertilisation,
adjustment of hearing aids, cosmetic interventions, gender change, convalescence,
research/trials or alternative medicine.
Patients may only utilize extended freedom of choice when they have been referred to a hospital
and it has become clear that the patient cannot be treated within the time limit. This means that,
unlike in free choice of public hospital, it is not possible for the GP to refer patients directly to a
private provider [20].
A homepage was established to provide patients with information about all the agreements
between the Danish regions and private providers to enable patients to investigate which providers
were relevant for the specific diagnosis, thereby providing patients with more information on
the providers they may choose among [21].
2003
By July 1 2003 the secretary for health was authorized by law to allow GPs and hospitals to refer
patients directly to highly specialized hospital departments, bypassing hospital departments in the
home county. Hitherto the GPs had had to refer patients to a local hospital for evaluation of the
patient’s need for treatment at highly specialized departments (see 1997). Formally this change had
no implications for patients’ rights, but it may have increased utilization of choice of highly
specialized hospital departments [22].
By November 2003 new legislation made it mandatory for the counties to employ the patient
advisors in the county’s administration rather than at hospitals – in an attempt to prevent that
hospitals’ interests interfered with the patient advisors’ obligations to assist the patients. This
strengthening of the patient advisors’ independence represented an attempt to make sure that
patients received more information on choice [23].
2004
By January 1 2004 the activity-based share of the hospitals’ budget was increased from 10 % to 20
%, strengthening the hospital departments’ financial incentive to accept patients
from other counties. Some counties distributed a greater share of the budget through activity
based financing in order to promote production and/or to make hospitals and department
managers accustomed to activity based financing [24].
55
2005
By January 1 2005 the rules on mandatory information to patients about freedom of choice were
tightened to provide patients with more information on choice: Within eight days from
reception of a referral, the hospital had to inform the patient about: 1) time and location for
examination or treatment at the hospital; 2) whether the hospital was able to provide
examination/treatment within two months [one month from 1 October 2007]; 3) the opportunities
for choice according to free choice and extended free choice of hospital; 4) waiting times at public
and non-profit hospitals; 5) the opportunity to receive information about hospital volume, and 6)
that the hospital is willing to re-refer the patient to another hospital, if the patients want to be
treated there [25].
In the spring of 2005 the Danish counties introduced a new precondition for funding databases on
clinical quality: the counties (regions) only fund databases which publish data on individual
departments’/hospitals’ performance, thereby providing patients with more information on
the providers they may choose among [26], if a database for clinical quality covers their
health problem.
In a letter of July 5 2005 to the counties the National Health and Medicines Authority pointed out
that a rule allowing psychiatrists to block psychiatric patients’ utilization of free choice of hospital
should only be utilized to avoid that treatment of individual patients was interrupted and in effect
hindered by repeated re-referrals of patients utilizing choice. The rule did not constitute a general
opportunity to ignore psychiatric patients’ right to choice. The letter did not change the rules on
choice, and no evidence is available on whether the letter influenced psychiatric patients’ utilization
of choice.
2006
By January 1 2006 two private centres for rehabilitation of torture victims were included in choice,
increasing the number (and range) of providers [27].
On October 30 2006 the Danish National Health and Medicines Authority started regular
publishing of proxy measures of quality and service measures aggregated to a “star-rating-system”,
thereby providing patients with more (proxy) information on the providers they may
choose among. The star-rating system was quietly suspended several years later.
2007
By January 1 2007 the activity based budgets’ share of the hospitals’ total budgets were raised to at
least 50 % [29], strengthening the hospital departments’ financial incentives to accept
patients from other counties. Some regions distributed a higher share of the budget (70 %)
through activity based financing in order to promote production and/or to make hospitals and
department managers accustomed to activity based financing29.
By June 1 2007 a departmental order the National Health and Medicines Authority made it
mandatory for hospital personnel who declined to let a psychiatric patient utilise choice, to
document their decision in the patient’s file. Presumably the departmental order reflected
continuing suspicion that psychiatric patients were barred from utilizing choice. The order did not
29 For example since its inception on January 1 2007 Region Zealand distributes 70 % of its budget to the hospitals by
way of activity based financing.
56
by itself change the rules on choice, and no evidence is available on whether the order influenced
psychiatric patients’ access to and utilization of choice [30].
From 1 October 2007 the time limit regarding extended freedom of choice was lowered from two
months to one, thereby offering choice to more patients [31].
2008
By August 1 2008 extended freedom of choice was extended to include patients referred to
examination (but not treatment) at departments and clinics of child and adolescent psychiatry,
allowing patients to choose a private provider, which the regions have entered into agreement with,
if the region could not provide examination/treatment within two month (not one month like
somatic patients) of referral at one of the region’s own departments or clinics or one of the public
providers which the county usually cooperated with. The initiative increased the range of
providers involved in freedom of choice, and more patient groups were offered extended
choice of hospital [32].
From November 7 2008 to June 30 2009 extended freedom of choice was suspended due to a
strike by nurses working at public hospital [33], causing a temporary reduction in the
opportunities for choice.
2009
By January 1 2009 extended freedom of choice was extended further to include patients referred to
treatment at departments and clinics of child and adolescent psychiatry, allowing patients to
choose a private provider, which the regions had entered into agreement with, if the region could
not provide examination/treatment within two months (not one month like somatic patients) at
one of the county’s own departments or clinics or one of the public providers which the county
usually cooperated with. Waiting time was measured from the day when a psychiatrist has decided
to refer the patient to treatment. The initiative extended choice to include more patient groups
and more providers [32].
2010
On January 1 2010 extended freedom of choice was extended to include adult patients referred to
examination or treatment at psychiatric departments, allowing patients to choose a private
provider, which the regions had entered into agreement with, if the region could not provide
examination/treatment within two months (not one month like somatic patients) at one of the
county’s own departments or clinics or one of the public providers which the county usually
cooperated with. The initiative increased the range of providers involved in freedom of choice
and more patient groups were offered extended choice of hospital [34;35].
2013
Effective from January 1 2013 utilization of extended freedom of choice of hospital for less serious
diseases was limited to patients waiting for two or more months, thereby reducing some patient
groups’ access to extended choice [36]. The National Health and Medicines Authority was
made responsible for drawing up criteria for distinguishing between “more serious” and “less
serious” diseases and conditions.
57
2014-15
In June 2013 the national and the government and the Danish regions entered into an agreement
on the conditions for the regions’ budgeting for 2014 including:
o A transparency reform involving publishing of yearly reports on the health care system’s
performance. By July 2014 data was published at the regional level only and so far did not
constitute another source of information on hospital quality [37].
o An obligation for the regions to develop a system to improve transparency regarding spare
hospital capacity and ease referrals between hospitals and between regions. Depending on
the facility’s final design, it may ease utilization of choice and provide patients with more
information on providers they may choose among [37].
References:
1. Callahan K. Citizen participation: models and methods. Int J Public Admin 2007; 30: 117996.
2. Hirschman AO. Exit, voice and loyalty. Response to decline in firms, organizations and
states. Cambridge: Harvard University Press, 1970.
3. Olejaz M, Juul Nielsen A, Rudkjøbing A, Okkels Birk H, Krasnik A, Hernández-Quevedo C.
Denmark: Health system review. Health Systems in Transition, 2012, 14(2):1 – 192.
4. Magnussen J, Vrangbæk K, Saltman RB, Martinussen PE. Introduction: the Nordic model
of health care. In: Magnussen J, Vrangbæk K, Saltman RB. Nordic Health Care Systems:
recent reforms and current policy changes. Maidenhead: Open University Press, 2009: 320.
5. Martinussen PE, Magnussen J. Health care reform: the Nordic experience. In: Magnussen
J, Vrangbæk K, Saltman RB. Nordic Health Care Systems: recent reforms and current
policy changes. Maidenhead: Open University Press, 2009: 21-52.
6. Vrangbæk K. Markedsorientering i sygehussektoren: Opkomst, udformning og
konsekvenser af frit sygehusvalg. PhD-afhandling. [in Danish: Market orientation in the
hospital sector. Introduction, design and consequences of patients’ freedom of choice of
hospital. PhD-thesis] København: Institut for Statskundskab, Københavns Universitet,
1999.
7. Le Grand J. Choice and competition in publicly funded health care. Health Econ Policy Law
2009; 4: 479-88.
8. Cooper Z, Gibbons S, Jones S, McGuire A. Does hospital competition save lives? Evidence
from the English NHS patient choice reforms. Econ J 2011; 121: F228-F260.
9. Propper C, Burgess S, Green K, Does competition between hospitals improve the quality of
care?: Hospital death rates and the NHS internal market. J Public Econ 2005; 88: 1247-72
10. Vrangbæk K, Østergren K. Patient empowerment and the introduction of hospital choice in
Denmark and Norway. Health Econ Policy Law 2006; 1: 371-94.
11. Vrangbæk K, Østergren K, Birk HO, Winblad U. Patient reactions to hospital choice in
Norway, Denmark and Sweden. Health Econ Policy Law 2007; 2: 125-52.
12. http://www.esundhed.dk/sundhedskvalitet/Sider/sundhedskvalitet.aspx (in Danish only,
accessed January 29 2015).
13. http://www.esundhed.dk/sundhedskvalitet/Sider/sundhedskvalitet.aspx (in Danish only,
accessed January 29 2015).
14. http://www.patientoplevelser.dk/center-patient-experience-and-evaluation (presentation
in English, accessed January 29 2015).
15. Act no. 1024 of December 19 1992 on amendments to the Hospital Act and the Public
Health Insurance Act (more free choice of hospital, treatment at private specialized
hospitals, concentration of services, prehospital care, patients ready for discharge, and
planning and cooperation).
58
16. Act no. 464 of June 10 1997 on amendment to the Hospital Act (strengthening of freedom
of choice of hospital, authority to perform experimental organizational changes, and
municipality services).
17. Agreement of 1997 between the Association of County Councils and the Ministry of Finance
for 1998.
18. Act no. 1044 of December 23 1998 on amendment to the Hospital Act (center for evaluation
of hospitals, strengthening of free choice of hospital, information about waiting times etc.).
19. Act no. 470 of May 31 2000 on amendment to the Hospital Act (hospices and free choice of
hospital).
20. Act no. 143 of March 25 2002 on amendment to the Hospital Act (higher limits to free
choice, state subsidies for hospital purposes, and free choice of private hospitals).
21. www.sygehusvalg.dk (accessed January 29 2015).
22. Act no. 383 of May 28 2003 on amendment to the Hospital Act (referral to highly
specialized departments etc.).
23. Act no. 428 of June 10 2003 on amendment to the Act on the National Health and
Medicines Authority etc. (changed representation in the Patients’ Complaints Board, set up
of advisory offices for patients at county level etc.).
24. Agreement of June 14 2003 between the Association of County Councils and the Ministry of
Finance (only available in Danish: http://www.fm.dk/publikationer/2003/regeringensaftale-med-amtsraadsforeningen-om-amternes-oekonomi-i2004/download/~/media/Files/Publikationer/2003/Download/ARF_aftale.ashx
(accessed January 29 2015).
25. Act no. 441 of June 9 2004 on amendment to the Hospital Act and the Act on public health
insurance (strengthening of freedom of choice of hospital etc.).
26. Basiskrav for kliniske databaser [in Danish only: basic requirements for databases on
clinical quality]. Copenhagen: Danske Regioner [Danish Regions], May 11 2007.
http://www.regionh.dk/NR/rdonlyres/C1C921D6-123C-45A1-A7E230DA2B27A050/0/11052007_basiskrav_for_landsdaekkende_kliniske_kvalitetsdatabaser
pdf.pdf (accessed January 29 2015).
27. Act no. 1395 of December 21 2005 on amendment to the Hospital Act and the Health Act
(extension of free choice of hospital to include DIGNITY – Danish Institute against
Torture).
28. Agreement of June 10 2006 between the association Danish Regions and the Ministry of
Finance on the regions’ finances in 2007 (available in Danish only: Finansministeriet.
Aftaler om den regionale og kommunale økonomi for 2007. København: Finansministeriet,
2006.
29. Departmental order of May 8 2007 on recording rejections of psychiatric patients’
utilization of freedom of choice of hospital.
30. Act no. 1556 of December 20 2006 on amendment to the Health Act; the Act on the Right to
Complain and Receive Compensation within the Health Service, and the Pharmacy Act (1
month waiting time limit, access to the Medicine Profile etc.).
31. Act no. 539 of June 17 2008 on amendment to the Health Act; the Act on the Right to
Complain and Receive Compensation within the Health Service, and the Act on regions and
the abolishment of the counties, the Greater Copenhagen Authority and the Copenhagen
Hospital Corporation (extended rights to treatment for children and young people with
mental disease, and physiotherapy provided by the municipalities). The act was
implemented in two steps: the right to treatment only took effect on January 1 2009
32. Act no 1064 of November 6 2008 on amendment to the Health Act (suspension of extended
freedom of choice of hospital).
33. Act no 530 of June 12 2009 on amendment to the Health Act (extended freedom of choice
of hospital for adult psychiatric patients).
34. Act no. 1521 of December 27 2009 on amendment to the Health Act (freedom of choice of
diagnostic examinations, board of arbitration, extended freedom of information etc.).
59
35. Act no. 1401 of December 23 2012 on amendment to the Health Act and to the Act on the
Right to Complain and Receive Compensation within the Health Service (right to early
examination and differentiated right to extended freedom of choice of hospital).
36. Aftale af 4. juni 2013 om regionernes økonomi for 2014 (In Danish. Agreement of June 4th
between the national government and the Danish regions on the framework for the regions’
budgeting for 2014. Link to the agreement (in Danish only): http://www.fm.dk/nyheder/pressemeddelelser/2013/06/aftale-om-regionernes-oekonomi-for2014/~/media/Files/Nyheder/Pressemeddelelser/2013/06/DR%20aftale/aftale%20om%2
0regionernes%20økonomi%20for%202014.pdf (accessed January 29 2015).
60
Appendix 2-4: Original papers
61
Health Policy 77 (2006) 318–325
Why do not all hip- and knee patients facing long waiting times
accept re-referral to hospitals with short waiting time?
Questionnaire study
Hans Okkels Birk ∗ , Lars Onsberg Henriksen
Roskilde County, Department of Health, Amtsgaarden, Koegevej 80, P.O. Box 170, 4000 Roskilde, Denmark
Abstract
Patients’ preferences are often assumed to be homogeneous and to favour hospitals with a short waiting time and high quality.
Due to long waiting times (6 months) for artificial hip or knee implantation a Danish county in 1999–2000 offered patients
on a waiting list a choice between remaining on the local hospital’s waiting list with the long waiting time, or re-referral to a
hospital outside the county with a shorter waiting time. Fewer patients than expected took advantage of the offer of re-referral
(“accepters”): 89 of 149 patients (60%). In 2003, we asked patients about the reasons for their choice: 87% of patients responded.
Respondents and non-respondents were similar by decision, choice of hospital, diagnosis and age; men were significantly more
likely to respond than women. Accepters and decliners were similar by age, sex, diagnosis and the presence of a car in the
household. Short distance, short transport time and previous experience with the nearby hospital were the most important
reasons for choosing that hospital. Some patients appeared to be willing to accept a long waiting time, if they were told exactly
when they would undergo surgery. The results of this study question the validity of the conventional wisdom, that patients are
willing to travel long distances in order to receive treatment with short waiting time.
© 2005 Elsevier Ireland Ltd. All rights reserved.
