ACCELERATING CHILD SURVIVAL AND DEVELOPMENT (ACSD)
Final Report
The Retrospective Evaluation of ACSD:
Benin
Submitted to UNICEF on 7 October 2008
Institute for International Programs
Johns Hopkins University Bloomberg School of Public Health
Baltimore, MD
Disclaimer:
This report was prepared by IIP-JHU under contract with UNICEF. All photos were taken by members of the
IIP-JHU evaluation team after requesting permission from those who were photographed. All text, data,
photos and graphs should be cited with permission from the authors and UNICEF.
ii
IIP-JHU | Retrospective evaluation of ACSD in Benin
Summary
Introduction
UNICEF implemented the Accelerated Child Survival and Development (ACSD) project between 2001
and 2005 in 11 countries in Africa with support from Canadian CIDA. The aim of ACSD was to reduce
mortality among children less than five years of age by working with governments and other partners to
increase coverage with a set of proven interventions. In the “high-impact” countries of Benin, Ghana, Mali
and Senegal, a total of 16 districts worked to deliver the full set of interventions grouped into three
packages: “EPI+” including vaccinations, vitamin A supplementation and the use of insecticide-treated
nets (ITNs) for the prevention of malaria; “IMCI+” including promotion of exclusive breastfeeding for six
months, timely complementary feeding, use of iodized salt and improved and integrated management at
the health facility and community levels of children suffering from pneumonia, malaria and diarrhea,
including home-based ORS use, treatment of malaria, and treatment of pneumonia with antibiotics; and
“ANC+” including intermittent preventive treatment of malaria with SP (Fansidar) for pregnant women
(IPTp), tetanus immunization during pregnancy to prevent maternal and neonatal tetanus and
supplementation with iron/folic acid during pregnancy and with vitamin A post-partum. An internal
evaluation by UNICEF estimated through modeling that the levels of coverage achieved through ACSD
were associated with about a 20 percent reduction in all-cause under-five mortality relative to comparison
districts in participating districts in four “high-impact” countries. This retrospective evaluation was
commissioned by UNICEF to confirm these findings and provide additional information that could be used
th
in planning effective programs to reduce child mortality and achieve the 4 Millennium Development Goal
(MDG-4) in poor countries in Africa.
The IIP evaluation team worked with ACSD managers at international and national levels to develop a
generic ACSD framework that defined the pathways through which ACSD activities were expected to lead
to reductions in child mortality and improvements in child nutritional status. The generic framework
served as the “backbone” of the evaluation design. The country-specific evaluations also addressed
equity across socioeconomic and ethnic groups, for urban-rural residence and for girl and boy children.
At the request of UNICEF, the evaluation does not include an economic evaluation or a full assessment of
the effects of ACSD on national policy.
Aim of the independent retrospective evaluation in Benin
The aim of the evaluation was to provide valid and timely evidence to child health planners and policy
makers about the effectiveness of ACSD in reducing child mortality and improving child nutritional status
in Benin, as a part of the larger retrospective evaluation designed to inform future programs intended to
reduce child mortality and accelerate progress toward MDG-4. Equity was also assessed.
Two questions served as a guide to the analysis and reporting of the evaluation findings:
a. Was ACSD implementation associated with improvements in coverage, nutrition and mortality
over time?
b. If so, was progress in the ACSD zones faster than that observed for the national comparison
area?
ACSD implementation in Benin
In March 2002, US$1.5 million in support from Canadian CIDA was transferred to UNICEF-Benin for
support of ACSD activities. UNICEF staff reported that ACSD activities have been supported since 2003
with funding from Belgium and UNICEF general resources. The Benin ACSD project aimed to build on
and complement existing child health activities in the country, with a specific focus on two health zones—
Djidja/Abomey/Agbangniizoun (DAA) in the department of Zou and Pobè/Adja-Oere/Kétou in the
department of Plateau. These six communes are referred to as “high-impact zones” or HIZs. Together
the two HIZs represented an estimated 482,838 people, or 7% of the population of Benin. ACSD
IIP-JHU | Retrospective evaluation of ACSD in Benin
i
implementation began in late 2002 in both HIZs. ACSD-Benin inputs and activities in the HIZs focused
on:
1) Providing essential drugs, supplies and equipment. ACSD-Benin a) provided 19 motorcycles
and four vehicles to the HIZs over the course of the project for outreach and supervision
activities; b) equipped health facilities with basic medical equipment including delivery kits in 2002
and a kit for newborn care in 2004, refrigerators for the cold chain, and computers for monitoring
activities; and c) supplied commodities including vaccines, vitamin A supplements and ITNs for
the prevention of malaria.
2) Training and supervising facility-based workers. ACSD-Benin trained 57 health workers in
2002 and 84 health workers in 2004-5 in EPI provision, monitoring and surveillance. Thirty-one
facility-based health workers were trained in standard IMCI and their facilities were reported to
have received three supervision visits per year between 2003 and 2006. Over 200 health
workers were trained in healthy child visits in 2003 and 2004. As part of the ANC+ component,
facility-based workers were trained on IPTp, focused ANC care, management of the neonate and
active management of the third stage of labor. There was some additional training on the
prevention of mother-to-child-transmission of HIV for facility-based staff in 2005. Supervision of
workers in provision of ANC was reported to be sporadic and no quantitative data on the
frequency or quality of supervision are available.
3) Training, equipping and supervising community health workers. ACSD-Benin provided
support for the training of approximately 400 community-based health workers (CHWs) in 200
villages to deliver key messages concerning family practices and to coordinate ITN distribution
and retreatment in 2003 and 2004. CHWs were supervised annually thereafter. These 200
villages in the HIZs received health kits containing chloroquine, antipyretics, antihelminthic drugs
for deworming, ORS, and iron to be distributed trained CHWs. As part of a pilot study in 2005,
additional training was provided to 40 of these community-based workers in community
management of pneumonia using cotrimoxazole.
ASCD also provided support to train
community-based women’s groups in the promotion of ITNs.
4) Supporting outreach and campaign activities. ACSD-Benin provided support for national and
local vaccination campaigns, catch-up vaccination activities, and bednet retreatment campaigns.
The major barriers to implementation of ACSD-Benin were reported to be delays in receipt of funds from
UNICEF headquarters, insufficient amounts of resources for implementation, and the absence of
sufficient incentives (whether monetary and non-monetary) to motivate facility- and community-based
health workers to apply the skills learned through training, especially after initial program funds were
exhausted. Additionally, MOH and UNICEF program managers reported that the detection of wild
poliovirus from Nigeria in 2003 meant that time and resources were diverted to organize an intensive
series of national immunization campaigns.
Evaluation design and methods
The IIP evaluation team worked with UNICEF-Benin, the Government of Benin and other partners to
adapt the generic ACSD evaluation design to local implementation characteristics. Relevant data were
identified and assessed based on minimum quality criteria. Baseline data on coverage and nutritional
status were obtained from the 2001 demographic and health survey (DHS) representative at the
provincial level. Endline data on coverage, nutritional status and mortality were obtained from the 2006
DHS. To ensure sufficient sample size in the HIZs, additional households were sampled in early 2007
using methods and procedures as similar to the 2006 DHS as possible.
Information on ACSD implementation and on contextual factors that could have affected the evaluation
results or their measurement was collected by independent IIP investigators. Methods included review of
documents and administrative reports and personal interviews. Inequities by socioeconomic status were
determined by comparing priority indicators across quintiles of household assets; results were also
stratified by sex of the child, urban/rural residence and ethnic group affiliation.
ii
IIP-JHU | Retrospective evaluation of ACSD in Benin
The initial analyses described levels and time trends in intervention coverage, nutritional status and
under-five mortality in the HIZs. Next, these results were related to levels and trends in a national
comparison area, which included the remainder of the Benin, except metropolitan Cotonou and the HIZs.
Additional analyses were conducted to explore alternative explanations for the results. All results and
interpretations were reviewed with representatives of the Government of Benin and UNICEF-Benin on
several occasions.
Results
Vitamin A supplementation for children 6-59 months and puerpera, and ITN utilization among children
and pregnant women increased significantly in both the HIZs and the comparison area. ITN use among
pregnant women increased more in the HIZs than in the comparison area. Neither the HIZs nor the
national comparison area showed significant improvements in coverage for vaccination, antenatal care,
delivery care, case-management or infant feeding over the study period. For postnatal visits, coverage
declined in the HIZs and remained stable in the comparison area.
For coverage, the answers to the two primary evaluation questions were:
(a) ACSD implementation was associated with increases in vitamin A supplementation and ITN
utilization among women and children. Most indicators of coverage and practices did not improve
over time in the HIZs, and coverage levels at the end of the implementation period were well
below the ACSD targets.
(b) Overall, there was no consistent increase in coverage in the HIZs relative to the rest of the
country.
The evaluation results on nutritional status and mortality were consistent with the lack of an effect on
most coverage indicators. For nutrition, the findings were:
a. The HIZs showed a reduction over time in underweight and wasting prevalence, but not in
stunting.
b. The lack of change in stunting and the reduction in wasting prevalence over the course of the
study period were similar for the HIZs and the comparison area. The reduction in underweight in
the HIZs was not observed in the rest of the country, but lack of progress in the national
prevalence of underweight was due to a food crisis in the north of the country; once this region is
excluded from the comparison area, time trends in underweight reflect what was observed in the
HIZs.
For mortality, the findings were:
(a) There was a non-significant reduction of 13% in under-five mortality in the HIZs between 1999
and 2006, half of the ACSD target of a 25% reduction.
(b) In the comparison area, the U5MR declined by 25% during the same period. Analyses of
mortality rates by age subgroups within the 0-59 month range also found no evidence that rates
fell more rapidly in the HIZs than in the rest of the country.
Analyses of inequalities in coverage in 2006-7 showed that poor populations are consistently being
underserved, both in the ACSD and comparison area. It was not possible to examine how inequalities
changed over time due to small sample sizes at baseline. Inequalities were present even for
interventions delivered through community-based outreach and/or campaigns, such as vitamin A
supplementation and ITNs. No coverage inequalities were found in 2006-7 by the sex of the child. As
observed globally, rural populations were systematically worse off than urban populations. There was
some evidence that this gap was smaller in the HIZs than in the comparison area. Differences in
coverage and impact between the two major ethnic groups – Fon and Yoruba – were not consistent and
could not be interpreted as systematic disparities.
IIP-JHU | Retrospective evaluation of ACSD in Benin
iii
Discussion and interpretation
In Benin, there were improvements in two important coverage indicators – vitamin A and ITNs for children
– but other important indicators were either stagnant or declined. There was no consistent evidence of an
improvement in nutritional status, and mortality declined by 13%, about half of the proposed target. There
were no difference in coverage, nutrition or mortality trends in the HIZs relative to the rest of the country.
We examined the possibility that external factors might explain the apparent lack of an impact of ACSD
through extensive reviews of existing data and interviews with key informants, and were unable to identify
any contextual factors that might account for the lack of impact.
The retrospective nature of the evaluation imposed important limitations. First, information on ACSD
implementation and contextual factors had to be reconstructed from available reports and the subjective
recall of program implementers. There is no way to confirm the validity of the results or the extent or
direction of possible bias in the documentation of activities. Second, data quality issues limited the
evaluation because some of the available data did not meet basic quality criteria. The evaluation team
did its best to overcome these limitations by working closely with the in-country implementation team and
reviewing all summaries and results with them on successive occasions.
How can the finding of “no effect” of the ACSD-Benin project be explained? Possible explanations
generated in collaboration with the in-country team included the following:
1) Weaknesses in program design. Given the cause-of-death profile in Benin, accelerated
mortality reduction is dependent on the effective prevention of deaths from malaria, pneumonia,
diarrhea and neonatal causes. The potential effect of ACSD-Benin was limited by three basic
design weaknesses. First, messages designed to improve infant feeding practices and
careseeking for suspected pneumonia were only included in late 2005. Only 32% of children with
suspected pneumonia (the second largest cause of child deaths in Benin) were taken to a trained
provider for care, and this did not change over time. Second, no effective antimalarial treatment
was available in the HIZs. National policy recommended artemisinin combination therapy (ACT)
as the first-line treatment for fever (presumed malaria). However, ACTs were not widely available
during the study period, and both facility and community workers prescribed chloroquine long
after resistance levels had reduced the drug’s antimalarial effectiveness Finally, no monetary
incentives were provided to the community health workers who were expected to deliver lifesaving interventions to mothers and children. ACSD program implementers highlighted this as a
major barrier to ACSD effectiveness.
2) The intensity of implementation was lower than needed to achieve coverage or change
family practices. The ACSD program in Benin received US$ 1.5 million in initial implementation
funds, transferred approximately one year after funds were received by the three other ACSD
“high-impact” countries. This represents approximately US$15 per child less than five years of
age in the HIZs over the life of the project. National implementers believe that this delay, in
combination with the low level of investment, did not allow coverage to increase to the levels
necessary for achieving an impact on mortality. For example, ITNs are a key component of
ACSD, but increased by only 20 percentage points in the HIZs, compared to 23 percentage
points in the rest of the country. Deaths due to diarrhea are also unlikely to have changed as a
result of ACSD implementation, because correct home treatment practices also failed to increase.
3) The implementation strategies failed to reach the poorest, among whom the impact would
have been greatest. The evaluation findings indicate that coverage of interventions, even those
delivered through campaigns or community-based delivery strategies, remained highly
inequitable in 2006-7.
The lack of a differential effect of implementation on coverage in the HIZs must be interpreted in light of
numerous concurrent initiatives in the rest of the country. These included initiatives to which ACSD
contributed directly or indirectly, particularly UNICEF-supported programs for promoting vitamin A
supplementation and ITN use by children under five. Despite these combined efforts, however, coverage
levels for most of these high-impact interventions remained below 60% in both the HIZs and the rest of
the country in 2006-7. The “acceleration” effects expected by ACSD planners did not occur.
iv
IIP-JHU | Retrospective evaluation of ACSD in Benin
Despite high levels of commitment and effort by the MOH, UNICEF, and other country partners, the
resources available for acceleration under ACSD were too little and too late to result in a major
acceleration in coverage for child survival interventions in the HIZs. The evaluation results on mortality,
undernutrition and equity are consistent with the lack of an effect on intervention coverage and family
practices. In comparison to the rest of the country, where several other initiatives were also being
promoted, ACSD was unable to accelerate progress towards child survival in Benin.
IIP-JHU | Retrospective evaluation of ACSD in Benin
v
Table of Contents
1. The external retrospective evaluation of ACSD in four countries .....................
1
2. Evaluation methods ...........................................................................................
5
3. Characteristics of “high-impact” and comparison area ......................................
11
4. ACSD as implemented in Benin .......................................................................
17
5. Coverage and family practices ..........................................................................
27
6. Nutrition ............................................................................................................
47
7. Mortality .............................................................................................................
53
8. Equity of coverage, nutrition and mortality ......................................................
59
9. Conclusions ......................................................................................................
65
References ...............................................................................................................
68
Appendices
A. Description of Benin and “high-impact” zones
B. Methodology for documentation of implementation activities and contextual factors
C. Documentation of ACSD implementation in “high-impact” zones
D. Definition of priority indicators
E. Comparison of survey questions used for priority coverage indicator calculation
F. Methodologies of surveys and other data in Benin 2001-2007
G. Tables presenting priority coverage indicators over time for ACSD “high-impact” zones
H. Tables presenting comparisons of priority coverage indicators over time in ACSD “high-impact” zones
and the comparison area
I.
Tables presenting 2006-7 results for key indicators in the ACSD “high-impact” zones by sociodemographic characteristics of the population
J.
Summary of contextual factors possibly associated with coverage outcomes
K. Description of methodological challenges
L. Tables presenting additional nutritional analyses
M. Tables presenting additional equity analyses
N. References for the appendices
O. Annotated list of documents reviewed in the ACSD evaluation (file available upon request)
vi
IIP-JHU | Retrospective evaluation of ACSD in Benin
Acknowledgements
This evaluation could not have been conducted without full participation of the representatives from the
Ministry of Health, the national statistics offices, and UNICEF-Benin, who formed the Benin ACSD
evaluation team. We thank them for their commitment to child survival, as reflected in their willingness to
share their time, as well as information and their personal opinions about the contributions and limitations
of the ACSD project. We specifically would like to thank Alban Quenum and Gilbert Vissoh from the
Ministry of Health who provided insights throughout the evaluation. The national statistics office (Institute
Nationale de la Statistique) carried out surveys integral to this evaluation; we especially thank Elise
Ahovey who collaborated throughout the evaluation process.
UNICEF-Benin staff were responsible for working with governments and partners to implement the ACSD
project and collaborate in activities related to the independent retrospective evaluation and we thank them
for their commitment to child survival and to the evaluation process as a means of improving program
effectiveness. Andrée Cossi was an essential resource throughout the evaluation, we are truly grateful
for her dedication to this evaluation process and her strong commitment to using data to improve
programs. We would also like to express our appreciation to Souleymane Diallo, Philippe Duamelle,
Marianne Clark-Hattingh, Paul Adovohekpe, Hortense Kossou, Dominique Robez-Masson, Arnaud
Houndeganme and Loukmane Agbo-Ola. UNICEF-Benin also provided financial support for the
supplemental survey and advanced technical assistance from Macro, International. This support was
essential, as without it there would have been few data to analyze. We would also like to thank UNICEF
staff at regional and global levels for their efforts to provide us with documentation about ACSD and the
values and conceptual frameworks that guided its implementation.
Additionally, we would like to thank the members of the IIP-JHU for their insights and help throughout the
evaluation, as well as Macro International and Trevor Croft for technical assistance. Suzanne van Hulle
provided valuable assistance in the documentation of ACSD implementation and contextual factors.
Finally, we thank the leadership of UNICEF and CIDA, for their continuing commitment to the importance
of independent evaluations and their efforts to see that this work was completed.
IIP-JHU | Retrospective evaluation of ACSD in Benin
vii
Acronyms
viii
ACSD
Accelerated Child Survival & Development Project
ACT
Artemisinin combination therapy for use in treating fever/malaria.
ANC
Antenatal care
ANC+
One of the ACSD intervention packages, consisting of antenatal care and the
intermittent prevention of malaria during pregnancy (IPTp)
BASICS
Basic Support for Institutionalizing Child Survival, a project supported by the United
States Agency for International Development.
CFA
Central and West African Francs, the currency used in Benin, Mali and Senegal. On 7
January 2007, Forex quoted the exchange rate as USD1 = CFA 504.
CDC
US Centers for Disease Control and Prevention
CHW
Community health worker
CIDA
Canadian International Development Agency
C-IMCI
Community component of Integrated Management of Childhood Illness
DAA
Djidja/Abomey/Agbangniizoun health zone (DAA) in the Zou Collines region of Benin;
one of the ACSD “high-impact” zones
DHS
Demographic and Health Surveys (DHS), supported by USAID.
DPT
Diphtheria, Pertussis, Tetanus immunization
EPI
Expanded Program on Immunization
EPI+
One of the ACSD intervention packages, consisting of the full EPI schedule as well as
the provision of vitamin A and deworming twice each year for children aged six to
59 months, and the provision of insecticide-treated nets for the prevention of
malaria.
F-IMCI
Facility component of Integrated Management of Childhood Illness, which includes
improving the skills of facility-based health workers as well as strengthening
aspects of the health system needed to provide appropriate care for children less
than five years of age.
Hib
Haemophilus influenzae type b immunization
HIZs
“High-Impact” Zones of ACSD implementation, including Djidja/Abomey/Agbangniizoun
health zone (DAA) in the Zou Collines region and Pobè/Adja-Oere/Kétou (PAK)
health zone in the Ouémé-Plateau region of Benin
IEC
Information, Education and Communication
IIP
The Institute for International Programs at JHU
IMCI
Integrated Management of Childhood Illness
IIP-JHU | Retrospective evaluation of ACSD in Benin
INSAE
Institute Nationale de la Statistique
IPTi
Intermittent preventative treatment for malaria in infancy
IPTp
Intermittent preventative treatment for malaria in pregnancy
ITN
Insecticide-treated net
JHSPH
The Johns Hopkins University Bloomberg School of Public Health
MBB
Marginal Budgeting for Bottlenecks, a tool developed by UNICEF and the World Bank
to support results-based planning for maternal, newborn and child survival in
developing countries.
MDG
Millennium Development Goal
MDG-4
The fourth millennium development goal, which aims to reduce mortality among
children less than five years of age by two-thirds from levels in 1990.
MICS
Multiple Indicator Cluster Survey designed by UNICEF
MOH
Ministry of Health
NID
National Immunization Days
NGO
Non-governmental organization
ORS
Oral Rehydration Salts, usually pre-packaged in a sachet
ORT
Oral Rehydration Therapy, can be either pre-packaged in a sachet or prepared in the
home
PAK
Pobè/Adja-Oere/Kétou health zone in the Ouémé-Plateau region of Benin; one of the
ACSD “high-impact” zones
PBT
Preceding birth technique – a simplified approach to estimating young child mortality
PMTCT
Prevention of mother-to-child transmission of HIV
PNLP
Programme National de Lutte contre le Paludisme; National malaria control program
pp
percentage points
PROLIPO
Projet de Lutte Integré contre le paludisme dans l’OuéméPlateau (MOH-led malaria
control project in OuéméPlateau region in Benin, implemented by CDC with
support from USAID.
PSI
Population Services International
RHF
Recommended home fluids, for the treatment of childhood diarrhea
SP
A combination of two drugs, sulfadoxine and pyrimethamine. This drug combination is
commonly known as Fansidar.
TT
Tetanus toxoid vaccination
U5MR
Under five mortality rate
IIP-JHU | Retrospective evaluation of ACSD in Benin
ix
x
UN
United Nations
UNICEF
United Nations Children’s Fund
USAID
United States Agency for International Development
IIP-JHU | Retrospective evaluation of ACSD in Benin
1.
The external retrospective evaluation of ACSD in four countries
UNICEF implemented the Accelerated Child Survival and Development (ACSD) project between 2002
and 2005 in 11 countries in Africa with support from Canadian CIDA and other partners. The main
objective was to use results-based planning techniques to increase coverage with three packages of
high-impact interventions known to reduce child mortality (see Box 1). In Benin, Ghana, Mali and
Senegal, 16 “high-impact” districts worked to deliver all three packages; in the remaining countries, the
focus was on the “EPI+” package that included
vaccination, Vitamin A and insecticide-treated
Box 1:
nets (ITNs) for the prevention of malaria.
ACSD High-Impact
Internal UNICEF evaluations in 2003 and 2004
Implementation Packages*
showed increases in coverage for the EPI+
package in all countries; UNICEF modeled the
Immunization plus (EPI+)
associated reductions in mortality using the
 Routine immunization and periodic measles
"Marginal Budgeting for Bottlenecks " (MBB)
catch-up and mop-up
tool and estimated an overall mortality
 Vitamin A supplementation bi-annually
reduction of 20% in the “high-impact” districts in
the four countries, relative to comparison
 Distribution and promotion of Insecticide Treated
1
districts.
Nets for all children who are fully vaccinated as
well as pregnant women, and re-dipping of
bednets every six months
UNICEF and the evaluation team recognized
the limitations of a retrospective evaluation,
including the difficulties associated with
reconstructing
project
assumptions
and
activities on a post hoc basis, and making the
best possible use of available data and
information
despite
their
shortcomings.
Readers are reminded to treat the results with
2
caution.
Improved management of pneumonia, malaria
and diarrhea (IMCI+)
 Promotion of exclusive breastfeeding for six
months, timely complementary feeding

The aim of the evaluation is to provide valid
and timely evidence to child health planners
and policy makers about the effectiveness of
ACSD Phase I in reducing child mortality and
improving child nutritional status. The specific
objectives are:
Improved and integrated management (at the
health facility, community and family levels) of
children suffering from ARI, malaria and
diarrhea, including home-based ORS use,
treatment of malaria with anti-malarial blisters,
and treatment of ARI with antibiotic blisters

Promotion of household consumption of iodized
salt
Antenatal Care (ANC+)

Intermittent preventive treatment (IPT) of
malaria with SP (Fansidar) for pregnant women
mortality and nutritional status among children
under five.

Tetanus immunization during pregnancy to
prevent maternal & neonatal tetanus
2. To document the process and intermediate

Supplementation with iron/folic acid during
pregnancy and with vitamin A post-partum.
1. To evaluate the impact of ACSD on
outcomes of ACSD and results-based planning
as a basis for improved planning and
implementation of child health programs.
3. To
document the contextual factors
necessary for effective implementation of
efforts to reduce child mortality in order to be
able to extrapolate evaluation findings to other
settings.
______
*
UNICEF grouped these interventions into
paragraphs in different ways at various points
during the project; we have adopted the grouping
used in the final report from UNICEF to CIDA for
1
the ACSD project in 2005.
4. To assess the process, outcomes and impact of ACSD and results-based planning on socioeconomic, ethnic and gender inequities.
IIP-JHU | Retrospective evaluation of ACSD in Benin
1
Achievement of these objectives should also expand regional and global capacity for large-scale
effectiveness evaluations of strategies, programs and interventions designed to improve child health in
low-income countries.
1.1
Evaluation design
Geographic focus.
The global retrospective evaluation covers the four countries within which UNICEF defined selected
districts or zones as “high-impact” for the ACSD project. Within each country, the evaluation focuses on
these “high-impact” zones (HIZs).
Development of a generic impact model for ACSD.
The first step in any evaluation is to define what those implementing the project expect to happen as a
consequence of project activities. The evaluation team developed an impact model that specifies the
pathways through which UNICEF and implementing countries expected ACSD activities to result in
3
reductions in child mortality. Figure one presents the generic ACSD impact model in two parts. Figure
1A shows the “top” of the framework describing expected ACSD inputs and processes from the point of
introduction at national level in a country through the definition of the three packages of interventions
recommended for accelerated implementation (see Box one for a description of the three packages). We
e.g.,4
and discussions with ACSD implementers
derived the “top” of the framework from ACSD documents
at all levels. Figure 1B shows the “bottom” of the framework, defining the pathways through which each of
the three packages was expected to result in reductions in under-five mortality and improvements in the
nutritional status of infants and young children. ACSD documents did not describe the pathways in the
“bottom” of the model in detail, but made reference to other sources where the effects of the interventions
5,6
1
are defined and quantified. For the internal evaluation, UNICEF utilized the estimates of effectiveness
published in these sources and changes in intervention coverage as the basis for modeling the impact of
ACSD on child mortality.
A central tenet of the evaluation is that the coverage, family practices and impact reflected in the “bottom”
of the framework cannot be attributed to ACSD alone. UNICEF and country partners designed ACSD to
reinforce existing activities in child survival by the government of each country and its partners.
Therefore, increases or decreases in coverage and mortality must be understood as the result of a
combined implementation effort, tempered by contextual factors. A key challenge for the current
evaluation is to arrive at a qualitative assessment of ACSD’s role as a part of this overall effort.
Quantified attribution of the results to ACSD alone is not warranted given the implementation approach.
Definition of priority indicators for coverage and family practices.
Priority coverage indicators address the prevalence of key family practices and intervention coverage for
each of the elements defined in the “bottom” of the framework. Although some of these indicators reflect
behaviors—such as exclusive breastfeeding and complementary feeding—rather than intervention
coverage, these will be referred to as coverage indicators throughout the text. Priority indicators of
coverage utilized in the evaluation are defined in Appendix D. Whenever possible, the ACSD priority
coverage indicators are consistent with those supported by a consensus of United Nations (UN) agencies
7,8
Where no international
and multi- and bi-lateral partners for tracking progress toward MDG-4.
consensus indicator exists, the evaluation team contacted technical experts in the topical area to obtain
advice on selection of a valid coverage indicator that could be calculated using the data available in
Benin.
Definition of priority indicators of impact (nutrition and mortality).
The main objective of the ACSD project was to reduce mortality among children less than five years of
age. The primary impact indicator is the under-five mortality rate, expressed as the probability of dying
between birth and exact age five years. Additional priority indicators include neonatal and infant mortality.
Some ACSD project documents described expected improvements in child nutritional status, reflecting
2
IIP-JHU | Retrospective evaluation of ACSD in Benin
9
the synergy between undernutrition and infectious disease. Thus, priority impact indicators include
prevalence of stunting, wasting and underweight. Appendix D presents the definitions of the priority
indicators for mortality and nutritional status.
Equity.
As part of the evaluation, we examine inequity in coverage and impact indicators, including socioeconomic status, sex of the child, place of residence (urban or rural) and ethnic groups.
Documenting contextual factors.
Contextual factors are defined as variables that can confound the association between the delivery of
10
interventions and their health impact, or modify the effects of the approach. We documented contextual
indicators in the HIZs and comparison area, including: (1) indicators of implementation-related contextual
factors such as characteristics of the health system (e.g., utilization rates), child health policy, drug policy,
and availability of drugs; and (2) indicators of impact-related contextual factors including baseline levels
10
and patterns of child morbidity and mortality that can affect the potential magnitude of program impact.
Economic evaluation.
At the request of UNICEF, the evaluation does not include an economic component.
IIP-JHU | Retrospective evaluation of ACSD in Benin
3
ACSD impact model:
“Top” model showing inputs and processes
Figure 1A
Results-based approach
Selection of
effective
interventions
Choice of delivery
channels
Building upon
what exists
Establishing
partnerships
Procurement
of supplies
HW training and
supervision
(facilities)
CHW training
and supervision
IEC activities
Immunization +
Antenatal care +
IMCI +
“Bottom” model showing interventions to impact
Figure 1B
ACSD impact model:
“Bottom” model showing interventions to impact
“Top” model showing inputs and processes
Antenatal
care +
Immunization +
DPT, Hib, measles
vaccines
Pneumonia
Vitamin A
supplementation
Insecticide treated
nets
IPT for malaria
Iron/folic
acid
Tetanus
toxoid
Post-partum
Vitamin A
High attendance at facilities/outreach sessions;
deployment at community level
High attendance at facilities/outreach sessions
Increased coverage
Increased coverage
Measles
Meningitis / sepsis Diarrhea
Preterm delivery Neural tube defects Neonatal tetanus
Malaria
?????
Spillover effect (co-morbidity)
IMCI +
Improved nutrition
Reduced mortality
Malaria treatment
ORT
Reduced mortality
? Improved nutrition?
Pneumonia
treatment
Breastfeeding
promotion
Deployment of interventions at community level
Increased coverage
Malaria
Diarrhea
Pneumonia
Other infections
Spillover effect (co-morbidity)
Reduced mortality
Improved nutrition
4
IIP-JHU | Retrospective evaluation of ACSD in Benin
2.
Evaluation Methods
2.1
Evaluation design
Overall design.
The overall design was retrospective, drawing on existing population-based surveys with over-sampling in
the two “high-impact” zones (HIZs), commissioned for the purpose of this evaluation. We re-analyzed
data sets whenever possible to ensure that the indicator definitions were correct and consistent.
Preliminary results were reviewed in meetings of the evaluation team with representatives of the Benin
Ministry of Health, the National Statistics Unit and the UNICEF country office in Cotonou, Benin in August
2007, Dakar, Senegal in October 2007 and Bamako, Mali in June 2008.
Coverage and family practice indicators.
We reanalyzed existing household surveys to calculate the ACSD priority coverage and family practice
indicators. As described above, these indicators are consistent with those used internationally for
7,8
monitoring progress toward the Millennium Development Goals and are presented in Appendix D.
Appendix E provides the specific survey questions utilized for indicator calculations.
Nutrition and mortality indicators.
We reanalyzed existing household surveys to calculate the priority nutrition indicators using the 2006
11
Appendix L and section 6 present more details on these methods. For
WHO Growth Standards.
calculation of priority mortality indicators, the evaluation team analyzed mortality retrospectively, using
direct child mortality estimates based on full birth histories.
Intervention areas.
The two health zones selected for ACSD “high-impact” implementation served as the intervention areas.
They are: 1) Djidja/Abomey/Agbangniizoun (DAA) in the Zou department and 2) Pobè/Adja-Ouere/Kétou
(PAK) in the Plateau department. Throughout the body of the report we refer to these zones as “highimpact” zones (HIZs).
Comparison groups.
The main comparison group is the remainder of Benin excluding Cotonou. We have excluded Cotonou
because access to services and living conditions in this metropolitan area differs considerably from the
mostly rural HIZs.
Intervention activities.
We documented the timing and scale of intervention activities using information collected from key
informant interviews and document review, such as administrative and supervision reports and monitoring
data. (See Appendix B for details).
Equity.
To examine inequities, we performed analyses of selected intervention coverage indicators and impact
measures stratified by sub-groups of the population, including household assets (expressed in quintiles),
sex of the child, place of residence (urban/rural) and ethnic group.
IIP-JHU | Retrospective evaluation of ACSD in Benin
5
Contextual factors.
We collected standard information on contextual factors, defined above, in order to assist in interpretation
of the results and the potential contributions of ACSD. Certain elements, such as economic status,
ethnicity and access to clean water were reanalyzed for HIZs and comparison area using existing
household survey data. Field visits to the HIZs, key informant interviews and document review provided
contextual information not available in existing surveys. Appendix B provides further details on the
methods used to collect contextual factors.
2.2
Data sources and methods
Tables 1a and 1b summarize the different types of information used in the evaluation. The primary data
sources for estimates of intervention coverage and nutritional status were the demographic and health
surveys (DHS) conducted in 2001 and 2006, and a supplemental survey linked to the 2006 DHS that
over-sampled in the two ACSD “high-impact” zones (table 1a). These surveys used highly comparable
methodologies with data quality controlled by Macro International, although the sample size in 2001 was
much smaller than that in 2006. We used the DHS 2006-7 to estimate child mortality both before and
after ACSD implementation. The full-birth history method used to collect mortality data allows the
calculation of period estimates of mortality ranging from the previous 12 months to 10 or more years in
the past. The oversampling of the DHS 2006-7 allowed for more precise child mortality estimates.
Section 7 describes the mortality analysis methods in detail.
Other survey data were available, but given lesser prominence in the analyses, because they did not fully
meet the quality criteria established for the evaluation. These criteria were: 1) full data sets and
documentation, including sampling weights, available to the evaluation team so that the data could be
reanalyzed using the standard definitions for priority indicators; and 2) no more than 5% missing values
on key socio-demographic variables (e.g., child age) or the variables needed to construct the priority
indicators. We did not use the data from the Benin Cooperation Coverage 2005 and CDC-ACSD 2003
surveys in the primary analyses because they did not fulfill these criteria; data from the 2003 and 2005
surveys are presented to explore time trends between 2001 and 2006-7. Appendix F provides
descriptions of the methodology and conduct of all surveys presented in the evaluation, as well as
overviews of other data not utilized in the evaluation. Full documentation of 2003 ACSD-CDC survey
data quality issues is available upon request from JHU evaluation team.
Table 1b presents sources of information used in the documentation of intervention activities and
contextual factors. We collected information through: 1) key informant interviews; 2) review of
documents, including administrative and monitoring reports; and 3) searches and review of published and
grey literature. Available information regarding ACSD expenditures from 2002 to 2004 was utilized to
confirm the documentation of ACSD implementation. Technical staff at UNICEF-Benin provided input
and revisions throughout the process of documentation.
Due to the retrospective nature of the evaluation, it was difficult to collect complete and standardized
information on ACSD implementation and other health activities in the HIZs. Furthermore, the
collaborative nature of ACSD made it difficult to distinguish which activities were: 1) carried out as part of
the ACSD program; 2) carried out with only partial technical and/or financial support from the ACSD
program; or 3) carried out by ACSD partners, but independent of the ACSD program. Documents
providing full descriptions of the ACSD activities were not available for all activities; in many instances, we
relied on summative reports and presentations for this information. Sometimes, although not often,
information in one document conflicted with information found in other sources.
6
IIP-JHU | Retrospective evaluation of ACSD in Benin
Table 1a: Data sources for the independent retrospective evaluation of ACSD in Benin - populationbased surveys.
TYPE OF DATA
DESCRIPTION
USE IN EVALUATION
Population-based
surveys that met
inclusion criteria
DHS 2001: Nationally
representative household survey
conducted from June to December
12
2001.
Used to establish baseline
levels of priority coverage and
nutrition indicators
DHS 2006: Nationally
representative household survey
conducted from August to
13
December 2006.
Used to estimate coverage,
nutrition and mortality
indicators in comparison area
in 2006.
Supplemental survey in DAA and
PAK zones: Additional 1540
households in HIZs surveyed using
DHS methods in May 2007.
Used in combination with data
collected in DAA and PAK
zones by DHS 2006 to
estimate coverage, nutrition
and mortality indicators in HIZs
CDC-ACSD 2003: Household
survey of 2610 households in HIZs
carried out from August to
14
September 2003.
Reported, but given limited
weight in analysis due to
concerns about data quality.
Benin Cooperation Coverage
Survey 2005: Household survey in
Ouémé-Plateau & Zou-Collines
departments, including 1097
households in HIZs, carried out in
15
January 2005.
Reported, but given limited
weight in analysis due to
concerns about comparability.
Other populationbased surveys
IIP-JHU | Retrospective evaluation of ACSD in Benin
7
Table 1b: Data sources for the independent retrospective evaluation of ACSD in Benin - routine data,
administrative reports and key informant interviews.
TYPE OF DATA
DESCRIPTION
USE IN EVALUATION
Routine health
information system
data
Routine data collected through health facilities
pertaining to intervention coverage, compiled at
local, regional and national levels
Documentation of MOH
and ACSD activities
Administrative
reports
Training and workshop reports: Over 35
summative reports pertaining to training of
trainers, training of health providers and
community workers, and workshops to develop
strategies, materials and capacity.
Documentation of
ACSD and partners’
activities.
Supervision and monitoring reports: Over 20
summative reports describing supervision and
monitoring activities and findings.
Administrative and routine activity reports: Over
50 documents pertaining to ACSD and MOH
planning and activities, including notes from
routine meetings, ACSD consultant activity
reports, outbreak investigations, etc.
Compiled ACSD expenditure data 2002-2004:
List of UNICEF-Benin ACSD expenditures
2002-2004 compiled as part of the internal
evaluation.
8
Job aids and tools
Job aids and tools, such as visual aids and
register books, used in the implementation of
ACSD were collected and reviewed where
possible
Documentation of
ACSD and partners’
activities.
Summary reports &
presentations
Over 30 reports and presentations compiled by
UNICEF and partners summarizing the
activities, results and challenges of ACSD and
other child survival activities
Documentation of
ACSD activities.
Survey reports,
maps & other
documents
Over 30 survey reports, maps, and other
documents pertaining to contextual factors (e.g.,
16
2002 Census report ) and child survival
activities in Benin were collected during field
visits and through literature searches
Documentation of
contextual factors.
Key informant
interviews
Over 15 interviews and focus groups during
field visits conducted with key informants in
PAK, DAA and Cotonou during field visits and
missions; see list of informants in Appendix B.
Documentation of
ACSD activities and
contextual factors.
Working
discussions
Collaborative discussions in Cotonou (Aug
2007), Dakar (Oct 2007), and Bamako (June
2008) to review preliminary results and refine
analyses with UNICEF-Benin staff, MOH
officials, and Benin statistical agency (INSAE)
staff.
Discussion and
documentation of
ACSD activities and
contextual factors;
interpretation of results.
IIP-JHU | Retrospective evaluation of ACSD in Benin
2.3
Analysis
17
We have employed the Habicht et al. framework for “real-life” evaluations. Starting with an adequacy
evaluation, we assessed whether trends in coverage, nutrition and mortality indicators moved in the
expected direction within the ACSD areas, and whether goals were met. Next, we carried out a
plausibility evaluation, in this case a controlled, non-randomized study that assessed whether observed
impact could be attributed to program implementation. ACSD in Benin was a combination of separate
interventions – vaccines, mosquito nets, vitamin A supplementation, etc. – that are highly efficacious if
delivered at optimal coverage. This evaluation does not assess the efficacy of these interventions, but
instead focuses on their impact when delivered under routine conditions. We worked with national
i
counterparts to conduct the analysis of coverage and nutrition in four steps. Each step is explained
below.
Step 1: Generating indicator levels for each survey in the analysis.
Objective: To describe levels of priority indicators for coverage and nutrition in all surveys included in the
analysis, overall and for specific subsets of children defined by age, sex, geographic location of the
household, mothers’ education and socioeconomic status, where sample sizes permit. We applied
standard indicator definitions to the reanalysis of all datasets to ensure the comparability of indicators
over different surveys. For each indicator, only data for women and children with known responses were
included in the analyses; cases with missing or unknown data were excluded. The point estimates of
indicators presented here may therefore differ slightly from those calculated using standard DHS and
MICS tabulation programs, which do not exclude missing cases from the analysis.
Step 2: Comparing rates of change over time within each ACSD zone (“time trends”).
Objective: To determine whether there are statistically significant differences in indicator levels within
HIZs from before ACSD was implemented to after ACSD was implemented in ACSD areas, with a midpoint during the process of implementation where adequate data are available, overall and for specific
subsets of children. This step refers to the adequacy evaluation.
Step 3: Comparing rates of change between ACSD and non-ACSD zones within each country
(“time trend with comparison”).
Objective: To determine whether there were statistically significant differences in the rates of change for
indicator levels between the HIZs and a defined comparison area where ACSD was not implemented (the
comparison area is comprised of the rest of Benin, excluding Cotonou and the HIZs), overall and for
specific subsets of children.
Step 4: Attributing improvements to ACSD and related child survival activities at country level.
Objective: To determine whether any statistically significant changes in indicator levels can be attributed
to ACSD activities, including activities implemented by others in collaboration with ACSD and the national
child survival plan, overall and for specific subsets of children. Steps 3 and 4 refer to the plausibility
evaluation, assessing whether progress was greater in the HIZs than in the comparison area, and
whether or not external factors can account for these differences.
For all comparisons of coverage and prevalence of undernutrition across time and geography, we initially
calculated a simple chi-square statistic of difference. The simple chi-square statistic does not take into
account the design effect of the survey, thus it under-estimates the variance. If no statistical differences
were observed using the simple chi-square statistic, we assumed that none would be observed after the
design effect was taken into consideration (adding to the variance) and that the groups were therefore not
statistically different from one another. For comparisons with a significant chi-square, we calculated
standard errors and 95% confidence intervals that take into account the survey design effect, using the
Taylor Linearized Variance method. We used a “difference-in-differences” approach to compare whether
i
Section 7 explains the analysis of mortality in more detail.
IIP-JHU | Retrospective evaluation of ACSD in Benin
9
the change in each indicator over time differed significantly between the HIZs and comparison area for
the plausibility analyses.
10
IIP-JHU | Retrospective evaluation of ACSD in Benin
3. Characteristics of the “high-impact” zones and comparison area
This section presents pertinent characteristics of Benin as a whole and the HIZs and comparison area.
We emphasize differences between the HIZs and comparison area, as well as factors that have changed
over the evaluation period to help guide the interpretation of evaluation results. Some of the quantitative
results (table 2) presented here are based on our reanalyzes of available survey data, because these
provide the most recent information disaggregated by the HIZs and comparison area. Appendices A and
J present additional information on the geographic, socio-demographic, economic, health and health
service factors in Benin and the HIZs.
Figure 2: Map of Benin and its
3.1
The Benin context
neighbors
The Republic of Benin has an estimated population of about seven
16
million, divided into 12 departments and 77 communes, with an
estimated under-five mortality rate of 125 and an infant mortality rate
18
of 67 in the five year period preceding the 2006 DHS survey.
Benin (then called Dahomey) became a territory of France in 1946,
and declared independence in 1960. A succession of military
leaders ruled Benin until Major Mathieu Kerekou seized power in
1972 and gradually restored civilian rule. Political newcomer
Thomas Boni Yayi, former head of the Togo-based West African
Development Bank, won the second round of presidential elections
in March 2006, gaining more than 74% of the vote.
There have been no major armed conflicts involving Benin since independence. To the north, there have
been sporadic clashes along Benin's border with Burkina Faso primarily due to land disputes between
rival communities on either side of the border.
Thousands
of Togolese refugees fled to Benin in
Box 2:
2005 following political unrest in their homeland.
Overview of child health in Benin
Benin requested and received international aid to
help shelter and feed the exiles.
Causes of under-five deaths in Benin*
Pneumonia
21%
Malaria
27%
Injuries
2%
HIV/AIDS
2%
Measles
5%
Although Benin’s economy has grown over the past
few years and it is now one of Africa's largest cotton
producers, it ranks among the world's poorest
countries. The economy relies heavily on trade with
its eastern neighbor, Nigeria. Benin is a member of
the African Financial Community, and exports cotton
and palm oil. The World Bank estimated the GNI
21
Appendix A
per capita in 2006 as US$510.
provides more details about Benin and the HIZs
where ACSD was implemented.
Diarrhea
17%
Neonatal
25%
3.2
1990
Child health in Benin
2006
The population of children under age five in Benin
was estimated at about 1.4 million in 2000, and had
Under-five
185
148
22
The under-five
increased to 1.5 million by 2006.
Infant
88
mortality rate has decreased from 185 deaths per
Prevalence of undernutrition***
1000 live births in 1990 to 148 deaths per 1000 live
Stunting (% mod + severe)
44
births in 2006, a reduction of 20%. This rate of
Underweight (% mod + severe)
20
reduction falls short of that needed for Benin to
19
20
13
achieve the two-thirds reduction from 1990 levels
Sources: *WHO, 2006 ; **SOWC ; ***DHS 2006
defined by the fourth Millennium Goal (62 per 1000
live births). Box two shows the major causes of under-five deaths in Benin in 2003, as reported by
Mortality rates (per 1000 live births)**
IIP-JHU | Retrospective evaluation of ACSD in Benin
11
19
WHO. The major causes are malaria (27%), pneumonia (21%) and diarrhea (17%), with relatively low
proportions of deaths due to measles (5%) and HIV/AIDS (2%). One-quarter of all under-five deaths
occur in the neonatal period, Among these deaths, infections account for approximately one-third (34%)
with the remainder attributed to preterm births (28%), asphyxia (19%) and other causes representing less
22
than one in ten neonatal deaths. Child undernutrition is also a major problem in Benin. Estimates from
2006 using the new WHO growth standards indicate that 44% of children are either moderately or
severely stunted, 9% are wasted, and 20% are underweight. Appendix A includes the full profile of
22
maternal, newborn and child health in Benin from the Countdown to 2015: 2008 report.
3.3
Selection of the ACSD “high-impact” zones in Benin
Two health zones (zones sanitaires), equivalent to health districts in other countries, were selected for
“high-impact” implementation of ACSD: Djidja/Abomey/Agbangniizoun (DAA) health zone in the Zou
department and Pobè/Adja-Ouere/Kétou (PAK) health zone in the Plateau department. Each of these
health zones is comprised of
Figure 3: Map of Benin showing “high-impact” zones and geographic
three communes. Thus, the
comparison area.
“high-impact” zones (HIZs)
cover six communes out of
77 communes at the
national level in Benin and
in 2004 represented an
estimated 515,515 people
or
7%
of
the
total
23
population
of
Benin.
Figure 3 presents the HIZs
and comparison area.
UNICEF-Benin reports that
the two HIZs were selected
due to high levels of underfive mortality and poor
access to health services.
UNICEF
was
already
supporting activities in two
communes (Kétou and
Djidja) in these zones,
which
was
another
consideration.
12
IIP-JHU | Retrospective evaluation of ACSD in Benin
3.4
Socio-economic and demographic factors
Figure 4 presents the incidence of non-monetary poverty in the country as a whole and in the
24
departments which include the HIZs, as measured in Benin in the 2002 census. Within the department
of Zou, Djidja and Agbangniizoun are among the poorest communes in the country, while the more
urbanized Abomey is much wealthier (shown in the call-out box). The Plateau department is better off
compared to the rest of the country; however, the PAK zone is poorer than the department as a whole,
with Adja-Ouere commune among the poorest communes in the country.
Figure 4: Map of non-monetary poverty in Benin showing “high-impact” zones and geographic
comparison area.
24
Source: Vodounou et al. Carte de pauvreté non monétaire au Benin.2006.
Table 2 presents selected characteristics of these areas at two points in time: 1) in 2001, before the
implementation of the ACSD project had begun; and 2) in 2006-7, after the implementation of the project.
Additional details on the characteristics of the HIZs are available in Appendix A. Appendix table D2
presents the definitions and calculation of selected contextual variables presented in table 2.
The results show differences between the HIZs and the geographic comparison area that need to be
considered in the analysis and interpretation of the results. Households in the HIZs tended to be less
poor than in the comparison area based on their household assets, although this difference was not
statistically significant. A higher proportion of households in the HIZs were of the Fon ethnic group. A
secondary analysis of the characteristics of the Fon relative to the other ethnic groups in the 2001 sample
showed that they are significantly more likely to fall into the least poor quintile of the population. This is
consistent with the overall pattern of results presented in Table two, suggesting that households in the
HIZs had somewhat more resources than those in the comparison area. Education and literacy among
women was similarly low in both the HIZs and comparison area. Approximately one-third of women
reported any education in 2001 and 2006-7 and less than one-fourth of women reported literacy during
this same period.
IIP-JHU | Retrospective evaluation of ACSD in Benin
13
The HIZs and comparison area have similar proportions of rural residences, with about 70% of
households rural and 30% urban (table 2). Households in the HIZs tend to have better water supplies
(p=0.05) and sanitation facilities (p>0.05), which is likely due to Guinea worm projects that have focused
on improved water sources and hygiene in these areas.
Table 2: Selected characteristics of “high-impact” zones and comparison area (all other areas of the
country excluding the HIZs and Cotonou) as measured in the DHS 2001 and the DHS 2006-7 and
supplemental surveys, Benin.
2001 DHS
INDICATORS
HIGH IMPACT
ZONES
n*
%
2006/7 DHS
COMPARISON
AREA¥
n*
%
P
HIGH IMPACT
ZONES
n*
%
COMPARISON
AREA¥
n*
%
P
Ethnicity
Adja
Fon
Yoa & Lokpa
Yoruba
Other
Wealth quintiles
Poorest
Poorer
Poor
Less poor
Least poor
Education among
women
None
Primary school
Secondary school+
Literacy among women
Hygiene **
Improved water source
Improved sanitation
463
422
463
3.2
74.9
0.0
20.7
1.2
13.4
22.9
17.2
21.5
25.1
66.9
22.2
4844
4624
4844
11.0
15.1
40.2
4.5
11.4
28.9
21.7
21.2
19.1
18.3
19.7
70.7
19.8
<0.01
2702
>0.10
2943
>0.10
2731
9.6
2.1
63.7
0.1
33.1
1.1
24.0
18.4
20.3
20.3
17.0
70.9
18.7
14116
14060
14633
10.4
19.2
41.3
3.9
10.0
25.7
20.5
18.9
19.5
19.6
21.7
68.3
18.3
<0.01
>0.10
>0.10
13.4
458
20.7
4811
19.3
>0.10
2715
21.4
14064
24.4
>0.10
422
81.6
4623
59.5
0.05
2943
75.2
14060
65.6
<0.01
418
10.3
4612
7.1
>0.10
2943
14.8
14055
11.0
<0.01
422
71.0
4624
71.9
>0.10
2943
69.7
14080
66.3
>0.10
Rural residence
¥ Excluding Cotonou and High Impact districts
*Weighted
3.5
Environmental characteristics
Table 3a presents selected environmental characteristics of the HIZs and comparison area; appendix A
provide further details. The HIZs are located in the central transitional region of Benin, falling between
the tropical, wet climate of the south and the semi-arid tropical and Sahelian regions to the north. The
HIZs experience similar rainfall as the rest of Benin, and malaria transmission is seasonal eight to nine
25
months of the year, similar to the rest of the country, except in the drier northern regions. Before and
26-29
Appendix table J4
during ACSD implementation, resistance of malaria parasites to chloroquine grew;
provides full details on antimalarial resistance trends.
Table 3a: Selected environmental characteristics of “high-impact” zones, PAK and DAA, and
comparison area, Benin.
14
IIP-JHU | Retrospective evaluation of ACSD in Benin
HIZS
CHARACTERISTICS
DAA
PAK
COMPARISON AREA¥
Climate
Transitional
between tropical
wet & semi-arid
tropical
Semi-arid tropical
Tropical wet, Semi-arid tropical
Main Geographic
Characteristics
Palm plantations
Grasslands
Savanna & the
semi-deciduous
forest
Palm plantations
Grasslands; Savanna & the
semi-deciduous forest; Sahel
900-1200
1000-1400
900-1500
8-9 months/year
8-9 months/year
5-9 months/year, with shortening
transmission season in north
Annual rainfall (mm)
Months of malaria
30
transmission
¥Rest of Benin, minus Cotonou and HIZs
Source: INSAE, 2007.
Our investigations did not reveal any natural disasters, famines or other emergencies in the HIZs over the
primary evaluation period. Key informants reported that Djida commune in the DAA zone experienced
sporadic periods of instability. In 2005, the regions of Aribori and Atakora in the north of Benin did
31
experienced food insecurity, associated with the famine in neighboring Niger.
3.6
Baseline health conditions
Section 3.2 presents a profile of child health in Benin as a whole, including the cause of death profile.
Cause of death information is not available disaggregated by HIZs and comparison area. We present
and consider baseline levels of undernutrition in section six and under-five mortality in section seven.
3.7
Health service characteristics
Availability of health services.
Table 3b summarizes available information on health facilities in the HIZs and comparison area as
32
Appendix J presents more details on the coverage of
reported by the Benin Ministry of Health in 2006.
health facilities over time. On average and over time, DAA health zone had higher per capita coverage of
health facilities than the comparison area, while PAK zone has fewer health facilities than DAA or the
comparison area.
Table 3b: Health system characteristics in PAK and DAA zones and comparison area in 2006, Benin
HIZS
CHARACTERISTICS
DAA
PAK
COMPARISON AREA¥
Total health facilities
31
30
744
Average # of health centers per commune
10
10
11.5
7030
8797
7493
Private or NGO health facilities
5
2
127
Hospitals
1
1
22
Population per facility
¥Rest of Benin, minus Cotonou and HIZs
32
Source: Annuaire des statistiques sanitaires 2006
Changes in health policies.
IIP-JHU | Retrospective evaluation of ACSD in Benin
15
A number changes in national policies, implemented in both the HIZs and comparison area, influencing
child health took place between 2001 and 2006 (see Appendix J for further details), including:
•
Introduction of Hib vaccination into national policy and inclusion in routine EPI vaccination schedules
in June 2005;
•
Change in first-line antimalarial policy from chloroquine to ACTs in 2004, with no generalized
availability of ACTs until 2008;
•
Importation of polio from Nigeria to Benin, with two cases in late 2003,and six cases in early 2004
necessitating the organization of national immunization days;
•
Distribution of Vitamin A supplementation twice a year nationally, coupled with polio campaigns when
they were organized, starting in 2002;
•
Change in the national pricing policy of insecticide-treated nets (ITNs) from approximately US$7 in
2002 for all nets to US$1 for targeted populations in 2005.
3.8
33
Other projects that may impact child health
Health and development projects in the HIZs.
In addition to the ACSD program and routine government services, other health and development
projects supported activities related to child health during the period under study in the HIZs. The Benin
Malaria Control Program, local health services, Africare and CDC implemented activities to prevent and
treat malaria and other childhood illness through the African Integrated Malaria Initiative (AIMI, also
known as PROLIPO in French) with support from USAID. These activities took place from 2001 to 2005
in Ouémé and Plateau departments, which include the PAK health zone. Other projects taking place in
2002 to 2007 focused on clinical improvements, especially in maternal and neonatal health, including
support and care for AIDS orphans and vulnerable children, as well as prevention of mother-to-child
transmission of HIV (PMTCT). Local and international NGOs in DAA also carried out nutritional
rehabilitation and education in selected communes. Appendix J provides further details about activities,
geographic coverage and timing of other health and development programs in the HIZs.
Other external partners and donors in whole of Benin.
A multitude of other donors and external partners provided support for activities in the rest of Benin, which
is our comparison area excluding Cotonou. USAID supported health activities throughout Benin, with
annual budgets for child survival and infectious disease ranging between US$3 and US$4 million in fiscal
years 2004 to 2007. The Global Fund to Fight AIDS, Tuberculosis and Malaria supported national-level
malaria control activities through a US$2.4 million grant issued in 2003, of which a large proportion was
utilized for ITN distribution and treatment activities. Other large-scale health and development partners in
the rest of Benin included various local and international NGOs, WHO, UNFPA, France, Denmark,
Germany, Belgium, The Netherlands, Canada, Switzerland, the World Bank, the European Union, and the
African Development Fund.
16
IIP-JHU | Retrospective evaluation of ACSD in Benin
4. ACSD as implemented in Benin
This section provides an overview of the ACSD activities in the HIZs. Funding, adaptation of the generic
ACSD package, and the timeline of activities are considered; the activities and inputs for each ACSD
component are then briefly described. Appendix C includes further textual description on implementation
and detailed timelines for ACSD activities.
4.1.
Funding
UNICEF headquarters transferred US$1.5 million to UNICEF-Benin in March 2002 to support ACSD
34
activities in Benin. This was about one year later than the transfer of funds to the other three “highimpact” countries. UNICEF staff report that ACSD activities have been co-funded and sustained since
2003 through support from Belgium, UNICEF-Benin general resources, and general program funds.
4.2.
Adaptation of the generic ACSD intervention package and approach
The generic ACSD strategy was adapted at the national level, starting with a planning workshop held in
Bohicon in January 2002. In the Abomey commune, community IMCI (C-IMCI) communication messages
and materials were adapted for use in the DAA zone at a two-week workshop held with community
leaders—such as teachers, community health workers (CHWs), and health staff—in November 2002.
Participatory community situation assessments and communication plan development workshops utilized
materials developed for use in all ACSD countries. A similar workshop took place in February 2003 in
34
Pobè to adapt C-IMCI materials for use in the PAK zone.
4.3
Results-based planning
ACSD implementers at the international level chose the package of interventions to be implemented in
the four “high-impact” countries based on evidence and cost-effectiveness, using the Marginal Budgeting
for Bottlenecks (MBB) tool. The ACSD strategy set specific targets for each package and UNICEF
monitored results at the zone and commune levels. We did not find evidence that ACSD in Benin
included performance contracts or other innovations linking results to specific incentives.
4.4
Timeline for ACSD activities in the “high-impact” zones
UNICEF officially launched the ACSD project in February 2002 in collaboration with the Government of
Benin. A national steering committee, put in place by the Ministry of Health (MOH) in early 2002, held its
34
first meeting in April 2002. The remainder of 2002 was devoted to strategic planning, development of
tools and materials, and logistic preparations with the MOH and partners. During this time, UNICEF hired
two consultants to assist with implementation of ACSD in each health zone. Implementation of most
ACSD-supported activities began in 2003. Figure 5 presents a summary timeline for the acceleration or
implementation of the ACSD intervention packages, as well as for household surveys conducted in the
two HIZs. These timelines are based on information obtained from key informants and document
reviews. Table 4 provides additional information about the timing of specific activities. Appendix C
includes more details about the timing and content of ACSD activities.
IIP-JHU | Retrospective evaluation of ACSD in Benin
17
Figure 5: Timeline for implementation of ACSD packages of interventions and surveys conducted to
evaluate intervention coverage, 2001–2007.
Surveys
Activities
EPI+
ITN
IPTp
IMCI
Facility
Facility Community
Community
May 01
2001
2002
DHS
Nov 02
Dec 02
2003
July 03
July 03
Jan 04
2004
ACSD-CDC
2005
No ITNs Available
2006
2007
Figure 5 Key:
18
DHS
Supplem ental
Survey
First bar represents PAK zone and second bar represents DAA zone
Grey bars represent implementation before “acceleration” through ACSD
IIP-JHU | Retrospective evaluation of ACSD in Benin
Table 4: Start times for accelerated implementation of ACSD intervention packages in the ACSD
“high-impact” zones of Benin.
INTERVENTION PACKAGE
~ START IN PAK
~ START IN DAA
EPI
Routine EPI on-going
prior to ACSD;
th
acceleration in 4 quarter
2002
Routine EPI on-going
prior to ACSD;
th
acceleration in 4 quarter
2002
ITNs
Late 2002 – Facility
July 2003 - Community
November 2002-Facility
July 2003 - Community
October 2003 (Training of
Trainers)
October 2003 (Training of
Trainers)
June 2001
December 2002
IPTp
*
Facility IMCI
rd
rd
C-IMCI: management of illness, EPI
defaulter tracing and ITN re-treatment,
promotion of exclusive breastfeeding,
complementary feeding, vitamin A, etc.
3 quarter 2003
3 quarter 2003
Mid 2004
Mid 2004
Strengthening of nutrition & feeding
elements
Mid 2005
Mid 2005
ITN stock-outs
3 quarter 2005 to mid 2007
rd
*
IPTp began in PAK and DAA with the training of trainers in October 2003, followed immediately by training of
service providers. Administration of SP to pregnant women presenting to health facilities began in 2004.
4.5
ACSD activities in the “high-impact” zones
EPI+.
Vaccination and vitamin A supplementation
Vaccination activities and the provision of vitamin A supplements to children 6-59 months of age were
already in place through the MOH system prior to the introduction of ACSD. The health system in Benin
delivers vaccines to children in three principal ways: 1) routine, facility-based vaccination; 2) routine
outreach activities; and 3) vaccination campaigns. Routine facility and outreach vaccination activities
were supported throughout the evaluation period. Measles campaigns took place in early 2003 and late
33
2005. After importation of wild poliovirus from Nigeria was detected in Benin in 2003, the quality and
quantity of national-level supplemental immunization days (NIDs) for polio were reinforced. Primarily, the
NIDs and other isolated campaigns delivered vitamin A twice a year, starting in 2002 with support from
UNICEF.
Table 5 summarizes available information about ACSD inputs intended to reinforce EPI+ activities
extracted from administrative and summary reports. Appendix table C2 presents further description of
EPI+ activities and details on exact timing.
IIP-JHU | Retrospective evaluation of ACSD in Benin
19
Table 5: Description of inputs related to vaccination and vitamin A supplementation in the ACSD “highimpact” zones of Benin.
DESCRIPTION OF
ACTIVITY
Provision of basic
medical materials,
refrigerators for cold
chain, computers for
monitoring and data
collection activities and
commodities
Provision of new
motorcycles for outreach
activities
Provision of 4x4 vehicle
for supervision
Training of facility-based
health workers in EPI
provision, monitoring and
surveillance
Support for NIDs, local
vaccination campaigns,
catch-up vaccination,
supervision and
monitoring activities
Provision of vitamin A
capsules
TIMING
GEOGRAPHIC
AREA
INTENSITY
OF ACTIVITY
COVERAGE
ESTIMATE
2002 - 2003
PAK, DAA and
expansion area of
Zou-Collines and
Ouémé-Plateau
Quantitative data incomplete
2002
DAA
4 motorcycles
4 motorcycles for 29
arrondissements
2002 - 2003
PAK
15
motorcycles
15 motorcycles for 17
arrondissements
2002
Ouémé
department
4 vehicles
DAA
57 health
workers
trained
PAK
Quantitative data incomplete
2004 - 2005
DAA
84 health
workers
trained
2002 - 2006
PAK and DAA
Other Quantitative data incomplete
2004
DAA
4 rounds catch-up EPI & 2 tetanus
campaigns;
2002 - 2006
PAK, DAA, ZouCollines, OuéméPlateau, nationally
Detailed quantitative data presented in
appendix C
2002
1.3 trained workers per
1000 children U5*
1.8 trained workers per
1000 children U5*
*Population of children under-five extracted from 2002 census, Cahier des villages; (population estimated at 56,609
children U5 in PAK and 45,624 children U5 in DAA)
20
IIP-JHU | Retrospective evaluation of ACSD in Benin
Insecticide-treated nets (ITNs).
ACSD in Benin utilized different strategies for the provision and promotion of utilization of ITNs in OuéméPlateau and Zou-Collines, including the PAK and DAA health zones, respectively. In the Ouémé-Plateau
departments, including the PAK zone, bednets were sold and their use promoted by women’s groups and
in maternity centers through an agreement with Africare, an international NGO. The CHWs promoted,
distributed and re-treated ITNs starting in mid-2003. In the departments of Zou-Collines, including the
DAA health zone, bednets were sold through social marketing techniques at town markets, as well as at
health centers; they were also sold and promoted by CHWs in villages. Population Services International
(PSI) was a key partner and implementer of this activity. Mosquito nets were retreated through periodic
community-based treatment campaigns in the four departments, including the PAK and DAA zones.
UNICEF and other partners, such as the national malaria control program (PNLP) supported provision of
insecticide treatment at no cost.
Table 6 summarizes available information about ACSD contributions to the promotion, distribution and
treatment of mosquito nets extracted from administrative and summary reports. To provide rough
guidance on the potential coverage of these activities, we present several of the indicators as ratios per
1,000 children under-five years, even though it is recognized that the ITNs were also targeted to pregnant
women and may have been used by non-targeted members of the population. As a result, the coverage
estimates below are likely overestimated. Appendix table C3 presents more details about ITN distribution,
promotion and treatment.
IIP-JHU | Retrospective evaluation of ACSD in Benin
21
Table 6: Description of inputs related to the distribution, promotion and treatment of bednets in the ACSD
“high-impact” zones of Benin.
DESCRIPTION OF
ACTIVITY
Training of women’s
groups in ITN
promotion
Training and
deployment of
CHWs to promote,
distribute and treat
bednets
TIMING
2002
270 women
trained in 90
villages
5.5 trained women per
1000 children under 5*
PAK
200 CHWs trained
4.1 trained CHWs per
1000 children under 5*
DAA
200 CHWs trained
5.0 trained CHWs per
1000 children under 5*
DAA
1900 bednets
distributed
42 bednets per 1000
children under 5*
PAK
9765 bednets sold
at subsidized price
199 bednets sold per
1000 children under 5*
DAA
22,575 bednets
distributed
495 bednets per 1000
children under 5*
8,375 bednets
distributed
184 bednets per 1000
children under 5*
7,150 bednets
distributed
157 bednets per 1000
children under 5*
22,575 bednets
distributed
93 bednets per 1000
children under 5*
2003
2005
DAA
2006
2004 –
2006
COVERAGE ESTIMATE
PAK
2004
ii
PAK
Quantitative data incomplete
PAK
9,330 bednets
treated
164 bednets treated per
1000 children under 5*
2,892 bednets
treated
10,883 bednets
treated
65 bednets treated per
1000 children under 5*
192 bednets treated per
1000 children under 5*
DAA
4,509 bednets
treated
101 bednets treated per
1000 children under 5*
2004
PAK
9,295 bednets
treated
164 bednets treated per
1000 children under 5*
2005
PAK
19,795 bednets
treated
350 bednets treated per
1000 children under 5*
2005
DAA
55,378 bednets
treated
1213 bednets treated per
1000 children under 5*
Early
2003
DAA
Community
campaigns for
treatment of bednets
INTENSITY OF
ACTIVITY
2003
2002
Distribution of
bednets
GEOGRAPHIC
AREA
Late
2003
PAK
*Population of children under-five extracted from 2002 census, Cahier des villages; (population estimated at 56,609
children U5 in PAK and 45,624 children U5 in DAA)
ii
Summary presentations given by UNICEF staff indicate that 44,250 ITNs were distributed at a subsided price
between 2002 and 2006 in the DAA health zone.
22
IIP-JHU | Retrospective evaluation of ACSD in Benin
IMCI+.
Integrated case management of child illness (IMCI) and promotion of improved feeding practices were
carried out in both facilities and the community through the ACSD strategy. Standard 11-day training of
facility-based health workers had already taken place in the PAK zone in June 2001. UNICEF and ACSD
supported the MOH to carry out IMCI training in DAA in late 2002. Periodic supervisory visits and review
of monitoring data collected from IMCI-compatible health registers were carried out to reinforce IMCI
implementation. In addition to IMCI, facility-based health workers in PAK and DAA received training in
“healthy child consultations” and the minimum package of nutrition activities, which focused on
micronutrient supplementation and infant feeding practices.
The PAK and DAA health zones were pilot zones for community IMCI in Benin. As explained above,
UNICEF supported a series of workshops and community situation analysis exercises with MOH officials,
local health zone staff and community members and leaders to assist in planning of community IMCI
strategies. In mid-2003, 404 community health workers (CHWs) were selected in 202 remote villages in
PAK and DAA (102 in PAK and 100 in DAA), chosen by local leaders. CHWs and community leaders
received a series of initial and refresher trainings addressing:
•
Vaccination promotion and defaulter tracing;
•
Treatment of bed nets;
•
Home management of malaria and diarrhea and medicine management;
•
Promotion of infant feeding practices, vitamin A and appropriate management of child illness;
•
Hygiene; and
•
Recording of data, including birth registration.