Keywords: Waiting time; Choice behaviour; Patient mobility; Waiting lists
1. Introduction
A large majority of Danish and English healthy citizens report that they would be willing to travel to distant
hospitals for treatment with short waiting time [1–3]:
waiting time being viewed as a performance indicator
for hospitals [4], and short waiting time being viewed as
∗ Corresponding author. Tel.: +45 46 30 38 46;
fax: +45 46 32 04 83.
E-mail address: [email protected] (H.O. Birk).
an indisputable and important good in patients’ choice
of hospital. Meanwhile, a growing number of studies
of patients’ choice performed in various countries and
in different health care systems demonstrate inverse or
negative relationships between distance to health care
and its utilisation [5–7]. This study concerns a natural experiment, where elective patients were offered a
trade-off between short waiting time and short distance
to hospital.
Due to long waiting times (6 months) for artificial hip or knee replacement a Danish county offered
0168-8510/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.healthpol.2005.08.002
H.O. Birk, L.O. Henriksen / Health Policy 77 (2006) 318–325
319
patients on a waiting list a choice between remaining
on the local hospital’s waiting list with the long waiting time or re-referral to a hospital outside the county
with a shorter waiting time, assuming that almost all
patients would accept re-referral to reduce their waiting time. However, fewer patients (59%) than expected
took advantage of the offer of re-referral to cut waiting
time. Other Danish counties have carried out similar
initiatives, but we have not found any scientific studies
of these projects. Two English studies presented different patient groups with a simple trade-off between
a long waiting time and a long distance to hospital and
found very different results: roughly half of patients
waiting for surgery chose re-referral to a distant hospital [8], while 95% of parents of clinically relevant
children with recurrent tonsillitis chose a distant hospital [9]. New legislation allows Danish patients to
choose among a larger number of public and private
hospitals, and hospitals’ financial incentives to receive
patients have been strengthened, making it more difficult to perform a prospective study of a simple trade-off
between short distance and short waiting time. Therefore, we have investigated why patients accepted or
declined the offer of earlier treatment at a distant hospital (“accepters” and “decliners”), testing the following
hypothesis:
responsible for basic examinations and treatments. The
GPs are self-employed and responsible for their own
facilities and never carry out their work in a hospital
but may refer their patients to admittance or out-patient
services at any public hospital for specialised services.
The private Danish hospital sector owns less than 1%
of the Danish hospital beds.
Roskilde County is a mixed urban/rural county
in eastern Denmark with 230,368 inhabitants as of
July 1, 1999 (area: 891.5 km2 , population density:
258 inhabitants/km2 ), concentrated in one large town,
Roskilde, and a suburban area along the county’s Baltic
coast, including the Town of Koge. In 1999, 41% of the
county’s economically active workforce worked in the
Danish capital of Copenhagen, east of the study area.
A single public hospital (A) in the study area (Koge)
hosts an orthopaedic department, which performs hip
and knee replacement surgery.
Travel distances within the study area are small by
international standards. No point in the county is farther
than 66 km by road from any of the hospitals included
in this study. Roskilde and the Baltic coast are closely
connected to Copenhagen and with each other by highways and public transport (buses and railways).
On average accepters were younger than decliners.
Men were more likely to accept re-referral than
women.
Hip- and knee patients were equally likely to accept
re-referral.
Decliners were more likely to regret their decision
than accepters.
Short distance was the most important factor behind
decliners’ choice.
Short transport time was more important to decliners
than short distance.
The source group was made up of a population of
patients on a waiting list for hip or knee replacement
at A. The patients were placed on A’s waiting list after
an examination at the hospital had established, that a
replacement was indicated. A notified the patient about
the date of surgery shortly before their hospitalisation. Therefore, the patients did not know when they
would be operated on, except that the waiting time
from the examination to hospitalisation was at least
4–5 months.
The study group included all the patients who
accepted re-referral or preferred to remain on the waiting list. Patients who did not answer or who asked to
be deleted from the waiting list were excluded from the
study.
2. Material
2.2. Participants
2.1. Setting
2.3. Intervention
In Denmark, examination and treatment by the general practitioner (GP) and at public hospitals is taxfinanced and provided free of charge by the patients’
home county. Each Dane chooses a local GP, who is
Roskilde County bought 89 hip or knee replacements at 2 public hospitals in Copenhagen: B, a hospital in the eastern part of Copenhagen, performed hip
320
H.O. Birk, L.O. Henriksen / Health Policy 77 (2006) 318–325
replacements but not knee replacements; C, a hospital
in the western/central part of Copenhagen, performed
hip and knee replacements.
The county sent an invitation to patients on the waiting lists in the autumn of 1999, explicitly stating that
they were invited to be re-referred to earlier treatment
at B or C, and asking whether they preferred re-referral
or remaining on the waiting list at A. No specific date
of surgery at B or C was reported in the invitation, but
the patients were informed that any re-referred patient
would experience a shorter waiting time than patients
who remained on A’s waiting list. In order to ease
patients’ access, patients eligible to free transport to
A were offered free transport to B or C as well. No
specific official data were available to the county; the
GPs; the patients; or the hospitals on the quality of
treatment or patient satisfaction at the three hospitals.
B was the object of some media interest in the summer
of 1999 because of complaints from patients regarding
the cleanliness of its wards.
Patients who did not want to remain on the waiting
list or did not respond to the invitation were deleted
from A’s waiting list as reported in the invitation. When
fewer patients than expected chose re-referral, more
patients with shorter expected waiting time than in the
originally selected group were invited, until 89 patients
had accepted re-referral. The hip and knee replacements at B and C were performed during the autumn
of 1999 and the winter of 1999–2000.
2.4. The questionnaire
The members of the study group received a questionnaire developed by the authors. In an attempt to
increase the questionnaire’s relevance to each participant, and thereby increase the response rate, we prepared different questionnaires for accepters and decliners. The questionnaires were validated by interviews
with five members of the study group who were asked to
explain their understanding of the content of each question; which answers they missed; which topics ought
to be included in the study. The validation was performed using an interview-guide developed by Unit of
Patient Evaluation, Denmark, for validation of questionnaires regarding patient satisfaction. The validation
led to slight changes of the answers and more openended questions were added. In the absence of major
changes the questionnaire was not re-tested.
Data on all patients’ age, sex, diagnosis, offer of
hospital and decision were obtained from an administrative database used for billing purposes. All respondents were asked whether they regretted their choice
and whether there was a car in the household. The
accepters were asked about their previous experience
with A and the hospital they chose, and how important the following factors were for their choice: waiting
time; access for visitors; their impression of and experience with A. We asked the decliners about their ability
to get to B or C, when they were offered re-referral,
and about the influence of the following factors on
their choice: transport expenses; pain during transport;
transport time; distance to the hospital offered; a scheduled day of surgery at A; preferences for a hospital
the patient had been treated at before; recommendations of A from relatives or friends; access for visitors,
and a bad impression of or negative experience with B
or C. Questions on waiting time or about severity or
duration of disease were not included, assuming that
patients would find it difficult to present valid detailed
responses 3 years after the experience.
The questionnaires were sent in June 2003. In order
to increase the response rate we enclosed a freepost
return-envelope and a letter from the head of the county
council’s committee for health, emphasising that the
ultimate objective of the study was to tailor the county’s
health services to the public’s needs, and that the study
was carried out in cooperation with the University of
Copenhagen. In August 2003, we sent one reminder to
the non-respondents.
2.5. Statistical analysis
Data were fed into a database (EPIINFO Version
3.2.2—April 14, 2004). Respondents/non-respondents
and accepters/decliners sex, diagnosis and decisions
were compared by use of a chi-square-test. We used
the t-test to compare mean age in respondents/nonrespondents and in patients who regretted and did not
regret their decision. We compared the presence of a
car in responding accepters’ and decliners’ household
by use of a chi-square-test. Responding patients’ likelihood of regret was compared by use of a chi-square-test
(decision, sex, diagnosis and hospital) and a t-test (age).
Accepters and decliners were compared by sex, diagnosis and hospital by use of multivariate unconditional
logistic regression.
H.O. Birk, L.O. Henriksen / Health Policy 77 (2006) 318–325
321
Table 1
Descriptive statistics for respondents and non-respondents
Respondents
Non-respondents
Total, n
n
%
n
%
125
87
19
13
144
Sex
Women
Men
70
55
81
95
16
3
19
5
86
58
Hospital
B
C
47
78
85
88
8
11
15
12
55
89
Diagnosis
Hip
Knee
86
39
87
87
13
6
13
13
99
45
Decision
Accepters
Decliners
76
49
88
84
10
9
12
16
86
58
Mean age, years (standard deviation)
64.6 (11.1)
Total
p
0.02
0.71
0.97
0.50
66.6 (14.1)
0.49
N = 144. Univariate analysis: ANOVA parametric test (age) and chi-square, uncorrected, two-tailed.
3. Results
The source group consisted of 184 patients. Forty
patients (22%) did not respond to the invitation or
asked to be deleted from the waiting list. The 86
accepters and 58 decliners made up the study group
and received a questionnaire (n = 144). One hundred
and twenty-five (87%) filled in and returned the questionnaire. Univariate and multivariate analysis of the
whole study group showed that choice of local hospital
(decliners) versus choice of distant hospital (accepters)
was not associated with sex, diagnosis or hospital
(Table 2). Older patients were slightly more likely to
accept re-referral, but this tendency was not significant (coefficient 0.0203 (p = 0.18), constant −0.9172
(p = 0.35)). Univariate analysis showed that responding
was not associated with diagnosis or hospital, but men
were significantly more likely to respond than women
(Table 1). Multivariate analysis likewise found that men
Table 2
Acceptance of re-referral by patients’ characteristics
were significantly more likely to respond (p = 0.02).
Responding accepters and decliners were equally likely
to have a car in the household (accepters: 59%; decliners: 41%, p = 0.42). Patients treated at B were more
likely to regret than patients treated at C, while regret
was not associated with decision, sex, diagnosis or age
(Table 3). Among accepters more patients re-referred
to B criticised the cleanliness and the access to rehabilitation than patients re-referred to C.
Table 3
Association of regret with other variables
Regret
Total
n
%
n
%
Decision
Accepters
Decliners
12
5
16
10
64
44
84
90
76
49
Sex
Women
Men
11
6
16
11
59
49
84
89
70
55
13
4
15
10
73
35
85
90
86
39
11
6
23
14
36
72
77
86
47
78
Patients’ characteristics
Odds ratio
95% CI
p
Diagnosis
Hip
Knee
Hospital offered (C/B)
Diagnosis (knee/hip)
Gender (male/female)
1.9152
1.3867
0.8223
0.8485–4.3233
0.5680–3.3856
0.4092–1.6524
0.12
0.47
0.58
Hospital
B
C
N = 144. Multiple unconditional logistic regression.
No regret
p
0.37
0.44
0.46
0.01
N = 125. Univariate analysis: chi-square, uncorrected, two-tailed.
322
H.O. Birk, L.O. Henriksen / Health Policy 77 (2006) 318–325
Table 4
Patients’ reasons for accepting or declining re-referral from A
Very important
Important
Quite important
Total (%)
Reasons to decline re-referral
Short distance to A
Short transport time to A
Prior experience with A
Visitors’ easy access to A
Surgery scheduled at A
Relatives recommend A
Transport too expensive
Transport too painful
Bad experience with B or C
Bad impression of B or C
22
17
22
11
20
10
1
2
2
0
9
11
6
11
4
6
2
3
0
0
1
3
2
3
0
4
4
1
3
1
32 (65)
31 (63)
30 (61)
25 (51)
24 (49)
20 (41)
7 (14)
6 (12)
5 (10)
1 (2)
Reasons to accept re-referral
Short waiting time
Bad impression of A
Visitors’ easy access
Bad experience with A
53
3
1
2
15
3
2
1
2
3
3
2
71 (91)
9 (12)
6 (8)
5 (6)
N = 49 decliners and N = 78 accepters.
Short distance, short transport time and prior experience with A were the most important factors behind
decliners’ choice (Table 4). Some participants had
already been informed of their date of surgery at A,
when they received the offer of re-referral. Fifty percent of the decliners stated that this information had
been important for their choice. We do not know how
many decliners or acceptors had been notified of their
date of surgery. Short waiting time was the most important factor behind accepters’ choice. Nine patients
attributed their choice to a negative impression of A, of
which six patients had been hospitalised at A. Thirtyseven accepters reported that they were satisfied with
a previous experience at A. Decliners who mentioned
transport distance as a factor behind their choice were
equally likely to mention transport time and attributed
similar weight to that factor (p < 0.0001).
4. Discussion
Our study was performed in a country, where much
emphasis has been put on providing patients with a
choice of hospital and achieving short waiting times
to surgery, and where conventional wisdom states that
patients are willing to travel long distances in order
to receive treatment with short waiting time elsewhere
in the country. However, patients’ choice of hospital
is also influenced by the distance to hospital, the hos-
pitals’ quality and reputation and the patients’ social
network. Patients’ choice may also be influenced by
their age, but most of the participants were recruited
from a narrow age group. This group is highly relevant
because it is responsible for most of the utilisation of
health care but our results may not be applicable to
other age groups.
4.1. Waiting time
Waiting time was the dominant reason for accepters’
choice. Half of the decliners already knew when they
would be operated on if they remained on A’s waiting list, when they received and declined the offer
of earlier treatment. Decliners’ willingness to accept
a longer waiting time at the local hospital than at a
distant hospital, if they were told exactly when they
would be operated on, indicates that uncertainty about
the length of the waiting time was more important to
these patients than the numerical length of waiting time.
This hypothesis could not be tested in our study and
the finding may not apply to waiting times, which are
longer than in this study. In two English studies, the
uncertainty about the waiting time at the local hospital
was greater, and the patients were more likely to accept
to travel much farther, suggesting that uncertainty was
important to patients’ choice [8,9]. However, studies of
patients’ views on waiting time for surgery indicate that
the numerical length of waiting time indeed influences
H.O. Birk, L.O. Henriksen / Health Policy 77 (2006) 318–325
patients’ satisfaction negatively: cataract patients generally accept waiting times of 3 months or less, while
waiting times of 6 months or more are likely to be perceived as too long [10,11]. Orthopaedic patients’ waiting time is a significant predictor of patients’ acceptance of waiting time [12]. Cataract patients perception
of waiting time for cataract surgery being too long is
not associated with patients’ demographic characteristics [11], but in younger patients employment may
strengthen patients’ preferences for early treatment. A
study in New Zealand of patients on a waiting list for
hip or knee joint replacement found a moderate inverse
correlation between severity categories like bodily pain
and physical functioning and acceptable waiting time
[13], but the present study included no measures of
severity.
4.2. Short distance and/or transport time to the
hospital
Short distance/transport time to the closest hospital was patients’ primary reason for remaining on A’s
waiting list, and this finding is consistent with findings in other studies of patients’ movements [14,15].
Unlike in several other studies the patients in our
study were free to choose, were conscious of the
alternative, and had plenty of time to consider the
choice.
Inspired by studies, which found that use of transport time or transport distance as measures of accessibility produced different results, or that transport
time was more important to patients than distance
[16], we asked patients about the importance of transport distance as well as transport time. Patients did
not distinguish between the two factors, indicating
that it is sufficient to ask patients about transport
distance, which is easier to measure than transport
time.
4.3. Quality and reputation
A few patients chose a hospital outside the county,
because they had gained a negative impression of
the closest hospital. These patients attributed the bad
impression to their friends rather than to media reports,
indicating that the media are less important for patients’
views on specific hospitals or departments than the
experience of patients’ friends and relatives.
323
No specific official data were available on quality
of care at A, B and C, but even when data on quality at
hospital- or surgeon level are available, few patients
display much interest in it [17]. The importance of
relatives’ and patients’ own previous experience with
specific hospitals may reflect that patients use their relatives’ and friends’ comments as a proxy measure for
service and/or quality at the hospital.