CHWs were issued bednets and a medicine box with ORS, chloroquine, paracetamol, mebendazole, and
iron in 2003 and visual aids for health promotion in 2005. The medications were to be sold to the families
of sick children in the villages at reasonable prices; a small margin of benefit for the CHWs and
community committees was planned. In 2004, a UNICEF-supported operational research project trained
40 CHWs in the Kétou commune in PAK and 40 CHWs in the Djidja commune in DAA to manage
pneumonia with cotrimoxazole at the community level.
Table 7 summarizes available information about ACSD contributions to facility and community IMCI
extracted from administrative and summary reports. Again, we present selected indicators as ratios per
1,000 children less than five years of age to better assess potential coverage. Appendix table C4 further
describes IMCI+ activities and exact timing.
IIP-JHU | Retrospective evaluation of ACSD in Benin
23
Table 7: Description of inputs related to the implementation of the IMCI+ intervention package in the
ACSD “high-impact” zones of Benin.
DESCRIPTION OF
ACTIVITY
Standard facility IMCI
training
On-going IMCI
supervision and
monitoring
“Health child visit” /
“minimum package of
nutrition” training
Training of CHWs and
community leaders in
promotion of EPI, ITNs,
infant feeding practices,
hygiene and correct
management of ARI,
malaria, diarrhea §
TIMING
GEOGRAPHIC
AREA
COVERAGE
ESTIMATE
2003
DAA
24 health workers trained; denominator of
health workers not available
2004
PAK
7 health workers trained for a cumulative total
of 38 health workers out of 42
20032006
PAK and DAA
4 supervisions per year
2003
PAK
44 health workers
trained
2004
DAA
184 health workers
trained
0.8 health worker per
1000 children under
5*
4.0 health worker per
1000 children under
5*
200 CHWs trained
3.5 CHWs per 1000
children under 5*
40 arrondissement
leaders trained
40 leaders of 29
arrondissements
200 CHWs trained
4.5 CHWs per 1000
children under 5*
20 arrondissement
leaders trained
20 leaders of 17
arrondissements
PAK
2003
DAA
Deployment of village
drug kits (chloroquine,
antipyretics,
deworming, ORS, and
iron) managed by
CHWs
INTENSITY OF
ACTIVITY
2003
PAK
102 villages with
drug kits
102 villages / 128
villages in PAK
2004
DAA
100 villages with
drug kits
100 villages / 155
villages in DAA
Kétou (PAK)
40 CHWs trained
Djidja (DAA)
39 CHWs trained
PAK and DAA
Throughout ACSD
period, at least yearly
Training of CHWs for
community pneumonia
management using
cotrimoxazole
2005
Supervision and
monitoring of CHW
activities through field
visits and meetings
20032005
1.8 CHWs per 1000
children under 5* in
Kétou
2.0 CHWs per 1000
children under 5* in
Djidja
See Appendix table
C4 for details and
timing
*Population of children under-five extracted from 2002 census, Cahier des villages
§ Visual Aids for promotion of correct illness management and feeding practices were distributed to CHWs in 2005
24
IIP-JHU | Retrospective evaluation of ACSD in Benin
ANC+.
Antenatal care interventions supported under the ACSD approach of “Focused ANC+” in Benin included:
1) focused antenatal care iii; 2) utilization of ITNs; 3) intermittent preventive treatment for malaria for
pregnant women (IPTp) with a combination of sulfadoxine-pyrimethamine (SP); 4) prevention of motherto-child transmission of HIV (PMTCT); 5) deworming; and 6) supplementation in iron and folic-acid. In
late 2005, an “ANC kit” was introduced that contained a bednet, iron/folic acid supplements, SP for IPT of
iv
malaria, and mebendazole for de-worming. In the DAA zone, this “ANC kit” was provided in a special
sachet to facilitate distribution; the different elements are sold separately to pregnant women in PAK. A
radio communication system was put into operation with UNICEF support in the DAA zone in 2004 to
facilitate evacuation of obstetrical emergencies. UNICEF reported 95% coverage of health facilities and
maternities in 2006. Full installation of the radio system was complete in PAK in 2004.
Table 8 summarizes available information about ACSD inputs related to ANC activities extracted from
administrative and summary reports. Appendix C and table C5 provide further description of ANC+
activities and timing.
iii
“Focused ANC” reorients ANC care to treat all pregnancies as “at risk”. Starting in the first ANC visit, this strategy
is intended to encourage: 1) women to plan for the delivery; 2) planning logistically and financially for evacuation in
the case of complications; and 3) husbands to assist at least one ANC visit to help with this planning. Increasing the
decision power of pregnant women is at the heart of this strategy.
iv
The ANC kit is sold for CFA 1000 (~ USD 2.00) to a woman at her first antenatal visit to a community health center
or maternity. According to health officials, the total cost of the kit contents to the health center is more than the price
of CFA 1000. Health centers lose money with the sale of kits at the subsidized price, and cannot recover their costs.
Thus, strong disincentives exist to promote or make kits available and issues remain with re-supplying the
commodities included in the kits.
IIP-JHU | Retrospective evaluation of ACSD in Benin
25
Table 8: Description of inputs related to the implementation of the ANC+ intervention package in the
ACSD “high-impact” zones of Benin.
DESCRIPTION OF
ACTIVITY
Provision of medical kits and
supplies
Introduction of IPTp and
focused ANC
Training of maternity
personnel in management of
neonate ( 2 sessions)
Training of nurses and
midwives in Active
Management of the Third
Stage of Labor (2 sessions)
Training of Drs, midwives,
nurses, lab techs and social
workers in PMTCT
Supervision and monitoring
activities for ANC activities,
including IPTp, PMTCT,
neonatal care
26
TIMING
GEOGRAPHIC
AREA
INTENSITY OF
ACTIVITY
2002
PAK and DAA
All health centers equipped with
basic kit
2004
PAK and DAA
Kits for newborn care
PAK
Quantitative data incomplete
DAA
37 nurses/
midwives trained
PAK
Quantitative data incomplete
DAA
32 nurses/
midwives trained
32 trained
nurses for 21
maternities
DAA
30 nurses and
midwives trained
30 trained
nurses for 21
maternities
PAK
Quantitative data incomplete
DAA
61 personnel trained
PAK and DAA
Sporadically
throughout
ACSD period
2004
COVERAGE
ESTIMATE
37 trained
nurses for 21
maternities
2004
20042005
2005
20042006
See Appendix
table C5 for
details and
timing
IIP-JHU | Retrospective evaluation of ACSD in Benin
5. Coverage and family practices
This section of the report presents the results and interpretation of priority coverage and family practices
indicators. Section 2 describes the methodology for the analysis of coverage and family practices, with
priority indicators defined in appendix D. We present results in graphical form for selected priority
coverage indicators within each intervention package. Two graphs are presented for each package. The
first shows time-trends in indicator levels in the HIZs. We present data from surveys conducted in 2003
and 2005 in shades of grey and without confidence limits because, as explained in the methods section,
these estimates are of lesser quality and should be interpreted with caution.
The second graph for each intervention component presents indicator levels in 2001 (baseline) and 20067 (endline) in both HIZs and the comparison area. The bars in these graphs represent the 95%
confidence limits. We carried out differences-in-differences statistical tests for these comparisons and
they are presented in the text. Appendices G and H present the full results for HIZs and comparison
groups; further detail on coverage levels by zone, urban-rural residence, education of the mother and
wealth quintiles in the 2006-7 survey, as well as other descriptive tables, are presented in Appendix I.
For certain indicators and sub-populations, the results should be interpreted with caution due to the small
sample sizes for some cells.
5.1
Results
EPI+.
Vaccinations and vitamin A supplementation.
Figure 6 shows time trends in measles and DPT vaccination coverage and in vitamin A supplementation
in the ACSD HIZs. There is no evidence of an increase in measles or DPT between 2001 and 2006-7.
Coverage levels for vitamin A supplementation increased dramatically in this period; the ACSD coverage
survey results suggest that the greatest part of this increase occurred between 2001 and 2003.
Not shown in graphical form are results on coverage of Hib vaccine. Hib vaccine was introduced to
children less than one year of age in June 2005. The 2006-7 survey results indicate that only 17% of
children aged 12-23 months had been vaccinated with the recommended three doses of Hib by their first
birthday in 2006-7, but coverage increases to 25% when examined among the cohort of children born
after June 2005, when the vaccination was officially introduced. Trends in EPI+ coverage were similar in
PAK and DAA health zones (Appendix table G2).
IIP-JHU | Retrospective evaluation of ACSD in Benin
27
Figure 6: Coverage levels for measles and DPT3 vaccination and the receipt of one vitamin A
supplement in the preceding six months in the ACSD “high-impact” zones as measured in DHS
surveys in 2001 and 2006-7, as well as coverage reported in the CDC-ACSD survey in 2003,
Benin.
100
DHS 2001
ACSD 2003
DHS 2006-7
80
Coverage (%)
63
51
45
49
61
60
60
54*
51
40
20
10
0
Measles
DPT3
Vitamin A
*Vitamin A coverage data available only for children aged 6-32 months in the 2003 ACSD survey
Note: Measles and DPT3 indicators are calculated based on MICS protocol, where the distribution of
children with card confirmed vaccination before 12 months is applied to all other children reported
as vaccinated.
Appendix tables I2 and I3 provide further information on coverage levels for vaccinations and vitamin A
supplementation in the HIZs in 2006-7. Vaccine coverage tended to be higher in DAA than PAK, with a
slightly greater differential in DPT3 coverage (p<0.001) than for measles vaccination (p=0.03).
Vaccination coverage also tended to be slightly higher in urban areas, although this difference was not
statistically significant. Children of more educated mothers were more likely to have been vaccinated for
measles (p=0.04) and to have received three doses of DPT (p=0.01) before their first birthday. A greater
proportion of boys than girls were vaccinated against DPT3 (p=0.05). Children residing in the poorest
quintile of households were significantly less likely to be vaccinated for measles (p<0.001) or DPT3
(p<0.001) than children residing in less poor households. For example, reported measles vaccination
was 29% in the poorest quintile and 65% in the least poor quintile; coverage for DPT3 was 37% and 72%
for these two quintiles, respectively.
For vitamin A coverage, again DAA performed better than PAK (p<0.001), and higher levels of maternal
education were associated with higher levels of coverage (p<0.001). Despite delivery through
campaigns, often thought to promote equity, coverage for vitamin A supplementation was relatively
inequitable in the HIZs (p<0.001): it reached only 45% of children 6-59 months in the poorest quintile of
households compared to 70% among the least poor. There were no urban/rural or sex differentials in
vitamin A supplementation.
28
IIP-JHU | Retrospective evaluation of ACSD in Benin
Figure 7: Coverage levels for measles and DPT3 vaccination and receipt of one vitamin
A supplement in the preceding six months in ACSD “high-impact” zones and comparison
area as measured in DHS surveys in 2001 and 2006-7, Benin.
Absolute change in percentage points
+6
-3
-3
-2
100
High-impact zones
Comparison area
+46
+51
Coverage (%)
80
60
40
20
0
51 54
49
60
63 67
60 64
61 63
10
2001
2006-7
Measles
2001
DPT3
2006-7
17
2001
Vitamin A
2006-7
Figure 7 shows coverage levels for vaccinations and vitamin A supplementation in the HIZs and the
comparison area in 2001 and 2006-7. Measles vaccine coverage remained stable between 2001 and
2006-7 in HIZs, but there was a non-significant increase in comparison area. The difference between
time trends in the HIZs and the comparison area was not statistically significant. DPT3 coverage
decreased slightly in both the HIZs and the comparison area; neither these changes nor the difference-indifferences between the HIZs and the comparison area were statistically significant over time. Vitamin A
supplementation increased by 51 percentage points (pp) in the HIZs and 46 pp in the comparison area.
Although increases over time in both the HIZs and comparison area were statistically significant
(p<0.001), the difference between the two rates of increase was not statistically significant.
IIP-JHU | Retrospective evaluation of ACSD in Benin
29
Insecticide-treated nets (ITNs).
Figure 8 shows time trends in the use of ITNs in the HIZs. There were increases in the proportion of
children sleeping under an ITN, but only one in four children was protected in 2006-7. Results for
pregnant women are more difficult to interpret because the baseline survey did not assess whether the
nets used by pregnant women were impregnated with insecticide. There does appear to be an increase
in coverage for any net use. However, in 2006-7 only one out of every four to five pregnant woman
reported using an ITN the night before the survey.
Figure 8: Coverage levels for insecticide-treated nets in the ACSD “high-impact” zones as
measured in DHS surveys in 2001 and 2006-7, as well as coverage reported in the CDC-ACSD
survey in 2003 and a coverage survey in 2005, Benin.
100
DHS 2001
ACSD 2003
Coverage survey 2005*
DHS 2006-7
60
43
38
34
27
26
20
14
6
0
Child slept under ITN
Pregnant woman slept
under ANY net*
Data not available
31
Data not available
40
Data not available
Coverage (%)
80
22
Pregnant woman slept
under ITN
*Data concerning insecticide treatment for women’s nets not collected in 2001; any net used as a proxy NOTE:
All measurements of ITNs in ACSD survey 2003 for nets treated in previous 6 months (as compared to previous
12 months in DHS and coverage surveys)
Appendix tables I4 and I5 provide further information on coverage levels for ITNs in 2006-7 by health
zone, commune, urban/rural residence and child’s age or women’s education. Coverage among
pregnant women and young children was very similar in the DAA and PAK health zones. Levels of ITN
use among children were slightly higher in urban versus rural areas, although this trend was not
significant. Younger children also had higher proportions of reported utilization of ITNs (p=0.03) than
older children. The proportion of children sleeping under an ITN the night previous to the survey ranged
from only 14% in the lowest wealth quintile to 41% in the highest (p<0.001). Use of an ITN among
pregnant women was also positively associated with wealth (p=0.02).
30
IIP-JHU | Retrospective evaluation of ACSD in Benin
Figure 9: Coverage levels and absolute change in percentage points for insecticide-treated nets in
ACSD “high-impact” zones and comparison area as measured in DHS surveys in 2001 and 2006-7,
Benin.
High-impact zones
Comparison area
Comparison
area
~+17
Absolute change in percentage points
+23
100
+20
+11
+20
~+22
60
40
6
5
26 28
31
34
42
14
0
2001
2006-7
Child slept under ITN
2001
2006-7
Pregnant woman slept
under ANY net
~0
20
~0
Coverage (%)
80
2001
22 17
2006-7
Pregnant woman slept
under ITN
Figure 9 shows reported ITN use in the HIZs and comparison area in 2001 and 2006-7. The proportion of
children sleeping under an ITN increased by 20 pp in the comparison area and 23 in the HIZs (both
p<0.001). The rates of increase over time did not differ significantly between the HIZs and comparison
area. Women sleeping under any net —used as a proxy due to lack of data in 2001—increased by 20 pp
in the HIZs and by 11 pp in the comparison area; the difference in percentage point change was not
significant. If we assume that ITN use was zero in 2001, coverage among pregnant women by 2006-7
would have increased 22 pp in the HIZs and 17 points in comparison area (p< 0.01).
IMCI+.
The ACSD IMCI+ strategy implemented in Benin included two packages of interventions, to be delivered
by trained health workers. The first package is comprised of treatments for childhood illnesses, the
second package contains interventions designed to improve child feeding. We report coverage levels for
these interventions in this section.
Case management of childhood illnesses.
Figure 10 shows time trends in the case management of childhood fever (presumed to be malaria in this
highly endemic country), appropriate careseeking for suspected pneumonia and oral rehydration therapy
with continued feeding for diarrhea. These indicators rely on mothers’ reports for children with these
symptoms in the two weeks prior to the survey. Appendices G and I provide additional data. The results
show no improvement in the case management of childhood illness between 2001 and 2006-7, although
the power of the before and after comparison is affected by the small sample sizes at baseline. About
two-thirds of children with reported fever received an antimalarial, but over 90% of children received only
chloroquine (Appendix table I7). This drug is no longer effective in Benin, where the national policy
29
recommends ACT (see Appendix table J4 for trends in antimalarial resistance and policy). In both 2001
IIP-JHU | Retrospective evaluation of ACSD in Benin
31
and 2006-7, approximately one-third of children with probable pneumonia were taken to a health facility,
and a similar proportion of those with diarrhea received oral rehydration and continued feeding. Further
details on the management of diarrhea (Appendix table I8) and care seeking for pneumonia (Appendix
table I9) over time are provided in Appendix I, but should be interpreted with caution due to limited
sample sizes in some cells.
Figure 10: Coverage levels for case management indicators in the ACSD “high-impact” zones
as measured in DHS surveys in 2001 and 2006-7, as well as coverage reported in the CDCACSD survey in 2003 and a coverage survey in 2005, Benin.
DHS 2001
100
Proportion of ill children managed for illness (%)….
ACSD 2003
Coverage survey 2005
DHS 2006-7
80
70
60
70
67
53
46
38
40
32
31
26
43
34
30
20
0
Fever*
Careseeking for suspected
pneumonia
Diarrhea
*Fever management is with any antimalarial, regardless of policy
32
IIP-JHU | Retrospective evaluation of ACSD in Benin
Appendix table I6 provides a breakdown of case-management indicator levels estimated from the 2006-7
survey by health zone, sex, commune, urban/rural residence and child’s age and mother’s education. A
slightly higher proportion of children of more educated women received an antimalarial (p=0.08) than
children of women with less education. Pneumonia care seeking was marginally more common for boys
(p=0.13) than girls, and better in PAK than in DAA (p=0.13). Older children tended to receive better
diarrhea management (p=0.07) than younger children. Children residing in better off households were
more likely to receive any antimalarial (p=0.04) than children in poorer households. There were no
significant associations between household wealth and pneumonia care seeking or diarrhea
management, but it should be noted that the sample sizes for these indicators, within each wealth
quintile, were very small.
Figure 11: Coverage levels and absolute change in percentage points for the proportion of sick
children correctly managed at home in ACSD “high-impact” zones and comparison area as
measured in DHS surveys in 2001 and 2006-7, Benin.
High-impact zones
Comparison area
Comparison
area
-1
Absolute change in percentage
+3
-8
100
-3
-2
-4
Coverage (%)
80
60
40
70 66
20
0
67
58
32 32
2001
2006-7
Fever
30 35
2001
2006-7
Careseeking for
suspected pneumonia
38 43
2001
34 42
2006-7
Diarrhea
NOTE: Fever management is with any antimalarial, regardless of national policy
Figure 11 shows coverage levels for case management in the HIZs and the comparison area in 2001 and
2006-7. Levels of treatment with any antimalarial for fever decreased slightly in both the HIZs and the
comparison area; the difference-in-differences test was not statistically significant. However, if we define
the indicator as “treatment of fever with an effective and nationally recommended antimalarial” there
was a precipitous drop in coverage in both HIZs and national comparison area (Appendix table I7). Care
seeking for pneumonia and correct home management practices for diarrhea remained stable, with no
statistical differences over time or between rates of change in the HIZs and comparison area.
IIP-JHU | Retrospective evaluation of ACSD in Benin
33
Feeding, including breastfeeding.
IMCI+ as recommended by ACSD also included promotion of appropriate infant and young child feeding
practices (Box 1). Figure 12 shows the prevalence of selected feeding behaviors reported by mothers of
children less than one year of age at the time of the survey. Breastfeeding behaviors tend to be stable
over time, so the apparent fluctuations should be interpreted with caution because they may reflect
differences in how the questions were posed or the answers recorded. Additional data are available in
Appendix G, H and I.
There is no evidence of improvement over time among the indicators assessed. Exclusive breastfeeding
for infants less than six months of age is least adequate, with only 27% of mothers reporting this practice
in the 2006-7 survey.
Figure 12: Prevalence of infant feeding behaviors as reported by mothers in the ACSD “high-impact”
zones as measured in DHS surveys in 2001 and 2006-7, as well as coverage reported in the CDCACSD survey in 2003 and a coverage survey in 2005, Benin.
DHS 2001
ACSD 2003
Coverage survey 2005
DHS 2006-7
80
78
80
59
60
40
20
0
59
47
45
40
32
Data not available
Prevalence of behavior as reported by mothers (%)
100
25
24
27
0
Initiation of breastfeeding Exclusive breastfeeding to
within one hour of birth
six months of age
Timely and appropriate
complementary feeding
Appendix table I10 presents the breakdown of infant feeding practices in 2006-7 by socio-demographic
characteristics. Reported exclusive breastfeeding levels were marginally higher in DAA than in PAK
(p=0.09); other breastfeeding practices were not significantly different in PAK and DAA. Mothers with
higher levels of education were more likely to report complementary feeding at ages 6 to 9 months than
mothers with less education (p=0.14). Breastfeeding of children at 20-23 months was significantly more
common in rural (76%) than in urban areas (42%); (p<0.001). Infant feeding practices were not
significantly different among children residing in households of different socio-economic status, as
measured by wealth quintiles.
34
IIP-JHU | Retrospective evaluation of ACSD in Benin
Figure 13: Coverage levels and absolute change in percentage points for infant feeding behaviors as
reported by mothers in ACSD “high-impact” zones and comparison area as measured in DHS
surveys in 2001 and 2006-7, Benin.
High-impact zones
Absolute change in percentage points
Comparison area
Prevalence of behavior as reported by mothers (%)
+7
?
-3+75
+2
-8
+44
+7
Comparison
area
-2
+50
-13
-2
100
80
60
80
40
20
0
45 47
47
66
54
2006-7
2001
Initiation of breastfeeding
within one hour of birth
40 38
78 73
45
27
2001
2006-7
Exclusive breastfeeding to six
months of age
2001
2006-7
Timely and appropriate
complementary feeding
Figure 13 shows the prevalence of selected infant feeding behaviors in the HIZs and the comparison area
in 2001 and 2006-7. Initiation of breastfeeding within one hour of birth did not change significantly in the
HIZs; in the comparison area it increased by seven pp (p<0.01). Exclusive breastfeeding up to six
months of age showed a non-significant decrease in HIZs, while increasing slightly in the comparison
area. Complementary feeding from six to nine months of age remained relatively stable, again with slight
increases in the comparison area. Changes over time in the three feeding indicators were not
significantly different between HIZs and comparison area.
ANC+.
The ANC+ package as implemented in Benin included interventions in both the antenatal and perinatal
periods. We present coverage levels for antenatal interventions and then coverage with interventions
designed to improve maternal and neonatal health during delivery and the post-natal period in this section
of the report.
Antenatal care.
Figure 14 shows the time trends in coverage of antenatal care in the HIZs. Appendices G and I present
further details. A high proportion of mothers reported three or more ANC attendances, but there was no
evidence of an increase across the study period. Intermittent presumptive treatment (IPTp) for malaria
during pregnancy increased slightly. Reported coverage with SP during pregnancy was 28% in 2001;
however, this was before the IPTp intervention was available in Benin. Technical staff at the statistical
agency implementing the survey (INSAE), as well as health implementation staff, thought this was a
measurement error. Nonetheless, only seven percent of pregnant women reported IPTp with SP in 20067. Tetanus toxoid (TT) vaccination during pregnancy increased slightly in this period, but this trend was
not significant. The use of iron supplements during pregnancy increased from 33% to 55% during the
study period (p<0.001) (Appendix table G4).
IIP-JHU | Retrospective evaluation of ACSD in Benin
35
Figure 14: Coverage levels for antenatal indicators in the ACSD “high-impact” zones as
measured in DHS surveys in 2001 and 2006-7, as well as coverage reported in the CDC-ACSD
survey in 2003 and a coverage survey in 2005, Benin.
100
80
71
DHS 2001
ACSD 2003
Coverage survey 2005
DHS 2006-7
74
60
55
52
20
0
3+ antenatal visits
18
~ 0...
40
Data not available
44
Data not available
Coverage (%)
64
7
1
IPTp with SP*
2 TT doses
*Any dose of SP during pregnancy (not limited to two doses, due to data limitations)
NB: IPTp with SP incorrectly measured as 28% in 2001 in the HIZs, before IPTp with SP was
available
Appendix tables I11 and I12 provide further information on coverage levels of antenatal care in the HIZs
in 2006-7. Coverage was significantly higher for women residing in the DAA zone, in urban residences
and among those with higher levels of education. In the DAA zone, 74% of women reported three or
more antenatal visits during their previous pregnancy versus 52% in PAK (p<0.001). IPTp and TT2
vaccination, closely linked to ANC visits, were also significantly more frequent in DAA (p<0.001 for both
comparisons). Women reporting a birth in the 6 to 11 months preceding the survey were significantly
more likely to report three or more antenatal care visits (p=0.01) than those with a more recent birth.
Higher levels of women’s education were also associated with three or more ANC visits (p<0.01), TT2
(p=0.05), IPTp (p<0.001) and iron supplementation (p<0.001). Coverage of ANC interventions was highly
inequitable in 2006-7 in the HIZs. More than double the proportion of women in the wealthiest
households reported three or more ANC visits (90%) as compared to those in the poorest households
(40%); (p<0.001). Similar inequities were observed for IPTp (p<0.001), TT2 vaccination (p<0.001) and
iron supplementation (p<0.001)).
36
IIP-JHU | Retrospective evaluation of ACSD in Benin
Figure 15: Coverage levels and absolute change in percentage points for antenatal indicators in
ACSD “high-impact” zones and comparison area as measured in DHS surveys in 2001 and 2006-7,
Benin.
High-impact zones
Comparison area
Comparison
area
+44
+4
Absolute change in percentage points
+2
?
100
-3
+75
+7
-2
-8
-7
+50
+11
Coverage (%)
80
60
40
71
65
64 67
20
2001
2006-7
3+ antenatal visits
7
55 51
~0
~0
0
7
44 47
2001
2006-7
IPTp with SP*
2001
2006-7
2 TT doses
*Any dose of SP during pregnancy (not limited to two doses, due to data limitations)
NB: IPTp with SP incorrectly measured as 28% in the HIZs and 6% in the comparison area in 2001
DHS, before IPTp with SP was available
Figure 15 shows reported antenatal care in the HIZs and the comparison area in 2001 and 2006-7. There
were few differences between the rates of change in the HIZs and comparison area between 2001 and
2006-7. Changes over time did not differ significantly between HIZs and comparison area for three or
more ANC visits, IPTp, TT2 vaccination or iron supplementation.
IIP-JHU | Retrospective evaluation of ACSD in Benin
37
Assisted delivery and postnatal care.
Figure 16 shows trends in assisted deliveries and postnatal care as reported by women having a birth
within 12 months before the survey. Assisted deliveries include those attended by a doctor, nurse or midwife, but not those assisted by assistant nurses or auxiliary mid-wives. Additional data concerning these
indicators are available in Appendices G, H and I. Sample sizes are limited in 2001 with only 88 women
having a live birth within the previous 12 months. Assisted delivery and postnatal visits were quite high at
baseline and there is no evidence of a consistent improvement. Supplementation with vitamin A within 40
days after birth improved significantly over the period from 2001 to 2006-7 (p< 0.001).
Figure 16: Coverage levels for assisted deliveries and postnatal care in the ACSD “high-impact”
zones as measured in DHS surveys in 2001 and 2006-7, as well as coverage reported in the CDCACSD survey in 2003 and a coverage survey in 2005, Benin.
100
94
DHS 2001
84
80
76
DHS 2006-7
74
60
40
Data not available
Coverage (%)
70
ACSD 2003
20
38
21
5
0
Assisted delivery*
Postnatal visit
Postnatal Vitamin A
*Includes only deliveries assisted by doctor, nurse or midwife
Appendix table I13 shows the breakdown of delivery and postnatal care in 2006-7 in the HIZs by sociodemographic characteristics. Women in the DAA zone had significantly higher proportions of assistance
during delivery (p<0.001), postnatal visits (p<0.001) and postnatal supplementation with vitamin A
(p<0.001). Urban dwellers had higher proportions of assisted deliveries than those living in rural areas
(p<0.01), but not higher coverage levels for postnatal care or vitamin A supplementation. Mothers with
higher levels of education were also significantly more likely than mothers with less education to report
assisted delivery (p<0.01) and postnatal care (p<0.01). Assisted delivery and postnatal care were much
less frequent in the poorest household than in less poor household. Women in the highest wealth quintile
were almost twice as likely to have a delivery assisted by a skilled provider (96%) than women in the
poorest households (48%); (p<0.001). Coverage results for postnatal visits within three days (p<0.001)
and postnatal vitamin A supplementation (p<0.01) also showed significant inequities.
38
IIP-JHU | Retrospective evaluation of ACSD in Benin
Figure 17: Coverage levels and absolute change in percentage points for assisted delivery
and postnatal care indicators in ACSD “high-impact” zones and comparison area as
measured in DHS surveys in 2001 and 2006-7, Benin.
High-impact zones
Comparison area
Comparison area
+44
+23
Absolute change in percentage points
-3 +2
5
?
+12
100
+33
+50
-10
-2
-8
-2
Coverage (%)
80
60
94
40
76
63
78
74 75
84 80
20
38 40
5
17
0
2001
2006-7
Assisted delivery*
2001
2006-7
Postnatal visit
2001
2006-7
Postnatal vitamin A
*Includes only deliveries assisted by doctor, nurse or midwife
Figure 17 shows reported coverage of assisted deliveries and postnatal care in the HIZs and the
comparison area in 2006-7. Deliveries assisted by skilled providers were relatively common at baseline in
the HIZs (76%) and remained stable over time. In the comparison area, women reporting an assisted
delivery increased from 63% to 75% in 2006; the difference in differences test was not statistically
significant (p=0.10). Postnatal visits were also very high at baseline, and decreased somewhat in the
HIZs, while remaining stable in the comparison area. The change in pp over time were significantly
different between the HIZs and the comparison area (p=0.04). Supplementation with vitamin A within 60
days of birth increased significantly in the HIZs, and this increase was significantly greater in HIZs than
the in the comparison area (p=0.03).
IIP-JHU | Retrospective evaluation of ACSD in Benin
39
Indicators in the expansion area.
The departments of Ouèmè, Plateau, Zou and Collines were part of the ACSD expansion phase and are
included in our national comparison area, excluding the HIZs. These departments represented 38% of
the comparison area sample in the 2006 survey. Figure 18 presents key coverage indicators for three
groups: 1) the HIZs, 2) the ACSD expansion area, excluding the HIZs, and 3) the rest of the country
excluding the ACSD expansion area, HIZs and Cotonou). For all indicators except breastfeeding,
coverage in the HIZs was similar to that in the comparison area after excluding the expansion area. In
fact, coverage in expansion area was slightly higher than the HIZs for ITNs (p<0.001), DPT3 (p=0.03),
skilled delivery (p<0.001) and exclusive breastfeeding (p=0.03). In short, removing the expansion zones
from the national comparison area did not change the conclusion that the HIZs did not perform better than
the rest of the country in terms of the outcomes under study. Appendix M provides further coverage
results for the expansion area.
Figure 18: Coverage levels for selected indicators in ACSD “high-impact” zones, ACSD
expansion area, and national comparison (excluding HIZs, expansion area and Cotonou) as
measured in DHS 2006-7, Benin.
HID
Expansion areas (excluding HIZ)
100
Nat'l Comparison (excluding HIZs, Exp. Areas & Cotonou)
…..
86
80
61
Coverage (%)
74
71
64 65
60
72
64
60
49
36
36
40
26
26
27
20
0
Vitamin A for
children 6-59m
40
ITN for children 0- DPT3 for children Skilled delivery for
Exclusive
59m
12-23m
births in previous breastfeeding in
12m
infants 0-5m
IIP-JHU | Retrospective evaluation of ACSD in Benin
5.3
Summary and interpretation of results
Table 9 summarizes the main results of the analyses of time trends in coverage. Most indicators did not
improve in either the HIZs or comparison area. Indicators showing no significant improvement included
vaccination, antenatal care, delivery care, case-management and infant feeding. IPTp with SP for women
increased only slightly in the HIZs and comparison area. Two sets of indicators improved rapidly in both
HIZs and comparison area: vitamin A supplementation among post-partum women and children 6-59
months and ITN utilization among children and pregnant women. Significantly different trends were found
between HIZs and the comparison area only for ITN use, postnatal care and postnatal vitamin A
coverage. For ITN use, coverage for pregnant women increased significantly in both HIZs and
comparison area, with significantly greater gains in the HIZs For postnatal visits, coverage declined in the
HIZ areas and remained stable in the comparison area. For postnatal vitamin A, coverage increased
significantly faster in the ACSD than in the comparison area. The findings for postnatal visits and
postnatal vitamin A seem likely to be due to chance, because the level of statistical significance was
borderline (p=0.04 and 0.03, respectively). These results, taken together, suggest that ACSD as
implemented in the HIZs in Benin did not have an effect on coverage levels for the interventions targeted
for accelerated implementation.
A technical team from Benin reviewed and discussed these preliminary results in October 2007. Team
members included those directly involved either in ACSD implementation or in the collection and analysis
of the data used in the evaluation. The interpretation presented below is largely based on these
discussions and the review of implementation and contextual documentation.
EPI+.
In 2006-7, coverage levels for measles and DPT3 were around 50 to 60%, well below the stated ACSD
target of 80% EPI coverage and universal child immunization goals. The stagnation in vaccination
coverage in both HIZs and nationally is compatible with the information provided by local officials.
According to child health program managers, routine vaccination activities—such as monitoring and
supervision—have received less priority nationwide in recent years due to the time and human resources
needed to organize national immunization days, specifically for polio. Between 2003 and 2007, Benin
has carried-out 13 national polio campaigns (Appendix table C2) and this may have contributed to the
stagnation in routine vaccination services. Administrative reports and key informants also indicated that
vaccination outreach activities have faltered, possibly due to the general deterioration of transportation
v
resources and changes in how motorcycles and per diems are managed for health workers. This may
have resulted in low health worker motivation and irregular outreach activities.
Coverage of vitamin A supplementation among children increased markedly between 2001 and 2006-7 in
both ACSD and comparison area. Vitamin A distribution was linked to national vaccination campaigns
starting in late 2003. The frequency of national vaccination campaigns, especially polio, is likely to have
contributed to this marked increase. UNICEF support and inputs of vitamin A at the national level may
explain increases in the comparison area. MOH officials reported that there have not been substantive
differences in vitamin supplementation activities between ACSD zones and the rest of the country.
UNICEF reported having supported these nationwide efforts through funding from the Canadian
government. UNICEF and the MOH noted in key informant interviews that increases in coverage with
vitamin A supplementation are considered one of the main achievements of ACSD. To our knowledge,
there have been no activities supporting vitamin A supplementation in the HIZs other than those
supported by UNICEF.
v
Before 2001, the motorcycles given for outreach activities became the property of the health workers after 3 years of use. The
motorcycles now become the property of the health center, which may have implications for transportation maintenance and health
worker motivation.