The patients’ own positive experience with A was
the third most important factor behind decliners’ choice
(and the second most important factor if transport time
and distance are viewed as a single factor), but half of
the accepters were satisfied with A as well. It was surprising that so many patients reported that they had
been treated at A before. Some respondents’ comments indicate that they referred to the examination
in the orthopaedic out-clinic before the implantation
rather than previous treatment at the hospital. Still,
it is noteworthy that patients put so much emphasis on their previous experience, even though only a
minority of the decliners could have had experience
with joint replacements at the department. Presumably
some decliners used their experience with other departments as a proxy measure for service or quality at the
orthopaedic department—a “halo”-effect.
4.4. Social network
Patients’ social network may strengthen their mobility, if the relatives are willing to drive them to and
from the hospital. In health care a predominantly local
social network may actually reduce patients’ willingness to travel for treatment, if visits play a major role for
patients while they are hospitalised. Potential visitors’
access to the hospitals was one of the factors which
the largest number of patients reported influenced their
choice, but on average they attributed little weight to
the factor, usually considering it an “important” rather
than a “very important” factor. Apparently, the factor influenced many patients’ choice of hospital but
was of secondary importance compared to other factors
like distance to hospital. One Danish study of citizens’
hypothetical choice of hospital found that 9% reported
that visitors’ access would influence their choice [18].
The low rating may reflect that the choice was hypothetical. Patients facing an actual choice, like in this study,
may put more emphasis on visitors’ opportunities for
access.
324
H.O. Birk, L.O. Henriksen / Health Policy 77 (2006) 318–325
4.5. Limitations of the study
The population was unselected, because all patients
in the source population were placed on a waiting list
in chronological order, and because the patients with
the longest expected waiting time were invited first. A
high response rate further reduced the risk of selection
bias, although men were more likely to respond than
women. Patients were not informed about their specific
expected waiting time, presenting all the patients with
the same choice based on standardised information
about their opportunities of choice, except that some
patients knew when they would be operated on at A.
The population was homogeneous with regard to
diagnosis and age, reducing the opportunities to generalise the conclusions to other patient groups. Patients
were not asked about their social background. Most of
the patients were pensioners but within this group educational level and marital status may influence patients’
choice. Data on health status were not included. This
was a drawback because patients’ views on acceptable waiting times for surgery is influenced by their
symptoms [12,13]. Recall bias is probably a serious
weakness in the study: the participants received the
questionnaire more than 3 years after they made their
choice. Recall bias will probably make respondents
simplify the choice they made, thereby exaggerating
the influence of the most important reasons for the
choice. Many respondents elaborated on their answers,
but we cannot say whether these comments reflect that
patients found it easy to remember the actual reasons
for their choice or an attempt to rationalise or even
defend their decision.
Patients’ preferences are often assumed to be homogeneous, favouring hospitals with a short waiting time
and documented high standards of quality. Therefore,
findings indicating that patients are most likely to
choose the hospital nearest to their home are often
viewed as symptoms of barriers to choice of hospital
or irrational patient behaviour.
This study indicates that elective patients’ preferences regarding hospitals are heterogeneous and elective patients’ choice of the hospital closest to their
home often represent a conscious selection of that hospital. Long waiting times are an important factor behind
many but far from all patients’ choice of hospital and
apart from the influence of waiting time per se it may
act as a confounder for uncertainty about the length of
the waiting time as well; some patients being willing
to make a trade-off between a long waiting time and
other factors, if they know exactly when they will be
treated. Other patients put more emphasis on choosing
a hospital they trust due to their own or other people’s
experience. Patients facing a choice of hospital have
few or no data they can base their choice on. If they
make a wrong choice, the consequences may be very
unpleasant for the patients. Many patients are averse
to uncertainty and try to reduce this uncertainty by
choosing a specific rather than an uncertain date of
surgery, and by using their own, friends’ and/or relatives’ experience with certain hospitals as an indicator
of the service and quality they will receive at a certain
hospital department.
The results of this study question the validity of the
conventional wisdom, that patients are willing to travel
long distances in order to receive treatment with short
waiting time. More emphasis should be put on providing care close to the patients’ home.
5. Conclusion
We found that accepters were slightly older than
decliners within the study population’s narrow age
range, female patients were slightly more likely to
accept re-referral than male patients, knee patients
were slightly more likely to accept re-referral than hip
patients, and accepters were slightly more likely to
regret their choice than decliners, but all of these results
were insignificant. Short distance to hospital was the
most important reason behind decliners’ choice, and
short distance was equally important to patients as short
transport time.
References
[1] PLS Consult. Det umuliges kunst—om prioriteringer i
velfærdsstaten belyst ved sygehusvæsenet [The art of the
impossible—on priority setting in the welfare states; case: the
hospital sector]. Aarhus: PLS Consult; 1999 [in Danish].
[2] Ryan M, McIntosh E, Dean T, Old P. Trade-offs between location and waiting times in the provision of health care: the case of
elective surgery on the Isle of Wight. Journal of Public Health
Medicine 2000;22:202–10.
[3] Ross M-A, Avery AJ, Foss AJE. Views of older people on
cataract surgery options: an assessment of preferences by conjoint analysis. Quality and Safety in Health Care 2003;12:13–7.
H.O. Birk, L.O. Henriksen / Health Policy 77 (2006) 318–325
[4] Newton JN, Henderson J, Goldacre MJ. Waiting list dynamics
and the impact of earmarked funding. British Medical Journal
1995;311:783–5.
[5] Shannon GW, Skinner JL, Bashshur RL. Time and distance:
the journey for medical care. International Journal of Health
Services 1973;3:237–44.
[6] Porell FW, Adams EK. Hospital choice models: a review and
assessment of their utility for policy impact analysis. Medical
Care Research and Review 1995;52:158–95.
[7] Place M. Concentration and choice in the provision of hospital services. The relationship between concentration, patient
accessibility and utilisation of services. CRD report 8, part III.
York: The University of York. Centre for Health Economics
York Health Economics Consortium, NHS Centre for Reviews
and Dissemination; 1997.
[8] Howell GP, Richardson D, Forester A, Sibson J, Ryan JM,
Morgans BT. Long distance travel for routine elective surgery:
questionnaire survey of patient’s attitudes. British Medical Journal 1990;300:1171–3.
[9] Nofal F, Moran MM. Long distance travel by children for
tonsillectomy: experience of the ORL Department at Princess
Alexandra Hospital (PAH), Royal Air Force, Wroughton, Swindon. Journal of Laryngology and Otology 1990;104:417–8.
[10] Conner-Spady BL, Sanmugasunderam S, Courtright P, McGurran JJ, Noseworthy TW. Steering Committee of the Western
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
325
Canada Waiting List Project. Determinants of patient satisfaction with cataract surgery and length of time on the waiting list.
British Journal of Ophthalmology 2004;88:1305–9.
Dunn E, Black C, Alonso J, Norregaard JC, Anderson GF.
Patients’ acceptance of waiting for cataract surgery: what makes
a wait too long? Social Science and Medicine 1997;44:1603–10.
Lofvendahl S, Eckerlund I, Hansagi H, Malmqvist B, Resch S,
Hanning M. Waiting for orthopaedic surgery: factors associated
with waiting times and patients’ opinion. International Journal
for Quality in Health Care 2005;17:133–40.
Derrett S, Paul C, Morris JM. Waiting for elective surgery:
effects on health-related quality of life. International Journal
for Quality in Health Care 1999;11:47–57.
Mooney C, Zwanziger J, Phibbs CS, Schmitt S. Is travel distance
a barrier to veterans’ use of VA hospitals for medical surgical
care? Social Science and Medicine 2000;50:1743–55.
Cohen MA, Lee HL. The determinants of spatial distribution of
hospital utilization in a region. Medical Care 1985;23:27–38.
McGuirk MA, Porell FW. Spatial patterns of hospital utilization:
the impact of distance and time. Inquiry 1984;21:84–95.
Schneider EC, Epstein AM. Use of public performance reports.
Journal of American Medical Association 1998;279:1638–42.
Hansen TB. Hvilke faktorer har betydning for valg af sygehus
ved tilskadekomst? [Which factors affect choice of hospital in
cases of trauma?]. Ugeskr Laeger 1994;156:652–5 [in Danish].
Birk et al. BMC Health Services Research 2011, 11:262
http://www.biomedcentral.com/1472-6963/11/262
RESEARCH ARTICLE
Open Access
Patients’ experience of choosing an outpatient
clinic in one county in Denmark: results of a
patient survey
Hans O Birk1,2*†, Rikke Gut3† and Lars O Henriksen1†
Abstract
Background: Research on patients’ choice of hospital has focused on inpatients’ rather than outpatients’ choice of
provider. We have investigated Danish outpatients’ awareness and utilisation of freedom of choice of provider;
which factors influence outpatients’ choice of hospital, and how socio-demographic variables influence these
factors in a single uptake area, where patients were free to choose any public hospital, where care was provided
free at the point of delivery, and where distance to the closest hospitals were short by international standards.
Methods: Retrospective questionnaire study of 4,232 outpatients referred to examination, treatment, or follow-up
at one of nine somatic outpatient clinics in Roskilde County in two months of 2002, who had not been
hospitalised within the latest 12 months. The patients were asked, whether they were aware of and utilised
freedom of choice of hospital.
Results: Fifty-four percent (2,272 patients) filled in and returned the questionnaire. Forty-one percent of
respondents were aware of their right to choose, and 53% of those patients utilised their right to choose.
Awareness of freedom of choice of provider was reported to be especially high in female outpatients, patients
with longer education, salaried employees in the public sector, and in patients referred to surgical specialties.
Female outpatients and students were especially likely to report that they utilised their right to choose the
provider. Short distance was the most important reason for outpatients’ choice, followed by the GP’s
recommendations, short waiting time, and the patient’s previous experience with the hospital.
Conclusions: Outpatients’ awareness and utilisation of free choice of health care provider was low. Awareness of
freedom of choice of provider differed significantly by specialty and patient’s gender, education and employment.
Female patients and students were especially likely to choose the clinic by themselves. Most outpatients chose the
clinic closest to their home, the GP’s recommendation and short waiting time being the second and third most
important factors behind choice.
Background
A common characteristic of public sector governance
reforms in the Nordic countries in the latest two decades is a gradual development from collective systems
towards an individual-based democracy model [1],
where individual citizens are viewed as autonomous
consumers rather than clients [2] and are expected to
set priorities and allocate resources by utilising consumers’ rights [3] to choose. Applied to health care,
* Correspondence: [email protected]
† Contributed equally
1
Region Zealand, Quality and Development, Alléen 15, 4180 Sorø, Denmark
Full list of author information is available at the end of the article
consumers - patients - may choose or be involved in the
choice of:
○
○
○
○
Treatment
Individual health professional
Appointment time/date
Provider [4]
In the Nordic countries the emphasis so far is on
patients choosing a provider more or less freely among
several competing providers. The interest in introducing
choice mostly builds on two fundamental arguments
© 2011 Birk et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
Birk et al. BMC Health Services Research 2011, 11:262
http://www.biomedcentral.com/1472-6963/11/262
[5,6], mirroring views of choice as an end in itself or a
means to an end [7]:
The ideological viewpoint
Providing citizens with an opportunity to choose a supplier is an objective in itself, as it strengthens personal
freedom [8,9].
The instrumental viewpoint
The public sector can improve its effectiveness, reduce
inequities in access to care and increase and increase its
responsiveness and quality of services by introducing or
strengthening choice, e.g. an opportunity for patients to
choose a health care provider [4]. In combination with
activity-based payments, where “the money follows the
patient” [10-13]choice constitutes a self-correcting allocation mechanism, which resembles the market mechanism in competitive markets [10,14], which communicate
consumers’ preferences to providers more efficiently
than by central planning [15], as consumers “punish”
irresponsive providers by exit [16], leaving providers
with the choice between improving the quality of their
services or go out of business. Thereby individual actors’
utility maximisation through rational choices, ideally,
leads to an optimal resource allocation in society at the
national and even international level [17,18].
The health care sector is one of the public sector areas
where the introduction of governance-tools has been
very important in countries following the Beveridgemodel, like England [19], Denmark and Sweden [20],
and to a lesser degree in some countries with health
care organized in accordance with the Bismarck-model,
like France, Germany and the Netherlands [21,22].
It is important for the effect of introduction of choice
of provider on the public sector, whether consumers - e.
g. patients - are aware of and utilise choice, and how
they choose among different providers. A large number
of studies describe inverse or negative relationships
between distance to health care and its utilisation in different countries and in different institutional set-ups
[23-25]. This persistent negative relationship between
distance and utilisation may reflect reduced mobility in
patients, judicial barriers (laws or administrative guidelines on patient referral), lack of performance data facilitating choice, and/or patients’ preferences for choice of
hospital.
Research into patients’ choice of hospital has focused
on inpatients’ choices, but a large and growing share of
patients are outpatients: from 2002 to 2009 the number
of outpatient visits at somatic Danish hospital departments rose by 34% from 4,917,000 to 6,612,000 while
the number of discharges from Danish somatic hospital
departments only rose by 12% from 1,126,000 to
1,257,000 [26]. If the factors determining outpatients’
Page 2 of 10
choice of hospital differ significantly from those behind
inpatients’ choice, the assumptions underlying management and planning in the health care sector may not be
valid.
On this background we investigated how outpatients
chose an outpatient clinic, specifically whether awareness and utilisation differed by socio-demographic variables. Building on previous studies of Danish inpatients’
choice of hospital we tested the following hypotheses,
most of which were based on previous studies of Danish
or Norwegian inpatients’ choice of hospital:
○ Outpatients’ awareness and utilisation of freedom
of choice varies by specialty [27]
○ The distance between patients’ home and hospitals
is the most important factor behind patients’ choice
of hospital - most patients prefer to be treated at the
hospital which is the closest to their home
[23,25-27]
○ The distance to hospital is of greater importance
to older than to younger outpatients, older outpatients being especially likely to choose the hospital
closest to their home [28]
○ The GPs’ advice strongly influences outpatients’
choice of hospital [27]
○ Patients’ previous experiences with a hospital
strongly influences their choice of hospital; significant others’ experience influence patients’ choice but
is of lesser importance [27]
○ Female patients are more likely than males to be
aware of and utilise choice [29]
The present study was performed in Denmark where
hospital care was provided free at the point of delivery
by a universal, tax-financed, public health care system
[30]. In the study period the citizen’s home county was
responsible for provision of health care performed by
GPs, specialists, the county’s hospitals, or other counties’
hospitals (chosen by patients utilising freedom of choice
or used by patients referred to hospitals performing
highly specialised interventions). Each citizen had to
register with a local GP, who was responsible for basic
examinations and treatments. GPs, acting as gatekeepers, could refer a patient to any public hospital or a
specialist for specialised services. In case of an emergency, the patient had direct access to a hospital but
could not choose the hospital by themselves. GPs were
self-employed and responsible for their own facilities
and never performed their tasks in a hospital.
The GPs were paid by the counties in proportion to 1)
the number of patients registered with them (capitation,
approx. 1/3 of GPs’ income), and 2) the number of services they provide to their patients (fee-for-service,
approx. 2/3 of GPs’ income). Specifically the payments
Birk et al. BMC Health Services Research 2011, 11:262
http://www.biomedcentral.com/1472-6963/11/262
to GPs were independent of the number of referrals or
by choice of hospital. The counties and the Ministry of
the Interior and Health published waiting time forecasts
for common elective treatments at the hospitals on the
Internet to ease patients’ choice of hospital, but data on
other aspects of service or clinical quality at clinics was
not published systematically.