IIP-JHU | Retrospective evaluation of ACSD in Benin
41
Table 9: Summary of ACSD coverage results in ACSD “high-impact” zones and comparison area as
measured in DHS surveys in 2001 and 2006-7, Benin.
COVERAGE INDICATOR
AREA
BASELINE
VALUE IN 2001
(%)
ABSOLUTE
CHANGE
2001-2006-7
(% POINTS)
DIFFERENCE
IN
DIFFERENCES
TEST
(p LEVEL)
HIZ
51
-2
>0.10
Comparison
54
+6
HIZ
63
-3
DPT
>0.10
Comparison
67
-3
HIZ
10
+51
Vitamin A to child
>0.10
Comparison
17
+46
HIZ
6
+20
ITN for child
>0.10
Comparison
5
+23
HIZ
14
+20
Net for pregnant woman
>0.10
Comparison
31
+11
HIZ
~0
+22
ITN for woman
<0.001*
Comparison
~0
+17
HIZ
70
-3
Any antimalarial for fever
>0.10
Comparison
66
-8
HIZ
32
-2
Careseeking for pneumonia
>0.10
Comparison
32
+3
HIZ
38
-4
Oral rehydration for
>0.10
diarrhea
Comparison
43
-1
HIZ
45
+2
Breastfeeding initiation
>0.10
Comparison
47
+7
HIZ
40
-13
Exclusive breastfeeding
0.09
Comparison
38
+7
HIZ
80
-2
Complementary feeding
>0.10
Comparison
66
+7
HIZ
71
-7
Antenatal care (3+ visits)
>0.10
Comparison
65
+2
HIZ
~0
+7
IPTp with SP
>0.10
Comparison
~0
+7
HIZ
44
+11
Tetanus toxoid in
>0.10
pregnancy
Comparison
47
+4
HIZ
76
-2
Skilled delivery
>0.10
Comparison
63
+12
HIZ
94
-10
Postnatal visit
0.04
Comparison
78
+2
HIZ
5
+33
Postnatal vitamin A
0.03
Comparison
17
+23
* Difference in end-line estimates only, assuming 0% coverage with ITNs at baseline in HIZs and
comparison area.
Measles vaccine
ACSD in Benin had an objective of achieving 80% coverage for all EPI+ interventions. In spite of the
significant increases, vitamin A levels remain at around 60% in all areas, and two in every five children
are not currently covered.
42
IIP-JHU | Retrospective evaluation of ACSD in Benin
The promotion, distribution and re-treatment of ITNs were large components of the ACSD strategy in
Benin and elsewhere. Increases in coverage with ITNs were observed between 2001 and 2006-7 in HIZs
and nation-wide. However, these increases were much less than the expected objective of 60%
coverage among children and pregnant women in the HIZs. Only one quarter of all children under age
five reported sleeping under a treated net the previous night in 2006. The relatively modest effect on
coverage is consistent with the report of widespread stock-outs in nets, starting in late 2005 and
persisting until the end-line survey, and the delay in re-treatment campaigns in the HIZs (carried out in
November 2006, just after the DHS data collection). Issues in retreating bednets, such as the onerous
logistics and problems with meeting the required periodicity, prompted the government to opt for longlasting bednets in 2005.
In fact, the 2003 and 2005 levels of ITN use were observed to be higher than those in 2006-7. Although
there are comparability problems with the 2003 and 2005 surveys, this trend is likely reflective of changes
vi
in pricing structures and subsequent ITN stock-outs. This finding may have important implications for
sustainability of such interventions. The fact that one out of four children slept under a net the night
before the survey suggests that the despite the stock-outs there were still some nets in the zone that
were re-treated and utilized.
IMCI+.
Efforts were made to introduce and strengthen facility-based and community IMCI in the HIZs. UNICEF
and other partners developed supervision and communication materials, and deployed, trained and
supported CHWs in 2004 and 2005 to try to strengthen this ACSD component. However, in 2006-7, case
management practices for common child illnesses remained unchanged in the HIZs and the comparison
area. Administrative and summary reports noted deficiencies in stock management, mobilization at the
community level, and lack of motivation of CHWs in both PAK and DAA zones throughout ACSD
implementation.
Although the indicator concerning treatment of fever presented here appears to have remained relatively
stable, this does not necessarily represent effective management of fever because there are high levels
29
of resistance to chloroquine in southern and central Benin. Despite the first-line policy change in 2004,
the more effective artemisinin combination therapy for use in treating fever/malaria is still unavailable in
Benin except in 12 communes in Mono and Couffo departments receiving support from Africare through
35
the Global Fund and the PNLP. No child with fever received an ACT in 2006-7 in the HIZs, and less
than one percent did so in comparison area. Most children are still treated with chloroquine (Appendix
table I7), which is no longer an effective treatment for malaria in Benin and could not have contributed
significantly to reductions in child mortality.
Only one-third of children were taken for care to an appropriate facility for presumptive pneumonia
throughout the period from 2001 to 2006-7. Another third did not receive any care and approximately
one-fifth received inappropriate treatments (médicaments par terre) from shops or ambulatory vendors.
The community case management of ARI is in a pilot phase and does not cover many villages; therefore,
it is not surprising that almost no child was reported to receive care from a community health worker for
pneumonia. In terms of home management practices for diarrhea, there were no improvements either in
the HIZs or comparison area over the period under evaluation, and current levels mean that fewer than
one half of children with diarrhea are being properly managed.
There was no evidence of improved infant and young child feeding practices in the HIZs over the course
of the study period. In fact, there were slight decreases in the proportion of women reporting
vi
When ACSD was first implemented, the official cost of ITNs was CFA 3500 (~ USD 7.00) for all. With the advent of
ACSD, the official price was reduced to CFA 1500 (~ USD 3.00) for pregnant women and children less than five years
of age. The price was further reduced in 2005 to CFA 500 (~ USD 1.00) for targeted populations. Program managers
report that this most recent price reduction led to a large increase in demand for bednets, which combined with
problems in the financing of the program, created widespread ITN stock-outs beginning in late 2005 to early-mid
2007.
IIP-JHU | Retrospective evaluation of ACSD in Benin
43
recommended practices in the HIZs, despite other small-scale projects with nutrition components reported
as operating in the HIZs. In the comparison area, there were small gains in breastfeeding and feeding
indicators between 2001 and 2006.
Training of health providers in the minimum package of activities for nutrition - including the promotion of
exclusive breastfeeding and appropriate feeding practices - took place in PAK and DAA in 2003 and
2004. IMCI also includes a focus on infant and young child feeding at both facility and community levels.
However, community-based promotion of exclusive breastfeeding and appropriate feeding practices
supported by UNICEF started intensively only in 2005, after the dissemination of results from a household
survey highlighted the low prevalence of appropriate infant feeding practices in the HIZs, especially PAK.
Approximately 400 community-base workers received training in promotion of case management,
breastfeeding and feeding practices; UNICEF provided visual aids to assist in their work. Of the three
breastfeeding practices assessed, early initiation appears to have responded to the interventions within
the study period, but exclusive breastfeeding and continued feeding remained stable and have been
found in other settings to require intensive behavior change interventions extending over longer periods.
It is therefore possible that more time is required to measure the impact of the efforts started in 2005 on
these two practices. The stagnation/decline in feeding practices may also reflect the intensive early
emphasis by ACSD-Benin on more “vertical” interventions—such as vitamin A and ITNs—relative to case
management and feeding practices.
ANC+.
There was no evidence of significant increases in coverage of ANC interventions in the HIZ over the
course of the study period. The ACSD program in Benin had a stated objective of 80% ANC coverage;
observed coverage of ANC interventions fell short of this objective. ANC intervention indicators tended to
be better in the DAA health zone before (2001) and during ACSD (2006-7). According to administrative
reports, ANC kits have been promoted in DAA since late 2005; however as noted above, issues
associated with recovery of health center costs associated with the ANC kits remain. Long-standing
stock-outs of ITNs have also meant that there have been difficulties in providing the complete set of
materials needed for a kit.
vii
However, the end-line estimates
IPTp was introduced in 2004 in the HIZs and 2005 nationally.
indicated coverage of less than 10 percent in both HIZs and in the comparison area.
Assisted deliveries and postnatal visits were relatively high at baseline and did not improve significantly in
the HIZs, although higher coverage was observed in the DAA than the PAK health zone for both
interventions. Postnatal supplementation with vitamin A was introduced in the HIZs between 2001 to
2006-7, and coverage improved in response.
Contextual Factors.
10
The contextual factors considered in the evaluation were based on those proposed by Victora et al for
child survival programs. Section 3 and appendices A and J provide a more comprehensive description of
contextual factors. Given that the findings on coverage do not suggest that ACSD had a significant effect
beyond what was happening in the rest of the country, the analysis of contextual factors here examines
two questions to better interpret the data:
2. Were there any major disruptions in the HIZs that could explain why ACSD did not lead to a more
marked effect on coverage levels?
3. Why did the interventions for which coverage increased substantially in the ACSD areas also
improve markedly in the comparison area?
vii
Baseline coverage estimate assumed to be zero for IPTp in 2001 in HIZs and comparison area; measured
coverage was 28% in HIZs and 6% in comparison area, although this was before introduction of the intervention
and likely to do survey measurement error.
44
IIP-JHU | Retrospective evaluation of ACSD in Benin
To our knowledge, no major events occurred in the HIZs that could have disrupted the deployment or
effects of ACSD. We have noted in the text above the implementation factors, such as changes in
delivery strategies, pricing and national policies, which may have affected the coverage of certain
interventions. We describe other development partner activities in the HIZs in section three and appendix
J; however, we would expect these to have a positive effect on coverage. A potential problem affecting
the HIZs only was that the introduction of health worker support (per diems and transport costs) for
activities such as outreach as part of ACSD, followed by their withdrawal, reportedly led to reduced health
worker motivation. However, this managerial decision must be regarded as part of the ACSD
intervention, and not as an external factor. In short, there is no evidence of any marked disruption
affecting exclusively the HIZs that could have offset a positive effect of ACSD on coverage.
Vitamin A supplementation and utilization of ITNs improved significantly in the HIZs as well as in the
national comparison area. As noted above, supplementation with vitamin A for children was coupled with
polio campaigns nationally starting in 2002. UNICEF, through ACSD and other programs, supported
these efforts at a national scale in Benin. The distribution, promotion and retreatment of bednets were
supported by national programs and other externally funded programs outside the ACSD areas. For
example, USAID supported similar ITN programs in Borgou-Alibori and Ouémé-Plateau departments and
the Global Fund to Fight AIDS, Tuberculosis and Malaria supported national-level malaria control
activities through a US$2.4 million grant issued in 2003, of which a large proportion was utilized for ITN
distribution and treatment activities. Finally, changes in policies concerning the price of ITNs, discussed
above, may have affected the financial accessibility of ITNs over time in both HIZs and nationally.
These activities in the country as a whole may help explain why the observed improvements in vitamin A
and ITNs were observed not only in the HIZs – where they were delivered through ACSD – but also in the
comparison area where other agencies promoted their delivery.
Methodological Challenges.
Here we present a very brief overview of the methodological challenges, noting how they may affect the
evaluation results of ACSD coverage. Complementing this section, Appendix L provides a more thorough
review of methodological challenges, Appendix F provides descriptions of surveys included in the
evaluation, appendices D and E provide indicator definitions and a list of the questions utilized for
indicator for each survey and Appendix Q compiles the questionnaires from each survey.
The retrospective nature of the evaluation, which necessitates relying on existing—even if imperfect—
2
data and information presented methodological challenges to evaluation team. The 2001 DHS survey
had limited sample sizes for calculation of baseline coverage indicators in the HIZs, especially those
indicators measured among limited subgroups. These small sample sizes affect the precision of point
estimates and the statistical power to detect small differences over time, even though they are
representative of the HIZs. The 2001 and 2006-7 DHS, used in both the adequacy and plausibility
comparisons, were very similar in methodology and conduct. Appendices F and K provide a review of
differences in the surveys and the associated methodological challenges; however, these differences
were minimal and we would not expect them to greatly affect the findings.
The data available in the 2001 DHS survey did always allow for calculation of the preferred indicators
7,8
In the 2001 DHS, the
used for monitoring progress toward the Millennium Development Goals (MDG).
treatment status of bednets used by pregnant women was not collected, bed net use by children was only
assessed through the women’s questionnaire, the timing of antimalarial administration for febrile children
was not available and women who reported a facility-based delivered were not asked about a postnatal
viii
For the evaluation of time trends between 2001 and 2006-7, we utilized indicator definitions that
visit.
could be calculated from the 2001 data to ensure comparability with indicator estimates in 2006-7 (see
Appendices D and E). These proxy indicator definitions were less stringent than the preferred indicator in
viii
It was assumed that women who delivered in a facility received a postnatal visit. In the 2006-7 DHS, all women
were questioned about postnatal consultation, regardless of place of delivery.
IIP-JHU | Retrospective evaluation of ACSD in Benin
45
all cases; coverage estimates from 2006-7 using the more stringent, MDG preferred coverage indicators
are presented in Appendices G, H and I.
Taken together, these methodological issues are not likely to influence the endline comparisons between
the HIZs and comparison area. Differences in the conduct of the survey, the DHS questionnaires and
interviewers’ style of asking questions may have introduced some bias into the comparison of coverage
levels between 2001 and 2006-7. However, these methodological challenges are not likely to change the
main evaluation findings or conclusions in any substantial way.
46
IIP-JHU | Retrospective evaluation of ACSD in Benin
6. Nutrition
In this section, we describe the differences in nutritional status of young children between the ACSD highimpact zones (HIZs) and comparison area; including the rest of the country with the exception of Cotonou
and the HIZs. As described in section 2, data from the comparison area were collected in the 2006 DHS.
Results from the HIZs represent a combination of the 2006 DHS results and those from a supplemental
survey carried out in 2007 using the same methodology, aimed at increasing the sample size in HIZs.
About half of the HIZ sample available for analyses comes from each of the two surveys.
Three indicators of undernutrition prevalence were calculated from the baseline (2001) and endline
(2006-7) surveys: prevalence of stunting (low length for age for children below 24 months; low height for
children 24-59 months of age), wasting (low weight for length/height), and underweight (low weight for
11
age). Based on the 2006 WHO Growth Standards, a cutoff of minus two z-scores was used to define
moderate or severe undernutrition and a cutoff of minus three z-scores was used to define severe
undernutrition. Mean z scores of the three indices were also calculated. Appendix L presents a
schematic of the inclusion and exclusion criteria for children included in the analysis.
Results are presented for all children less than five years of age. For stunting, results are also presented
36
for children aged 24-59 months, the age group with the highest prevalence of this condition. Wasting
results are described for children aged less than 24 months.
Presentation of the results follows the approach used in the section on coverage indicators. First, the
adequacy findings are discussed (time trends in the HIZs), followed by the plausibility results (comparison
between HIZs and the rest of the country). Appendix L presents full nutrition results for sub-groups in
both areas.
6.1
Results
Figure 19 shows that there was little change in the prevalence of stunting over time in the HIZs. The
prevalence of underweight declined from 26 to 20% (p=0.02), and the prevalence of wasting from 19 to
13% (p=0.04).
IIP-JHU | Retrospective evaluation of ACSD in Benin
47
Figure 19: Time trends in stunting (children 24-59 months), wasting (children 0-23 months) and
underweight (children 0-59 months) in the ACSD high-impact zones as measured in DHS surveys in
2001 and 2006-7, Benin.
100
DHS 2001
DHS 2006-7
Prevalence (%)
80
60
51
53
40
26
20
19
20
13
0
Stunting
(children 24-59m)
Wasting
(children 0-23m)
Underweight
(children 0-59m)
Table 10 and Figure 20 show results for the HIZs and comparison area at baseline and endline. There
was an increase in stunting over time of about two percentage points (pp) for children 24-59 months in
the HIZs, compared to a five pp increase in the comparison area. The increase in overall stunting
(moderate or severe) in both areas was due to a rise in the prevalence of severe stunting. Mean
height/length for age worsened in both areas. The increases are also present in analyses of all children
less than five years of age. There was no significant difference between the HIZs and the comparison
area.
There were important declines in wasting over time: six and five pp in the HIZs and comparison area,
respectively, among children less than 24 months of age. Most of the change can be attributed to
declines in moderate rather than severe wasting. Similar results were observed for all children under five,
and the analyses of mean weight for length/height were consistent with the prevalence results. There
were no significant differences by area.
There was a decline in the prevalence of underweight over time: six pp in the HIZ for all children under
five years of age, compared to stable levels in the comparison area. Most of the decline was due to
severe underweight, and the difference between HIZs and the comparison area was statistically
significant. However, there was no decline in mean weight for age, suggesting that the lower tail of the
distribution was affected but not the overall weight for age curve.
48
IIP-JHU | Retrospective evaluation of ACSD in Benin
n
20.1
5.9
-1.1 (1.2)
26.3
8.8
-1.2 (1.3)
3230
9.0
2.6
-0.3 (1.2)
11.5
3677
3.2
-0.3 (1.28)
15.1
-1.5 (1.5)
37.6
46.3
19.3
-1.9 (1.3)
%
16.0
4.7
-0.6 (1.4)
3139
1647
n
GEOGRAPHIC
COMPARISON¥
1582
19.3
5.1
-0.7 (1.4)
40.4
17.4
-1.7 (1.4)
51.1
21.3
-2.0 (1.3)
%
IIP-JHU | Retrospective evaluation of ACSD in Benin
% wasted (< -2 SD) 148
% severely wasted (< -3 SD)
mean (sd)
0-59 months
% wasted (< -2 SD)
352
% severely wasted(< -3 SD)
mean (sd)
Underweight (weight for age)
0-59 months
% underweight (< -2 SD)
342
% severely underweight (< -3 SD)
mean (sd)
0-23 months
% stunted (< -2 SD)
188
% severely stunted (< -3 SD)
mean (sd)
0-59 months
% stunted (< -2 SD)
331
% severely stunted (< -3 SD)
mean (sd)
Wasting (weight for height)
24-59 months
Stunting (height for age)
NUTRITIONAL INDICATOR
HIGH IMPACT
ZONES
2001 DHS
2165
2156
998
2041
1078
n
10726
4696
10283
5814
n
49
20.4
6.3
-1.0 (1.3)
8.2
2.5
0.06 (1.5)
11.0
3.3
0.01 (1.6)
24.3
-1.9 (1.6)
45.5
50.5
26.3
-2.1 (1.5)
%
GEOGRAPHIC
COMPARISON¥
20.2
11172
4.6
-1.1 (1.2)
8.2
2.1
-0.1(1.4)
13.2
3.7
-0.3 (1.5)
20.9
-1.8 (1.5)
44.7
52.8
25.5
-2.2 (1.4)
%
HIGH IMPACT
ZONES
2006/7 DHS
0.02
<0.001
0.07
>0.10
0.04
>0.10
>0.10
>0.10
>0.10
>0.10
>0.10
>0.10
>0.10
>0.10
<0.001
0.01
<0.001
<0.001
0.02
<0.001
0.01
0.01
>0.10
>0.10
>0.10
>0.10
>0.10
>0.10
>0.10
0.04
0.02
>0.10
>0.10
>0.10
>0.10
P VALUE
P VALUE
P VALUE DIFFERENCE
(BASELINE- (BASELINEIN
ENDLINE) ENDLINE) DIFFERENCE
HIZ
NC
S
Table 10: Summary of anthropometry results in ACSD “high-impact” zones and comparison area as measured in DHS surveys in 2001 and 20067, Benin.
Figure 20: Prevalence of stunting, underweight and wasting and absolute change in percentage
points in the ACSD “high-impact” zones and comparison area as measured in DHS surveys in 2001
and 2006-7, Benin.
High-impact zones
Comparison area
Comparison
area
+44 0
Absolute change in percentage points
-3 5 -5
?+5
-8
+2
-6
+50
-2-6
100
Prevalence (%)
80
60
40
20
51
46
53 51
19
0
2001
2006-7
Stunting
(children 24-59m)
2001
16
26
13
11
2006-7
Wasting
(children 0-23m)
20
2001
20 20
2006-7
Underweight
(children 0-59m)
5.2 Summary and interpretation of results
The results suggest that there was no differential impact of ACSD on nutritional status.
considered in this interpretation are presented below.
Factors
Stunting and wasting.
Stunting increased slightly over the period of 2001 to 2006-7 in both the HIZs and the comparison area.
13
This is compatible with the national trends reported by the DHS 2006 for all children under five years.
The active process of stunting, or growth faltering, occurs up to the age of 24 months, after which children
tend to grow parallel to the growth standards and prevalence remains constant up to five years of age.
The most sensitive indicator, therefore, is the prevalence of stunting among children age 24-59 months.
However, for an intervention such as ACSD to have an impact on stunting, children should be exposed to
it during their first two years of life when active faltering is occurring. For this reason, analyses were
repeated for children who were born in 2004 (when ACSD was fully implemented) and therefore spent
their first two years with fully implemented ACSD. Stunting prevalence was 53.5% in the HIZ and 53.4%
in the comparison area, confirming the absence of an impact. Stunting is primarily influenced by dietary
quality and quantity, as well as by the incidence and severity of infections. Because data on
implementation and coverage did not suggest differences between the HIZs and the comparison area in
terms of dietary, preventive or case-management interventions, the lack of impact in stunting is not
surprising.
50
IIP-JHU | Retrospective evaluation of ACSD in Benin
Because analyses of the endline survey (see section three) showed that the HIZs children tended to be
slightly poorer than children in the comparison area, we investigated whether these socioeconomic
differences might have affected our findings. We used direct standardization techniques to estimate the
prevalence of stunting among HIZs children, had their socioeconomic distribution been similar to that
observed in the comparison area. The standardized stunting prevalence in HIZ was equal to 51.7%,
virtually identical to that in the comparison area.
Wasting, on the other hand, decreased similarly in the HIZs and the comparison area, consistent with
13
what was reported by DHS for children under five years in the country as a whole.
Underweight.
The finding of a decline in severe underweight prevalence in the HIZs, in the absence of a similar decline
in the comparison area, was not compatible with the lack of differences in time trends of coverage
indicators. Because ACSD implementation was unlikely to explain the observed impact, we sought
alternative explanations for the findings. These included:
•
Presence of other nutritional interventions or programs in the HIZs. Although our analyses of
contextual factors (see annex J) showed that a number of programs were active in the HIZs, their
coverage relative to the total HIZs populations was limited, and it is unlikely that these programs
could have led to a generalized impact on the area as a whole.
•
Seasonality and differences between the main DHS and supplemental survey. The main DHS was
carried out in the HIZs and comparison area simultaneously, but the supplemental survey - restricted
to the HIZs - was carried out in a different season, about six to eight months later than the main
survey. When the HIZs results were stratified into the two data collection periods, the endline
prevalence of 20% in the HIZs was shown to be a combination of a prevalence of 17.4% in the main
DHS and of 23.5% in the supplemental survey. This implied that the observed decline in the HIZs
would be even greater had the analyses been restricted to the main DHS survey. In addition, the
effect of seasonality should be most marked for wasting (weight for length/height) - a better proxy for
acute undernutrition than weight for age - but this was not the case (Table 10). For these reasons,
neither seasonality nor differences in the two surveys can explain our findings.
•
Contextual factors. Documentation of contextual factors showed that the Djidja and Pobè zones
make up the cereal product granaries of their respective departments, with intense agriculatural
production (two rainy seasons). At the national level, there had been a food crisis in some Northern
zones in the comparison area between 2005 and 2006. This observation prompted a reanalysis of
the anthropometric data, stratifying the national comparison area into regions (Table 11). The
stagnation in underweight prevalence in the comparison area over time was shown to result from the
combination of a sharp increase in the Northern zones (from 25.1 to 33.0% in the baseline and
endline surveys, respectively), a sharp decline in the Central zones (from 25.4 to 19.4%) and near
stagnation in the Southern region (17.5 and 16.4%). The increase in the North is compatible with the
reported food crisis that was limited to this part of the country. Although the HIZs belong to the
Southern region, they are contiguous to the Central region (see map in section four) where
underweight declined by six pp in the period, the same magnitude of reduction observed in the HIZs.
IIP-JHU | Retrospective evaluation of ACSD in Benin
51
Table 11: Prevalence of underweight among children under five years of age, by region of the
country, as measured in DHS surveys in 2001 and 2006-7, Benin.
GEOGRAPHIC AREA
% SEVERELY
%
UNDERWEIGHT UNDERWEIGHT
(< -3 SD)
(<-2 SD)
DIFFERENCE IN ENDLINE BASELINE
DHS 2006/7
DHS 2001
n
% SEVERELY
% UNDERWEIGHT UNDERWEIGH
(<-2 SD)
T (< -3 SD)
n
% SEVERELY
% UNDERWEIGHT UNDERWEIGHT
(<-2 SD)
(< -3 SD)
ACSD "High Impact" Zones
26.3
8.8
342
20.2
4.6
2165
-6.1
-4.2
National Comparison area
20.1
5.9
3230
20.4
6.3
11172
0.3
0.4
12.0
6.1
4.4
2142
2829
6200
7.9
-6.0
-1.0
3.4
-3.2
0.0
National Comparison stratified by
region
Northern*
25.1
8.6
521
33.0
Central**
25.4
9.3
560
19.4
Southern***
17.5
4.4
2148
16.4
*Alibori, Atakora **Borgou, Donga, & Collines ***Couffo, Mono, Atlantique, Plateau, Oeume, Zou
The examination of alternative explanations suggests that the differences in time trends between HIZs
and the national comparison area are due to the food crisis in the Northern zones, which masked a
declining trend in the country as a whole.
Summing up, there was no evidence of an impact of ACSD on any of the three nutritional indicators
studied.
52
IIP-JHU | Retrospective evaluation of ACSD in Benin
7. Mortality
This section reports on changes in child mortality in the ACSD “high-impact” zones (HIZs) and in the
national comparison area, the latter having been defined earlier in this document. The comparisons in
this section differ from those presented previously because the full birth history data collected in the 2006
DHS and its 2007 extension as a supplemental survey to ensure adequate sample sizes in the HIZs,
considered here as a single survey, are used to estimate child mortality both before and after ACSD
became operational.
There are two reasons why we elected to use the 2006-7 survey as the basis for estimating mortality
throughout the evaluation period. First, the use of a full birth history allows the calculation of period
estimates of mortality from the previous year to 10 or more years in the past. The use of the same survey
greatly reduces non-sampling error, although use of this method for short periods may lead to large
sampling errors. The second reason for using a single survey to estimate mortality for the two time
periods – before and after ACSD implementation - is that both estimates are based on the same sample
of households. This reduces the sampling error of the difference in mortality between the two periods,
and hence enables smaller differences to be measured more precisely.
Our main comparison refers to two periods of about three years each, before and after ACSD
implementation in the HIZs. As shown in Figure 21, based on the documentation of ACSD
implementation, we defined the baseline period as July 1999 to June 2002, and the full implementation
period as January 2004 to December 2006, with a phase-in period between.
Figure 21: ACSD implementation time periods in Benin for the retrospective mortality analysis
using full-birth history data, based on documentation of ACSD implementation.
Y E AR (fro m fu ll b irth h isto ry)
T im e p e riods us ed in m o rta lity a n a lys is
1999
2000
2001
A . B AS E L IN E : be fore
im p lem e nta tio n AC S D
Ju n 199 9-J un 2002
2002
2003
2004
B . P H AS E -IN :
s tart o f
AC S D interve ntion s
J ul 2 002 -D ec 2003
C o m p are U 5M R
2005
C . E N D L IN E 2006
F ull
im p lem e nta tio n AC S D
Ja n 20 04-D e c 200 6
2007
A. BASELINE: No ACSD implementation: start of period chosen for symmetry with period C
B. PHASE-IN: Start reinforcement of EPI & cold chain; donation of vehicles, motos & refrigerators; outreach
C. ENDLINE = EPI + ITNs; Vitamin A; CHW training & deployment; IPTp introduced
IIP-JHU | Retrospective evaluation of ACSD in Benin
53
The under-five mortality rate (U5MR) is our
priority indicator for measuring changes in
mortality, because the goal of the ACSD
project was to reduce it by 25% by the end
4
One benefit of using U5MR
of 2006.
relative to other measures of child mortality
(see Box 3) is that it provides the largest
sample size, and is less sensitive to age
heaping than infant or neonatal mortality.
Although we present findings for specific
age groups within 0 to 59 months, we have
considered U5MR as the primary indicator
of mortality impact.
Box 3:
Measures of child mortality
(Expressed as deaths per 1,000 live births)
7.1 Results
Neonatal mortality
(NN)
The probability of dying between birth
and the first month of life
Post-neonatal
mortality
The probability of dying between the
exact age of one month and the exact
age of one year
Infant mortality (IMR)
The probability of dying between birth
and exact age one year
Child mortality (CMR)
The probability of dying between
exact ages one and five years
Figure 22 presents annual estimates of
Under-five mortality
The probability of dying between birth
U5MR in the HIZs and comparison area
(U5MR)
and exact age five years
from 1997 to 2006. Mortality over the last
10 years is declining in both areas, and no
statistically significant difference in the rate of decline between the HIZs and the comparison area given
the large sampling error for the yearly estimates.
Figure 22: Annual estimates of under-five mortality rates in the ACSD “high-impact” zones and
comparison area, 1997-2006, Benin.
250
U5MR (deaths per 1000 births)
200
150
100
50
High-impact
95% confidence bounds
National comparison
95% confidence bounds
0
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Year
54
IIP-JHU | Retrospective evaluation of ACSD in Benin
Figure 23: Under-five, infant and child mortality rates and absolute change in the ACSD “high-impact”
zones and comparison area, Benin.
High-impact zones
Comparison area
Comparison
+44-19 area
Absolute change in mortality
-20
-3 5-19
-36
?
-18
-8
Under-five mortality rate
200
July 1999 180 June 2002
160
-6
-2 -13
Jan. 2004 Dec. 2006
July 1999 June 2002
+50
Jan. 2004 Dec. 2006
July 1999 June 2002
75 59
58
58 73
Endline
Baseline
Jan. 2004 Dec. 2006
140
120
100
80
60
40
20
0
141 145
Baseline
123 109
Endline
Under-five
88 78
Baseline
Infant
52 54
Child
Endline
Figure 23 presents changes in under-five, infant and child mortality rates for the HIZs and the comparison
area in the periods before and after full ACSD implementation. The numbers of births on which these
estimates are based are 1,445 in the HIZs at baseline and 1,624 at endline; the corresponding numbers
are 7,887 and 7,805 in the comparison area. In every age range examined, mortality appears to have
declined faster in the comparison area than in the HIZs.
Table 12 presents age-specific mortality rates in the baseline and end line periods for the HIZs and the
comparison area, as well as reductions over time and 95% confidence limits for these estimates. Postneonatal mortality was the only indicator that did not seem to have fallen faster in comparison area than
in the HIZs. There was a reduction in the U5MR of 18 deaths per thousand, or 13%, between baseline
and endline for the HIZs; this reduction was not statistically significant at the 95% confidence level as the
limits include zero – going from a lower bound of minus five to an upper bound of 41. The national
comparison area had double the reduction (36 deaths per thousand, or 25%) of the HIZs, and this
reduction was statistically significant at both the 95% and 99% confidence levels. The difference in the
rate of under-five mortality decline in the national comparison area relative to the decline in the HIZs was
not statistically significant, nor were there statistical differences in the differences in declines for agespecific mortality indicators.
IIP-JHU | Retrospective evaluation of ACSD in Benin
55
Table 12: Priority and age-specific mortality rates in “high-impact” zones and comparison area before
and after ACSD implementation, Benin.
HIGH IMPACT ZONES
JUL 1999 JUN 2002
JAN 2004 DEC 2006
Priority Indicator
Under-five mortality (5q0)
140.8
95% CI (121.4-160.3)
Postneonatal mortality (PNN)
43.5
95% CI (32.0-55.0)
Infant mortality (1q0)
88.3
95% CI (73.2-103.4)
Child mortality (4q1)
57.6
95% CI (46.9-68.3)
122.9
145.2
(106.8-139.0) (135.5-155.0)
ACSD Phase-in period
Age-specific indicators
Neonatal mortality (NN)
44.8
95% CI (33.5-56.0)
JAN 2004 DEC 2006
JUL 1999 JUN 2002
40.4
(29.5-51.2)
38.3
(33.5-43.2)
34.7
39.8
(21.4-48.1)
(34.9-44.6)
HIGH IMPACT
ZONES
27.3
(23.1-31.5)
COMPARISON
AREA
DIFFERENCES
Difference, Baseline Difference, Baseline (A)
IN
- Endline (C)
(A) - Endline (C)
DIFFERENCES
ABSOLUTE p-value ABSOLUTE p-value
(P VALUE)
109
17.9
(100.8-117.3) (-5.1-41.0)
ACSD Phase-in period
MORTALITY MEASURES
COMPARISON AREA
>0.10
36.2
(25.5-46.9)
<0.001
>0.10
4.4
>0.10
(-10.3-19.1)
11
(5.0-17.1)
<0.001
>0.10
8.5
0.01
>0.10
<0.001
>0.10
<0.001
0.1
31.3
8.8
(27.1-35.5)
(-6.4-24.1)
75.1
78.1
58.6
13.2
(60.9-89.3)
(70.9-85.2)
(52.7-64.5)
(-4.2-30.7)
51.7
72.8
53.6
5.9
(40.8-62.6)
(65.9-79.8)
(48.1-59.1)
(-8.3-20.1)
>0.10
(2.4-14.5)
>0.10
19.5
(10.8-28.2)
>0.10
19.2
(11.9-26.6)
7.2 Summary and interpretation of results
Based on these findings, the U5MR in Benin has declined by 13% in the period from 1999 to 2006—from
141 to 123 per 1,000 live births— about half of the ACSD goal of 25%. However, U5MR declined at
almost double this speed in the comparison area where ACSD was not implemented. This finding held
true across all age subgroups with the exception of post-neonatal mortality, which declined at about the
same rate in both the ACSD HIZs and the comparison area.
Contextual Factors
We considered the possibility that contextual factors might have offset the impact of ACSD. Our
extensive review led to the identification of two contextual factors that might have distorted the findings:
socioeconomic status and ethnic composition. In both instances, we carried out additional analyses to
explore this possibility.