Elective patients could choose the clinic during the
visit to the GP or after the visit to the GP but before
going to the clinic. If one or more visits to the clinic
were indicated, the patients could choose another
clinic at any time before the last visit. Information to
patients about freedom of choice of hospital was provided in the media, in leaflets available at general practices, and libraries and other public buildings. If
patients did not make the choice by themselves, the
GP chose the clinic.
If hospital personnel found indication for one or more
check-ups after treatment at the hospital/outpatient
clinic the patient was free to choose a specific outpatient
clinic. Otherwise the hospital personnel chose the outpatient clinic.
Danish public hospitals were owned and managed by
the counties. The private Danish hospital sector owned
less than 1% of Danish hospital beds in the study period.
Danish hospitals provide in-patient as well as outpatient
care. If a patient was referred to a hospital outside the
home county, the home county/region paid a DRGcharge to the county/region which owned the hospital
performing the treatment, thereby creating a financial
incentive to treat patients living in other counties/
regions, but clinics were not obliged to accept elective
patients from other counties.
Methods
The Danish county of Roskilde, which was responsible
for provision of health care in the study area until the
introduction of an administrative reform by January 1
2007, performed a biannual survey of outpatients’
experience with the county’s outpatient clinics. In 2002
three of the 38 questions in the survey concerned outpatients’ awareness and utilisation of their freedom of
choice of hospital and their reported reasons for choosing the outpatient clinic.
The source group consisted of all outpatients referred
to examination, treatment (including surgery), or followup at one or more of the 11 somatic outpatient clinics
in Roskilde County in two months of 2002.
To eliminate influence of patients’ experience as inpatients, patients who had been hospitalized at any Danish
hospital within 12 months of attending the outpatient
clinic were excluded from the study. Therefore the
study included patients who attended an outpatient
clinic only once or a few times, and patients who
Page 3 of 10
attended the clinic for a regular check-up and whose
latest discharge took place more than one year before
the study period.
The 11 outpatient clinics included the following specialties: Internal medicine (2 clinics), general surgery
(2), orthopaedic surgery (1 clinic), rheumatology (1),
neurology (1), ophthalmology (1), paediatrics (1),
gynaecology and obstetrics (1), and ear, nose and
throat (1). The survey of patients’ experience aimed at
reflecting patients’ experience at all clinics, and therefore 400 patients from each outpatient clinic were randomly allocated to the study group. For clinics visited
by less than 400 patients in the two months all
patients were included in the study group. Patients
were only included in the study group once for each
outpatient clinic they attended.
Waiting time varied by specialty and by intervention
and data on the intervention was not included in the
questionnaire. Therefore we could not include patients’
expected waiting time in the study.
The study group received a standardised questionnaire
developed for use in a biannual nationwide survey of
inpatients’ experience with Danish public hospitals. The
original version of the questionnaire was validated for
readability and understanding by interviews with 80
inpatients and was used for two nationwide studies of
Danish inpatients’ experience. Unit of Patient Evaluation, Copenhagen County, Denmark, (UPECC, now
renamed “the Unit of Patient-Perceived Quality, Capital
Region of Denmark”) revised this questionnaire for use
by outpatients and validated the questionnaire by interviews with 12 patients from five outpatient clinics using
an interview guide developed by UPECC. See additional
file 1: Extract from the questionnaire.
Compared to the standardised questionnaire the language was adjusted to the outpatient-clinic setting, we
referred to the media rather than the clinic’s “reputation”, and we added family and friends’ experiences as
potential reasons for choosing the clinic.
Socio-demographic data on each patient in the study
group included:
○
○
○
○
○
Specialty
Gender
Age
Education
Employment
The questionnaire did not include questions concerning patients’ use of published data on clinics’ quality or
service level.
The study was performed anonymously. All members
of the study group received one reminder by mail.
Data were entered into a database (SAS).
Birk et al. BMC Health Services Research 2011, 11:262
http://www.biomedcentral.com/1472-6963/11/262
We weighted the responses from each specialty in
accordance with the specialty’s share of the number of
outpatients which attended the clinics during the study
period and met the inclusion criteria.
Respondents’ and non-respondents’ specialty, gender,
and age were compared by a univariate chi2-test.
Respondents’ awareness and utilisation of free choice
of hospital was analysed by gender, education, and
employment by a univariate chi-square test.
Respondents aware of and utilising free choice of hospital were compared by gender, age (0-60 vs. 61+ years),
referring doctor, education (none/short, medium and
long), and specialty category (surgical vs. medical specialty but not by single specialties), using a logistic regression analysis, which did not control for other factors.
Level of significance: 5%. “Surgical specialties” included
general surgery, orthopaedic surgery, ophthalmology,
gynaecology and obstetrics, and ear, nose and throat,
while “medical specialties” included internal medicine,
neurology, and paediatrics.
Respondents’ reasons for choice of hospital were analysed by specialty category (surgical vs. medical specialties but not by single specialty), gender, education
(none/short, medium and long), employment (in
employment vs. other), and age (0-60 vs. 61+ years)
using a univariate chi-2-test.
The study was performed in accordance with the Helsinki Declaration. According to section eight in the Danish Act on a Biomedical Ethics Committee System and
the Processing of Biomedical Research Projects questionnaire studies like the present study are not notifiable
to the Danish research ethics committee system, if the
study does not include biological material [31].
Results
Respondents and representativeness
The study group included 4,232 patients; 2,272 (54%)
filled in and returned the questionnaire. The respondents did not differ significantly from the study group
but due to the recruitment method the unweighted
study group differed from a random sample of outpatients. Female patients (response rate: 56%), patients
attending a clinic of gynaecology/obstetrics (59%) and
patients aged 40-79 years (58%) were especially likely to
respond. Male patients (50%), patients aged 0-29 years
(41%), and patients attending a clinic of neurology (46%)
were the least likely to respond.
On average the respondents had attended the clinic
four times within the latest 12 months. When the
respondents filled in the questionnaire 20% had visited
the clinic only once, and 41% had visited the clinic two
or three times, within the latest 12 months.
Page 4 of 10
Patients’ reported awareness of freedom of choice
(weighted respondents)
Forty-one percent of the respondents reported that they
were aware of their right to choose the hospital (Table
1). Patients’ reported awareness differed significantly by
specialty, patients referred to clinics of ophthalmology,
ear, nose and throat, gynaecology/obstetrics and orthopaedics (the surgical specialties) being especially likely
to report that they were aware of their right to choose.
Female patients, patients with longer education and salaried employees in the public sector were significantly
more likely to be aware of their right to choose than
other patient groups.
In logistic regression analysis involving gender, age
and education female patients and patients with longer
education were significantly more likely to report that
they were aware of their freedom of choice, like in univariate analysis (Table 2).
Utilisation of choice among patients aware of freedom of
choice (weighted respondents)
Fifty-three percent of respondents, who reported that
they were aware of their right to choose, reported that
they utilised this right.
In univariate and logistic regression analysis female
patients were significantly more likely than men to
report that they chose the hospital (Table 1, Table 2).
The share of parents which utilised free choice of hospital on behalf of their children was markedly lower
than in other patient groups, even though the share of
parents who were aware of free choice was lower than
in other patient groups. Reported utilisation of free
choice was also low in patients referred to outpatient
clinics in neurology, while utilisation was high in
patients referred to ophthalmology (where awareness
also was high). Reported utilisation was no higher in
patients referred to surgery than in patients referred to
internal medicine.
Patients who had an education of long duration, and
patients who were self-employed or salaried employees
in the public sector were especially likely to choose the
hospital (Table 1). The statistically significant univariate
association between education and utilisation of choice
disappeared in logistic regression (Table 2), unlike the
association with awareness of choice.
Patients who were 20-39 years old were also especially
likely to choose the hospital, but age was not a statistically significant variable (Table 1, Table 2).
Reasons for choice of hospital
Distance to hospital was the factor which the greatest
number of patients reported to be important for their
Birk et al. BMC Health Services Research 2011, 11:262
http://www.biomedcentral.com/1472-6963/11/262
Page 5 of 10
Table 1 Weighted respondents’ characteristics and reported awareness of and utilisation of freedom of choice of
hospital.
Patients characteristics
Gender
Age (years)
Specialty
Referring doctor
Education
Employment
Response rate (%)
Respondents’ reported awareness
Reported use of choice
Men
58
38*
52**
Women
42
43*
63**
0-9
10-19
5
5
41
31
53
40
20-29
4
33
71
30-39
10
44
65
40-49
13
42
58
50-59
22
45
56
60-69
20
42
56
70-79
14
39
63
80+
Rheumatology
6
9
38
34**
66
63
Internal medicine
27
39**
58
Surgery
17
39**
56
Neurology
5
34**
46
Ophthalmology
6
54**
70
Ear, nose and throat
12
46**
57
Paediatrics
5
33**
36
Gynaecology/obstetrics
7
46**
57
Orthopaedics
Surgical specialties
13
55
45**
44**
62
60
Medical specialties
45
37**
56
GP
51
40
62
Specialist
27
43
55
Ambulatory
7
40
53
Other
14
40
54
Does not remember
1
22
75
Very short
Short
32
40
37*
44*
57
56
Medium or long
29
46*
63
Student
3
35*
78**
Non-skilled labor
4
42*
64**
Skilled labor
6
39*
61**
Salaried employees, priv.
17
37*
59**
Salaried employees, publ.
22
49*
58**
Self-employed
Unemployed
8
13
44*
39*
63**
38**
Pensioners
26
40*
66**
**: p < 0.01 *:p < 0.05
choice (Table 3), followed by the GP’s recommendation
and the waiting time’s length.
Twenty-two percent of outpatients reported that their
own experience influenced their choice of hospital,
while seven percent were influenced by their friends’
experience, and five percent by their family’s experience.
Media reports were only referred to by three percent of
the patients.
Female patients were significantly more likely than
men to choose the clinic closest to their home, while
male patients and patients referred to surgical specialties
were significantly more likely to make their choice
based on waiting time than female patients and patients
referred to medical specialties. Male patients and
patients out of employment (including pensioners) were
significantly more likely to make their choice based on
Birk et al. BMC Health Services Research 2011, 11:262
http://www.biomedcentral.com/1472-6963/11/262
Page 6 of 10
Table 2 Weighted respondents’ reported awareness/utilisation of freedom of choice of hospital, logistic regression,
adjusted odds ratios.
Factor
Awareness
Utilisation
Unit
Odds ratio
Lower 95%-conf.
Upper 95%-conf.
Female vs. male patients
1.00
1.234*
1.001
1.522
Age: 0-60 vs. 61+ years
1.00
0.960
0.768
1.201
Short/no education vs. long
Medium education vs. long
1.00
1.00
0.695**
0.902
0.532
0.704
0.907
1.156
2.049
Female vs. male patients
1.00
1.462*
1.043
Age: 0-60 vs. 61+ years
1.00
1.033
0.720
1.482
Short/no education vs. long
1.00
0.801
0.519
1.237
Medium education vs. long
1.00
0.835
0.566
1.234
*: p < 0.05: **: p < 0.01
their personal experience with clinics than female
patients and patients in employment. Friends’ experience was especially important to patients referred to
surgical specialties, and patients with longer education
and younger patients were especially likely to refer to
reasons which were not listed in the questionnaire.
Discussion
Awareness of freedom of choice
Questionnaire studies of Danish inpatients’ awareness
and utilisation of choice found a higher overall awareness of freedom of choice (more than 80%) than the
present study of outpatients [27,32-34], even though the
study was performed within the same health care system. The results regarding differences between specialties almost resembled results from a study of inpatients
performed in the same county: inpatients referred to
departments of internal medicine, ophthalmology, or
ear, nose and throat were especially likely to be aware of
choice, while awareness was the lowest in patients
referred to geriatric departments and in parents of children referred to a paediatric department [33].
Utilisation of freedom of choice
Patients’ utilisation of choice varied by specialty. Two
reasons for paediatric patients’ low utilisation of free
choice could be a low number of paediatric departments
and quite long distances between these departments,
and parents may have to pick up their child at a kindergarten or school before they can go to the clinic, leading
to an even longer transport distance for the parents,
making them especially sensitive to the distance to the
closest clinic. In the present study we did not disaggregate from specialty level to diagnoses or the severity of
the specific disease. Concern about severe diseases may
make patients more likely to utilise freedom of choice of
hospital, and in general in patients are more likely than
outpatients to suffer from serious diseases [4].
Table 3 Weighted respondents’ reported reasons for choice of hospital by specialty category, gender, education,
employment and age.
Respondents’
characteristics
Short
GP’s
Waiting
distance recommendation time
Patient’s
experience
Other
Friends’
reasons experience
Family’s
experience
Media
reports
Specialty
Surgical
Medical
40
50
23
27
30****
14****
21
22
11
16
10*
4*
7
4
2
4
Gender
Female
47*
26
19**
25*
15
8
6
2
Male
37*
22
33**
16*
10
5
6
3
None/very
short
35
25
23
23
10*
9
7
1
Short
47
24
22
17
12*
7
5
1
Medium/
long
42
24
25
24
21*
8
5
5
In
43
employment
25
23
17*
16
5
5
2
Other
43
23
26
27*
11
9
6
4
0-60
44
25
22
19
17**
6
5
3
61+
43
22
28
27
7**
9
7
2
44
24
24
22
13
7
5
3
Education
Employment
Age (years)
Total
*: p < 0.05; **: p < 0.01; ****: p < 0.001
Respondents could state more than one reason for choice.
Birk et al. BMC Health Services Research 2011, 11:262
http://www.biomedcentral.com/1472-6963/11/262
In univariate and logistic regression analysis female
patients were more likely than men to report that they
chose the hospital. Likewise Norwegian female inpatients are more likely to utilise choice than male inpatients [29]. The present study provides no explanation for
this difference, but women are hospitalized more often
than men due to childbirth, and a longer average lifespan etc. and this greater experience may facilitate utilisation of choice.
Although age was not a statistically significant factor
behind reported utilisation of choice in the present
study other studies indicate that younger patients may
need assistance to think through what is important to
them [35], and despite older patients’ reduced mobility
their greater experience with providers and choice of
provider may make them active choosers, although
accessing and utilising data on the internet may constitute a challenge [35].
US studies have found that patients’ travel distances
grow with their educational level, indicating a positive
association between education and mobility: patients
with higher education on average earn higher incomes
and may be more likely to own a car, and are more
likely to live in urbanized areas with access to public
transport [4]. In a French interview study pregnant
women with a higher educational level were especially
likely to choose a maternity unit with special technical
attributes [36].
We found no studies of an association between
employment and choice.
The Danish population is quite homogeneous with
regard to socio-demographic variables, which may make
it difficult to show statistically significant differences in
behaviour between social groups.
Reasons for choice of hospital
Short distance was the most important factor behind
choice of hospital in similar Danish studies [27,32-34]
and the present study where female patients were significantly more likely than men to make their choice based
on short distance. Many other studies have, by use of
different methodologies, found that the distance to alternative hospitals is very important for patients’ choice of
hospital [37], one US registry study finding that equal
shares of patients chose the hospital closest to their
home in rural and metropolitan areas, suggesting that
there is no lower threshold below which short distance
loses importance [38]. Distance interacts with other
patient- and disease-specific factors like patients’ age
[28], but in the present study gender was the only statistically significant factor behind choice. Institutional differences between health care services in the US and in a
Beveridge-system like Danish health care complicate
international comparisons.
Page 7 of 10
The GP’s recommendation and waiting time were the
second most important factor behind patients’ choice of
clinic.
In 2002 20% of Danish inpatients who chose a specific
hospital reported that the GP influenced their choice
[32]. The present study indicates that the GP plays an
even larger role for outpatients’ choice of provider than
for inpatients.
Danish studies have consistently found that (short)
expected waiting time is the fifth most important factor
behind inpatients’ choice [27,32-34]. The great importance of (short) waiting time to patients referred to
clinics of surgical specialties is not surprising as the
share of elective patients is higher in surgical specialties
than in medical specialties. Studies of patients’ acceptance of waiting time indicate that its numerical length
is very important to patients: cataract patients generally
accept waiting times of three months and less, while
waiting times of six months or more are likely to be
perceived as too long [39,40]. In British studies of
patients on a waiting list, with very long and uncertain
waiting times, who were offered early treatment at a distant hospital, all or a major share of the patients were
willing to travel far to reduce their waiting time [41-44].
It was not possible to distinguish between the effect of
waiting time itself and uncertainty about its length. In a
hypothetical study patients reported that for every additional hour of travel they would, on average, require a
reduction in waiting time of 2.3 months to take up the
offer of treatment at a distant hospital. A choice of a
hospital abroad required a reduction in waiting time of
around 5.9 months [44]. Cataract patients’ perception of
waiting times for surgery being too long is not associated with their demographic characteristics [40]. A
study of patients waiting for hip or knee implantation
found a moderate inverse association between on the
one hand severity categories like bodily pain and physical functioning, and acceptable waiting time on the
other [45].
The present study does not provide information on
whether the patients who point to their own previous
experience with the hospital have been treated at the
same department or at another department. A study
from the UK found that a previous negative experience
with a local hospital was the strongest predictor for willingness to choose a non-local hospital [4].
Family and friends’ experience played a minor role for
choice - probably partly because the hospital was chosen
during a visit to the GP where it is difficult to approach
other people and ask them about their experience,
before the visit is over, although patients were free to
call the GP’s secretary and ask for a re-referral if they
changed their mind after the visit to the GP. Studies of
clinically healthy citizens’ hypothetical choice of hospital
Birk et al. BMC Health Services Research 2011, 11:262
http://www.biomedcentral.com/1472-6963/11/262
indicate that patients’ experience and other people’s
reports of their experience with certain hospitals is
more important or just as important for the choice of
hospital as the GP’s advice and published information
about hospitals’ quality [46-48].
Positive media reports about the clinic played a minor
role for the patients’ choice. The patients may find it
difficult to remember media reports about each and
every hospital they are able to choose among. According
to a US study reports of single sensational events at a
hospital is more important for patients’ choice of a hospital than data on general mortality [49].
Implications
A major reason for the introduction of proxy markets,
like free choice of hospital in combination with activitybased financing is the assumption that this system will
push hospital managers to improve quality and service.
This management concept depends on that patients are
aware of and willing to choose the hospital based on
quality and service. The present study indicates, that
outpatients’ choice behaviour will send relatively weak
signals to hospitals compared to inpatients, because relatively few outpatients are aware of and utilise free choice
of hospital, and because a large share of the outpatients
choose the hospital based on which hospital is the closest, which is independent of hospitals’ quality or
service.
Waiting time influences patients’ choice, but the relatively small share of patients who choose the hospital
based on waiting time data will not be sufficient to level
out waiting times at different hospitals.
The study was performed at a time when only little
information was published about hospitals’ quality and
service levels. In such situations proxy-measures of service and quality like the GP’s, the patient’s, and family
and friends’ experience constitute important factors
behind choice, which means that outdated data on quality and service may play a major role for patients’
choice.
Assumptions on outpatients’ preferences for choice of
provider should build on studies of outpatients rather
than generalizations from studies of inpatients’ preferences and choices.
Further research into outpatients’ choice behaviour
and utilisation of data sources is warranted and should
distinguish between referrals of outpatients to a clinic
and outpatients attending a clinic for a check-up for an
ongoing medical condition or after hospitalisation.
Limitations of the study
A medium response rate increased the risk of selection
bias, because some patient groups may be especially
Page 8 of 10
likely to answer, e.g. patients with strong views on
choice of hospital probably were more likely to participate in the present study than other patients, and female
patients, who on average were more likely to be aware
of and utilise choice than male respondents, were also
more likely to respond. Therefore the present study may
exaggerate 1) the importance of any single factor for
choice, 2) patients’ likelihood to choose the hospital by
themselves, and 3) factors which are especially important to female patients.
The present study includes many statistical tests and
some of the statistically significant findings in univariate
analysis may be due to mass significance rather than
causality.
The study group received the questionnaire approximately three months after attending the outpatient
clinic. Recall bias may constitute a problem. This is
especially so because chronic patients attending a regular check-up may have chosen the clinic several years
before the present study was performed. Furthermore
respondents may have provided a simplified description
of the decision process, thereby exaggerating the influence of the most important reasons for their choice.
The respondents’ participation in the study may have
led them to describe a decision making process which is
more rational than their real choice behaviour - for
example by exaggerating their awareness and utilisation
of choice, and the influence of supposedly rational reasons for choice like short waiting time and the GP’s
advice, while underreporting reasons which may be considered to be less rational, like short distance to hospital
and informal information sources, like family and friends.
However, most of the questions in the questionnaire
concerned their experience (’patient satisfaction’); only
three of 38 questions concerned the patients’ choice of
outpatient clinic, and therefore we assume that the subjects’ participation in the study may only have had little
impact on their responses.
The study did not provide evidence on outpatients’
utilisation of published data on quality or service in
choice of provider.
The study was performed nine years after the introduction of freedom of choice of (public) hospital in
Denmark and before freedom of choice was extended to
include some private clinics. Danish patients have
become increasingly used to utilising freedom of choice.
Therefore the results from the present study may be
more representative of patients’ behaviour in public
health care systems (Beveridge- or NHS-health care systems) several years after the introduction of choice than
immediately after its introduction. The results and conclusions should only be generalized to other institutional
settings with caution.
Birk et al. BMC Health Services Research 2011, 11:262
http://www.biomedcentral.com/1472-6963/11/262
Conclusions
Nine years after the introduction of free choice of public
hospital/outpatient clinic in Denmark, outpatients’
awareness and utilisation of free choice of health care
provider was low. Awareness of freedom of choice of
provider differed significantly by specialty and patient’s
gender, education and employment. Female patients and
students were especially likely to choose the clinic by
themselves. Most outpatients chose the clinic closest to
their home, the GP’s recommendation and short waiting
time being the second and third most important factors
behind choice.
Additional material
Additional file 1: Extract from the questionnaire. The seven questions
concerning outpatients’ choice of hospital in the questionnaire used for
investigation of patients’ experience with outpatient clinics.
Page 9 of 10
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Acknowledgements and funding
The study was funded by the former Roskilde County, Region Zealand and
grants from “The Health Research Forum in East Denmark, Research
Programme for Promotion of Regional Cooperation on Medical Research”
and “The Health Research Foundation of the Counties in Eastern Denmark
(Region 3)”.
The authors are grateful to the two referees, Anna Dixon, director of policy,
The King’s Fund, London, and Roland Friele, professor, Tylburg University,
the Netherlands, for their valuable comments and suggestions. However, the
authors are solely responsible for remaining omissions and errors.
Author details
Region Zealand, Quality and Development, Alléen 15, 4180 Sorø, Denmark.
University of Copenhagen, Department of Public Health, Øster
Farimagsgade 5, P.O.Box 2099, 1099 København K, Denmark. 3Unit of
Patient-Perceived Quality, The Capital Region of Denmark, Frederiksberg
Hospital, Nordre Fasanvej 57, Hovedvejen indgang 13, 1. sal, 2000
Frederiksberg, Denmark.
1
2
Authors’ contributions
All authors conceived and designed the study; RG developed the
questionnaire, collected the data, and assisted in performing the statistical
analyses and in writing the manuscript. HOB analysed the data and wrote
the manuscript. LOH assisted in writing the manuscript. All authors read and
approved the final manuscript.
Competing interests
The authors declare that they have no competing interests.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Received: 29 December 2010 Accepted: 10 October 2011
Published: 10 October 2011
26.
References
1. Winblad U, Ringard Å: Meeting rising public expectations: the changing
roles of patients and citizens. In Nordic health care systems. Recent reforms
and current policy challenges. Edited by: Magnussen J, Vrangbæk K, Saltman
RB. Maidenhead: Open University Press; 2009:.
2. Le Grand J: Motivation, agency, and public policy. Of Knights & Knaves, Pawns
& Queens Oxford: Oxford University Press; 2003.
3. Vrangbæk K, Østergren K: Patient empowerment and the introduction of
hospital choice in Denmark and Norway. Health Economics, Policy and
Law 2006, 1:371-94.
27.
28.
29.
Dixon A, Robertson R, Appleby J, Burge P, Devlin N, Magee H: Patient
choice. How patients choose and how providers respond London: The King’s
Fund; 2010.
Perri G: Giving consumers of British public services more choice: what
can be learned from recent history? Jnl Soc Pol 2003, 32:239-70.
Thomson S, Dixon A: Choices in health care: the European experience. J
Health Serv Res Policy 2006, 11:167-71.
Kreisz FP, Gericke C: User choice in European health systems: towards a
systematic framework for analysis. Health Econ Policy Law 2010, 5:13-30.
Dowding K, John P: The value of choice in public policy. Public
Administration 2009, 87:219-33.
Wilmot S: A fair range of choice: justifying maximum patient choice in
the British National Health Service. Health Care Anal 2007, 15:59-72.
Porter ME, Teisberg EO: Redefining Health Care. Creating value-based
competition on results Cambridge, Harvard Business School Press; 2006.
Appleby J, Harrison A, Devlin N: What is the real cost of more patient choice
London: King’s Fund; 2003.
Vrangbæk K: The interplay between central and sub-central levels: the
development of a systematic standard based programme for governing
medical performance in Denmark. Health Econ Policy Law 2009, 4:305-27.
Tai-Seale M: Voting with their feet: patient exit and intergroup
differences in propensity for switching usual sources of care. J Health
Polit Policy Law 2004, 29:491-514.
Le Grand J: Choice and competition in publicly funded health care.
Health Economics, Policy and Law 2009, 4:479-88.
Hayek FA: The use of knowledge in society. American Economic Review
1945, 35:519-30.
Hirschman AO: Exit, voice and loyalty. Response to decline in firms,
organizations and states Cambridge: Harvard University Press; 1970.
Bernstein AB, Gauthier AK: Choices in health care: what are they and
what are they worth? Med Care Res Rev 1999, 56:5-23.
Martinsen DS, Vrangbæk K: The Europeanization of health care
governance: implementing the market imperatives of Europe. Public
Administration 2008, 86:169-84.
Lewis R: More patient choice in England’s National Health service.
International Journal of Health Services 2005, 35:479-83.
Vrangbæk K, Østergren K, Birk HO, Winblad U: Patient reactions to hospital
choice in Norway, Denmark and Sweden. Health Economics, Policy and
Law 2007, 2:125-52.
Berendsen AJ, de Jong GM, Schuling J, Bosveld HEP, de Waal MWM,
Mitchell GK, van der Meer K, Meyboom-de-Jong B: Patients’ need for
choice and information across the interface between primary and
secondary care: A survey. Patient Education and Counseling 2010, 79:100-5.
Lisac M, Reimers L, Henke KD, Schlette S: Access and choice - competition
under the roof of solidarity in German health care: an analysis of health
policy reforms since 2004. Health Economics, Policy and Law 2010, 5:31-52.
Shannon GW, Bashshur RL, Metzner C: The concept of distance as a factor
in accessibility and utilization of health care. Medical Care Review 1969,
26:143-61.
Porell FW, Adams EK: Hospital choice models: a review and assessment
of their utility for policy impact analysis. Med Care Res Rev 1995,
52:158-95.
Place M: The relationship between concentration, patient accessibility and
utilisation of services York: The University of York. Centre for Health
Economics. York Health Economics Consortium. NHS Centre for Reviews &
Dissemination; 1997.
Indenrigs- og Sundhedsministeriet: Sundhedsvæsenet i nationalt perspektiv
Copenhagen; 2010, In Danish: the Danish Ministry of the Interior and
Health. The health care system in a national perspective.
Unit of Patient Evaluation: Patienters vurdering af landets sygehuse 2000.
Spørgeskemaundersøgelse blandt 34.000 patienter [in Danish: Patients’
experiences with hospitals in Denmark 2000. Questionnaire study including
34,000 patients] 2001, Glostrup: Enheden for brugerundersøgelser i
Københavns Amts sundhedsvæsen [Unit of Patient Evaluation, Copenhagen
County.
Goddard M, Hobden C: Patient Choice: a review. Report to the Department of
Health York: Centre for Health Economics; 2003.
Godager G, Iversen T: Hvem bruker retten til fritt sygehusvalg? Resultater
fra Samordnet levekårsundersøkelse 2002. Økonomisk Forum 2004, 58(4/
Birk et al. BMC Health Services Research 2011, 11:262
http://www.biomedcentral.com/1472-6963/11/262
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
5):49-56, In Norwegian Who utilises the right of freedom of choice of
hospital? Results from the integrated study of living conditions 2002.
Strandberg-Larsen M, Nielsen MB, Vallgårda S, Krasnik A, Vrangbæk K,
Mossialos E: Denmark: Health system review. Health Systems in Transition
2007, 9(6):1-164.
Lov om et videnskabsetisk komitésystem og behandling af
biomedicinske forskningsprojekter [in Danish: Act on a Biomedical Ethics
Committee System and the Processing of Biomedical Research Projects. ,
http://www.cvk.sum.dk/English/actonabiomedicalresearch.aspx.
Unit of Patient Evaluation: Patienters vurdering af landets sygehuse 2002.
Spørgeskemaundersøgelse blandt 32.000 patienter [in Danish: Patients’
experiences with hospitals in Denmark 2002. Questionnaire study including
32,000 patients] 2003, Glostrup: Enheden for brugerundersøgelser i
Københavns Amts sundhedsvæsen [Unit of Patient Evaluation, Copenhagen
County].
Unit of Patient Evaluation: Patienters vurdering af sygehusafdelinger i Roskilde
Amt. Spørgeskemaundersøgelse blandt 3.400 patienter [in Danish: Patients’
experiences with hospital departments in Roskilde County 2002. Questionnaire
study including 3,400 patients] 2003, Glostrup: Enheden for
brugerundersøgelser i Københavns Amts sundhedsvæsen [Unit of Patient
Evaluation, Copenhagen County.
Unit of Patient Evaluation: Patienters oplevelser på landets sygehuse 2004.
Spørgeskemaundersøgelse blandt 26.300 indlagte patienter [in Danish: Patients’
experiences with hospitals 2004. Questionnaire study including 26,300
inpatients] , Glostrup: Enheden for Brugerundersøgelser i Københavns Amt.
[Unit of Patient Evaluation, Copenhagen County], 2005. Abstract in English
available at: http://www.patientoplevelser.dk/index.asp?id=210 [homepage
on the Internet] [cited on 3 July 2011].
Boyce T, Dixon A, Fasolo B, Reutskaja E: Choosing a high-quality hospital. The
role of nudges, scorecard design and information London: The King’s Fund;
2010.
Combier E, Zeitlin J, de Courcel N, Vasseur S, Lalouf A, Amat-Roze JM, de
Pouvourville G: Choosing where to deliver: decision criteria among
women with low-risk pregnancies in France. Soc Sci Med 2004,
58:2279-89.
Tai WTC, Porell FW, Adams EK: Hospital choice of rural Medicare
beneficiaries: patient, hospital attributes, and the patient-physician
relationship. Health Serv Res 2004, 39:1903-22.
Adams EK, Wright GE: Hospital Choice of Medicare Beneficiaries in a
Rural Market: Why not the Closest? J Rural Health 1991, 7:134-52.