Both the HIZs and the comparison area had similar U5MR at baseline, but important differences at the
endline. There was evidence that the HIZs may have worsened over time in socioeconomic terms
relative to the rest of the country (section 4), and mortality levels tend to be higher among the poor
(section 9). We used direct standardization to adjust mortality for endline differences in wealth, applying
quintile-specific mortality rates in the HIZ to the socio-economic distribution of the population in the
comparison area. The standardized U5MR in the HIZs was 122 per thousand, still substantially higher
than in the comparison area. This suggests that socioeconomic characteristics cannot account for the
observed differences in mortality.
The next section on equity examines further mortality impact by sex of the child, place of residence
(urban/rural) and socio-economic status.
56
IIP-JHU | Retrospective evaluation of ACSD in Benin
Methodological Challenges.
Three important methodological issues may have affected the results of the retrospective estimation of
the effect of ACSD on under-five mortality. The first was the potential effect of data quality issues on the
estimates, and specifically whether differences in survey procedures might have affected the
comparability of the data collected through the original DHS in 2006 and the supplemental survey
conducted in 2007 in the HIZs. We attempted to document survey procedures independently for both
parts of the survey, interviewing survey planners and surveyors and even participating in the training of
surveyors and early period of data collection for the supplemental survey (Appendices F and K). We did
document differences in survey procedures that may have affected data quality and biased the reported
estimates, but further analyses suggested that the overall trends and their interpretation were not
affected. Appendix K presents a more detailed discussion of these issues and how they were addressed
in analysis.
A second methodological challenge was the definition of the “before” and “after” periods of ACSD
implementation. Documentation of implementation is difficult in a retrospective evaluation, and is based
by necessity on records maintained for other purposes and the subjective recall of project implementers.
The two, three-year periods defined for the purpose of this evaluation were discussed and agreed to with
in-country teams composed of ACSD implementers and national counterparts, and we believe that they
do distinguish between times before ACSD was implemented and times during which ACSD was “fully
implemented” in the views of those responsible.
A third issue is the extent to which ACSD activities may have affected mortality in the remainder of Benin
(excluding Cotonou), either directly or indirectly. We address this in the conclusions of the report, which
begin on page 64.
IIP-JHU | Retrospective evaluation of ACSD in Benin
57
58
IIP-JHU | Retrospective evaluation of ACSD in Benin
8. Equity in coverage, nutrition and mortality
In addition to evaluating the impact of ACSD implementation on indicators of coverage, undernutrition and
mortality, it is also important to assess whether or not the strategy helped reduce inequities in health. In
this section, we describe within-population inequalities for the ACSD HIZs and the comparison area
across socioeconomic levels, sex of the child, place of residence and ethnic group. Socioeconomic level
was analyzed by wealth quintiles, obtained from an index based on ownership of household assets and
housing characteristics (Appendix D). The definition of urban or rural residence was based on the
sampling frame from the 2002 national census; this categorization was used as a basis for sampling the
2006 DHS. Ethnic groups included Fon, Yoruba and Adja; all other groups were pooled, as each
comprised less than five percent of the study population. Our analyses document how inequalities differ
between the HIZs and the national comparison area in 2006/07, after implementation of ACSD, because
the sample for the 2001 DHS was too small to support subgroup analyses. Here we present results for
both the HIZs and comparison area, but focus on six coverage indicators representing the three ACSD
components. For the EPI+ component, we present coverage for measles vaccine and vitamin A
supplementation for children. For the IMCI+ component, we present coverage for correct treatment of
child diarrhea (ORS, RHF or increased fluids plus continued feeding) and the use of an insecticidetreated bednet by children less than five years of age. For the ANC+ component, we present coverage
for three or more antenatal visits and the reported presence of a skilled attendant at delivery. We also
assessed equity for the two primary indicators of impact in the evaluation: stunting among children aged
24-59 months and under-five mortality. Appendix I presents the breakdown of all coverage indicators by
sex and by wealth quintiles within the HIZs 2006-7.
For the examination of inequities by socioeconomic status, we present measures of inequality using three
methods. First, we examine levels of selected indicators by wealth quintile, and present the results in
graphs. Second, we calculated two summary measures of inequality. The slope index of inequality
shows the absolute difference between the top and bottom of the wealth scale, based on a regression
approach that uses data from all quintiles rather than just the two extreme groups. Concentration indices
show the extent to which the outcome is equitably distributed across all wealth groups, as reflected in a
value between minus one and one. A value of zero indicates that the outcome is equitably distributed
across all wealth groups. A negative value indicates disproportionate concentration of the health variable
among the poor, for example in the case of disease or malnutrition, where the poor are more likely to be
affected. A positive value indicates that the poor are getting less than would be expected had the
ix
distribution been equitable, as often occurs for preventive and curative interventions.
8.1 Results
Socioeconomic inequalities.
Figures 24a-h and Table 13 summarize the equity results based on a subset of indicators; appendix M
presents the full results. Marked socioeconomic inequities were documented for most coverage
indicators in both the HIZs and the comparison area, with children in the poorest group having lower
levels of coverage than their better-off peers. For example, a slope index of 48 for measles coverage in
the “high-impact” zones indicates the difference in percentage points (pp) in coverage between the
poorest and least poor children. The size of the gap between the poorest and least poor is about 30-50
pp for most indicators, the exception being diarrhea management in the comparison area for which
inequality is negligible. The concentration indices for coverage indicators show similar trends, taking
values between 0.06 and 0.25, again with the exclusion of ORT that was remarkably equitable in the
comparison area.
ix
For more information see: (http://siteresources.worldbank.org/INTPAH/Resources/Publications/QuantitativeTechniques/health_eq_tn07.pdf)
IIP-JHU | Retrospective evaluation of ACSD in Benin
59
Table 13: Summary indices of socioeconomic inequalities for selected indicators in “high-impact”
zones and comparison area, Benin, 2006-7.
INDICATOR
SLOPE INDEX OF
CONCENTRATION INDEX
INEQUALITY*
ACSD
COMPARISON
ACSD
COMPARISON
Measles coverage
48.0
24.9
0.132
0.059
Vitamin A (children)
31.5
28.3
0.088
0.071
ITNs (children)
35.8
43.6
0.207
0.245
Diarrhea management
38.1
0.3
0.176
-0.002
ANC (3 visits)
50.3
55.6
0.125
0.131
Skilled delivery
50.1
52.5
0.110
0.112
Stunting
-41.6
-24.3
-0.115
-0.075
Under-five mortality
-35.0
-52.8
-0.037
-0.074
In the ACSD zones, vaccination coverage and Vitamin A supplementation are weak, even in wealthy
households. While prenatal visits (at least 3) ‘figure e’ and birthing assistance ‘figure f’, the levels were
near 90% for women from wealthy households. It is not surprising that indicators that require contact with
functional health services, such as antenatal or delivery care (Figures 24e-f), show the greatest
37
inequities, as this has been reported for several countries. It is harder to explain why care is so
inequitable for interventions that are delivered primarily through campaigns, outreach or community
approaches including measles vaccination, ITNs or vitamin A (Figures 24a-c) are so inequitable.
Contrary to intervention coverage indicators, the slopes for stunting and mortality (Figures 24 g-h) are in
the opposite direction than for coverage, that is, higher levels among the poor than the rich. The
summary indices (slope index of inequality and concentration index, table 13) take a negative sign under
these conditions.
In summary, there were no consistent patterns of greater or lesser inequities in coverage or impact in the
HIZs than comparison area, with the exception of diarrhea management, which is marginally less
equitable in the HIZs (p=0.03).
60
IIP-JHU | Retrospective evaluation of ACSD in Benin
Figure 24a-h: Socioeconomic inequalities, showing breakdown by wealth quintiles of selected
indicators in ACSD “high-impact” zones and the comparison area, Benin, 2006-7.
IIP-JHU | Retrospective evaluation of ACSD in Benin
61
Inequalities by sex of the child.
There is no evidence of preferential treatment for boys or girls, either in the HIZs or in the comparison
area (table 14). Sex inequalities were not analyzed for antenatal or delivery care, when the sex of the
baby was not yet known (assuming a low frequency of pregnancy ultrasound). For the impact indicators,
boys were more likely to be stunted than girls (p=0.01) but mortality rates were similar; these findings
were consistent for HIZs and the comparison area.
Table 14: Selected coverage and impact indicators by child’s sex in “high-impact” zones and
comparison area, Benin, 2006-7.
COVERAGE or
NUTRITIONAL
INDICATOR
2006 DHS + DHS supplemental 2006-7
Male
AREA
TOTAL
%
Female
n
%
n
Any measles
Innoculation (12-23m)
HIZs
59.3
57.0
257
62.0
228
Comparison
67.6
66.4
1283
68.8
1237
ITN use for under five
children
HIZs
26.4
27.4
1187
25.3
1163
28
28.6
5925
27.6
5860
61.2
61.0
1054
62.0
1026
63
64.0
5345
62.7
5272
HIZs
33.5
30.0
98
38.0
78
Comparison*
41.9
45.3
631
38.0
556
Moderate & severe
stunting (24-59m)
HIZs*
52.8
56.1
552
49.4
526
Comparison*
50.6
52.7
2964
48.4
2849
MORTALITY
AREA
U5MR
U5MR
Births
U5MR
Births
123
122
756
123
725
3937
108
3864
Vitamin A
supplementation of
children (6-59m)
Diarrhea management
Under-five mortality
Comparison
HIZs
Comparison
HIZs
Comparison
109
110
* p < 0.01 comparison of coverage by child's sex within area
Urban/rural inequalities.
About one quarter of the sample was urban in the HIZs, and one third in the comparison area. Urban
children show higher coverage than rural children for all interventions, in both areas, although not all
differences are statistically significant.
Urban/rural differentials tended to be less marked in the HIZs for most coverage indicators: measles
vaccine, ITNs, vitamin A and antenatal care (table 15). Coverage of diarrhea management showed small
urban/rural differentials, and skilled birth attendant showed large differentials in both areas. These
findings suggest that urban/rural differentials in coverage, as a whole, were less marked in the HIZs.
However, in terms of the impact indicators – stunting and mortality – the magnitude of the advantage of
urban children was similar in both HIZs and the comparison area.
62
IIP-JHU | Retrospective evaluation of ACSD in Benin
Table 15: Selected coverage and impact indicators by place of residence in “high-impact” zones and
comparison area, Benin, 2006-7.
COVERAGE or
NUTRITIONAL
INDICATOR
2006 DHS + DHS supplemental 2006-7
URBAN
AREA
RURAL
TOTAL
%
n
%
n
Any measles
Innoculation (12-23m)
HIZs
59.3
64.0
140
57.1
343
Comparison*
67.6
75.7
787
63.9
1733
ITN use for under five
children
HIZs
26.4
30.1
625
25.0
1724
28
36.2
3587
24.5
8197
61.2
63.0
558
61.0
1522
63
70.2
3257
60.3
7360
HIZs
33.5
36.0
44
33.0
132
Comparison
41.9
44.0
311
41.2
876
HIZs*
73.5
85.0
164
69.0
443
Comparison*
74.8
83.2
862
71.1
1983
HIZs
63.5
64.0
163
63.0
443
Comparison*
67.3
75.8
851
63.6
1948
Moderate & severe
stunting (24-59m)
HIZs*
52.8
44.0
285
56.0
793
Comparison*
50.6
42.2
1776
54.2
4038
MORTALITY
AREA
U5MR
U5MR
Births
U5MR
Births
123
106
343
129
1138
2378
115
5423
Vitamin A
supplementation of
children (6-59m)
Diarrhea management
Skilled birth attendant
3+ visits ANC care
Under-five mortality
Comparison*
HIZs
Comparison*
HIZs
Comparison
109
95
* p < 0.01 comparison of coverage by urban-rural within area
Ethnic group inequalities.
Table 16 presents analyses of ethnic group inequalities. Because the Adja are not numerous in the HIZs,
the main comparisons are between the Fon and Yoruba. For most indicators studied, the Yoruba have
lower coverage than the Fon in the HIZs, but higher or similar coverage in the national comparison area.
For stunting, similar levels are observed among the two groups in the HIZs, but in the comparison area
the Yoruba children are significantly less stunted than the Fon. In contrast to coverage, in both the HIZs
and the comparison area, U5MR is lower among Yoruba children than Fon children (p<0.01).
IIP-JHU | Retrospective evaluation of ACSD in Benin
63
Table 16: Selected coverage and impact indicators according to ethnic group in “high-impact” zones
and comparison area, Benin, 2006-7.
2006 DHS + DHS supplemental 2006-7
Adja
INDICATOR
AREA
HIZs**
Any measles
Innoculation (12-23m) Comparison*
ITN use for under five HIZs
children
Comparison*
HIZs**
Comparison**
HIZs
Diarrhea management
Comparison**
HIZs**
Skilled birth attendant
Comparison**
HIZs**
3+ visits ANC care
Comparison
Moderate & Severe
HIZs
Comparison**
Stunting (24-59m)
AREA
MORTALITY
HIZs
Under-five mortality
Comparison
Vitamin A
supplementation of
Fon
Yoruba
Other
%
n
%
n
%
n
%
n
59.3
¥
19
65.4
270
49.5
179
¥
5
67.6
26.4
68.9
40.1
468
55
68.5
27.1
1005
1330
77.4
24.5
214
893
62.7
36.2
755
27
28
61.2
63
33.5
41.9
73.5
74.8
63.5
67.3
52.8
50.6
U5MR
31.4
69.9
66.1
¥
37.9
¥
82.2
¥
72.6
¥
45.4
U5MR
2083
59
1873
3
140
17
481
17
472
20
1008
Births
30.9
67.3
63.4
35.9
33.8
83.4
90.4
73.4
79.1
53.7
52
U5MR
4461
1177
3921
84
387
347
1073
346
1063
637
2197
Births
36.9
52.7
80.4
30.7
48.8
58.5
81.4
47.8
78.7
52.4
43.4
U5MR
980
784
903
80
108
227
217
227
214
393
498
Births
20.8
¥
57.3
¥
47.7
¥
52.5
¥
49.2
¥
53.9
U5MR
3860
24
3571
6
510
5
961
5
940
14
1923
Births
123
109
±
±
±
±
140
120
920
3004
102
84
614
635
81
105
64
3900
TOTAL
* p < 0.05 comparison of coverage by Yoruba-Fon within area
** p < 0.01 comparison of coverage by Yoruba-Fon within area
¥ Sample size too small for estimate (< 25 children); ± Adja enthnicity combined with "Other" due to small sample size
64
IIP-JHU | Retrospective evaluation of ACSD in Benin
9. Conclusions
In this section, we summarize the findings of the evaluation, addressing two separate questions:
a. Was ACSD implementation associated with improvements in coverage, nutrition and mortality
over time?
b. If so, was progress in the ACSD zones faster than observed for the rest of the country?
Figure 25 summarizes coverage trends in the HIZs and comparison area during the study period. The
horizontal axis shows the change in coverage in the HIZs and the vertical axis the corresponding changes
in the comparison area. When the indicator increased or decreased to a similar extent in both areas, the
points are close to the diagonal. Indicators that are above the diagonal showed better performance in
HIZs than in comparison area. The reverse is true for those below the diagonal. Only four indicators –
two for vitamin A and two for ITNs increased by more than 15 percentage points (pp) in the HIZs—all
others showed lower increase or even declined. The same four indicators also increased more than 15
pp in the comparison area.
Figure 25: Summary of changes between 2001 and 2006-7 in coverage and family practices in “highimpact” zones and comparison area, Benin.
Absolute percentage change in coverage
Comparison area
-10
0
10
20
30
40
Vitamin A
(child)
40
+51 HIZs
+46 comp
Vitamin A
(postnatal)
30
ITN (women)
20
ITN (child)
TT2
HIZs
Absolute percentage change in coverage
-20
10
IPTp
0
Measles
AM for fever DPT3
ORT &
feeding
-10
BF within 1 hr
Skilled delivery
Complementary feeding
Careseeking pneumonia
ANC3+
Postnatal visit
EBF
-20
Key:
EPI+ interventions
Case management
Infant feeding
ANC+ interventions
Relative to the two questions posed at the beginning of the chapter, the answers for coverage indicators
are:
a. Most indicators of coverage and behaviors did not improve over time in the HIZs, and – even if
they did - levels at the end of the implementation period were well below the ACSD targets - all of
which were set between 50% and 80%;
b. As a whole, the HIZs did not perform better than the comparison area in terms of improvements
in coverage.
IIP-JHU | Retrospective evaluation of ACSD in Benin
65
Turning to nutrition again referring to our two basic questions, the conclusions are:
c.
The HIZs showed a reduction over time in underweight and wasting prevalence, but not in
stunting.
d. The lack of change in stunting and the reduction in wasting prevalence over the course of the
study period were similar for the HIZs and the comparison area. The reduction in underweight in
the HIZs was not observed in the rest of the country, but lack of progress in the national
prevalence level was due to a food crisis in the north of the country; once this region is excluded
from the comparison area, time trends in underweight reflect what was observed in the HIZs.
Reducing under-five mortality by 25% by 2006 was the primary goal of the ACSD strategy. Our analyses
showed that:
a. There was a non-significant reduction of 13% in under-five mortality in the HIZs, half of the stated
goal of 25%.
b. In the comparison area, the U5MR declined by 25% during the same period. Analyses of
mortality rates by age subgroups within the 0-59 month range also found no evidence that rates
fell more rapidly in the HIZs than in the rest of the country.
Analyses of inequalities in coverage indicators showed were limited to comparisons at the end of the
study period. Due to small sample sizes in the HIZs at baseline, it was not possible to study trends over
time. Our conclusions are:
a. Important socioeconomic and urban-rural inequities were present in the HIZs after ACSD
implementation, even for interventions distributed through strategies found to promote equity in
other contexts such as community-based outreach and/or campaigns for vitamin A
supplementation or ITNs. No coverage inequalities were found in 2006-7 by the sex of the child;
both boy and girl children in HIZs were equally likely to have received the target interventions.
Differences in coverage and impact between the two major ethnic groups – Fon and Yoruba –
were not consistent.
b. Levels of inequities in coverage, nutrition and impact indications did not vary significantly between
the HIZ and comparison area.
Explanations for the negative findings in the Benin ACSD evaluation include the following, alone or in
combination:
1) Despite effective strategies, the intensity (effort) of program implementation was insufficient;
2) The quality of implementation was lower than needed for changing practices – this could be
particularly true for case-management and feeding counseling; and/or
3) The program was implemented as planned, but the strategies and interventions did not work well
in this context.
The retrospective nature of the evaluation and incomplete documentation does not allow us to answer
these questions definitively; however, we briefly consider these explanations here. The ACSD program in
Benin received US$ 1.5 million in initial implementation funds; thus, ACSD was somewhat limited in the
financial inputs compared to other “high-impact” countries or even other programs operating in Benin,
such as the Global Fund which provided US$ 2.3 million during the period in 2003, or USAID-supported
programs with estimated inputs of approximately US$ 3-4 million annually throughout the ACSD period.
Delays and insufficient levels or resources were commonly cited in ACSD program reports as barriers to
full implementation, especially the absence of incentives (whether monetary or non-monetary) to motivate
facility- and community-based health workers. The retrospective evaluation documented that the
strongest implementation efforts were directed to vitamin A, ITNs and community IMCI – the latter mostly
as a strategy for distributing and treating bednets. Other components of ACSD, including vaccination,
correct case-management of infections, nutrition and antenatal care, were less strongly supported.
These findings are reflected in the coverage results presented here: vitamin A and ITNs increased, but
66
IIP-JHU | Retrospective evaluation of ACSD in Benin
interventions and practices
promoted through community
health
workers—such
as
community case management of
illness and infant feeding
practices—stagnated
or
declined.
ACSD program
documents and respondents
identified the need for future
programmatic efforts to increase
community mobilization, and
provide
better
support,
supervision and incentives to
community health workers. Box
4 presents key lessons learned
from the ACSD experience and
recommendations for future child
health programming, according
to counterparts in Benin.
Box 4:
The way forward: Lessons learned in ACSD according to
Benin counterparts
1. There was too little focus on interventions to improve undernutrition and
diarrhea management practices.
 Promotion of infant feeding practices and promotion of key family
practices were not well integrated until 2005; these efforts should be
continued and strengthened.
2.
CHWs were a promising delivery strategy to reach hard-to-access
populations, although CHW motivation and supervision were an on-going
challenge, especially in scaling up.
 CHW systems should be reinforced, with particular emphasis on the five
key family practices;
 In 2005, supervision systems were adapted to re-group CHWs at health
facilities for more frequent supervision; this strategy should be continued
and strengthened.
3.
Limited and discontinued funding may have hindered results.
 Adequate resources need to be provided to carry-out activities,
especially in support of outreach activities and motivation of health staff
and volunteers;
 Commodity security, such as ensuring no stock-outs of ITNs, will be
essential in future programming;
 Program managers need to plan for the discontinuation of funds, so that
there is not a gap in funding for activities.
The lack of a differential positive
effect on coverage in HIZs
relative to what was observed in
the comparison area must be
interpreted in light of numerous
concurrent initiatives - including
4. ACSD contributed to the uptake of effective interventions, such as ITNs
those
to
which
ACSD
and vitamin A, at the national level; as well as introducing promising
contributed - aimed at improving
strategies in Benin, such as ANC packages.
coverage for proven maternal
 The revision of national child survival policies should continue, with
and child health interventions
special attention to nutrition;
throughout the whole country.
 The experiences from ACSD in Benin should continue to contribute to
the policy dialogue.
This is particularly true for
initiatives
seeking
rapid
5. The retrospective nature of the evaluation across the four high impact
increases in coverage for
countries posed important challenges
vitamin A supplementation and
 ACSD in Benin was implemented later than in other countries, with only
ITN use by children under five.
3 yrs of implementation; Future evaluations should measure impact only
Despite these combined efforts,
after sufficient time for continued intensive implementation;
 Prospective evaluations will be needed that can plan to use the same
however, coverage levels for
methods throughout the evaluation process
most interventions were still
below 60% in both the ACSD
HIZs and the rest of the country in 2006-7. The “acceleration” effects expected by ACSD planners did not
occur in Benin.
We examined the possibility that external factors might explain the apparent lack of an impact of ACSD,
through extensive reviews of existing data and interviews with key informants. We were unable to identify
any contextual factors that might account for the lack of impact.
Results on mortality, undernutrition and equity are consistent with the lack of an effect of ACSD on levels
of intervention coverage and family practices. The ACSD strategy was implemented with suboptimal
resource levels and in comparison to the rest of the country where several other initiatives were also
being promoted, ACSD was unable to accelerate progress towards child survival.
IIP-JHU | Retrospective evaluation of ACSD in Benin
67
References
1.
UNICEF. Accelerating child survival and development: A results-based approach in high under-5
mortality areas. Final Report to CIDA. New York, 2005.
2.
Bryce, J., Gilroy, K., Black, R.E., Jones, G. & Victora, C.G. A Retrospective Evaluation of the
Accelerated Child Survival and Development Project in West Africa; Inception Report. Baltimore,
MD, Johns Hopkins University Institute for International Programs, 2007.
3.
Bryce, J., Victora, C.G., Habicht, J.P., Black, R.E. & Scherpbier, R.W. Programmatic pathways to
child survival: results of a multi-country evaluation of Integrated Management of Childhood
Illness. Health Policy Plan. 20 Suppl 1: i5-i17 (2005).
4.
UNICEF. Accelerating early child survival and development in high under-five mortality areas in
the context of health reform and poverty reduction: a results-based approach. UNICEF proposal
to Canadian CIDA. New York, 2002.
5.
Darmstadt, G.L., Bhutta, Z.A., Cousens, S., Adam, T., Walker, N. & de Bernis, L. Evidencebased, cost-effective interventions: how many newborn babies can we save? Lancet. 365 (9463):
977-88 (2005).
6.
Jones, G., Steketee, R.W., Black, R.E., Bhutta, Z.A. & Morris, S.S. How many child deaths can
we prevent this year? Lancet. 362 (9377): 65-71 (2003).
7.
Bryce, J., Terreri, N., Victora, C.G., Mason, E., Daelmans, B., Bhutta, Z.A. et al. Countdown to
2015: tracking intervention coverage for child survival. Lancet. 368 (9541): 1067-76 (2006).
8.
UNICEF. UNICEF/WHO Meeting on Child Survival Survey-based Indicators. Summary List of
Child Survival Indicators. New York, 2004. Available at www.childinfo.org.
9.
Bryce, J., Boschi-Pinto, C., Shibuya, K. & Black, R.E. WHO estimates of the causes of death in
children. Lancet. 365 (9465): 1147-52 (2005).
10.
Victora, C.G., Schellenberg, J.A., Huicho, L., Amaral, J., El Arifeen, S., Pariyo, G. et al. Context
matters: interpreting impact findings in child survival evaluations. Health Policy Plan. 20 Suppl 1:
i18-i31 (2005).
11.
W H O Multicentre Growth Reference Study Group & de Onis, M. WHO Child Growth Standards
based on length/height, weight and age. Acta Paediatr. 95 (s450): 76-85 (2006).
12.
INSAE & ORC_Macro. Enquête Démographique et de Santé au Benin 2001. Cotonou, Benin &
Calverton, MD, Institut National de la Statistique et de l’Analyse Économique (INSAE) & ORC
Macro, 2002.
13.
INSAE & MEASURE_DHS. Enquête Démographique et de Santé, Benin 2006 - Rapport Finale.
Cotonou, Benin et Calverton, MD, Institut National de la Statistique et de l’Analyse Économique
(INSAE) et ORC Macro, 2008.
14.
CERTI. Enquêtes de couverture de la Stratégie Accélérée pour la Survie et le Développement de
Enfant [Benin] 2003. Cotonou, Benin, CERTI, CDC & UNICEF, 2003.
15.
CEFORP. Etude de base dans les zones d'intervention du programme de coopération BeninUNICEF 2004-2008. Cotonou, Benin, Centre d'Etudes, de Formation et de Recherches sur la
Population (CEFORP), République du Benin & UNICEF, 2005.
68
IIP-JHU | Retrospective evaluation of ACSD in Benin
16.
République_du_Bénin. Troisième recensement général de la population et l’habitation (RGPH-3),
2002. Cotonou, Benin, Institut National de la Statistique et de l’Analyse Economique, Ministère
Chargé du Plan, de La Prospective et du Développement, 2004.
17.
Habicht, J.P., Victora, C.G. & Vaughan, J.P. Evaluation designs for adequacy, plausibility and
probability of public health programme performance and impact. Int J Epidemiol. 28 (1): 10-8
(1999).
18.
INSAE & MEASURE_DHS. Enquête Démographique et de Santé, Benin 2006 - Rapport
Préliminaire. Cotonou, Benin, Institut National de la Statistique et de l’Analyse Économique
(INSAE), 2007.
19.
WHO. World Health Statistics. Geneva, 2007.
20.
UNICEF. State of the World's Children 2008. New York, UNICEF, 2007.
21.
Mills, A., Brugha, R., Hanson, K. & McPake, B. What can be done about the private health sector
in low-income countries? Bull World Health Organ. 38 (3): 24-30, 41-4 (2002).
22.
Bryce, J., Requejo, J. & 2008_Countdown_Working_Group. Tracking progress in maternal,
newborn, and child survival: the 2008 report. (avilable at http://www.countdown2015mnch.org).
New York, UNICEF, 2008.
23.
MOH_Benin. Annuaires des statistiques sanitaires 2004. Cotonou, Benin, Ministère de la Santé
Publique, République du Bénin, 2004.
24.
Vodounou, C., Ahovey, E. & Hounkpodote, E. Carte de pauvreté non monétaire au Benin.
Cotonou, Benin, Institut National de la Statistique et de l’Analyse Économique (INSAE), 2006.
25.
Appawu, M., Owusu-Agyei, S., Dadzie, S., Asoala, V., Anto, F., Koram, K. et al. Malaria
transmission dynamics at a site in northern Ghana proposed for testing malaria vaccines. Trop
Med Int Health. 9 (1): 164-70 (2004).
26.
PNLP. Situation de la chloroquine-resistance au Benin, presentation at the ProLIPO Partners
Meeting in January 2002 by Dr. Dorothée KINDE-GAZARD (Coordinator of the National Malaria
Control Program in Benin). 2002.
27.
World_Bank. Project Appraisal Document for the Booster Program in Benin, 2006.
28.
PNLP. Politique et stratégies nationales de lutte contre le paludisme, presentation at CIEVRA in
July 2006 made by Dr. Hortense KOSSOU (Coordinator of the National Malaria Control Program
in Benin). 2006.
29.
Aubouy, A., Fievet, N., Bertin, G., Sagbo, J.C., Kossou, H., Kinde-Gazard, D. et al. Dramatically
decreased therapeutic efficacy of chloroquine and sulfadoxine-pyrimethamine, but not
mefloquine, in southern Benin. Trop Med Int Health. 12 (7): 886-94 (2007).
30.
MARA_ARMA. Benin: Duration of the Malaria Transmission Season, available at:
http://www.mara.org.za/pdfmaps/BenMonthsRisk.PDF, 2000.
31.
World_Food_Program. Where we work - Benin. Information obtained from:
http://www.wfp.org/country_brief/indexcountry.asp?country=204 accessed 15 February 2008.
32.
MOH_Benin. Annuaires des statistiques sanitaires 2006. Cotonou, Benin, Ministère de la Santé
Publique, République du Bénin, 2006.
IIP-JHU | Retrospective evaluation of ACSD in Benin
69
33.
Wild poliovirus importations--West and Central Africa, January 2003-March 2004. MMWR Morb
Mortal Wkly Rep. 53 (20): 433-5 (2004).
34.
UNICEF. Accelerated Child Survival and Development in Benin. Cotonou, Benin, Unpublished
report, 2006.
35.
Global_Fund_to_Fight AIDS, T., _and_Malaria. Grant Performance Report: Support for the fight
against malaria in the Mono and Couffo departments, Africare. Obtained from:
http://www.theglobalfund.org/programs/grantdetails.aspx?compid=598&grantid=291&lang=en&C
ountryId=BEN, accessed 17 November 2007, 2007.
36.
Shrimpton, R., Victora, C.G., Onis, M., Lima, R.C., Blossner, M. & Clugston, G. World wide timing
of growth faltering: implications for nutritional interventions. Pediatrics. 107 (5): E75 (2001).
37.
Victora, C.G., Wagstaff, A., Schellenberg, J.A., Gwatkin, D., Claeson, M. & Habicht, J.P. Applying
an equity lens to child health and mortality: more of the same is not enough. Lancet. 362 (9379):
233-41 (2003).
70
IIP-JHU | Retrospective evaluation of ACSD in Benin
ACCELERATING CHILD SURVIVAL AND DEVELOPMENT (ACSD)
Final Report
The Retrospective Evaluation of ACSD:
Benin
APPENDICES
Submitted to UNICEF Headquarters on 7 October 2008
Institute for International Programs
Johns Hopkins Bloomberg School of Public Health
Baltimore, MD
A. Description of Benin and “high-impact” zones
B. Methodology for documentation of implementation activities and contextual factors
C. Documentation of ACSD implementation in “high-impact” zones
D. Definition of priority indicators
E. Comparison of survey questions used for priority coverage indicator calculation
F. Methodologies of surveys and other data in Benin 2001-2007
G. Tables presenting priority indicators over time for ACSD high-impact zones
H. Tables presenting comparisons of priority indicators over time in ACSD high-impact zones and the
comparison area
I.
Tables presenting 2006-7 results for key indicators in the ACSD high-impact zones by sociodemographic characteristics of the population
J.
Summary of contextual factors
K. Description of methodological challenges
L. Tables presenting additional nutrition analyses
M. Tables presenting additional equity analyses
N. References for the appendices
O. Annotated list of documents reviewed in the ACSD evaluation (file available upon request)
APPENDIX A
Description of Benin and “high-impact” health zones
N IG E R IA
TOGO
3 5 0 0 62
Located in West Africa, the Benin Republic covers a surface of 114,763 square kilometers. It is
bordered in the north by Burkina Faso and Niger, in the east by Nigeria, in the west by Togo and in the
south by the Atlantic Ocean with a coastline spanning 120 kilometers. Geographically, Benin is made
up of a sandy coastal band in the
south, with the two plateau zones
Figure A1:
Benin 3
1 Map of regions,
2
4
of the Atacora in the north where
BENIN
all of Benin’s rivers originate.
REPARTITION POPULATION 2002 PAR DEPARTEMENT
Administratively, Benin is made up
NIGER
of 12 departments established
12
12
since January 15, 1999: Atacora,
Donga, Borgou, Alibori, Atlantic,
Littoral, Mono, Couffo, Oueme,
BURKINA FASO
Plateau, Zou and Collines. These
departments are divided into 77
Alibori
communes, including 3 with
521093
particular status: Cotonou, Porto11
11
Novo
and
Parakou.
These
Atacora
communes are subdivided in 546
arrondissements comprised of
villages and neighborhoods in
549417
towns. The Beninese population
is characterized by a plurality of
Borgou
ethnic groups and languages, and
10
10
contains about fifty ethnic groups.
724171
The informal sector continues to
Donga
develop in the country: according
to the third general Population and
Housing Census of February 2002
(RGPH3), 95% of people are
9
9
employed in the informal sector.
The secondary sector contributes
N
for 13% to the Gross Domestic
Product (GDP) with 35% for the
primary sector and 52% for the
Collines
tertiary sector. The process of
535923
administrative
decentralization,
8
Légende
8
aimed at promoting community
Littorral
lead development is currently
Borgou
underway.
Océan Atlantique
Alibori
Atacora
Collines
Couffo
Donga
Littoral
Mono
Oueme
Plateau
Zou
7
Plateau
599954
407116
Zou
524586
7
Couffo
#
#
360037
Oueme
Atlantique
Mono
Littoral
801683
730772
665100 0 20 40 Kilometers
OCEAN ATLANTIQUE
#
#
360 03 7
#
6
1
A2
2
3
In the last twenty three years, the
population of Benin has doubled in
size, growing from 3,331,210 in
1979 to 6,769,914 inhabitants in
2002; a population growth rate of
3.25% between the 1992 and
1
2002 censuses . This growth rate
is very high compared to the
average growth rate of other
6
4
IIP-JHU | Retrospective evaluation of ACSD in Benin
similarly developed countries. The population is 51.4% female, with 46% of women in reproductive
age (15-49 years), with 3.5% of the population under twelve moths and 17.4% under five years of age.
With 46.8% of the population under 15 years of age, the Beninese population is very young. Figure 1
depicts the population distribution by department in 20021.