Conner-Spady B, Sanmartin C, Johnston G, McGurran J, Kehler M,
Noseworthy T: Willingness of patients to change surgeons for a shorter
waiting time for joint arthroplasty. CMAJ 2008, 179:327-32.
Dunn E, Black C, Alonso J, Nørregaard JC, Anderson GF: Patients’
acceptance of waiting time for cataract surgery: what makes a wait too
long? Soc Sci Med 1997, 44:1603-10.
Howell GP, Richardson D, Forester A, Sibson J, Ryan JM, Morgans BT: Long
distance travel for routine elective surgery: questionnaire survey of
patients’ attitudes. BMJ 1990, 300:1171-3.
Nofal F, Moran MM: Long distance travel by children for tonsillectomy:
experience of the ORL department at Princess Alexandra Hospital (PAH),
Royal Air Force, Wroughton, Swindon. J Laryngol Otol 1990, 104:417-8.
Dawson D, Jacobs R, Martin S, Smith P: Is patient choice an effective
mechanism to reduce waiting times? Appl Health Econ Health Policy 2004,
3:195-203.
Burge P, Devlin N, Appleby J, Rohr C, Grant J: Do patients always prefer
quicker treatment?: a discrete choice analysis of patients’ stated
preferences in the London Choice Project. Appl Health Econ Health Policy
2004, 3:183-94.
Derrett S, Paul C, Moris JM: Waiting for elective surgery: effects on healthrelated quality of life. Int J Qual Health Care 1999, 11:47-57.
Wolinsky FD, Kurz RS: How the public chooses and views hospitals. Hosp
Health Serv Adm 1984, 29:58-67.
Marshall MN, Shekelle PG, Leatherman S, Brook RH: The public release of
performance data: what do we expect to gain? A review of the
evidence. JAMA 2000, 283:1866-74.
Robinson S, Brodie M: Understanding the quality challenge for health
consumers. The Kaiser/AHCPR Survey. Jt Comm J Qual Improv 1997,
23:239-44.
Page 10 of 10
49. Mennemeyer ST, Morrisey MA, Howard LZ: Death and reputation: how
consumers acted upon HCFA mortality information. Inquiry 1997,
34:117-28.
Pre-publication history
The pre-publication history for this paper can be accessed here:
http://www.biomedcentral.com/1472-6963/11/262/prepub
doi:10.1186/1472-6963-11-262
Cite this article as: Birk et al.: Patients’ experience of choosing an
outpatient clinic in one county in Denmark: results of a patient survey.
BMC Health Services Research 2011 11:262.
Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit
Birk and Henriksen BMC Health Services Research 2012, 12:126
http://www.biomedcentral.com/1472-6963/12/126
RESEARCH ARTICLE
Open Access
Which factors decided general practitioners’
choice of hospital on behalf of their patients in
an area with free choice of public hospital?
A questionnaire study
Hans O Birk1,2* and Lars O Henriksen1
Abstract
Background: Parts of New Public Management-reforms of the public sector depend on introduction of market-like
mechanisms to manage the sector, like free choice of hospital. However, patients may delegate the choice of
hospital to agents like general practitioners (GPs). We have investigated which factors Danish GPs reported as
decisive for their choice of hospital on behalf of patients, and their utilisation of formal and informal data sources
when they chose a hospital on behalf of patients.
Methods: Retrospective questionnaire study of all of the 474 GPs practising in three counties which constituted a
single uptake area. Patients were free to choose a hospital in another county in the country. The GPs were asked
about responsibility for choice of the latest three patients referred by the GP to hospital; which of 16 factors
influenced the choice of hospital; which of 15 sources of information about clinical quality at various hospitals/
departments were considered relevant, and how often were six sources of information about waiting time utilised.
Results: Fifty-one percent (240 GPs) filled in and returned the questionnaire. One hundred and eighty-three GPs
(76%) reported that they perceived that they chose the hospital on behalf of the latest referred patient. Short
distance to hospital was the most common reason for choice of hospital.
The most frequently used source of information about quality at hospital departments was anecdotal reports from
patients referred previously, and the most important source of information about waiting time was the hospitals’
letters of confirmation of referrals.
Conclusions: In an area with free choice of public hospital most GPs perceived that they chose the hospital on
behalf of patients. Short distance to hospital was the factor which most often decided the GPs’ choice of hospital
on behalf of patients. GPs attached little weight to official information on quality and service (waiting time) at
hospitals or departments, focusing instead on informal sources like feedback from patients and colleagues and their
experience with cooperation with the department or hospital.
Background
A common trait in public sector governance reforms in
the Nordic countries in the latest two decades is a gradual development from collective systems towards an
individual-based democracy model [1], where individual
citizens are viewed as autonomous consumers rather
than clients [2] and are expected to set priorities and
* Correspondence: [email protected]
1
Region Zealand, Quality and Development, Alléen 15, 4180 Sorø, Denmark
2
University of Copenhagen, Department of Public Health, !ster
Farimagsgade 5, P.O. Box 2099, 1099 København K, Denmark
allocate resources by utilising consumers’ rights [3] to
choose treatment, appointment times and/or providers
[4]. In general the interest in introducing choice is based
on two fundamental arguments [5,6]: an ideological
viewpoint, which views an opportunity for citizens to
choose a supplier as an objective in itself, as it strengthens personal freedom [7,8], and an instrumental viewpoint, which emphasizes that the public sector can
improve its effectiveness, quality, equity in access to care
and responsiveness by introducing or strengthening
choice, e.g. an opportunity for patients to choose a
© 2012 Birk and Henriksen; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the
Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.
Birk and Henriksen BMC Health Services Research 2012, 12:126
http://www.biomedcentral.com/1472-6963/12/126
health care provider [2,4]. In combination with activitybased payments, where “the money follows the patient”
[9-13] choice is assumed to constitute a self-correcting
allocation mechanism, which resembles the market
mechanism in competitive markets [9,13], as providers
who provide less than optimal care may be “punished”
by customers through exit [14]. Thereby, ideally, individual actors’ utility maximisation on the demand side as
well as the supply side results in an optimal resource allocation and production in society [4,15,16].
One “model of patient choice as a governance tool”
builds on several preconditions [4], including ten preconditions concerning patients’ and general practitioners’ (GPs’) knowledge, assumptions and behaviour:
! Patients are aware of their ability to choose
! Patients want to choose and think choice is
important
! Patients are offered choice of providers
! Quality is the primary discriminator in patients’
choice of which provider to attend
! Patients have access to relevant and appropriate
!
!
!
!
!
information on quality and are able to interpret the
data
GPs believe that choice is important to patients
GPs offer choice to all patients needing a referral
GPs involve patients in decision-making
GPs have access to information about the quality of
providers and convey this information to patients
GPs have time and resources to support patients to
make an informed choice [4]
Page 2 of 10
[24-26], and the GP is the most important source of information for a major share of the patients who choose
the hospital by themselves [24,26-28], patients being
even more sensitive to GPs’ warnings against specific
hospitals than to their recommendations of specific hospitals [28]. In 2004 87% of all Danish elective in-patients
were aware of freedom of choice of hospital, and 42% of
all elective patients chose the hospital by themselves (of
which 30% attributed major influence to the GP on their
choice). The GPs chose the hospital on behalf of 58% of
patients, who were not aware of choice or delegated the
choice to the GP [26].
An English study found varying support for choice
among GPs, and choice has not changed the GPs’ behaviour towards more emphasis on advice on choice [29].
GPs may be reluctant to provide advice to patients and
promote utilisation of choice, because this task competes
with other tasks; they do not consider this task a part of
their job [4,30]; consider this task too time-consuming
[4]; distrust data published by the providers [29], or want
to avoid being blamed by patients for presenting faulty
data [23]. Therefore GPs may choose the hospital on behalf of patients rather than provide advice to the
patients, thereby reducing patients’ influence on the
governance-effect of patient choice.
We investigated:
! Whether GPs considered the patients or themselves
to be responsible for choice of hospital?
! Which factors decided GPs’ choice on behalf of
patients?
! Which formal and informal data sources were
However, while choice in theory could be a driver for
improving quality and service in health care these preconditions from neoclassical microeconomic theory are
only fulfilled to some degree [17]. For example patients
appear to utilize information only if there is a single outcome of major importance and the data is easy to understand [18], many patients being insufficiently informed
to utilize data for choice resulting in market failure
[19,20] and thereby reducing the potentially positive impact of choice on quality and service [17]. Patients may
be reluctant to take responsibility for choosing the hospital in order to avoid regretting their choice [21-23],
preferring to enter into a principal-agent-relationship
with an intermediary. In an ideal principal-agentrelationship the agent (i.e. the GP) makes the decision,
which the principal (i.e. the patient) would have made, if
the principal had had the same information as the agent
about the expected effect of various interventions and
the quality of individual providers’ services. English,
Dutch and Danish surveys have shown that GPs choose
the hospital on behalf of a major share of patients, even
when patients’ awareness of their right to choose is high
utilised by GPs in choice on behalf of patients?
The study was performed in a setting, where patients
and GPs had one decade’s experience with free choice of
hospital; where patients’ awareness of choice was high,
and where the share of patients for whom GPs chose the
hospital had been stable at a high level for several years.
Hospital care was provided free at the point of delivery
by a universal, tax-financed, public health care system.
In the study period the citizen’s home county was responsible for provision of health care performed by GPs,
specialists, the county’s hospitals, or other counties’ hospitals (by patients utilising choice or by patients referred
to hospitals performing highly specialised interventions).
Each citizen had to register with a local GP, who was responsible for basic examinations and treatments. GPs,
acting as gatekeepers, decided whether a patient should
be referred to hospital for elective care, and could refer
a patient to any public Danish hospital or specialist for
specialised services. In case of emergency, patients had
direct access to hospital but could not choose the
hospital by themselves. GPs were self-employed and
Birk and Henriksen BMC Health Services Research 2012, 12:126
http://www.biomedcentral.com/1472-6963/12/126
responsible for their own facilities and never performed
their tasks in a hospital, unlike in the US [31].
The GPs were paid by the counties in proportion to 1)
the number of patients registered with them (capitation,
approx. 1/3 of GPs’ income), and 2) the number of services they provided to their patients (fee-for-service,
approx. 2/3 of GPs’ income). The payments to GPs were
independent of the number of referrals and the choice of
hospital [31].
Elective patients could choose the hospital during the
visit to the GP or after the visit to the GP but before
going to the hospital. If more than one hospitalisation
was indicated, the patients could choose another hospital at any time before the last hospitalisation. If the patient did not make the choice by themselves, the GP
chose the hospital, by filling in a referral form on paper
(no computerised facility like the English “Choose and
Book” was available). Filling in the form took equally
long time whether the patient was referred to a local
hospital or a hospital in another Danish county. GPs did
not receive any kind of incentive payment for advising
patients on choice of hospital [31].
Patients referred to hospital were responsible for
transportation arrangements and costs. However a patient was entitled to transportation or a refund of his/
her transportation costs by the county, if the patient was
a pensioner, lived more than 50 km by road from the
nearest hospital which could perform the procedure, or
could not utilise public transport for health reasons. If a
patient was entitled to a refund of transportation
costs due to the distance criterion and chose a more
distant hospital, the refund was calculated based on
the distance to the closest hospital capable of performing
the procedure.
Danish public hospitals were owned and managed by a
regional political/administrative level: the counties. The
private Danish hospital sector owned less than 1% of
Danish hospital beds in the study period. Danish hospitals provided in-patient as well as out-patient care. If a
patient was referred to a hospital outside the home
county, the home county/region paid a DRG-charge to
the county which owned the hospital performing the
treatment, thereby creating a financial incentive for the
counties to attract patients from the county as well as
patients from other counties to the county’s hospitals,
but hospitals were not obliged to accept elective patients
from other counties. Hospitals’ and hospital departments’ income grew with the production of DRG points
up to a certain level. To avoid discrimination against
patient groups, hospitals and departments received the
same payment for treatment of patients independent of
where the patients lived [31].
Elective patients could choose any public hospital, a
broad majority in the Danish parliament having introduced
Page 3 of 10
“free choice of [public] hospital” in Denmark in 1993 with
several parallel objectives in mind including a view on
choice as a patients’ right, to level out waiting times,
strengthen patients’ influence on the hospital sector, improve hospitals’ treatment results and improve patients’ satisfaction [32]. However, significant limitations on patients’
rights and on hospitals’ financial incentives to accept
patients from other uptake areas were introduced due to
fear that a more demand-driven health care sector would
lead to budget overruns. Gradually the government
extended patients’ freedom to choose and strengthened
hospitals’ financial incentives to accept patients, thereby
creating a common market at the national level for elective,
public health care. In 1991, before the introduction of free
choice at the national level, three counties in Eastern
Denmark independently introduced free choice within their
own area (the study area).
The counties and the Ministry of the Interior and
Health published waiting time forecasts for common
elective treatments at hospital level and results of biannual
surveys of patients’ experience with individual hospitals
(but not with individual departments). The ministry published data on individual departments’ volume for common surgical interventions as a proxy for quality on the
assumption that department volume was associated with
experience and thereby clinical quality. Data on other
aspects of service or clinical quality at clinics was not
published systematically, but some departments and
medical societies published data on individual departments’ performance as part of quality development or
clinical research.
Methods
The present study was performed in the three mixed
urban/rural counties of Roskilde, Storstrøm and
Vestsjælland (801,452 inhabitants on January 1 2004 in
total) in Eastern Denmark. The counties in the study
area provided hospital treatment at 13 public hospitals
evenly distributed within the region. No point in the
study area was more than 30 km from the nearest public
hospital in a bee line. Each specialty represented in the
study area was available at two hospitals or more, except
for dermatology and plastic surgery, which were only
available at one hospital each. Patients were free to
choose treatment, paid by the home county, at hospitals
in other Danish counties. Patients in need of care at tertiary hospitals were referred to hospitals outside the
study area from the counties’ own hospitals. Each public
hospital was obliged to accept any referral from any GP
in any of the three counties. Patients who were entitled
by law to free travel to the hospital closest to their home
were offered free travel to any public hospital in the
study area, thereby strengthening patients’ opportunities
to utilise their freedom of choice.
Birk and Henriksen BMC Health Services Research 2012, 12:126
http://www.biomedcentral.com/1472-6963/12/126
The three counties jointly published waiting timeprognoses for common surgical procedures at hospitals
within the study area. At first these prognoses were
mailed to each GP on paper, later they were published
on a website managed by the Danish counties and accessible to the public (www.sundhed.dk) and a national
website maintained by the Ministry for the Interior and
Health (www.venteinfo.dk). Departments held regular information meetings for the GPs about the interventions
provided at the department and the department’s
procedures.
The study group included all of the 483 GPs registered
as practising in the study area. The names and addresses
of the GPs were found by use of the Danish counties’
website, www.sundhed.dk.
The study was performed as a questionnaire study.
The retrospective design was chosen to avoid influencing
the GPs’ choice behaviour.
The questionnaire was developed after a review of the
literature [33-35] and face-to-face discussions with GPs
from two of the three counties. The questionnaire was
validated by interviews with three GPs: two GPs practising in a large and a small town in the study area, respectively, and one affiliated with the University of
Copenhagen’s Section of General Practice. The three
GPs were asked whether the questions were unambiguous, and whether the predefined answers were sufficient.
The GPs’ interpretation of the questions was compared
with the authors’ intentions. Based on the GPs’
responses several open and closed responses were added
to the questionnaire, for example hospitals’ confirmation
of reception of referrals and clinical reports to the GPs
after discharge was included as a source of proxy information on waiting time. The GPs also emphasized the
importance for choice of hospital of hospitals’ attitude to
the GPs and the cooperation between hospital and GP,
and these reasons for choice of hospital were added to
the questionnaire.
The final questionnaire included the following questions (Additional file 1):
! For the GP: gender and year of birth.
! For each of the latest three somatic patients referred
to hospital (department or out-patient clinic) by the
GP for treatment: gender, year of birth, and the
specialty the patient was referred to. Who chose the
hospital in the GP’s opinion (the patient, the GP or
the patient’s relatives)? Which of 16 factors
influenced the choice of hospital strongly in the GP’s
opinion (see Table 1 for the list of factors; the GPs
could tick off as many factors as they found
relevant)? How many factors influenced the choice?
The GP could add comments on responsibility for
choice and on the 16 factors.
Page 4 of 10
Table 1 Factors deciding 216 GPs’ choice of hospital on
behalf of their most recent patient referred to hospital
Decisive factor
The department was the closest to the patient’s home
The department takes the GP’s referrals seriously
Number of
GPs (%)
187 (85%)
60 (27%)
Excellent cooperation between GP and department
56 (26%)
Comments from patients referred to the department
by the GP
54 (25%)
The patient had been treated at the hospital before
47 (21%)
The patient had been treated at the department before
44 (20%)
The hospital takes the GP’s referrals seriously
44 (20%)
Comments from patients referred to the hospital
by the GP
41 (19%)
The department provides detailed clinical reports
39 (18%)
The hospital provides detailed clinical reports
33 (15%)
The department sends clinical reports soon
after discharge
24 (11%)
The hospital sends clinical reports soon after discharge
20 (9%)
The GP´s experience as a trainee at the hospital
11 (5%)
The GP´s experience as a trainee at the department
15 (7%)
Waiting time was shorter than at other departments
11 (5%)
Total
686
Number of referrals
216
Number of factors quoted/referrals
3.2
! For the GP: which of 15 sources of information on
quality at department level did the GP in general
consider most relevant (see Table 2 for the list of
sources of information; the GPs could tick off as
many factors as they found relevant)? How often did
the GP use six specified sources of information on
expected waiting time at hospital departments
(routinely (4 points); often (3 points); rarely (2
points); not at all (1 point); see Table 3 for a list of
the sources of information)? The GP could add
comments on the sources of information on quality
as well as on waiting time.
A patient and a GP may share the choice of hospital
[36], but our objective was to establish which person the
GP considered to have the greatest influence on choice,
and how they chose the hospital on behalf of patients rather than investigate shared decision-making. Therefore
the GPs could not respond that they shared the decision
with the patient.
To increase the GPs’ response rate the questionnaire
was limited to four A4-pages, the questionnaire was
mailed by first-class-post, and a stamped return envelope was enclosed [37]. The questionnaire was mailed to
the study group in December 2003. GPs who did not respond within a month received a single reminder. We
Birk and Henriksen BMC Health Services Research 2012, 12:126
http://www.biomedcentral.com/1472-6963/12/126
Page 5 of 10
Table 2 General practitioners’ sources of information on quality at hospital departments (number and share of
240 GPs)
Source of information
Number and share of respondents
All GPs
Patients’ comments on the department
Female GPs
160 (66%)
51 (68%)
Male GPs
109 (66%)
Other GPs’ comments on the department
131 (54%)
42 (56%)
89 (54%)
Patients’ comments on the hospital
128 (53%)
40 (53%)
88 (53%)
The GP’s acquaintance with hospital personnel
91 (38%)
23 (31%)
68 (41%)
Other GPs’ comments on the hospital
89 (37%)
31 (41%)
58 (35%)
Official information from the department
85 (35%)
39 (52%)****
46 (28%)****
Clinical reports from different departments
74 (31%)
23 (31%)
51 (31%)
Information meetings in hospital departments
74 (31%)
23 (31%)
51 (31%)
Official information from the hospital
71 (30%)
32 (42%)**
39 (23%)**
The GP’s trainee experience at the hospital
54 (22%)
21 (28%)
33 (20%)
The GP’s trainee experience at the department
47 (20%)
16 (21%)
31 (19%)
The hospital’s description of its quality standards
7 (3%)
4 (5%)
3 (2%)
The department’s description of its quality standards
6 (3%)
3 (4%)
3 (2%)
Media reports about the hospital
5 (2%)
3 (4%)
2 (1%)
Media reports about the department
4 (2%)
2 (3%)
2 (1%)
1,027
354
673
240
75
165
4.3
4.7
4.1
Number of sources ticked off by the GPs
Number of respondents
Sources/respondent
**: p < 0.01. ****: p < 0.001.
compared respondents and non-respondents by county,
number of years since graduation (available from the Danish MDs’ Association’s “Who’s Who”) and gender
(deduced from the GPs’ names).
The following data was recorded: the number of
patients where the hospital was chosen by the GP, the
patient or the patient’s relatives. For GPs who had
chosen the hospital on behalf of one patient or more we
recorded the reasons for the choice on behalf of the latest patient referred to hospital, for which the GP
reported that he or she made the choice. Each GP was
only included once in the study of reasons for choice to
avoid mutually dependent observations.
Data was recorded in a database (EPIINFO Version
3.2.2. April 14, 2004). Respondents were compared with
the study population by univariate analyses of gender,
county (chi²) and number of years since graduation
(t-test). This analysis was repeated for GPs who had
chosen the hospital on behalf of at least one patient.
For GPs who had chosen the hospital on behalf of one
or more patients we recorded the GP’s reasons for
choice on behalf of the most recent patient to minimize
recall bias. GPs’ reasons for choice were compared by
univariate analysis for gender (chi²) and years since
graduation (t-test), and by logistic multiple regression
analysis with the GPs’ gender and years since graduation
as the independent variables. We tested for correlation
between the number of factors for choice and the GPs’
gender and years since graduation by use of a multiple
regression analysis. The GPs’ use of information sources
on quality and expected waiting time at various hospitals
were compared by univariate analyses for association
Table 3 General practitioners’ use of various sources of information on expected waiting time at hospitals (n = 241)
Source of information
Routinely (4)
Often (3)
Rarely (2)
Not at all (1)
Average (1–4)
Confirmations/clinical reports
45 (19%)
84 (35%)
67 (28%)
44 (18%)
2,5
The counties’ prognoses (paper)
28 (12%)
64 (27%)
73 (30%)
75 (31%)
2,2
Calls to the departments
2 (1%)
29 (12%)
141 (59%)
68 (28%)
1,9
12 (5%)
36 (15%)
78 (32%)
114 (48%)
1,8
www.venteinfo.dk
7 (3%)
31 (13%)
84 (35%)
119 (49%)
1,7
www.sundhed.dk
2 (1%)
9 (4%)
63 (26%)
166 (69%)
1,4
The counties’ prognoses (web)
Birk and Henriksen BMC Health Services Research 2012, 12:126
http://www.biomedcentral.com/1472-6963/12/126
with gender (chi²) or years since graduation (t-test) and
by multiple logistic regression analysis with gender and
years since graduation as the independent variables.
The study was performed in accordance with the
Helsinki Declaration. According to section eight in the
Danish Act on a Biomedical Ethics Committee System
and the Processing of Biomedical Research Projects questionnaire studies were not notifiable to the Danish research ethics committee system, if they did not include
biological material [38].
Results
A questionnaire was sent to the 483 GPs listed in the
database. Nine of the registered GPs in the database
represented data errors, reducing the study population
to 474 GPs. Two hundred and forty GPs (51%) returned
a filled-in questionnaire. Male and female GPs’ response
rates were 50% and 54%, respectively. Respondents
did not differ significantly from the study population
with regard to county, gender or number of years since
graduation.
Responsibility for choice of hospital
Among the 240 respondents 183 (76%) reported that in
their view they chose the hospital on behalf of the latest
patient referred to hospital, 35 (15%) reported that the
patient made the choice, two (1%) reported that the
patients’ relatives made the choice, and 20 (8%) did not
state, who chose the hospital or ticked off several categories. Several of these GPs commented that they chose
the hospital in cooperation with the patient or that the
patient agreed with the GP. One GP commented that he
always asked whether the patient wanted to be referred
to another hospital than the one proposed by the GP,
and another GP reported that he asked whether the
patient wanted to be referred to a specific department.
Reasons for GPs’ choice of hospital
Ninety-two percent of the respondents (220 of 240)
reported that they had chosen the hospital on behalf of
one patient or more, while 20 GPs attributed all patients’
choice to the patients, their relatives or referral guidelines limiting free choice of hospital for the specific
patients (these guidelines did not interfere with patient’
rights to choose a hospital in another county). The 220
respondents did not differ significantly from the study
population with regard to county, gender or number of
years since graduation. Four GPs did not tick off any
reasons for choice of hospital and were excluded from
this part of the study. Eighty-seven of the 216 GPs (40%)
reported that a single factor decided their choice for the
patient, short distance to the hospital being the decisive
factor for 75 of the 87 GPs (86%). Ninety-five GPs
reported that 2–5 factors were very important for their
Page 6 of 10
choice, 25 quoted 6–9 factors, and nine GPs quoted ten
or more factors.
Short distances to hospital, the department’s serious
consideration of referrals from the GP, and comments
from previous patients referred to the department were
the most common factors behind GPs’ choice of hospital
on behalf of patients (Table 1).
The importance of each factor behind choice of hospital
was not associated with the GP’s gender (data not shown),
and multiple regression analysis showed no significant association between the number of reasons for choice and
the number of years since the GP’s graduation (β = 0.109,
p = 0.26) or gender (p = 0.96). In univariate analysis of each
factor and the number of years since graduation, GPs who
based their choice on their personal experience with the
department as employees were significantly younger (on
average 20.1 years since graduation) than GPs who did
not (on average 24.2 years since graduation) (p < 0.05).
One GP commented that problems associated with
transport and rehabilitation after hospitalisation posed
greater challenges than waiting time, and therefore
patients were only referred out of the county if the quality
of care within the county was very bad.
Sources of information on quality or service
The most frequently used sources of information on
quality at hospital departments were reports from
patients referred to the department or the hospital by
the GP previously (Table 2), and other GPs’ comments
on the department. Univariate analysis showed that female GPs were significantly more likely than male GPs
to consider official information from departments an important source of information, and younger GPs were
significantly more likely than older GPs to quote their
experience as trainees at a department (average number of years since graduation 19.2 and 25.1 respectively,
p < 0.001) or a hospital (average number of years since
graduation 21.4 and 24.7 years respectively; p < 0.05) or
comments from patients previously referred to a department (average number of years since graduation 23.2
and 25.3 years respectively; p < 0.05) as important
sources of information on quality. Multivariate analyses
confirmed that there were statistically significant negative associations between the number of years after the
GPs’ graduation and GPs’ utilisation of information from
previously referred patients (β = −0.05; p < 0.05), and an
association between gender and use of official information from the department (odds ratio 0.37; p < 0.01) or
the hospital (odds ratio 0.35; p < 0.01), female GPs being
more likely to quote official information as a source of information. Multivariate analysis found no association
between GPs’ age or gender and their quoting experience from employment at department or hospital as
sources of information on quality (data not shown).
Birk and Henriksen BMC Health Services Research 2012, 12:126
http://www.biomedcentral.com/1472-6963/12/126
One GP commented pointedly that she spent her
scarce time on the patients rather than reading official
information about quality at various departments. Several GPs repeated in their response to this question their
answers to another question that previous patients’
reports were the most important sources of information
on quality, two GPs underscoring that they attributed
greater weight to 20 years’ experience with a hospital
than to the hospital’s description of its quality, one GP
contrasting “action” with “words”.
The hospitals’ letters of confirmation of referrals were
the GPs’ most important source of information on waiting times (Table 3). Information available from websites
was used less often than information on paper. Univariate analysis found no association between GPs’ age or
gender and their utilisation of various sources of information, but multiple, logistical regression analyses
showed that male (p < 0.01) and younger GPs (p < 0.01)
were especially likely to use the counties’ waiting time
prognoses on the internet.
Eight GPs reported that they often asked patients to
call the county’s patient advisor to discuss which department they would like to be referred to, one GP adding
that in some cases she recommended accepting a long
waiting time if the department was an especially good
one. One GP commented that in her opinion it was a
task for the hospital to inform the patients about waiting
times, while another GP had delegated collecting of and
information about data on waiting times to her secretary. Two GPs reported that if patients wanted a shorter
waiting time, they asked the patients to look for waiting
time data elsewhere. One GP likewise stated that she
informed the patient about the right to free choice but
asked them to investigate the opportunities on their
own. One GP reported that she did not use any official
information about expected waiting times, because waiting time prognoses were outdated as early as at the time
of publishing. Several GPs described an intention to
utilize data on the web in the future, although some GPs
found that utilization of data on the internet was a very
time consuming and complicated process. One GP
found that calling departments likewise was time consuming because it usually took a long time to find somebody who could answer questions on expected waiting
time.
Discussion
Responsibility for choice of hospital
In the present study the GPs reported that they chose
the hospital on behalf of 76% of patients. This result
appears to contradict results from national Danish surveys of patients’ experience with hospitals: in 2004 46%
of in-patients treated in the study area reported, that
they chose the hospital; among elective in-patients 89%
Page 7 of 10
were aware before being hospitalised that they were free
to choose, and 52% of these patients chose the hospital
by themselves [26]. The divergent findings may be interpreted as an indicator of shared decision making. When
30% of patients reported that their GP’s recommendation influenced their own choice of hospital [26], the
GPs may have perceived that they chose the hospital on
behalf of the patient.
The results indicate that patients choose the hospital
to a lesser degree than policy makers (politicians and
administrators) want them to do to improve management of the public health care sector by introducing a
proxy for the market mechanism. Other studies have
found that GPs appear to question whether choice is
valuable to patients [29], and whether patients really
want to choose the hospital [4]. GPs’ choice behaviour
varies by GP [29] and by the patients’ diagnoses [39],
English GPs being more likely to offer choice to patients,
who are in need of a routine intervention, elective
patients, and patients who are relatively healthy [4].
One study distinguished between ‘choice enthusiasts’,
‘choice sceptics’ and ‘choice paternalists’ [29]. The
present study did not enable us to divide GPs into such
subgroups, but confirming results from other studies
[4,30] several GPs expressed reluctance to provide advice
to patients, because they did not consider this task a part
of their job [30,40]; considered this task too timeconsuming for a consultation [4], or distrusted data published by the providers [29] and wanted to forestall
blame for presenting faulty data [23]. This behaviour
may reflect an attempt to minimize the length of each
visit to the GP. However, at a more general level GPs’
behaviour may reflect a ‘logic of care’ rather than a ‘logic
of choice’ [41] - GPs making choices based on their professional views on patients’ needs and wants, rather than
as agents acting in a market place enabling patients to
make informed choices in line with the neoclassical
standard model.
Factors determining GPs’ choice of hospital on behalf of
patients
Short distance to hospital was the most important factor
behind GPs’ choice of hospital. Numerous other studies
of GPs’ actual referral pattern and patients’ choices in
structurally different health care systems likewise indicate that short distance strongly influences patients’ and
GPs’ choice of hospital [26,33,42]. Studies of GPs’ hypothetical referrals and patients’ hypothetical choices have
led to other results with GPs emphasizing the importance of short waiting time and the GP’s impression of
quality at the alternative departments [35,43], while
patients facing a hypothetical choice emphasized the importance of data on structure quality and attributed little
weight to waiting time [44].
Birk and Henriksen BMC Health Services Research 2012, 12:126
http://www.biomedcentral.com/1472-6963/12/126
Different findings in studies of GPs’ and patients’
choice behaviour may reflect differences between how
GPs and patients think they ought to choose the hospital
and how they actually make the choice, one study finding significant differences between GPs’ response to
hypothetical case stories and their actual referral pattern
[45]. Another reason could be international institutional
differences with regard to subsidization of transport
costs and the length of waiting times.