Cultural and educational aspects
The population of Benin is characterized by a plurality of ethnic groups and languages; approximately
fifty ethnics groups exist, but French remains the working language. For most of Benin, access to
school has notably progressed during the past ten years, however recently it has become relatively
stagnated. From 2003-2004 to 2005-2006, the gross schooling in primary education changed from
96.4%, with boys at 108% and girls 84.3%, to 95.6%, with boys at 104.8% and girls 86.1 %. The rate
of completion of primary education has increased from 37% in 1998-1999, with boys at 51% and girls
24%, to 54% in 2004-2005, with girls at 42%. The proportion of school aged children aged 6-14 is
currently 56.2% as of 2006.
Health situation in Benin
Since 1996, the population has been characterized with a high fertility rate, at approximately 6.3
children per woman. Fertility has since decreased, but remains relatively high, estimated at 5.6
children per woman in 2001 and 5.7 children per woman in 2006. Like other developing countries,
Benin is characterized by high child mortality rates; however, the estimated IMR decreased from 83
per 1000 live births in 1991-1996 to 67 per 1000 live births in 2001-2006, and the U5MR decreased
2
The maternal
from 151 per 1000 population to 125 per 1000 population for the same period.
mortality ratio, estimated at 498 per 100.000 live births in 1996 has remained stable and was
estimated at 474 per 100.000 live births in 2002. According to the Expanded Programme of
Immunization (EPI), all children should receive all vaccinations before their first birthday. In 1996,
49% of the 12-23 month old children were fully immunized; however 15% of children in this age group
did not receive any vaccinations. In 2001, 59% of 12-23 month old children were fully vaccinated and
7% of children did not receive any vaccinations. In 2006, the rates were 47% were fully immunized
and 7% did not receive any vaccinations. According to the MOH 2005 routine health information
system data, the most frequent child consultations were for malaria (41%), acute respiratory infections
ARI (20%) and diarrheal diseases (8%). Case fatality rates for malaria are relatively high among
children. Figure A2 presents the overall profile of child, neonatal and maternal health from the most
3
recent Countdown to 2015 report.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A3
A4
IIP-JHU | Retrospective evaluation of ACSD in Benin
Figure A2: Benin country profile of maternal, newborn and child survival extracted from Countdown to 2015, 2008 Report.3
Context of DAA and PAK health districts
Demographic characteristics of DAA and PAK.
DAA
The health district of Djidja-Abomey-Agbangnizoun occupies the Western North of the Zou
department. It is bordered in the west by Togo and the health district of Aplahoue-Djakotomey-Dogbo
(Couffo department); in the north, by the health district of Savalou Bante (Collines departement); in the
east by the health district of Cove Zangnanado-Ouinhi; and in the south by the health district of
Bohicon-Zogbodomey-Zakpota (Zou department). The distance from Cotonou to Abomey is
approximately 135 km. The geography presents obstacles to communication within the health district,
in particular in Djidja (e.g. rock outcroppings subdividing the district of Setto). On the administrative
level, DAA is comprised of 29 sub-levels (“arrondissements”) distributed as follows: Djidja (12),
Abomey (7) and Agbangnizoun (10). In 2002, DAA was estimated to have a population of 217,932
inhabitants of which 39% resided in Djidja, 36% in Abomey and 25% in Agbangnizoun1. Population
growth is estimated at an annual rate of 2.4% compared to 3.3% at the national level. The DAA health
zone comprises 77% of the population of the Zou department and 40% of the population of
Zou/Collines departments. DAA extends 69 km from west to east, and 72 km from north to south; it
covers a surface area of 2,570 km2, which is 13% of the total area of Zou/Collines and 49% of Zou
department. Djidja accounts for 85% surface area of DAA while Abomey accounts for 5.5% and
Agbangnizoun for 9.5%.
Figure A3: Map of DAA and PAK high impact zones, with commune boundaries, Benin
N
D ép t . C o lline s
D jid ja
Nigéria
Ké to u
Ab o m e y
Ag b a n gn izo u n
Po b e
ffo
ou
C
t.
ép
D
D ép t. A tlan t iq u e
Ad ja- -O u re
Lég end e
Z one d'é tud e
IIP-JHU | Retrospective evaluation of ACSD in Benin
A5
The population density of DAA is approximately 85 inhabitants per km2, of which Abomey is 552
inhabitants per km2, Agbangnizoun 225 inhabitants per km2, and Djidja 39 inhabitants per km2. Djidja
is a zone of high migration, with men leaving to search for employment in the bordering districts or
elsewhere. In Djidja, development activities are few and the zone is sparsely populated, strongly
isolated, and access is difficult, especially during the rainy season when several villages remain
inaccessible and roads unsecure.
PAK
The PAK health district is composed of the districts of Pobè, Adja-Ouèrè and Kétou. In 2002, the
population was estimated at 264,906 inhabitants, with 38% of the population in Kétou, 31% in Pobè
and 31% in Adja-Ouèrè, and an estimated annual growth rate of 4.1%. PAK accounts for 16% of the
population of the department of the Plateau and 45% of the departments of Ouémé/Plateau. The
current population in PAK came primarily from migration from Ife and Oyo in Nigeria. This population is
made up predominantly (68%) of the Yoruba ethnic group and related ethnic groups, such as Nago.
The Protestant religion is most dominant within the health district. On the administrative level, PAK is
comprised of 17 sub-levels (“arrondissements”) distributed as follows: Pobè (5), Adja-Ouèrè (6) and
Kétou (6). The area of PAK covers 2,202 km2, and accounts for 48% of the total surface of
Ouémé/Plateau and 67% of the Plateau department. Kétou accounts for 55% of the surface of the
PAK health district, while Pobè accounts for 12% and Adja-Ouèrè for 13%. Population density of the
PAK zone is estimated at approximately 83 inhabitants per km2; 207 inhabitants/km2 in Pobè, 196
inhabitants/km2 in Adja-Ouèrè, and 57 inhabitants/km2 in Kétou.
The health district is located in the Plateau department and is limited in the north by the communes of
Savè and Dassa, both in the Collines department, in the east by Nigeria, and in the west by the Zou
department (districts of Zangnanado and Ouinhi) and Ouémé department (district of Bonou). Ketou is
located approximately 138 km from Cotonou.
Table A1: Demographic and administrative features of PAK and DAA and comparison areas, Benin
CHARACTERISTICS
ACSD HI ZONES
COMPARISON AREA
DAA
PAK
REST OF BENIN MINUS
COTONOU & HIZS
217,000
265,000
5,574,000
2,570
2,202
109,912
85
83
51
Estimated population growth rate
2.39%
4.13%
3.25%*
Primary ethnic/language group(s)
Fon
Yoruba
Nago
Fon Yoruba
Nago
Number of Communes
3
3
70
Number of arrondissements
29
17
487
Population 20021 (to nearest 1000)
2
Area (km )
Population density /km
2
*National level data, not excluding high impact zones or Cotonou
Environmental characteristics of DAA and PAK.
DAA
DAA has a transitional climate between the subequatorial, tropical wet climate and the SudanoGuinean climate of northern Benin. The subequatorial climate in the south of Benin is characterized
by two rainy seasons, from April to June, and September to October, and by two dry seasons. The
A6
IIP-JHU | Retrospective evaluation of ACSD in Benin
semi-arid tropical climate of the north is characterized by one relatively long dry season from
November to May, accompanied by the dry winds of the harmattan, and by one relatively long rainy
season from June to October. Annual rainfall in DAA varies between 900 and 1200 millimeters. The
vegetation is dominated by natural palm plantations and grass lands, with small areas of classified
forests. Soils in this zone are “terres de barre” or the ferralitic type argilo-sandy and are highly
degraded but well drained, with a low water holding capacity. On the plateau of Abomey, soils are the
highly prized prairies. The zone is transected by 293 kilometers of rivers originating from the Couffo
basin. During the rainy season, in the districts of Agbangnizoun and especially of Djidja, ponds and
pools of water block the access of certain villages and hamlets from health services, supply chains,
supervision of health centres and medical evacuations. The climatic, hydrographic, telluric and
ecological characteristics of the zone, combined with the essentially agro-pastoral human activities,
create favorable conditions for many vectors/reservoirs, such as snails, flies, mosquitoes, rats,
monkeys. This maintains the endemicity or periodic and seasonal outbreaks of infectious and
parasitic diseases like: cholera, dracunculosis, onchocercosis, malaria, yellow fever, and meningitis).
*
Djidja is one of the most strongly endemic zone of dracunculosis in Benin .
PAK
The climate in PAK is of Sudano-Guinean type (i.e. two rainy seasons) with the annual rainfall ranging
between 1000 and 1400 millimeters. The vegetation is comprised of raised/shrubby savanna and the
semi-deciduous dense forest. The geography of PAK is characterized by the plateau of Kétou, the
valley of Issaba and the plateau of Pobè. Several types of soil are found in PAK: the ferruginous
tropical ones, the “terres de barre” on the “continental terminal profound” and the very argillaceous and
humus-bearing fertile soils, but these are often very difficult to work with rudimentary tools. This health
district is irrigated by the Ouémé River and its tributaries.
Table A2: Environmental characteristics of PAK and DAA and comparison areas, Benin
CHARACTERISTICS
Climate
Main Geographic
Characteristics
Annual rainfall (mm)
Months of malaria
transmission4
ACSD HI ZONES
COMPARISON AREA
DAA
PAK
REST OF BENIN MINUS
COTONOU & HIZS
Transitional
between
tropical wet &
semi-arid
tropical
Semi-arid
tropical
Tropical wet, Semi-arid
tropical
Palm
plantations
Grasslands
Savanna &
the semideciduous
forest
Palm plantations
Grasslands; Savanna & the
semi-deciduous forest; Sahel
900-1200
1000-1400
900-1500
8-9
months/year
8-9
months/year
5-9 months/year, with
shortening season in north
*
Case decrease from 85% of Guinea Worm cases registered in Benin in Zou/Collines (27% in Djidja) in 1998, to 64% and
26% respectively in 2000, 0.02% % in 2002, and 0 % since 2003, for Zou/Collines and Djidja.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A7
Economic aspects of DAA and PAK.
DAA
Agriculture is the dominant economic activity in DAA and accounts for approximately 47%, followed by
29% trade activities. The retail trade is dominated by women and based on distribution of foodstuffs.
The principal food crops are corn, yam, cassava, and beans, with export products of cotton,
groundnut, cabbage, and palm products, such as palm oil. A study on the living conditions of rural
households carried out in 1999-2000 showed that in the Zou/Collines department, the proportion of
poor households with more than 6 people is high (52.5%). The proportion of vulnerable† households
having more than 6 people is 59.8%. The proportion of non working people in charge of the household
is higher in poor households (54%) and vulnerable (50%) that in non poor households (48%). In the
department, as in the whole of Benin, the number of non working people in a household is about the
same than the number of active people. The average population size of a rural household in Zou is of
7 people, varying from 5 people in non-poor households to 8 people in poor households. The results of
the 1st EMICOV‡ survey in 2006 showed that the average annual expenditure per household in Zou
counts for 56% of those of the national level and the average annual expenditure per capita is 62% of
those of the national level. The structure of this expenditure shows that the households of Zou spend
approximately 50% their expenditure in food items compared to 41% at the national level, and less
than a third in the Littoral department (Cotonou). However it is known that the more the household
devotes its expenditure to food, the more impoverished it is. In this expenditure structure, health items
occupy only 14% and education 1.3%. Incidence of the monetary poverty is estimated at 37.5% in
DAA (Abomey 37.7%, Agbangnizoun 34.0% and Djidja 39.9%), mirroring 37.4% at the national level.
PAK
A study carried out in 2000 on the living conditions of rural households showed that in the Plateau
department, the proportion of non-working people in the charge of active people in the household is
higher in the non-poor households (46%), than in the vulnerable (43%) and in poor households (41%).
The average size of the rural household in the Plateau is about 5 people, varying from 5 in the non
poor to 6 in the poor households. The results of the 1st passage of EMICOV in 2006 shows that the
average annual expenditure per household in the Plateau accounts for 63% of those of the national
level and the average annual expenditure per capita, 70% of those of the national level. The structure
of this expenditure shows that the households of the Plateau spend approximately 50% of their
expenditure in food items, compared to 41% at the national level, and less of the third in the Littoral
department. In this expenditure structure, health items occupy only 4% and education 1%. The
incidence of monetary poverty in Plateau department is 40.2%, with Adja-Ouèrè at 38.39%, Pobè
42.4% and Kétou 41.2%, one of the third highest incidences, after Alibori 42.97% and Couffo 40.56%.
Cultural and educational aspects of DAA and PAK.
DAA
The population of DAA is mainly composed of the Fon ethnic group. Apart from the agglomeration of
Abomey, the population of DAA lives in general with average material conditions and are strongly
attached to their traditions. In the whole Zou department, access to school notably progressed during
the ten last years and from 2003-2004 to 2005-2006, the gross schooling rate in primary education
went from 101.90% for Zou/Collines (boys: 116.66 %; girls: 86.89%) to 104.06 % in Zou (boys:
†
Considered as non poor household of which average real expenditures are between 100 and 150% of poverty
line concerned;
‡
EMICOV : Integrated modular survey on the living conditions of households with 4 passages in the year (1st
passage made in 2006, simultaneously with 2006 DHS). Incidence of monetary poverty is measured by the level
of expenditures per capita compared to a poverty line.
A8
IIP-JHU | Retrospective evaluation of ACSD in Benin
119.58%, girls: 88.99%), and to 95.07% for the Collines department (boys: 103.79%, girls: 86.02 %). A
basic inquiry in 2004 showed that the gross school attendance rate was 100% and the net rate was
67%. It should be noted, however that the schooling of younger generation has improved
considerably. The proportion of the 6-14 years olds currently at school was 58.6% for Zou in 2006; it is
higher in Abomey and Agbangnizoun, respectively at 70% and 67%, whereas it is particularly weak in
Djidja (39%).
PAK
The population of PAK is mainly composed of the Yoruba and the related Nago, followed by Fon
ethnic groups. The population of PAK lives in general under average material conditions and is
strongly attached to their traditions. For the whole Plateau department, access to school notably
progressed during the ten last years and from 2003-2004 to 2005-2006, the gross schooling rate
(TBS) in primary education passed from 107.06% for Oueme/Plateau departments (boys: 124.63 %;
girls: 89.34%) to 88.45% in the only Plateau, boys: 104.22% and girls 72.92 %, and to 111.02% for
Oueme (boys 124.19 %, girls: 97.86 %).
A 2004 basic survey showed that the gross school attendance rate was 77% and the net rate 53%.
The proportion of the 6-14 years olds currently in school was 51.4% for the Plateau in 2006; these
proportions are the lowest in Pobè and Kétou (respectively 46% and 48%) whereas it is higher than
the average in Adja-Ouèrè (56%).
Health context of DAA and PAK.
DAA
DAA consists of three districts: Djidja, Abomey and Agbangnizoun, which are part of the Zou/Collines
§
department . From 1996 and 2001, it can be assumed that these three communes had a health
situation similar to that observed in this department. On this basis, the population of DAA would thus
be characterized by a high level of fertility similar to that of the Zou/Collines department: an estimated
6.5 children per woman in 1996 and 6.1 children per woman in 2001. After the territorial reform of
decentralization, DAA is part of the new Zou department which comprises 9 health zones. Data is not
available at the level of health zone; however, the Zou/Collines departments have high child mortality
rates. The IMR in Zou/Collines departments was estimated at 102 per 1000 live births for 1986-1996,
which placed it in 4th position of the highest quotients of mortality, after Atacora/Donga, Borgou/Alibori
and Mono/Couffo, and U5MR at 202 per 1000 population, the highest rate of mortality after
Atacora/Donga 203 per 1000 population. From 1991-2001, these rates were estimated at respectively
120 and 190 per 1000 population, the highest rates of all the departments.
Immunization coverage in Zou/Collines was estimated at 58% of 12-23 month old children were fully
immunized compared to 8% children who did not receive any vaccinations in 1996. This increased to
68% and 4.5% in 2001; and in 2006, for the new Zou department, the coverage rates were 44% and
3.5% respectively. This immunization coverage must be interpreted with caution, as the Zou-Collines
department in 1996 and 2001 may not be comparable to the 2006 level in Zou department. A survey
carried out in 2004 showed that 28% of 15-49 year old women were pregnant or already had a child.
Among these women, the median age of first pregnancy was 19 years. In the 15 days preceding the
2004 survey, 24% of children under five had reported fever, including 44% with at least one sign of
severity; 12% had reported diarrhoea including 55% with at least one sign of severity; 33% of children
had reported symptoms of ARI.
PAK
The 3 districts of PAK (Pobè, Adja-Ouèrè and Kétou) have a health situation considered as similar to
that observed in the departments of Ouémé/Plateau. Ouémé/Plateau department, including PAK, is
characterized by a high level of fertility, although a slight decline is ongoing: 5.9 children per woman
§
(in the past called Zou, made up of 15 communes).
IIP-JHU | Retrospective evaluation of ACSD in Benin
A9
(15-49 years) in 1996 and 5.0 in 2001. Today, PAK is administratively attached to the Plateau
department which is comprised of 5 districts. Mortality estimates do not exist at the health zone level;
however, Oueme/Plateau department is characterized by high child mortality rates. The IMR was
estimated at 88 per 1000 live births for 1986-1996 and at 82 per 1000 live births for 1991-2001. In
1996, it was estimated that 58% of the children between 12-23 months old of the Oueme/Plateau
departments were fully immunized and 8% did not receive any vaccinations. In 2001, the rates were
68% and 4%. In 2006, the Plateau department recorded 65% fully immunized children and 16% with
no immunization. A survey carried out in 2004 showed that the fertility in this department begins early:
36% of women 15-19 years were pregnant or had already a child. Among the 15-49 years old women,
the median age for the first pregnancy was of 17 rears. As for u5 children, in the 15 days preceding
the 2004 survey, 33% children reported fever including 35% with at least one sign of severity; 21%
reported diarrhea including 43% with at least one sign of severity; 28% reported symptoms of ARI.
Table A3: Health system characteristics in PAK and DAA zones and comparison areas 2006, Benin
CHARACTERISTICS
ACSD HI ZONES
COMPARISON AREA
DAA
PAK
REST OF BENIN MINUS
COTONOU & HIDS
Total health facilities
31
30
744
Average # of health centers
per commune
10
10
11.5
7030
8797
7493
5
2
127
1
1
22
Population per facility
Private or
facilities
Hospitals
NGO
health
SOURCE: ANNUAIRE DES STATISTIQUES SANITAIRES 2006
A10
IIP-JHU | Retrospective evaluation of ACSD in Benin
APPENDIX B
Methodology for implementation of ACSD activities and contextual factors
Various techniques were employed to collect information retrospectively about the implementation of
ACSD activities and contextual factors in the “high-impact” zones. Much information was gathered
from colleagues at the UNICEF-Benin field office, who have been collaborating on the retrospective
evaluation throughout the process. Field visits, key informant interviews and working meetings to
review of the preliminary coverage results all provided information pertaining to details of ACSD
implementation and contextual factors. Details of these discussions are provided in table B1. During
these encounters, the JHU evaluation team requested any documents providing additional details on
ACSD and other partner’s activities.
Over 150 documents pertaining to ACSD implementation and other project activities were reviewed;
the types of documents reviewed are presented in the text of the report.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A11
A12
IIP-JHU | Retrospective evaluation of ACSD in Benin
DESCRIPTION
st
Benoit Adsikpe (1 relais); Felieien
nd
Djidago (2 relais); village
groups/leaders
Village visit
Saluidji village (in
Setto
arrondisement)
Dec 2006
Paul Adovohekpe; Andree
Cossi;
Loukmane Agbo-Ola
Gilbert G. Vissoh
Interview &
discussions
Interview &
discussions
Interview &
discussions
All ACSD and contextual factor information, especially ACSD
implementation in DAA zone
All ACSD and contextual factor information, especially ACSD
implementation in PAK zone
All ACSD and contextual factor information
Pilot site for treatment of ARI in community; review of CHW
management of ARI and associated materials
Facility IMCI functioning; outreach activities; functioning of maternity &
ANC activities; review of previous birth technique and records
Xavier Tokpo (head dispensaire) ;
Estelle Zadsi (head maternity) ;
Sosthene Codji Zakpame ; Florence
Wanou ; Guilleme Gwawgoven ;
Jeanne Kinosietha Sononkindjieha ;
Henri Lokossi (PSI)
Site visit Centre
de Santé de
Arrondisement
Setto
Key informant interviews / discussions
Dec 2006
TOPICS COVERED
Clinical IMCI, ITN promotion distribution retreatment and stock-outs,
ANC packages and functioning of maternity; monitoring and monitoring
tools; follow up of vaccination through outreach; pharmacy and prices of
medicines and interventions (ITNs & ANC bags; functioning & activities
of mother’s groups; Involvement of community leaders (village chief.
King); review of counseling cards; review of village drug kits & contents;
(credit for meds, etc)
PARTICIPANTS
Joseph Zinsou (Health coordinator
Pobé) Hamidou Iroleke (head nurse
Ipkinle); Health staff and Community
workers in Pobé and Ipkinlé;
Mother’s groups, village chief,
Community Health Workers; King of
Issaba
Field Visit to
Pobé Ipkinle
health center &
Igbo-Oro (village)
in Ipkinle
arrondissement &
Issaba
Field Visits
DATE
CV, GJ, KG
CV, GJ, KG
CV, GJ, KG
CV, GJ, KG
+UNICEF-Benin
team
CV, GJ, KG
+UNICEF-Benin
team
CV, GJ, KG
+UNICEF-Benin
team
PARTICIPANTS
TEAM
EVALUATION
Table B1: Description of field visits, key informant interviews, and work sessions carried out to document ACSD implementation activities and
contextual factors
IIP-JHU | Retrospective evaluation of ACSD in Benin
A13
Oct 2007
Sept
2007
DATE
Interview &
discussion
Interview &
discussion
Interview &
discussion
Interview &
discussion
Interview &
discussion
Dr Gilbert Vissoh, MOH
(previously ACSD consultant in
DAA zone)
ACSD activities in DAA zone and external partners in DAA zone
External partners in PAK zone
External partners in DAA zone
Chief medical officer at the
CIPEC (Centre d'Information de
Prospective d'Education et de
Conseils pour le SIDA) – part of
the National AIDS Programme
(PNLS).
Joseph Zinsou - Chief medical
officer for PAK district
External partners in DAA and PAK health zone
External partners in DAA zone
Seraphin Vissoh, Chief of Health
Dept at Catholic Relief Services,
Benin)
Gabriel Sayi – Chief medical
officer for DAA zone
SvH
SvH
SvH
SvH
SvH
CV, GJ, KG
+UNICEF-Benin
team
MOH involvement in ACSD activities; Policy implications of
ACSD; challenges and successes of ACSD per MOH
Alban Quenum – director,
Direction de Santé Familiar
(DSF)
Interview &
discussion
CV, GJ, KG
Community-based situation analysis and community
IEC/promotion activities/materials in Benin and other ACSD
countries
Madame Osseni, prior ACSD
Consultant
Interview &
discussion
CV, GJ, KG
+UNICEF-Benin
team
PARTICIPANTS
TEAM
CV, GJ, KG
Faustin Onikpo; Odje Adeichan
PROLIPO
PROLIPO
presentation &
discussion
Staffing; advantages & difficulties of SASDE approach; stock out
of ITNs
TOPICS COVERED
PROLIPO activities in PAK region; preceding birth technique &
findings from this technique
Joseph Zinsou Zone Sanitaire
Pobé
PARTICIPANTS
Interview &
discussion
DESCRIPTION
EVALUATION
A14
IIP-JHU | Retrospective evaluation of ACSD in Benin
DESCRIPTION
PARTICIPANTS
Meetings to
review
preliminary
results
Regional
workshop to
review &
interpret
preliminary
results (Dakar,
Senegal)
Aug 2007
Oct 2007
Paul Adovohekpe; Andree
Cossi; Alban Quenum; Gilbert
G. Vissoh; Elise Ahovey
Paul Adovohekpe; Andree
Cossi; Alban Quenum;
Dominique Robez-Masson;
Arnaud Houndeganme; INSAE
staff involved in DHS &
supplemental surveys
TOPICS COVERED
Revision of preliminary results, discussion of contextual factors
and methodological issues
Revision of preliminary results, discussion of contextual factors
and methodological issues
Work sessions to review and interpret preliminary results
DATE
JB, CV, GJ, KG,
SvH
KG
PARTICIPANTS
TEAM
EVALUATION
APPENDIX C
Documentation of ACSD implementation activities in “high-impact”
zones
The ACSD implementation activities are described in more detail here, expanding on the information
provided in the main body of the report. Each ACSD component is described in more depth, and
timelines of implementation activities for each ACSD component are presented in tabular format. For
clarity, much of the information included in the main report narrative is repeated here. The timelines
and information presented here should not be regarded as comprehensive, as gaps in information are
inevitable given the retrospective nature of the evaluation. Some of the information presented here is
based solely on key informant interviewers and could not be independently confirmed.
The implementation of ACSD contained many components of training for clinical personnel. Table C1
presents an overview of the types of training health agents received between 2002 and 2006 in the
DAA health zone. The activities for the implementation of ACSD are presented in tables C2-C5.
Table C1: Overview of health agent training in DAA health zone during ACSD implementation
NUMBER OF
TOPIC
HEALTH AGENTS
%
TRAINIED
Clinical IMCI
52
100
Normal care and neonatal resuscitation in the delivery
room
32
100
Minimum Package of Activities in Nutrition
188
100
Refocused ANC and IPTp
32
100
EPI+
92
100
Medicine management
44
95
PMTCT
74
98
Extracted from a presentation on health provider training in DAA zone by P. Adovohekpe, 2006
IIP-JHU | Retrospective evaluation of ACSD in Benin
A15
Vaccination and vitamin A supplementation (EPI+).
Vaccination activities and supplementation of children 6-59 months with vitamin A were already in
place through the MOH system prior to the introduction of ACSD. The health system in Benin delivers
vaccines to children in three principle ways: 1) routine, facility-based vaccination; 2) routine outreach
activities; and 3) vaccination campaigns. Vitamin A supplementation is linked to vaccination
campaigns, as well as recommended to be given during IMCI visits for sick children. However, key
informants noted this system of delivery was only carried out in health centres with vitamin A stocks
leftover from national vaccination days, since 2006 stocks of vitamin A for both campaigns and routine
distribution are supported by UNICEF and CAME. Vitamin A is not distributed through routine or
outreach vaccination activities. Child health cards are to be completed and kept at health facilities
include a section to report dates of the child’s vitamin A supplementation. The completion and use of
this vitamin A information on the health cards is unknown.
Reinforcement of existing EPI activities was one of the earliest ACSD activities in Benin. According to
administrative and summary reports of late 2002 UNICEF provided: 1) basic medical materials; 2)
four-by-four vehicles and motorcycles for supervision and outreach; 3) refrigerators for the cold chain;
and 4) computers for monitoring and data collection activities to the “high-impact” zones and their
corresponding health departments (Zou-Collines and Ouémé-Plateau). The MOH, in collaboration
with UNICEF, trained facility-based health agents in PAK and DAA periodically between 2002 and
2006 to reinforce capacity in EPI related activities, such as vaccine policy, stock management,
secondary effects, outreach, active defaulter tracing, cold chain management, monitoring, and
surveillance (appendix table C2). ACSD staff collaborated on local vaccination catch-up campaigns
for all vaccinations and active defaulter tracing in 2003 and 2005 in PAK and DAA zones, as well as
participating in measles epidemic investigations and response in 2005 in Djidja. Administrative reports
from 2004 onward noted challenges in implementing the vaccination portion of the EPI+ package,
such as irregular outreach activities, low health agent motivation and at times a lack of catch-up
vaccination activities. Supervision and monitoring reports, as well as key informants, also noted
resistance to vaccination in certain areas of the PAK zone.
In 2003, ACSD supported the selection and training of over 400 community health workers (CHWs) in
approximately 200 remote villages in PAK and DAA. Promotion of vaccination and defaulter tracing in
the community were among the topics covered during initial CHW training sessions in mid-2003. Key
informant interviews and observations during field visits revealed that defaulter tracing for vaccination
may have also been strengthened through collection of monitoring data. Each child registered in the
health facility catchment area should have a health record card maintained at the health facility.
These cards are issued for children at birth or through outreach activities for children not born in
facilities. In some health centers, facility-based staff use these cards to determine which children are
behind in vaccinations to better target outreach activities, and to rely on community health workers
(CHWs) to help locate these children in their villages.
Measles campaigns took place in early 2003 and late 2005. After the importation of wild polio virus
5
from Nigeria was detected in Benin in 2003, the quality and quantity of national-level supplemental
immunization days (NIDs) for polio were reinforced (appendix table C2). Vitamin A supplementation is
done twice a year, linked to NIDs and other isolated campaigns since 2002, with support from
UNICEF. ACSD incorporated a de-worming strategy into campaigns in the "high-impact" zones in
2003. The Government of Benin introduced de-worming with national campaigns in 2005 using its
own financing.
A16
IIP-JHU | Retrospective evaluation of ACSD in Benin
Year
Oct-Dec
Jul-Sept
Apr-Jun
Jan-Mar
Oct-Dec
Jul-Sept
Apr-Jun
Jan-Mar
Oct-Dec
Jul-Sept
Baseline Jan-Jun
Trimester
Polio campaign
+ Vit A
Polio campaign
9
+ Vit A
Measles
7
campaign
vitamin A
campaign
(CIDA funding)
vitamin A
campaign
(CIDA funding)
National
campaigns
Intro of Hep B
vaccination
(1) ACSD adds
yellow fever and
Vitamin A
deficiency to the
list of illnesses to
be combatted.
(2) 2 doses of
Vit. A are added.
(3) De-worming
campaigns and
ITN distribution
in HIZ
Other
national /
general
EPI+
3 day training of 19 providers (public + private) in EPI+;
vaccine mgmt, policy, stocks, 2ndary effects, surveillance,
15
quality control, equipment mgmt, supervision
3 day (Mar) training for health providers on nutrition (incl.
Vit. A)
Supervision of 7 health centers; 5/7 distributing vitamin A;
10
6/7 cold chains working
Catch-up vaccination activities in Monsourou, Agouna &
13
Houto, Djidja
CHW training in techniques for bed net impregnation and
correct management vaccination cards.
nd
Monitoring of activities/results 2 semester 2002, review
8
of results
Door-to-door immunization campaigns as a results of
8
supervision visits that concluded low coverage rates
20 motorcycles provided to the Zou; 4 motos given to DAA
6
for outreach activities
Training in all health centers for EPI, including: 1) Cold
chain maintenance & 2) Plotting monthly immunization
curves
Situation of EPI: Immunizations coverage low; Cold chain
not functioning in all health centres (HC); No outreach: HC
have not filled in their immunization coverage curves, as
well as lost cases; Immunization plan implemented, but no
monitoring.
EPI+ activities in DAA
Table C2: Timeline of implementation of EPI+ activities in DAA and PAK health zones 2002 to 2007, Benin
2002
2003
2004
IIP-JHU | Retrospective evaluation of ACSD in Benin
A17
8
7
7
10 days of supervision of EPI activities in 17
14
health centers
Regional campaign for ITN trt, vit A distribution
11, 12
& de-working; 37,174 children de-wormed
Catch-up vaccination (all) in villages of Adja7
Ouere
Catch-up vaccination (all) in Ketou & Pobé
st
Monitoring of 1 semester 2003
11 motorcycles for PAK, refrig. for cold chain
Training of qualified health providers in EPI &
7
outreach activities
20 motos for Ouémé, 4 for PAK zone;
6
supervision vehicles for the whole department
Situation of EPI: No outreach implemention
due to transport resources
EPI+ activities in PAK
Year
2005
2006
A18
IIP-JHU | Retrospective evaluation of ACSD in Benin
Oct-Dec
Jul-Sept
Apr-Jun
Jan-Mar
Oct-Dec
Jul-Sept
Apr-Jun
Polio campaign
+ Vit A
Polio campaign
+ Vit A
Polio campaign
+ Vit A .
Polio campaign
Measles
campaign +
deworm
16,423 children vaccinated for polio during NID
Polio campaign
+ Vit A +
deworm
Intro of Hib
vaccine
16,152 children vaccinated for polio during NID
Polio campaign
17
200 CHW trained in EPI promotion
3 day training of auxiliary nurses (aide soignants) in EPI+;
vaccine admin stocks, 2ndary effects, surveillance, quality
19
control, equipment mgmt
17
17
9,172 vaccinated for measles (Campaign coupled with
distribution of albendazole & paracetamol to mobilize
population)
15,936 children vaccinated for polio during NID
Measles epidemic in Djidja with 16+ cases in U5s;
response with investigation & catch-up vaccination
18
campaign in villages in Djidja (623 U5 vaccinated)
15,767 children vaccinated for polio during NID
17
17
3 day training of 20 providers (public + private) in EPI+;
vaccine mgmt, policy, stocks, 2ndary effects, surveillance,
16
quality control, equipment mgmt, supervision
CHW (20) training in EPI management (Jan)
Jan-Mar
EPI+ activities in DAA
National
campaigns
Trimester
Other
national /
general
EPI+
nd
Monitoring activities for the 2 semester for
20
2005 in all health centers; review of results
EPI+ activities in PAK
Insecticide-treated nets (ITNs).
ACSD in Benin utilized different strategies for the provision and promotion of utilization of ITNs in the
l’Ouémé-Plateau and Zou-Collines, including the PAK and DAA health zone.
In the Ouémé-Plateau departments, including the PAK zone, bednets were sold and their use
promoted by women’s groups and through maternity centers. At the end of 2002, UNICEF signed an
agreement with the international NGO, Africare, to support these strategies. Africare had been in
collaboration with a local malaria control project, PROLIPO, to engage women’s groups in malaria
control activities; 270 women in 90 villages were trained in ITN promotion in late 2002. The promotion,
distribution and re-treatment of ITNs were also carried out by over 200 CHWs selected and trained by
UNICEF and the MOH starting in mid-2003. In some cases, but not always, the CHWs were also
members of the women’s groups. Activity reports show that the first ACSD-supported insecticide
treatment campaign in PAK zone treated 9,330 bednets in 2003. Re-treatment campaigns of similar
magnitude have been carried out periodically over the ACSD period in the Ouémé-Plateau
departments, including PAK zone.
In the departments of Zou-Collines, including the DAA health zone, bednets were sold through social
marketing techniques in towns, at larger town markets, as well as at health centers; they were also
sold by intermediaries—community volunteers—in villages. Population Services International (PSI)
was a key partner in this activity. UNICEF and PSI started collaborating in late 2002 with the
promotion of ITNs for pregnant women; Bonne Maman (good mother) bednets were launched in
November 2002 in DAA health zone. Similar to PAK, the strategy in the DAA health zone promoted,
distributed and treated bednets through approximately 200 trained CHWs in 100 villages. Retreatment campaigns were carried out in the community throughout the period of 2003 to 2006.
According to summary presentations given by UNICEF staff, 44,250 ITNs have been distributed at a
subsided price between 2002 and 2006 in the DAA health zone.
Mosquito nets were retreated through periodic community-based re-treatment campaigns in the four
departments, including the PAK and DAA zones; insecticide treatment were also provided at no cost
with support from UNICEF and other partners, such as the national malaria control program (PNLP).
More details about ITN distribution and re-treatment are presented in Annex table C3.
When ACSD was first implemented, the official cost of ITNs was CFA 3500 (~ USD 7.00) for
everyone. The official price was then reduced to CFA 1500 (~ USD 3.00) in 2004 for pregnant women
and children less than five years of age. The price was further reduced in 2005 to CFA 500 (~ USD
1.00) for targeted populations. Program managers report that this most recent price reduction led to a
large increase in demand for bednets, which combined with problems in the financing of the program,
created widespread ITN stock-outs beginning in late 2005 to early-mid 2007. Issues in retreating
bednets, such as the onerous logistics and problems with meeting the required periodicity, prompted
the government to opt for long-lasting bednets in 2005.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A19
Oct-Dec
Jul-Sept
Apr-Jun
Oct-Dec
Jan-Mar
Jul-Sept
Apr-Jun
Jan-Mar
Oct-Dec
Jul-Sept
Jan-Jun
Year Trimester
Price of ITNs at
1500cfa for U5 &
pregnant women
(2500cfa for all
others)
General ITN
information
6
6
6
Continuing
provision of
8
bednets
2nd bed net retreatment campaign8
Free impregnation campaigns (4,509 ITNs trt)
PSI ensured availability of bed nets in HC: 1500 bed nets
(50 MII x 29 centres de santé) for 1500 CFA (approx. $3)
for pregnant women and children less than five years old
were provided to all 29 HC. Local HC restock their bed
nets from Abomey HC. After the sale, the health workers
were used proceeds for a second round of bed net
distribution/provision.
Sale of 4,207 ITNs after CHW training
ITN promotion, distribution & retrt by CHWs (trained)
6
throughout 100 villages
Free impregnation campaigns (2,892 ITNs trt)
Launch of “Bonne Maman” ITNs for pregnant women in
6
Agouna in DAA
Signing of 3 yr. agreement btw. PSI & UNICEF to
6
promote ITNs for pregnant women in Zou-Colllines
No activities - researching conditions
ITN activities in DAA
7
Continuing
provision
8
of bednets
Jan 04 to Jun 04, 1367 ITNs sold through
11
health centers in PAK
Door-to-door ITN retrt campaign with CHW
involvement & BCC thru town criers; 9295 ITNs
trt in PAK (100,000 Ko-Tabs given to Oeume11, 12
Plateau 3 days after start of campaign)
Through 12/31/03, 9765 ITNs sold thru health centres,
7, 9
women’s groups & in villages via ACSD
ITN re-treatment campaign; 10,883 ITNs treated (881 1 place;
7
6206 2 place; 3796 3 place) in 5616 households
ITN promotion, distribution & retrt in village drug kits
throughout 102 villages (30 ITNs of 3 places given to each
locality to sell at 3,500cfa (then reduced to 2500cfa), along
6
with 30 K-Othrine
Sale of 2,300 PNLP ITNs at reduced price of 2000cfa
Free door-to-door impregnation campaigns; 9,330 ITNs trt. in
6, 7, 21
5574 households (fully paid by UNICEF 44mil cfa)
Signing of agreement btw. Africare & UNICEF to promote
ITNs through women's groups & maternities in Oeume6
Plateau ; 270 women trained through PROLIPO in 90 villages
7
for ITN promotion
No activities - researching conditions
ITN activities in PAK
Table C3: Timeline of implementation of bednet activities in DAA and PAK health zones 2002 to 2007, Benin
2002
2003
2004
A20
IIP-JHU | Retrospective evaluation of ACSD in Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A21
Oct-Dec
Jan-Mar
Apr-Jun
Jul-Sept
Oct-Dec
Jul-Sept
Apr-Jun
Jan-Mar
Year Trimester
2005
2006
ITN Stock-outs
1. Price of ITNs
drops to 500cfa
(~1usd)
2. Policy changed
for retreatment
every 12m (from
every 6m)
General ITN
information
ITN Stockouts
CHW
trainings
preceded ITN
treatment
campaigns.
ITN retreatment campaign (cost shared
by PNLP & UNICEF)
ITN activities in DAA
ITN Stock-outs
ITN retreatment campaign; IEC with 30,379 persons, and
22
19,794 bednets retreated
ITN activities in PAK
Case management of childhood illnesses and feeding, including breastfeeding.
Integrated case management of child illness and promotion of improved feeding practices were carried
out in both facilities and the community through the ACSD strategy. The Plateau department,
including the PAK health zone, was one of the first to introduce facility-based IMCI in Benin in June
2001, before ACSD. The DAA health zone was part of the IMCI extension phase, with training for
facility-based workers occurring in late 2002, for doctors and supervisors, and early 2003 for providers.
According to administrative and summary reports, the UNICEF and ACSD teams helped the MOH
coordinate this early IMCI training in DAA. Standard IMCI monitoring tools are utilized, some with
enhancements developed through the PROLIPO project, implemented in collaboration with the US
Centers for Disease Control and Prevention (CDC) and UNICEF in July 2002. Periodic supervisory
visits and review of monitoring data collected from IMCI-compatible health registers have been carried
out to reinforce IMCI implementation. Many of the local monitoring and supervision reports reviewed
by the evaluation team focused on financial and stock management; fewer included quality of care
assessments. It should be noted that the IMCI focuses on stock availability and quality of care, rather
than financial management.
In addition to IMCI training, approximately 180 health trainers and providers in Ouémé-Plateau,
including 44 providers in the PAK zone, were trained in healthy child consultations in mid-2003. These
trainings comprised 6 modules pertaining to the minimum package of nutrition activities: 1) iron and
folic acid supplementation for pregnant women; 2) exclusive breastfeeding for children less than 6
months; 3) complementary feeding and continued breastfeeding for children aged 6-24 months; 4)
vitamin A supplementation for children aged 6-59 months and postnatal supplementation of women; 5)
nutritional management of sick children; and 6) consumption of iodized salt. Child health cards kept at
health facilities included sections to collect information and dates about feeding practices, nutritional
counseling, vitamin A supplementation, and de-worming. It is unknown to what extent these sections
are completed in practice. In the DAA zone, approximately 50 health providers were also trained in
the minimum package of nutrition activities in 2004.
The PAK and DAA health zones were pilot zones for community IMCI. UNICEF organized a series of
workshops and supported community situation analysis exercises to assist in the planning of the
community IMCI strategies and activities and development of materials in late 2002 and early 2003.
MOH officials, local health zone staff and community members and leaders attended the workshops
and participated in the data collection for the situation analysis and planning of activities. According to
MOH officials, the experiences in these zones served as lessons to improve and expand C-IMCI,
which is currently implemented in ten other health zones in Benin.
In mid-2003, 102 remote villages in PAK and 100 remote villages in DAA (the majority in Djidja) were
chosen by local health agents and other local partners, in association with UNICEF, as sites for the
installation of community health workers (CHWs). Local community officials selected two CHWs in
each site. CHWs, 204 in PAK and 200 in DAA, received 5 days of initial training in vaccination
promotion, defaulter tracing, home management of malaria, and re-treatment of bed nets soon after
their selection. Administrative reports describing the training reveal that over 50 local leaders also
participated in this training. In late 2003, CHWs and community leaders received further training for 3
days, on: 1) promotion of exclusive breastfeeding and supplementary feeding; 2) prevention and
home-based management of diarrhea; 3) elimination of stools and hand washing practices; 4)
recording births; 5) medicine management; 6) promotion of vitamin A supplementation; and 7)
communication techniques. Around this time, the CHWs were issued bednets and a medicine box
with ORS, chloroquine, paracetamol, mebendazole, and iron. The medications were to be sold to sick
children in the villages at reasonable prices; a small margin of benefit for the CHWs and community
committees was planned.
Visual aids, also known as “image boxes,” were finalized and pre-tested in mid-2004. These visual
aids included modules to promote appropriate malaria, diarrhea and pneumonia management, as well
as appropriate infant feeding and vitamin A promotion. In mid 2005, the CHWs were also responsible
A22
IIP-JHU | Retrospective evaluation of ACSD in Benin
for promoting appropriate case management, exclusive breast-feeding, complementary feeding and
good nutrition practices, using the new visual aids.
In 2004, a UNICEF-supported operational research project supported by UNICEF allowed the training
of approximately 40 CHWs in the Ketou commune of PAK and 40 CHWs in the Djidja commune of
DAA to manage pneumonia with cotrimoxizole at the community level. Officials in Benin report that
the MOH is currently reviewing this experience and discussing whether community-based treatment of
pneumonia will be adopted as national policy.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A23
Year
JanMar
OctDec
JulSept
1 IMCI supervision of
8
22 health providers
st
Initial 11 day IMCI
training for 24 health
6
providers in DAA
Training of supervisors
to monitor IMCI-trained
6
providers
OctDec
AprJun
C-IMCI activities
DAA
Training of 24 CHW trainers; training of 200
CHWs & 20 arrondisement mayors in promotion
of ITNs, vaccination (2 days training, 2 days
6
promotion in different communities) (majority of
CHWs—140—in Djidja)
3 day training of 192 CHWs & 20 mayors (8
training sites) in breastfeeding, feeding, diarrhea
prevention & management, handwashing,
hygiene, medicine & ITN management, birth
25
registration
3 day workshop in Porto-Novo to develop & adapt
6, 23
CHW training materials
Workshop to develop C-IMCI materials for
Abomey commune (~50 community leaders);
6
community assessment for every locality
No activities: researching situation of health zones
Facility-IMCI
activities
IMCI training for 10
IMCI trainers (doctors);
initial 11 day IMCI
training for 24 health
providers in Zou6
Collines
JulSept
JanJun
Trimester
C-IMCI activities
PAK
Training of 6 trainers
for "health child visits"
training; Training of 44
providers for health
6
child visits in Pobé
102 villages are equipped with village drug
kits (CQ, AAS, Paracetamol, Mebendazole,
ORS, Fe); Training of 320 CHWs &
community leaders in diarrhea home
management, hand-washing, minimum
Training of 30 CHW trainers & supervisors ;
5 day training of 240 CHWs & 40
arrondisement mayors in promotion of ITNs,
vaccination, & home malaria trt. (3 days
training, 2 days promotion in different
6, 7
communities
24
3 day workshop in Porto-Novo to develop &
6, 23
adapt CHW training materials
Identification of 102 CHW sites, community
6, 7
committees & 240 CHWs
10 day workshop/situation analysis to
develop C-IMCI materials for PAK done with
6, 7
community leaders
Facility training for IMCI (2 sessions of 11 days for 24 providers)-early 2001
No ACSD activities: researching situation of health zones
Development of FIMCI supervision tools,
6
with PROLIPO
Pre-test of health child
8
visit tools
Facility-IMCI
activities
Table C4: Timeline of implementation of IMCI+ activities in DAA and PAK health zones 2002 to 2007, Benin
2002
2003
A24
IIP-JHU | Retrospective evaluation of ACSD in Benin
Year
2004
2005
IIP-JHU | Retrospective evaluation of ACSD in Benin
A25
JanMar
3 day supervision activities of CHWs managing
34
pneumonia in Djidja
4 day training session for 10 CHWs from Djidja on
management of pneumonia, danger signs,
33
practices, etc
OctDec
28
7 days CHW supervision done in all communes
by health zone coordinators, Drs, nurses,
midwives, etc; Monitoring of job tools, registers,
management of drug stocks & finances; 1 day
feedback mtg. for CHWs in Djidja & Agbangnizou
32
to go over recommendations & CHW needs
Launching of CHW activities & drug kits in Djidja
CHW supervision by 24 district directors and
other health workers
C-IMCI activities
JulSept
3 day training for 180
healthcare providers (in
4 sessions); minimum
activity package for
nutrition, including iron,
feeding practices,
vitamin A, iodine,
breastfeeding,
management of sick
27
child
Facility-IMCI
activities
DAA
22 participants for 4 day finalization of CHW
visual aids for diarrhea, ARI, & malaria in 3
languages; visual aids for MAP nutrition Porto14, 30
Novo, DAA, PAK, ABD
5 day pre-test of
malarial, ARI, & diarrhea visual aids in 20
31
communities
AprJun
JanMar
Trimester
38/42 health agents
7
trained in IMCI
7 new nurses trained
in facility IMCI
5 days of IMCI
supervision in health
centers
Facility-IMCI
activities
14
2 day group CHW & community supervision
by 2 district medical officers, 19 nurses, & 13
others at health centers; Monitoring of
community participation, job tools, registers,
management of drug stocks & finances;
89/102 communities participated; 654 child
29
treated & 51 child referred since start
5 days supervision of CHWs
22 participants for finalization & pre-test of
CHW visual aids for diarrhea, ARI, & malaria
in 3 languages; visual aids for MAP nutrition
14, 30
– Porto-Novo, DAA, PAK, ABD
; 5 day
pre-test of malarial, ARI, & diarrhea visual
31
aids in 20 communities
11 days supervision of CHWs to organize
on-going supervision, organize community
surveillance committees, give feedback on
BCC activities, control finance mgmt, monitor
7, 9 29
record completion, find target pop.
Planning for situation analysis, &
9
development of case management strategy
nutrition package & completion of record7, 26
books & monitoring tools
C-IMCI activities
PAK
Year
2006
A26
IIP-JHU | Retrospective evaluation of ACSD in Benin
JulSept
OctDec
JanMar
AprJun
OctDec
JulSept
AprJun
Trimester
2 supervisors over 5
days
Training for health
providers in Abomey for
40
micro-insurance
Facility-IMCI
activities
3 day training of 20 participants for breastfeeding
& infant nutrition support groups in Abomey
41
commune
Supervision activities of CHWs managing
38
pneumonia in Djidja
3 day training for 28 trainers on BCC tools/job
36
aids for diarrhea, malaria, ARI, & nutrtion
4 day training session for 11 CHWs from Djidja
on management of pneumonia, danger signs,
37
practices, etc
4 day training session for 9 CHWs from Djidja on
management of pneumonia, danger signs,
35
practices, etc
C-IMCI activities
DAA
Facility-IMCI
activities
KAP survey & presentation of results in each
26
commune of PAK
CHW field supervision for BCC activities,
curative care, surveillance, medicine mgmt,
record completion; coupled with a KAP
survey for 7 days in 92 /102 villages with 274
26, 39
CHW & community members
C-IMCI activities
PAK
Antenatal, delivery and postnatal care.
Antenatal care interventions supported under the ACSD approach of “Focused ANC+,” in Benin
included: 1) focused antenatal care; 2) utilization of ITNs; 3) intermittent preventive treatment for
malaria for pregnant women (IPTi) with a combination of sulfadoxine and pyrimethamine (SP), also
commonly known as Fansidar; 4) prevention of mother-to-child transmission of HIV (PMTCT); and 5)
deworming, and 6) supplementation with iron and folic-acid. “Focused ANC” reorients ANC care to
treat all pregnancies as “at risk”. Starting in the first ANC visit, this strategy is intended to encourage:
1) women to plan for the delivery; 2) planning logistically and financially for evacuation in the case of
complications; and 3) husbands to assist at least one ANC visit to help with this planning. Increasing
the decision power of pregnant women is at the heart of this strategy.
IPTi was introduced for the first time in the “high-impact” zones in late 2003, at the same time as
“focused ANC.” IPTi was implemented in 2004 after a study that compared the effectiveness of
chloroquine versus SP. Prevention of mother-to-child transmission of HIV (PMTCT) was also
introduced at maternity centers in the high impact health zones in 2005 and integrated with “focused
ANC” in 2007. In late 2005 an “ANC kit” was introduced that contained a bednet, iron/folic acid
supplements, SP for IPT of malaria, and mebendezole for de-worming. In the DAA zone this “ANC kit”
was developed to be provided in a special sachet containing all the elements to facilitate distribution.
The ANC kit is sold for CFA 1000 (~ USD 2.00) to a woman at her first antenatal visit to a community
health center or maternity. In the PAK zone these different elements are sold separately to pregnant
women. According to health officials, these kits are sold at a loss, which causes problems with stock
and re-supply.
In 2004, a radio communication system was put into operation with UNICEF support to facilitate
evacuation of obstetrical emergencies in the DAA zone. In PAK zone, the radio system installation
began in 2001 and was completed in 2004.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A27
Oct-Dec
Jul-Sept
Apr-Jun
Jan-Mar
Oct-Dec
Jul-Sept
Apr-Jun
Jan-Mar
Oct-Dec
Jul-Sept
Jan-Jun
Trimester
AMTSL training
6 day training of 16 nurses on
active management of the third
stage of labor (AMTSL) done by 1
gynecologist and 3 nurses in Zou
3-day training for maternity
personnel in management of the
st
neonate (1 session); funded by
42
Unicef
3-day training for maternity
personnel in management of the
nd
neonate (2 session); funded by
42
Unicef
43
3 days of supervision of IPTp ; 2 day
supervision of 18/20 maternities in refocused
44
ANC & IPTp
Development of micro-insurance in Abomey
(care and support for emergency obstetrical
complications).
IPTp starts in Abomey & Agbangnizoun
28
maternities, with maternal cards
Mop-up campaign for TT vaccination among
8
women of reproductive age, Djidja
Provision of medical kits & supplies
8
Community-based & on-going
activities
No activities: researching situation of health zones
Trainings & other
activities
DAA
PAK
Communitybased & ongoing activities
5 day workshop to train 58 participants
(media & community leaders) in
HIV/AIDS and PMTCT
3-day training for maternity personnel
(9 midwives, 9 nurses, 1 Dr.) in
11
management of the neonate
Receipt of materials for
maternities
2 days of supervision of
43
IPTp
10 days of supervision
of SONU & EPI
activities in 17 health
14
centers
Supervision of ANC
activities, with emphasis
11
on IPTp
IPTp introduced into all
9
health centers in PAK
Pilot zone for IPT; Comparative IPTp study between SP in HIZs &
in neighbour zones that use CQ. Study was delayed & results did
not come out before the national launching & adoption of IPT.
Facility training for IMCI --- mid-2001
No ACSD activities: researching situation of health zones
Trainings & other
activities
Table C5: Timeline of implementation of ANC+ activities in DAA and PAK health zones 2002 to 2007, Benin
Year
2002
2003
2004
A28
IIP-JHU | Retrospective evaluation of ACSD in Benin
Year
2005
2006
IIP-JHU | Retrospective evaluation of ACSD in Benin
A29
Oct-Dec
Jul-Sept
Apr-Jun
Jan-Mar
Oct-Dec
Jul-Sept
Apr-Jun
Jan-Mar
Trimester
Training of mid-wives in use of
40
ventouse
Workshop to adapt PMTCT
messages & visual aids for DAA
49
zone, led by Unicef consultants,
50
validation of messages, and
51
contracts with media
5 day training of 16 midwives &
nd
47
nurses in AMTSL (2 session)
4 day training (2 session) for HZ
Drs, midwives, social workers,
46
nurses, lab techs etc in PMTCT
nd
5 day training of 28 nurses,
midwives, lab techs, nurses, Drs in
45
PMTCT & HIV+ neonatal care
Trainings & other
activities
Supervision of ANC activities, IPTp,
40
reanimation of neonate
Introduction of ANC kit for 1000 CFA (~$2).
Includes ITN, SP, de-worming, Folic Acid &
40
Iron suppl.)
Promotion/involvement
of royal courts in development of PMTCT
52
strategies(2 days prep, 3 days field)
2 day workshop with 46 participants from
DAA & Zou to review & validate obstetrical &
neonatal references from health centers to
48
Zou hospital
Community-based & on-going
activities
DAA
Trainings & other
activities
PAK
Monitoring activities for
nd
the 2 semester for
2005 in all health
centers; review of
20
results
Communitybased & ongoing activities
APPENDIX D
Definition of priority indicators in the evaluation of ACSD
A30
IIP-JHU | Retrospective evaluation of ACSD in Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A31
Child
Child
Child
Percentage of children aged
12-23 months who received 3
doses of DPT vaccine before
first brithday.
Percentage of children aged
12-23 months received full
(3x doses) HIB vaccination
before first brithday.
Percentage of children 6 59m who received at least
one high dose vitamin A
supplement within the last 6
months
DPT3
immunization
coverage §
Hib3
immunization
coverage §
Coverage of
vitamin A in last 6
months §
2
3
4
Eligible children receiving
vitamin A in previous 6m
according to mother's report or
immunization card
Eligible children received Hib3
before 12 months of age;
according to immunization
card
3
EXCLUDE CASES: Missing
mother's report and no entry on
vaccination card
EXCLUDE CASES:
Unknwon/missing mother's
report and no card data
IMPUTE TIMING 2:
Missing/invalid date on card:
impute timing with distribution of
known vaccination dates
IMPUTE TIMING 1: Missing card
or vaccination on card: use
mother's report & impute timing
with distribution of known
vaccination dates
PROTOCOL FOR
MISSING/UNKNOWN
DATA
To estimate the children without a card to have rec'd vaccine before 12m of age, the proportion of vaccinations given in the first year is assumed to be the same as the
proportion of children with an immunization card who rec'd the vaccine before 12m of age (MICS manual)
All children 6-59m, still
alive, include MB
All children 12-23m,
still alive, include MB
All children 12-23m,
still alive, include MB
Eligible children received
DPT3 before 12 months of
age; according to
immunization card or mother's
3
report
²
All children 12-23m,
still alive, include MB
DOMINATOR ¹
Eligible children received
measles innoculation before
12 months of age; according to
immunization card, mother's
report or reciept of vaccination
3
during national campaign
NUMERATOR
² CDC 2003 - due to data quality issues, all children with valid data for indicator variables were included
¹ MB=Multiple birth: include all multiple birth children
§ International Consensus Coverage Indicator
Child
DATAFILE
1
INDICATORS
Percentage of children aged
12-23 months who received
measles vaccine before first
birthday
EPI+
ACSD TARGET
Measles
immunization
coverage §
NO.
Table D1: Definition of priority coverage indicators and protocols for missing data
A32
IIP-JHU | Retrospective evaluation of ACSD in Benin
ORS/RHF/increased
fluids for children
with diarrhoea +
continued feeding §
Continued feeding
Increased fluids (IF)
ORT/RHF
Rec'd somewhat less,
about the same or more
(MICS)
Rec'd more (MICS)
recommended home fluids
Eligible children received
ORS, RHF or increased fluids
AND continued feeding
Eligible children were seen at
appropriate health care facility:
excluding pharmacy and other
drug vendors
Eligible children received any
antimalarial mediciation during
illness in previous two weeks
Eligible children received
appropriate antimalarial
medication according to
national policy in previous two
weeks
NUMERATOR
² CDC 2003 - due to data quality issues, all children with valid data for indicator variables were included
¹ MB=Multiple birth: include all multiple birth children
§ International Consensus Coverage Indicator
8
Percentage of children
aged 0-59 months with
diarrhoea receiving ORS
OR RHF OR increased
fluids AND continued
feeding
Care seeking
pneumonia §
7
Child
Child
Percentage of children
aged 0-59 months with
suspected pneumonia
taken to an appropriate
health care facility.
6
ORS packets
Child
Percentage of children
aged 0-59 months with
fever receiving any
antimalarial drugs
Case management
malariaprogrammatic
(programmatic)
ORS
Child
DATAFILE
5
INDICATORS
Percentage of children
aged 0-59 months with
fever receiving appropriate
antimalarial drugs
IMCI+
ACSD TARGET
Case management
malaria (effective)
NO.
EXCLUDE CASES: Reported
treatment of child but missing for
specific location of treatment
Children (0-59) with
cough AND labored
breathing in previous
two weeks, include MB,
exclude deceased
Children (0-59) with
reported diarrhoea in
previous two weeks,
include MB, exclude
deceased
EXCLUDE CASES: Reported
treatment of child but missing for
specific medications used
Children (0-59) with
reported fever in
previous two weeks,
include MB, exclude
deceased
EXCLUDE CASES 2: Reported
treatment of child but
positive/missing for ORS, RHF or
IF and missing for continued
feeding
EXCLUDE CASES 1: Reported
treatment of child but missing for
ORS, RHF and IF and
positive/missing for continued
feeding
EXCLUDE CASES: Reported
treatment of child but missing for
specific medications used
²
Children (0-59) with
reported fever in
previous two weeks,
include MB, exclude
deceased
DOMINATOR ¹
PROTOCOL FOR
MISSING/UNKNOWN
DATA
IIP-JHU | Retrospective evaluation of ACSD in Benin
A33
Percentage of infants
aged 6-9 months who
are breastfed and
receive complementary
food (solid or semisolid
foods)
Percentage of children
aged 20-23 months who
are currently
breastfeeding
Breastfeeding
and
complementary
feeding (6-9
months) §
Continued
breastfeeding
(20-23 months)
§
Consumption of
iodized salt
Consumption of
iodized salt
11
12
13
14
Eligible HH has salt with
>=15ppm iodine
Eligible children still
breasfeeding
Eligible children still
breastfeeding and received
solid/semisolid foods in the
previous 24hr
Eligible children still
breastfeeding and did not
receive any liquids or foods in
previous 24h
Women iniatiated
breastfeeding within the first
hour after delivery
NUMERATOR
² CDC 2003 - due to data quality issues, all children with valid data for indicator variables were included
¹ MB=Multiple birth: include all multiple birth children
HH
HH
Child
Child
Child
Women
DATAFILE
§ International Consensus Coverage Indicator
Percentage of
households consuming
iodized salt: exclude HH
with no salt
Percentage of
households consuming
iodized salt: include HH
with no salt
Percentage of infants
aged 0-5 months who
are exclusively breastfed
10
9
Exclusive
breastfeeding
through 6
months (0-5m) §
INDICATORS
Percentage of newborns
put to the breast within
one hour of birth; most
recent live birth previous
12m
IMCI+
ACSD
TARGET
Timely initiation
of breastfeeding
§
NO.
²
All HH with completed
surveys
All HH with completed
surveys and salt
available for testing
Children (20-23): most
recently born (include
only one MB) still alive
& living with mom.
Children (6-9): most
recently born (include
only one MB) still alive
& living with mom.
Children (0-5): most
recently born (include
only one MB) still alive
& living with mom.
Women with a birth in
previous 12m
DOMINATOR ¹
EXCLUDE CASES: Missing salt test
EXCLUDE CASES: Missing salt test
and HH with no salt
EXCLUDE CASES: Missing for still
breastfeeding
EXCLUDE CASES 2: Postive/missing
for all feeding variables AND missing
for still breasfeeding
EXCLUDE CASES 1: Missing for all
feeding variables AND postive/missing
for still breastfeeding
EXCLUDE CASES 2: Negative/
missing for all feeding variables AND
missing for still breasfeeding
EXCLUDE CASES 1: Missing for all
feeding variables AND postive/missing
for still breastfeeding
EXCLUDE CASES: Reported ever
breastfeeding, but missing timing of
initiation
PROTOCOL FOR
MISSING/UNKNOWN DATA
A34
IIP-JHU | Retrospective evaluation of ACSD in Benin
Eligible pregnant woman slept
under an ITN mosquito net last
night
Eligible child slept under an
ITN mosquito net last night
Eligible pregnant woman slept
under a mosquito net last night
NUMERATOR
² CDC 2003 - due to data quality issues, all children with valid data for indicator variables were included
¹ MB=Multiple birth: include all multiple birth children
§ International Consensus Coverage Indicator
17
Woman
16
Percentage of pregnant
women sleeping under an
insecticide treated mosquito
net last night (Use trt'd <=12
months due to heaping at
12m)
Woman
Percentage of children aged
0-59 months sleeping under
an insecticide treated
mosquito net (Use trt'd <=12
months due to heaping at
12m)
Effective use of
bednets by
children < 5yr §
Effective use of
bednets by
pregnant women
Woman
DATAFILE
Percentage of pregnant
women sleeping under any
mosquito net last night
INDICATORS
Use of bednets by
pregnant women
ITNs
ACSD TARGET
15
NO.
²
All eligible pregnant
women
All children under five,
still living
All eligible pregnant
women
DOMINATOR ¹
EXCLUDE CASES 1: Missing
ITN data
(a) Net obtained <=12m prior
AND missing if treated when
obtained (b)Treated net
obtained AND missing months
ago obtained (c) Treated the
net after obtaining but missing
months ago treated
EXCLUDE CASES 2:
Unknown/missing for a,b & c
and positive/missing/unknown
for slept under a net last night
EXCLUDE CASES 3:
Positive/missing for a, b & c
AND unknown/missing for slept
under a bed net last night
EXCLUDE CASES:
Unknown/missing for slept
under a bed net last night
PROTOCOL FOR
MISSING/UNKNOWN
DATA
IIP-JHU | Retrospective evaluation of ACSD in Benin
A35
DATAFILE
Women
Women
Women
Women
Women
Women
Women
Percentage of pregnant
women who report at least 3
prenatal visits to a trained
worker
Percentage of pregnant
women receiving intermittent
preventative treatment for
malaria during pregnancy**
Percentage of newborns
protected against tetanus:
Mother rec'd at least 2 doses
of TT during pregnancy
Percentage of pregnant
women receiving 3 months of
iron supplementation.
Percentage of births
attended by skilled health
personnel
Percentage of newborns
receiving a postnatal visit by
a trained worker within 3
days of delivery. (<3 days)
Percentage of women
receiving vitamin A
supplementation within 2
months of birth
Intermittent
malaria
treatments in
pregnancy
TT2 coverage
during pregnancy
§
Pregnant women
take 3 months iron
supplements
Skilled attendant
at delivery* §
Postnatal visit
within 3 days of
delivery, skilled
HCW*
Postnatal
supplementation
with Vitamin A §
19
20
21
22
* Skilled Health Care Worker
(a)Doctor or Nurse/Midwife
All eligible
women with a
pregnancy
resulting in a live
birth in the
previous 12m
¹²
DOMINATOR
EXCLUDE CASES: Unknown/missing if
received vitamin A
EXCLUDE CASES 2: Noninstitutional
delivery and positive/missing skilled
HCW and positive/missing received
postnatal care
EXCLUDE CASES 1: Unknown/missing
place of delivery and no data for
postnatal care
EXCLUDE CASES: Unknown/missing
data for birth attendant
EXCLUDE CASES: Unknown/missing if
received iron or received iron but for
unknown time period
EXCLUDE CASES: Unkwown/missing if
received TT or received TT but
unknown dosage
EXCLUDE CASES: Received medicine
during pregnancy for malaria but
unknown.missing type of medicine
EXCLUDE CASES: Unkonwn/missing
for number prenatal visits AND
positive/missing for skilled HCW
PROTOCOL FOR
MISSING/UNKNOWN DATA
(b) Doctor, Nurse/Midwife or Aux. midwife
preg
** IPT for pregnant women: at least 1 dose of SP during pregnancy
Eligible women received
vitamin A supplementation
within 2 months of delivery
(a) Eligible women delivered at
an institutional facility (nondomestic)
(b) Eligible women who
delivered domestically
received at least one postnatal
checkup within 3 days of
delivery with a trained health
care worker
Eligible women delivered with
a trained health care worker.
Eligible women received iron
supplementation daily for at
least 90 days
Eligible women received at
least two doses of tetanus
toxoid during the pregnancy
Eligible women received at
least one dose of SP during
the pregnancy
Eligible women received 3+
prenatal care visits with a
trained health care worker
² CDC 2003 - due to data quality issues, all children with valid data for indicator variables were included
¹ MB=Multiple birth: include all multiple birth children
§ International Consensus Coverage Indicator
24
23
NUMERATOR
Most recent live birth within previous 12m
INDICATORS
3+ prenatal visits,
skilled HCW*
ANC+
ACSD TARGET
18
NO.
A36
IIP-JHU | Retrospective evaluation of ACSD in Benin
Household
Household
Stunting (low height for
age) among children 2459 months of age*
Wasting (weight for
height) among children 023 months of age*
Underweight (weight for
age) for children 0-59
months of age*
Under-five mortality rate
1
2
3
4
§ See appendix L for more details
* Also calculated for children 0-59 months of age
Birth
history
extracted
from
women’s
file
Household
DATAFILE
INDICATORS
NO.
Table D1: Definition of priority impact indicators
The probability of dying
between birth and exact
age five years
N/A
Cases with improbable values for
weight-for-age are excluded from
analysis; improbable defined as
+/- 4 standard deviations of Z
score relative to the overall
median Z score value from the
crude datafile
Children aged 0-59 months who:
1. Have a reported (nonmissing) birth month & year
2. Have a valid (non-missing)
anthropometric measure
3. Slept in the house the night
before the survey
4. Live with biological mother
Expressed as 1000 live births
Cases with improbable values for
weight-for-height are excluded
from analysis; improbable
defined as +/- 4 standard
deviations of Z score relative to
the overall median Z score value
from the crude datafile
Cases with improbable values for
height-for-age are excluded from
analysis; improbable defined as
+/- 4 standard deviations of Z
score relative to the overall
median Z score value from the
crude datafile
PROTOCOL FOR EXCLUSION
OF CASES§
Children aged 0-23 months who:
1. Have a valid (non-missing)
anthropometric measure
2. Slept in the house the night
before the survey
3. Live with biological mother
Children aged 24-59 months
who:
1. Have a reported (nonmissing) birth month & year
2. Have a valid (non-missing)
anthropometric measure
3. Slept in the house the night
before the survey
4. Live with biological mother
Moderate and Severe:
Children with <-2 z scores
for height for age based on
the 2006 WHO growth
curves53
Severe: Children with <-3 z
scores for height for age
based on the 2006 WHO
growth curves53
Moderate and Severe:
Children with <-2 z scores
for weight for height based
on the 2006 WHO growth
curves53
Severe: Children with <-3 z
scores for weight for
height based on the 2006
WHO growth curves53
Moderate and Severe:
Children with <-2 z scores
for weight for age based on
the 2006 WHO growth
53
curves
Severe: Children with <-3 z
scores for weight for age
based on the 2006 WHO
growth curves53
DOMINATOR§
NUMERATOR
IIP-JHU | Retrospective evaluation of ACSD in Benin
A37
SOURCE OF
DEFINITION
DHS standard
calculation of wealth
quintiles
(http://www.childinfo
.org/mics/mics3/doc
s/DHS%20Wealth%
20Index%20(DHS%
20Comparative%20
Reports).pdf)
CONTEXTUAL
VARIABLE
Wealth quintiles
For the calculation of
wealth quintiles for the
ACSD evaluation,
Cotonou was removed
and the indices calculated
for households in the HIZ
and comparison area only.
All household assets and
utilities are dichotomized
into indicator variables.
Principle components
analysis is performed
using all the indicator
variables to standardize
the weights of the
variables using the first
principle factor. Each
household is then
assigned a weighted index
value, based on its
reported assets and
utilities. Households are
then divided into quintiles
based on their index
value.
DESCRIPTION OF
DEFINITION
Table D3: Definition of contextual variables used in the ACSD evaluation
M27b (hv214)
M27c (hv215)
M26 (hv226)
Wall material
Roofing material
Cooking fuel
Main Lighting
source
Possession of
electricity
radio
television
refrigerator
bicycle
mobylette/moto
Car/truck
M25
(hv206)
(hv207)
(hv208)
(hv209)
M28(hv210)
(hv211)
(hv212)
M26a(sh26a)
M27 (hv213)
Flooring material
Toilet/latrine
M21 (hv021)
M23 & M24
(hv205&hv 225)
Water source
Assets
2001
Recode variable
(questionnaire
number)
Possession of
Car, moto or bike
Stereo
Refrigerator
Iron
Improved Stove
Bed or mattress
Phone
Radio
DVD or VCR
Sofa
Sewing machine
Fan
Generator
Television
Land ownership
Cooking fuel
Main Lighting
source
Roofing material
Wall material
Flooring material
Toilet/latrine
Water source
Assets
M25
M25
M25
M25
M25
M25
M25
M25
M25
M26A
M26
M27C
M27B
M27
M23 & M24
M21
Questionnaire
number
2006/7
APPLICATION OF DEFINITION TO DATAFILES
A38
IIP-JHU | Retrospective evaluation of ACSD in Benin
Improved
Water Source
CONTEXTUAL
VARIABLE
MDG water and
sanitation definitions
(http://www.unicef.o
rg/wes/mdgreport/d
efinition.php)
SOURCE OF
DEFINITION
Unimproved drinking
water sources include:
1) Unprotected well, 2)
Unprotected spring, 3)
Rivers or ponds, 4)
vendor provided water, 5)
Bottled water, 6) Tanker
truck water
Improved drinking
water sources include:
1) Household connection,
2) Public standpipe, 3)
Borehole, 4) Protected
dug well, 5) Protected
spring, 6) Rainwater
collection
DESCRIPTION OF
DEFINITION
piped into
yard/plot
public tab
EAU COURANTE AILLEURS
ROBINET PUBLIC/BORNE FONTAINE
No label
No label
AUTRE
tanker truck
rainwater
EAU EN BOUTEILLE
CAMION CITERNE
AUTRE EAU DE PLUIE
n/a
√
NO
NO
NO
√
NO
NO
NO
NO
rainwater in
a cistern
EAU DE PLUIE DANS LA CITERNE
NO
NO
river, stream
NO
RIVIERE/MARIGOT/MARE
SOURCE AMENAGEE
NO
NO
NO
open well
PUITS NON PROTEGE
EAU DE SURFACE
spring
√
√
protected
well
FORAGE EQUIPE DE POMPE MANUEL
PUITS BUSE OU PROTEGE
√
√
√
√
HID DHS 2006/7
√
√
√
√
DHS
2001
manual
pumped
water
EAU DE PUITS
piped into
dwelling
EAU COURANTE A LA MAISON
EAU DU ROBINET
Questionnaire (m21);
DHS
datafile
label
(hv201)
APPLICATION OF DEFINITION TO DATAFILES
IIP-JHU | Retrospective evaluation of ACSD in Benin
A39
Improved
Sanitation
Facilities
CONTEXTUAL
VARIABLE
MDG water and
sanitation definitions
(http://www.unicef.org/w
es/mdgreport/definition.
php)
SOURCE OF
DEFINITION
Improved sanitation
facilities include:
1) Connection to a
public sewer,
2) Connection to a
septic system,
3) Pour-flush latrine,
4) Simple pit latrine,
5) Ventilated
improved pit latrine.
.
Unimproved
sanitation facilities
include:
1) Public or shared
latrine,
2) Open pit latrine,
3) Bucket latrine.
DESCRIPTION OF
DEFINITION
AUTRE
PAS DE TOILETTES /NATURE
shared
not shared
no facility,
bush, field
shared
LATRINES À TINETTE
unventilate
d, open pit
latrine
LATRINE À FOSSE NON VENTILÉE
LATR. SUSPENDUE/SUR PILOTIS
RÉSEAU D’ÉGOUT
TOILETTE À CHASSE
LATRINE À FOSSE VENTILÉE
Questionnaire (m23 & m24)
DHS
datafile
variable
label
ventilated
latrine shared or
not shared
flush toilet
- shared or
not shared
NO
√
NO
NO
√
NO
NO
NO
√
√
DHS 2006-2007
NO
NO
NO
√
√
DHS
2001
APPLICATION OF DEFINITION TO DATAFILES
APPENDIX E
Comparison of survey questions utilized for calculation of priority coverage
indicators
Please note that the questionnaires used in surveys analysed as part of the retrospective
evaluation are available from the IIP-JHU evaluation team upon request.
Table E1: Questions utilized for priority indicator calculation from DHS 2001, ACSD survey 2003,
DHS/supplemental survey 2006-7
NO.
ACSD TARGET
DHS Questionnaire
2001
ACSD Questionnaire
2003
DHS Questionnaire
2006/2007
EPI+
Measles immunization
coverage
Have vaccination card
Have vaccination card (q458);
Have vaccination card
(q404); Measles innoc. on
Measles innoc. on card (q460);
(q458); Measles innoc. on
card (q405); Rec'd other Rec'd other vaccines (q462); Mom
card (q460); Rec'd other
vaccines (q407); Mom
report of measles innoc (q463G);
vaccines (q462); Mom report
report of measles innoc
rec'd vaccine during campaign
of measles innoc (q463G)
(q414)
(q465)
2
DPT3 immunization
coverage
Have vaccination card
(q458); DPT3 on card
(q460); Rec'd other
vaccines (q462); Mom report
of DPT(q463E); number of
doses(q463F)
Have vaccination card
(q404); DPT3 on card
(q405); Rec'd other
vaccines (q407); Mom
report of DPT(q412);
number of doses(q413)
Have vaccination card (q458);
DPT3 on card (q460); Rec'd other
vaccines (q462); Mom report of
DPT(q463E); number of
doses(q463F)
3
Hib3 immunization
coverage
N/A
N/A
Have vaccination card (q458); Hib3
innoc. on card (q460)
4
Coverage of vitamin A in
last 6 months
1
Have vaccination card
Have vaccination card
Have vaccination card (q458); VitA
(q458); VitA on card (q460); (q404); VitA on card (q405); on card (q460); Mother's report
Mother's report (q457)
Mother's report (q403)
(q457)
IMCI+
A40
5
Case management
malaria (effective)
Had fever(q515); gave
meds (q517); what meds
Had fever(q466); gave meds
(q518); prescribed meds
(q466A); what meds (q466B)
(q523); what meds
prescribed(q524)
6
Case management
malaria-programmatic
(programmatic)
Had fever(q515); gave
meds (q517); what meds
Had fever(q466); gave meds
(q518); prescribed meds
(q466A); what meds (q466B)
(q523); what meds
prescribed(q524)
7
Care seeking pneumonia
Suspected pneum. (q467 &
q468); consulted for
treatment (q470); where
consulted (q471)
Had fever(q466); gave meds
(q473); what meds (q474)
Had fever(q466); gave meds
(q473); what meds (q474)
Suspected pneum. (q511 &
Suspected pneum. (q467 & q468);
q512); consulted for
consulted for treatment (q470);
treatment (q513); where
where consulted (q471)
consulted (q514)
IIP-JHU | Retrospective evaluation of ACSD in Benin
NO.
ACSD TARGET
ORS/RHF/increased
fluids for children with
diarrhoea + continued
feeding
8
DHS Questionnaire
2001
Had diarrhea (q475)
Had diarrhea (q475)
ORS ORS (q478a)
ORS (q506a)
ORS (q478a)
ORT/RHF RHF (q478b)
RHF (q506b)
RHF (q478b)
Increased fluids (q504)
Increased fluids (q476)
Continued feeding (q505)
Continued feeding (q477)
Continued feeding Continued feeding (q477)
10
DHS Questionnaire
2006/2007
Had diarrhea (q501)
Increased fluids (IF) Increased fluids (q476)
9
ACSD Questionnaire
2003
Timely initiation of
breastfeeding
Ever breastfed (q323);
Ever breastfed (q440);
Timing of BF initiation
Timing of BF initiation (q441)
(q3243)
Exclusive breastfeeding Still breasfeeding (q445);
Still breasfeeding (q326);
through 6 months (0liquids in last 24h (q492a-e); liquids/food in last 24h
5m)
food in last 24h (q493a-j)
(q331b-g)
11
Breastfeeding and
complementary feeding
(6-9 months)
12
Continued
breastfeeding (20-23
months)
13
Consumption of iodized
salt
14
Consumption of iodized
salt
Still breasfeeding (q445);
food in last 24h (q493a-j)
Still breasfeeding (q445)
Iodized salt (q35)
Still breasfeeding (q326);
food in last 24h (q331g)
Still breasfeeding (q326)
Ever breastfed (q440); Timing of
BF initiation (q441)
Still breasfeeding (q445); liquids in
last 24h (q492a-e); food in last 24h
(q493a-j)
Still breasfeeding (q445); food in
last 24h (q493a-j)
Still breasfeeding (q445)
Iodized salt (q29)
ITNs
15
16
17
Use of bednets by
pregnant women
Pregnant (q226); Slept
under net last night (q494)
Pregnant (q221); Slept
under net last night (q602)
Pregnant (q226); Slept under net
last night (q494)
Effective use of
bednets by children <
5yr
Child slept under net last
night (q465C); How long ago
was net obtained (q465E);
Was a treated net obtained
(q465F); Was the net ever
treated(q465G); How long
ago treated (q465H)
Child slept under net last
night (q526); Was the net
ever treated(q527); How
long ago treated (q528)
Child slept under net last night
(q465C); How long ago was net
obtained (q465E); Was a treated
net obtained (q465F); Was the net
ever treated(q465G); How long
ago treated (q465H)
Effective use of
bednets by pregnant
women
N/A
IIP-JHU | Retrospective evaluation of ACSD in Benin
Pregnant (q221); slept
under net last night (q601);
Was the net ever
treated(q602); How long
ago treated (q603)
N/A (data in HH file)
A41
NO.
ACSD TARGET
DHS Questionnaire
2001
ACSD Questionnaire
2003
DHS Questionnaire
2006/2007
ANC+
18
3+ prenatal visits,
skilled HCW
Prenatal care and who did
Prenatal care (q303);who
you consult (q407); Number did you consult (q304);
of visits (q409)
Number of visits (q306)
19
Intermittent malaria
treatments in
pregnancy
Took meds for malaria
(q421); Which meds (q422)
20
TT2 coverage during
pregnancy
Rec'd TT (q415); number of Rec'd TT (q308); number of Rec'd TT (q415); number of doses
doses (q416)
doses (q309)
(q416)
21
Pregnant women take 3
Rec'd iron (q417); Number
months iron
of days took iron (q418)
supplements
Rec'd iron (q313); Number Rec'd iron (q417); Number of days
of days took iron (q314)
took iron (q418)
22
Skilled attendant at
delivery
Assisted with birth (q320)
23
Postnatal visit within 3 Location of delivery (q427);
days of delivery, skilled Rec'd postnatal care if nonHCW
institutional delivery (q429)
Assisted with birth (q426)
Took meds for malaria
Took meds for malaria (q421);
(q223); Which meds (q224) Which meds (q422)
N/A
Days after delivery rec'd
care (q430); who performed
care (q431)
24
A42
Postnatal
supplementation with
Vitamin A
Rec'd vitamin A (q433)
Prenatal care and who did you
consult (q407); Number of visits
(q409)
Assisted with birth (q426)
Location of delivery (q427); Rec'd
postnatal care if non-institutional
delivery (q429)
Days after delivery rec'd care
(q430); who performed care (q431)
Rec'd vitamin A (q322)
Rec'd vitamin A (q433)
IIP-JHU | Retrospective evaluation of ACSD in Benin
APPENDIX F
Methodology and implementation of household surveys
in Benin 2001 to 2007
The methodologies and implementation of households surveys re-analyzed for the ACSD
retrospective evaluation are presented in table F1. Less documentation of the methods and
implementation were available for the ACSD-CDC 2003 survey and the Benin-UNICEF Cooperation
Baseline coverage 2005 survey. These surveys are presented in the body of the report, but should
be interpreted with caution due to questions about the data quality and the exact methodologies
utilized. A full report describing data quality issues in the ACSD-CDC 2003 survey is available on
request from the JHU evaluation team. Table F2 presents a general review of the surveys and data
sources that were not utilized in the main ACSD retrospective evaluation, extracted from available
documents. The note at the bottom of the table provides an explanation of non-inclusion in the
main ACSD retrospective evaluation. For further data sources (including those outside the health
sector), an excellent review of all studies and surveys carried out in Benin and supported by UNICEF
between 2001-2006 is available.54
Comparability between surveys pertinent to the evaluation is highlighted below: 1) the
comparability of the Benin DHS 2006 and the ACSD supplemental DHS in 2007, which were
combined for endline estimates, and 2) the comparability of the Benin 2001 DHS and the Benin
2006 DHS, from which estimates which are utilized for the before-after and differences-indifferences comparisons.
Comparability of the Benin DHS 2006 and the ACSD supplemental DHS 2007.
The data from the supplemental surveys carried out in the HIZs in May of 2007 were merged with
the data from households surveyed during the DHS 2006. Although these surveys were
methodologically similar, there are differences that should be noted, as they might impact the
calculation of certain indicators:
•
The questionnaires in the 2007 supplemental survey were much shorter than standard
DHS surveys making the questionnaire administration easier for both interviewers and
respondents
•
Interviewer performance was better because of experience with the DHS 2006,
additional training and shorter questionnaires
•
Stronger supervision in the 2007 supplemental survey, with constant supervision
•
Overall better quality of the supplemental survey which built on lessons from the errors
in the 2006 DHS survey
•
The period of the data collection in the 2001 and 2006 DHS surveys was between
August and September which spans a part of the dry season and a part of the rainy
season. The supplemental survey was carried out in May of 2007, which was the start
of the rainy season.
•
In the national DHS surveys in 2001 and 2006, antimalarial medicines, ORS sachets,
iron supplements and vitamin A capsules examples were shown to women interviewed.
These medication samples were not available during the data collection for the 2007
supplemental survey until the 2nd and 3rd weeks of data collection (approximately ½ data
collection period).
Comparability of the Benin DHS 2001 and the Benin DHS 2006.
The 2001 and 2006/7 DHS utilized for comparison in the evaluation are very similar in
methodology and conduct, including: the sampling strategy, technical assistance provided, field
agents recruited, and data processing procedures. These two surveys did have different greatly
sample sizes and levels of stratification. In 2006, the DHS survey was conducted in conjunction
IIP-JHU | Retrospective evaluation of ACSD in Benin
A43
with an economic survey that was representative at the commune level. Interviewer training,
especially for the vaccination and bed net modules, was more intensive in 2006.
A44
•
In 2006, there was a concurrent survey (EMICOV) conducted in conjunction with the
DHS. This survey required data collection in 77 communes and in 95% of the
arrondissements in order to ensure representative data at the commune level. There
was no survey conducted in conjuction with the 2001 DHS
•
In 2001, Benin was divided into 6 departments (regions) which were utilized for
stratification; in 2006 there were 12 administrative departments in Benin;
•
The number of households and women interviewed was much greater in 2006 compared
to the 2001 DHS: 5769 households in 2001 and 17,511 households in 2006, 6219
women in 2001 compared to 17,794 women in 2006 ;
•
Interviewer training was more intensive in 2006 for vaccination due to the introduction of
the pentavalent vaccine in 2005. Training was also more intensive for identification of
types of nets and the types of re-treatment kits.
IIP-JHU | Retrospective evaluation of ACSD in Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A45
General
Yes
Sampling methods /
size; Sampling frame;
Revised questionnaire;
Datafile for analysis;
Report of data analyses
Yes
Sampling methods /
size; Sampling frame/
selection/weights;
Revised questionnaire
Training manual;
Interviewer manual;
Supervisor manual;
Datafile for analysis;
Report of data analyses
Datafile
available for
reanalysis
Survey
documentation
available
CERTI
INSAE (Macro)
Sampling methods/
size; Revised
questionnaire Datafile
for analysis; Report of
data analyses
Yes
CEFORP
"High Impact" zones
PAK & DAA
Implementing
Agency (& TA)
31 communes in 5
departments (Zou,
Collines, Oeume,
Plateau, Borgou)
(includes ACSD
"HIZ")
National
BENIN-UNICEF
COOPERATION BASELINE
COVERAGE SURVEY 2005
CDC-ACSD 2003
Geographic
Coverage
SURVEY COMPONENT
DHS 2001
DHS 2006
Sampling methods /
size; Sampling frame/
selection/weights;
Revised questionnaire;
Training manual;
Interviewer manual;
Supervisor manual;
Datafile for analysis;
Report of data analyses
Yes
INSAE (Macro)
National
BENIN – SURVEYS 2001 - 2007
Table F1: Methodology and implementation of household surveys in Benin 2001 to 2007 presented in the ACSD evaluation report
Sampling methods
/ size; Sampling
frame/
selection/weights;
Revised
questionnaire
Training manual;
Interviewer manual;
Supervisor manual;
Supervisory field
report; Datafile for
analysis; Report of
data analyses
Yes
INSAE (Macro)
"High Impact"
zones PAK &
DAA
ACSD
SUPPLEMENTARY
DHS 2007
A46
IIP-JHU | Retrospective evaluation of ACSD in Benin
Sampling &
enumeration
Done at same time as
questionnaire
administration
Done by survey
teams in field
systematically from
updated lists
Benin Report &
Standard CDC
Protocol: Listing of
HHs done before the
survey
Benin Report:
Randomly from listed
households Standard
CDC Protocol: 30
contiguous HHs
selected (using random
starting point)
Standard systematic
sample from listed
HHs
Mapping/ listing
Household
selection
25 (1097 HHs)
Standard systematic
sample from listed
HHs
Household listing done
in selected clusters by
the team leaders before
survey start
24
National: 750 clusters
total HID: 29 PAK, 30
DAA, 59 Total
Departments: 259
clusters HID: 20
PAK, 25 DAA, 45
Total
HH listing done in
selected clusters by
trained team leaders &
other agents in May &
June 2001;
Segmentation for ZD
larger than 399 HHs
30 (2610 HHs)
Number of
clusters
~25
HID: 44 PAK, 43
DAA, 87 total
National: 247
clusters HID: 8 PAK,
9 DAA, 17 Total
Stratification &
sampling of
clusters
2 stage sampling,
cluster selection
stratified by commune
and urban/rural
residence; clusters
chosen from census
2002
DHS 2006
2 stage sampling
stratified by health
zone; clusters chosen
from census 2002
BENIN-UNICEF
COOPERATION BASELINE
COVERAGE SURVEY 2005
Number of
households per
cluster (total HH
in HIDs)
2 stage sampling,
clusters selection
stratified by health
zone; clusters chosen
from census 2002
CDC-ACSD 2003
2 stage sampling,
cluster selection
stratified by 6
departments (+
Cotonou) and
urban/rural residence
(13 strata); clusters
chosen from census
1992
SURVEY COMPONENT
DHS 2001
BENIN – SURVEYS 2001 - 2007
Standard
systematic
sample from
listed HHs
Same as DHS
2006
35 (1540 HHs)
HID: 40 clusters
total
Clusters not
included in DHS
2006 using same
methodology as
DHS 2006 - done
by INSAE & revised
by Macro
ACSD
SUPPLEMENTARY
DHS 2007
IIP-JHU | Retrospective evaluation of ACSD in Benin
A47
Questionnaires
French; Fon; Yoruba
Household, women's
Socio-demographic
Info; Reproduction;
Contraception;
Pregnancies, ANC, &
breastfeeding;
Vaccination; Child
health & illness;
Hygiene, marital status,
work of women
Pre-tested 4 days in
Aug 2003 in 2 zones
Abomey & Ketou (for
interviewer &
respondent
understanding), w/
revisions after
French; Adja; Bariba;
Fon; Yoruba;
Ditamari
Household, women's
[men's]
Socio-demographic
Info; Reproduction;
Contraception;
Pregnancies, ANC, &
breastfeeding;
Vaccination & child
health and nutrition;
Marriage and sexual
activity; Fertility
preferences; Work of
women & partner;
HIV/AID and STIs;
Female genital cutting
Pre-tested in 4 urban
& rural clusters
(interviewer &
respondent
understanding) 73
HH, 300 women’s,
100 men questionn.
administered;
Training, revision &
pre-test from 15 May
to 5 June 2001
Language of
questionnaire
Questionnaires
used
Modules
included in
women's/child
questionnaire
Pre-test / pilot
SURVEY COMPONENT
CDC-ACSD 2003
DHS 2001
Pre-tested for 6 days
in 4 rural & urban
clusters (done by 27
interviewers)
None – as
modules used in
DHS 06
Socio-demographic
Info; Reproduction;
Pregnancies, ANC,
Breastfeeding,
Vaccination, child
health and nutrition
Socio-demgraphic Info;
Reproduction;
Contraception;
Pregnancies, ANC, &
breastfeeding;
Vaccination & child
health and nutrition;
Marriage and sexual
activity; Fertility
preferences; Work of
women & partner;
HIV/AID and STIs;
Female genital cutting;
Maternal mortality
Women's: Sociodemographic Info;
Micronutrients & IPTp;
KAP child illnesses;
KAP HIV/AIDS; KAP
child trafficking & labor;
Female Genital Cutting;
Exposure to IEC Child:
Birth certificate; vitamin
A; breastfeeding;
prevalence & care for
diarrhea, fever, & ARI;
de-worming
Pre-tested 2 days in
Dec 2004 in 4
different language
zones (for interviewer
& respondent
understanding) 253
HH, 324 women’s,
113 child questionn
administered
Household,
women's
French
ACSD
SUPPLEMENTARY
DHS 2007
Household, women's
[men's]
French
DHS 2006
Household, women's,
children's
French
BENIN-UNICEF
COOPERATION BASELINE
COVERAGE SURVEY 2005
BENIN – SURVEYS 2001 - 2007
A48
IIP-JHU | Retrospective evaluation of ACSD in Benin
Training
Unknown - but see
pre-test
Yes; 3rd week of
training for French &
local language
simulations
Training content
Practice survey
admin in field
Logistics &
timing
1 day of simulations
in local languages
with translation of key
words
Training took place 59 Aug 2003 in
Abomey for
interviewers & editors
Training took place 9
- 30 July
(interviewers, health
agents) - many had
University degree
Knowledge tests
used to choose field
agents
Yes
Yes - but unknown
CDC-ACSD 2003
Manuals
SURVEY COMPONENT
DHS 2001
Yes - see pre-test
Based on training
manual
6 days of training
done for interviewers,
editors and data entry
agents
Yes - but unknown
BENIN-UNICEF
COOPERATION BASELINE
COVERAGE SURVEY 2005
DHS 2006
Unknown
Based on standard
DHS modules
Training took place
23 may - 13 June
2006
Standard DHS
manual (longer
version than suppl-07
manual)
BENIN – SURVEYS 2001 - 2007
interviewers were
the same as DHS
2006 and were
experienced with
questionnaire
None; as
Supervisor and
interviewer
manual available
Led by El-Aarbi
Housni (Macro)
from April 23-27
with INSAE;
Training
shortened as
teams same field
agents as DHS06
Interviewer manual;
also manual for
team supervisor
ACSD
SUPPLEMENTARY
DHS 2007
IIP-JHU | Retrospective evaluation of ACSD in Benin
A49
Field
organization /
work
2 Aug to 13 Nov 2001
Period of field
work
Number of
teams
Survey start-up
9 teams
Survey team
composition
2 months into survey,
teams were rearranged due to
attrition of 4
interviewers
1 team leader; 1
controler/editor; 2
interviewers; 2 male
interviewer; health
agent; driver
SURVEY COMPONENT
DHS 2001
Slight delay btw. Pretest & survey start -X-mas holiday
4 Jan to 8 Feb 2005
08 Aug to 05 Sept
2003
27 (?) teams
8 teams (4 teams per
health zone)
Done at same time
(perhaps overlapped)
as training
Total of 12
supervisors; 27 team
leaders; 259 guides;
54 interviewers
BENIN-UNICEF
COOPERATION BASELINE
COVERAGE SURVEY 2005
1 interview for
household quest; 1
controler/editor; 2
interviewers for
women's quest
CDC-ACSD 2003
DHS 2006
April 29th to June
1st (start of rainy
season)
There was a 6 week
pause in between
training and field
collection -- 3 days of
refresher training was
given to interviewers
3 August to 18 Nov
(all 3 months in same
regions - dry season)
4 teams (2 in
PAK and 2 in
DAA)
ACSD
SUPPLEMENTARY
DHS 2007
1
supervisor/editor
(2 teams with
male, 2 teams
with female)
4 interviewers
Driver
1. Teams started
in more remote
areas (Adja-Oere
& Djidja) to have
access before
rains
2. El-Arbi Housni
accompanied H
Togouni (director
of survey) 30th
April/May 1st for
supervision (no
pause in start-up)
25 teams
1 team leader
(EMiCov); 1
controler/editor; 1
health agent; 2-5
interviewers; 1 male
interviewer; driver
BENIN – SURVEYS 2001 - 2007
A50
IIP-JHU | Retrospective evaluation of ACSD in Benin
Data
processing
Supervision
Editing of
questionnaires
5 editors/agents were
responsible for
editing & recoding
Every 2 weeks with
technical team from
INSAE
Yes - specified in
report
Verification
interviews (done
by supervisors
for ages, etc)
Technical team
supervision
Specified in report editors observed
women's question
admin
Observation of
interviews (by
supervisor or
controller/
editor)
SURVEY COMPONENT
DHS 2001
1 responsible for
survey; 2 supervisors
( 1 in each zone; 8
editors (1 per team)
Unknown
Unknown
CDC-ACSD 2003
4 editors/agents were
responsible for
editing & recoding
5 cadres of CEFORP
supervision in each
department
Unknown
Unknown
BENIN-UNICEF
COOPERATION BASELINE
COVERAGE SURVEY 2005
DHS 2006
Likely similar
Every 2 week
technical supervision
by INSAE team
20 agents were
reponsible for coding
& editing
Not specified in
manual; Done by
male team 2 in
PAK
Specified in
manual; Never
done in teams
with male
supervisor; done
irregularly in
women-led teams
ACSD
SUPPLEMENTARY
DHS 2007
Done almost
continuously in
PAK & DAA
regions by 1
INSAE technical
officer in each
zone; unknown
focus on
interview
observations, readministering
questionnaires,
etc
Unknown
Specified in manuals
BENIN – SURVEYS 2001 - 2007
IIP-JHU | Retrospective evaluation of ACSD in Benin
A51
References
DHS Benin 2001
Done by technical
committee and Macro
consultants
DHS 2006
Preliminary report58
Unknown
Survey
documents,
discussions &
observations
Done according to
standard DHS
Done by
technical
committee and
Macro
consultants
similar
Data edited/cleaned for
internal consistency by
INSAE, with TA from
Macro
Enquête de
couverture de la
SASDE 200356
Imputation of
birth dates
55
Done according to
standard DHS
Finalization of
data
Data edited/cleaned for
internal consistency by
CEFORP
None - survey too
short to have
feedback loop
Etude de base dans
les zones
d'intervention du
Progr.amme de
Coopération BeninUNICEF 200557
Recoding done in
SPSS
Recode file done by
INSAE & Macro to
include "other" codes,
imputed values, and
sampling weights
Data editing
Unknown
Done according to
standard DHS
Unknown
Data edited/cleaned for
internal consistency by
2 DP agents at INSAE,
with TA from Macro
Quality control
loop
1. Double data entry
in CSPro done by
data entry agents,
under supervision of
data managers
2. H Koche (Macro)
set up data entry
procedures &
supervised data
entry start-up
1. Double data entry in
CSPro done by 40 data
entry agents (20
additional agents in
Dec/Jan), under
supervision of 4 data
managers, 1 programmer
& assistant
2. H Koche (Macro) set
up entry procedures &
supervised data entry
Data completeness and
errors detected after
data entry sent back to
each team & interviewer
through the supervision
ACSD
SUPPLEMENTARY
DHS 2007
DHS 2006
Not in final datafile
Unknown
Data completeness and
errors detected after
data entry sent back to
each team & interviewer
through the supervision
Data entry
procedures
1. 6 Data entry agents
trained for 3 days
2. Database in EpiInfo
3. 2 1/2 months of entry
4. Control for
inconsistencies (but not
double entry)
BENIN-UNICEF
COOPERATION BASELINE
COVERAGE SURVEY 2005
Not in final datafile
1. 5 trained Data entry
agents
2. Entry
from 15 Aug to 13 Sept
2003
3. Database in CSPro
adapted from standard
CDC base
CDC-ACSD 2003
1. Double data entry in
ISSA done by 11 data
entry agents, under
supervision of 2 data
managers
2. Verification and editing
after double entry
SURVEY COMPONENT
DHS 2001
BENIN – SURVEYS 2001 - 2007
A52
IIP-JHU | Retrospective evaluation of ACSD in Benin
Survey
Elements
Sample
General Info
Respondents
Survey
documentation
available
Report of findings
1200 HHs (654 in
Zou & 555 in
Collines)
HH heads, Women
and Women of U5s
Done according to WHO
cluster methodology; Selection
of villages unclear; Total of 90
clusters sampled in
departments
Caregivers/mothers of children
U5; women with pregancy last
yr.
Through routine data
collection at health centers
Reports of findings
Data available at health zone
level 2003-2006
Ministry of Health
R Adjimon, C
Sessou, A Tollegbe,
M Gigigaye
No
National
ROUTINE MONITORING DATA
Zou & Collines
Departments
SURVEY
SURVEY
Unavailable
2006 PSI KAP
2004 PSI
Report of findings
No
UNICEF Consultant
Implementing
Agency (& TA)
Datafile available
for reanalysis
Zou, Collines, Oeume, Plateau
Departments (includes ACSD
"HIZ")
2003 UNICEF CLUSTER
SURVEY (DEC 03)
Geographic
Coverage
SURVEY COMPONENT
BENIN
Table F2: Methodology and implementation of other surveys and data in Benin 2001 to 2007
IIP-JHU | Retrospective evaluation of ACSD in Benin
A53
Notes
Sources
Field work
Period of field
work
Organization
Training
Information
collected
SURVEY COMPONENT
Not utilized due to
comparability of sampling
methodology & survey conduct
Mid-term survey
2006 for ITNs, PSIBenin60
Not utilized due to
geographical
coverage, unknown
sampling
methodology, &
comparability of
indicators
UNICEF coverage survey of
vaccination, ITN, & Vitamin A
200359
unknown
unknown
unknown
Documentation
not available
SURVEY
SURVEY
Socio-demographic
characteristics, KAP
of ITN use
2006 PSI KAP
2004 PSI
22 & 23 Dec 2003
1 day of training with a "pretest"
30 interviewers in each health
zone (60 total in PAK & DAA)
with 1 coordinator & 2
supervisors per zone
Vaccination, ITN & vitamin A
forms
2003 UNICEF CLUSTER
SURVEY (DEC 03)
BENIN
Not utilized in evaluation due
to difficulties in estimating
reference population &
comparability of active versus
passive surveillance
“Annuaire” of health statistics
2003-2006 61-64
On-going
Routine
Routine
Epidemiological surveillance
(malaria, diarrhea, ARI, EPI
dx); malnutrition; vaccination;
anemia; ANC & PNC; sick
child visits; deliveries
ROUTINE MONITORING DATA
APPENDIX G
Tables presenting priority indicators over time for ACSD high impact
zones
A54
IIP-JHU | Retrospective evaluation of ACSD in Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A55
359
62
P e r c e n ta g e o f c h ild r e n a g e d 0 - 5 9
m o n th s s le e p in g u n d e r a n in s e c tic id e
tr e a te d m o s q u ito n e t ( IT N : tr t'd < = 6 m )
( w o m a n q u e s t.)
P e r c e n ta g e o f p r e g n a n t w o m e n
s le e p in g u n d e r a m o s q u ito n e t (w o m e n
q u e s t)
14
5
6
N o d a ta
0
2
2
0 .1 - 2 7
1 - 10
1 - 12
5 - 29
3 - 17
n /a
n /a
95% CI
235
236
1984
627
372
27
N o d a ta
43
31
N o d a ta
54
1
1
7
4
7
8
m is s in g ( % )
N o d a ta
51
45
%
2 0 0 3 A C S D -C D C
368
n
795
833
467
n
N o d a ta
27
38
84
N o d a ta
%
6
1
27
m is s in g
(% )
320
320
328
2348
2349
820
2080
496
493
483
n
17
22
34
18
26
63
61
17
60
49
%
1
1
0
2
2
3
2
0 .2
1
3
m is s in g
(% )
13 - 21
18 - 26
29 - 39
15 - 20
23 - 30
58 - 68
57 - 65
n /a
n /a
n /a
95% CI
2 0 0 6 /7 D H S & S u p p le m e n t a l s u r v e y
*All vaccination indicators calculated based on MICS protocols (where distribution of children reported vaccination before 12m in card s applied to all children reported as
vaccinated).
** ITN = Insecticide treated net defined as treated within 12 months before the survey or long-lasting net.
P e r c e n ta g e o f p r e g n a n t w o m e n
s le e p in g u n d e r a n in s e c tic id e tre a te d
m o s q u ito n e t (IT N : < = 6 ) ( h o u s e h o ld
q u e s t)
P e r c e n ta g e o f p r e g n a n t w o m e n
s le e p in g u n d e r a n in s e c tic id e tre a te d
m o s q u ito n e t (IT N : < = 1 2 ) ( h o u s e h o ld
q u e s t)
359
P e r c e n ta g e o f c h ild r e n a g e d 0 - 5 9
m o