The small influence of waiting time on choice may be
considered to be remarkable, as the media and politicians at the national level consistently focus on waiting
times as a major performance measure and challenge,
but other studies of choice of hospital likewise found
only a small influence of waiting time on choice. Cataract patients generally accepted waiting times of three
months and less, while waiting times of six months or
more were perceived as too long [46,47]. In a hypothetical study patients reported that for each additional hour
of travel time they would, on average, require a reduction of in the waiting time of 2.3 months [48]. The
results of these studies and the present study may partly
explain why differences between waiting times at hospitals persisted more than a decade after the introduction
of free choice of hospital, but they may also reflect, that
a minority of patients are treated as elective patients.
In the present study we focused on the influence of
GPs’ sources of information about departments/hospitals
and factors commonly found to influence the GPs’
choices. However, GPs’ choices on behalf of patients may
be influenced by other agendas independent of the individual patient, i.e. GPs may refer patients to a local hospital to contribute to its continuing existence [30].
GPs’ use of sources of information on quality and service
In the present study GPs were less likely to use official
information on quality and waiting time than proxymeasures from informal sources like their own and other
GPs’ and patients’ experience with regard to quality and
waiting time. This result was consistent with other studies of GPs’ or patients’ utilisation of sources of information, which have found very little utilisation of such
sources [49] and refer to GPs as having “a sort of
‘mental filing cabinet’ of informal information or soft
intelligence”[28]. The GPs’ experience with cooperation
with various departments or hospitals was very important for the GPs’ choice. GPs’ responses indicated that
their experience with specific departments was the most
important factor, but many GPs attributed their choice
to their experience with a hospital in general rather than
the individual department, thereby indicating that they
generalised their experience from one or more departments at a hospital to other departments at the hospital
as a whole – a kind of ‘halo’-effect.
Page 8 of 10
The strong influence of informal data sources like
patients’ previous experience on choice and advice on
choice may reflect lack of official information on quality
or waiting time or that GPs are suspicious of published
data on performance, viewing such data as “spin” [29].
Several respondents commented that use of web-based
information was too time-consuming compared to data
on paper; their memory of previous referrals, and asking
the patient to call one or more hospitals or the county’s
patient’s advisors for information.
Some GPs wrote that they intended to use web-based
information more in the future. Such statements may reflect expectations that more experience and improved IT
will ease their access to the web or lack of experience.
When the present study was performed approx. 86% of
Danish general practices had access to the internet, and
a little less than half of the practices used the access
each day [50].
Implications
Further research is warranted on the interaction between
GP and patient in choice of hospital, preferably by direct
observation of the referral process followed by interviews with the GP as well as the patient about their
views on the referral process including their experience
of responsibility for the choice.
The findings in the present study support results from
studies of patients’ choice behaviour which indicate that
patients and their agents do not act as the autonomous
customers assumed in market-resembling models for
management of the public sector. When agents act on
patients’ behalf they tend to utilise informal sources of
information – even when systematically collected and
published information on service is available. One implication of the major influence of previous experience
with hospital departments may be a tendency to inertia
in referral patterns.
Limitations of the study
The response rate in the present study was 52%, which
appears to be quite normal for studies performed in
general practice.
The choice of study method meant that we did not
observe the process of choice, and only reasons we were
aware of beforehand were included in the study, but the
questionnaire was validated, and the respondents were
offered the opportunity to comment on the reasons
and did not refer to reasons not mentioned in the
questionnaire.
The respondents could report any number of reasons
and we did not ask them to quantify the importance of
each reason, because this would complicate the data collection and probably reduce the response rate. We
assumed that the cumulative importance of a reason for
Birk and Henriksen BMC Health Services Research 2012, 12:126
http://www.biomedcentral.com/1472-6963/12/126
choice of hospital was proportional to the frequency it
was quoted by the GPs, but this may not necessarily be
the case: a comparison of two Dutch studies published
recently may indicate that frequency of reporting may
give results which differ from estimations of importance
by way of a choice experiment [25,44].
Respondents did not differ from non-respondents with
regard to age, gender or county, but GPs with a stronger
than average interest in subjects concerning choice of
hospital may be especially likely to participate in the
study. Therefore the study may exaggerate the impact of
each individual factor on choice of hospital.
Usually studies should be performed prospectively to
reduce bias, but in the present study we chose a retrospective design in order not to influence the GPs’ choice
behaviour. Our choice of design increased the risk of recall bias, and the GPs may have reported factors which
they thought ought to have influenced their choices rather than the decisive factors. For example GPs may
have hesitated to quote media reports as an important
source of information. GPs probably are very conscious
about their use of some sources of information like websites, while the importance of some sources may be
underestimated, because their utilisation is more nebulous, like feedback from patients or media reports. Presumably patient characteristics influenced the GPs’
choices but not their willingness to participate. Therefore patient characteristics presumably did not introduce
bias in the study.
The study included a large number of statistical tests.
Some of the statistically significant associations in univariate analysis may be due to mass significance rather
than causality.
The study was performed thirteen years after the
introduction of free choice of public hospital within the
study area and eleven years after the introduction of free
choice of public hospital at the national level. Patients’
awareness of their right to choose was high. Therefore,
even though the study was performed at a specific time
in the process of introducing free choice of hospital, we
find it most likely that studies performed a few years before or after the present study would not have led to
results which were very different from those of the
present study.
Conclusions
In an area with free choice of public hospital GPs
strongly influenced patients’ choice of hospital by choosing the hospital on their behalf. Short distance to hospital was the factor which most frequently decided the
GP’s choice of hospital on behalf of patients. GPs focused on informal sources like feedback from patients
and colleagues and their experience with cooperation
with the department or hospital, attaching little weight
Page 9 of 10
to official information on quality and service (waiting
time) at hospitals or departments.
Additional file
Additional file 1: The questionnaire. The questionnaire used for
collecting data for the present study.
Competing interests
The authors declare that they have no competing interests.
Acknowledgements
The present study was supported and financed by Roskilde County, Region
Zealand, ” The Health Research Forum in East Denmark, Research
Programme for Promotion of Regional Cooperation on Medical Research”
and ” The Health Research Foundation of the Counties in Eastern Denmark
(Region 3)”.
Authors’ contributions
Both of the authors conceived and designed the study; HOB developed the
questionnaire, collected and analysed the data and wrote the manuscript.
LOH assisted in writing the manuscript. Both authors read and approved the
final manuscript.
Received: 10 October 2011 Accepted: 25 May 2012
Published: 25 May 2012
References
1. Winblad U, Ringard A: Meeting rising public expectations: the changing
roles of patients and citizens. In Nordic health care systems. Recent reforms
and current policy challenges. Edited by Magnussen J, Vrangbæk K, Saltman
RB. Maidenhead: Open University Press; 2009.
2. Le Grand J: Motivation, agency, and public policy. Of Knights & Knaves, Pawns
& Queens. Oxford: Oxford University Press; 2003.
3. Vrangbæk K, !stergren K: Patient empowerment and the introduction of
hospital choice in Denmark and Norway. Health Econ Policy Law 2006,
1:371–394.
4. Dixon A, Robertson R, Appleby J, Burge J, Devlin N, Magee H: Patient choice.
How patients choose and providers respond. London: The King’s Fund; 2010.
5. Perry G: Giving consumers of British public services more choice: what
can be learned from recent history? J Soc Policy 2003, 32:239–270.
6. Thomson S, Dixon A: Choices in health care: the European experience.
J Health Serv Res Policy 2006, 11:167–171.
7. Dowding K, John P: The value of choice in public policy. Public
Administration 2009, 87:219–233.
8. Wilmot S: A fair range of choice: justifying maximum patient choice in
the British National Health Service. Health Care Anal 2007, 15:59–72.
9. Porter ME, Teisberg EO: Redefining Health Care. Creating value-based
competition on results. Cambridge: Harvard Business School Press; 2006.
10. Appleby J, Harrison A, Devlin N: What is the real cost of more patient choice.
London: King’s Fund; 2003.
11. Vrangbæk K: The interplay between central and sub-central levels: the
development of a systematic standard based programme for governing
medical performance in Denmark. Health Econ Policy Law 2009, 4:305–327.
12. Tai-Seale M: Voting with their feet: patient exit and intergroup
differences in propensity for switching usual sources of care. J Health
Polit Policy Law 2004, 29:491–514.
13. Le Grand J: Choice and competition in publicly funded health care.
Health Econ Policy Law 2009, 4:479–488.
14. Hirschman Bernstein AB, Gauthier AK: Choices in health care: what are
they and what are they worth? Med Care Res Rev 1999, 56:5–23.
15. Bernstein AB, Gauthier AK: Choices in health care: what are they and what
are they worth? Med Care Res Rev 1999, 56:5–23.
16. Martinsen DS, Vrangbæk K: The Europeanization of health care
governance: implementing the market imperatives of Europe. Public
Administration 2008, 86:169–184.
17. Fotaki M, Roland R, Boyd A, McDonald R, Scheaff R, Smith L: What benefits
will choice bring to patients? Literature review and assessment of
implications. J Health Serv Res Policy 2008, 13:178–184.
Birk and Henriksen BMC Health Services Research 2012, 12:126
http://www.biomedcentral.com/1472-6963/12/126
18. Marshall MN, Shekelle PG, Leatherman S, Brook RH: The public release of
performance data: what do we expect to gain? A review of the
evidence. JAMA 2000, 283:1866–1874.
19. Akerlof GA: The market for “lemons”. Quality uncertainty and the market
mechanism. Q J Econ 1970, 84:488–500.
20. Arrow KJ: Uncertainty and the welfare economics of health care. Am Econ
Rev 1963, 53:941–973.
21. Loomes G, Sugden R: Regret theory: an alternative theory of rational
choice under uncertainty. Econ J 1982, 92:805–824.
22. Bell D: Regret in decision making under uncertainty. Oper Res 1982,
30:961–981.
23. Barnett J, Ogden J, Daniells E: The value of choice: a qualitative study.
Br J Gen Pract 2008, 58:609–613.
24. Rosen R, Curry Florin D: Public views on choices in health and health care.
London: King’s Fund; 2005.
25. Djis-Elsinga J, Otten W, Versluijs MM, Smeets HJ, Kievit J, Vree R, van der
Made WJ, Marang-van de Mheen P: Choosing a hospital for surgery: the
importance of information on quality of care. Med Decis Making 2010,
30:544–555.
26. !sterbye T, Gut R, Blæsbjerg C, Freil M: Patienters oplevelser på landets
sygehuse 2004 [in Danish: Patients’ Experience at Danish Hospitals 2004].
Glostrup: Enheden for Brugerundersøgelser; 2005.
27. Department of Health: Report on the National Patient Choice Survey. February
2010 England [online]. Available at http://www.dh.gov.uk/prod_consum_dh/
groups/dh_digitalassets/documents/digitalasset/dh_117096.pdf [homepage
on the Internet] [cited on 1 April 2012].
28. Burge P, Devlin N, Appleby J, Gallo F, Nason E, Ling T: Understanding
patients’ choices at the point of referral. Cambridge: RAND Europe; 2006.
29. Rosen R, Florin D, Hutt R: The anatomy of GP referral decisions. A qualitative
study of GPs’ views on their role in supporting patient choice. London: King’s
Fund; 2007.
30. Wiinblad U: Do physicians care about patient choice? Soc Sci Med 2008,
67:1502–1511.
31. Strandberg-Larsen M, Nielsen MB, Vallgårda S, Krasnik A, Vrangbæk K,
Mossialos E: Denmark: Health system review. Health Syst Transit 2007,
9(6):1–164.
32. Vrangbæk K: Markedsorientering i sygehussektoren [in Danish: Market
orientation in the hospital sector]. PhD-thesis. Copenhagen: Institute of
Political Science, University of Copenhagen; 1999.
33. Mahon A, Whitehouse C, Wilkin D, Nocon A: Factors that influence general
practitioners’ choice of hospital when referring patients for elective
surgery. Br J Gen Pract 1993, 43:272–276.
34. Odell A: A study of patient referrals. Publ Hlth Lond 1983, 97:109–114.
35. Kennedy F, McConnell: General practitioner referral patterns. J Public
Health Med 1993, 15:83–87.
36. Ringard A, Rico A: Introducing patient choice of hospital in National Health
Systems – A comparison of the UK and Norway. Working Paper. Oslo: Health
Organization Research Norway (HORN); 2006.
37. Edwards P, Roberts I, Clarke M, DiGuiseppi C, Pratap S, Wentz R, Kwan I:
Increasing response rates to postal questionnaires: systematic review.
BMJ 2002, 324:1183–1191.
38. Lov om et videnskabsetisk komitésystem og behandling af biomedicinske
forskningsprojekter [in Danish: Act on a Biomedical Ethics Committee System
and the Processing of Biomedical Research Projects]. Available in English from:
http://www.cvk.sum.dk/English/actonabiomedicalresearch.aspx [homepage
on the Internet] [cited on 1 April 2012].
39. Ringard Å: Why do general practitioners abandon the local hospital? An
analysis of referral decisions related to elective treatment. Scand J Public
Health 2010, 38:597–604.
40. Dixon A, Robertson R, Bal R: The experience of implementing choice at
point of referral: a comparison of the Netherlands and England. Health
Econ Policy Law 2010, 5:295–317.
41. Mol A: The logic of care. Health and the problem of patient choice. Abingdon:
Routledge; 2008.
42. de Mheen PJ Marang-van, Dijs-Elsinga J, Otten W, Versluijs M, Smeets HJ,
Vree R, van der Made WJ, Kievit J: The relative importance of quality of
care information when choosing a hospital for surgical treatment: a
hospital choice experiment. Med Decis Making 2011, 31:816.
43. Adams EK, Wright KE: Hospital choice of Medicare beneficiaries in a rural
market: why not the closest? J Rural Health 1991, 7:134–152.
Page 10 of 10
44. McArdle PJ, Whitnall M: The referral practice of general medical
practitioners to the surgical specialties: implications for the future.
Br J Oral Maxillofac Surg 1996, 34:394–399.
45. Morrell DC, Roland MO: Analysis of referral behaviour: responses to
simulated case histories may not reflect real clinical behaviour.
Br J Gen Pract 1990, 40:182–185.
46. Conner-Spady B, Sanmartin C, Johnston G, McGurran J, Kehler M,
Noseworthy T: Willingness of patients to change surgeons for a shorter
waiting time for joint arthroplasty. CMAJ 2008, 179:327–332.
47. Dunn E, Black C, Alonso J, Nørregaard JC, Anderson GF: Patients’
acceptance of waiting time for cataract surgery: what makes a wait too
long? Soc Sci Med 1997, 44:1603–1610.
48. De Groot IB, Otten W, Smeets HJ, Marang-van de Mheen PJ: Is the impact
of hospital performance data greater in patients who have compared
hospitals? BMC Health Serv Res 2011, 11:214.
49. Burge P, Devlin N, Appleby J, Rohr C, Grant J: Do patients always prefer
quicker treatment?: a discrete choice analysis of patients’ stated
preferences in the London Choice Project. Appl Health Econ Health Policy
2004, 3:183–194.
50. Praksistælling 2003 [in Danish: ” General Practices in 2003”]. Copenhagen: The
Danish General Practitioners Organisation; 2003.
doi:10.1186/1472-6963-12-126
Cite this article as: Birk and Henriksen: Which factors decided general
practitioners’ choice of hospital on behalf of their patients in an area
with free choice of public hospital? A questionnaire study. BMC Health
Services Research 2012 12:126.
Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit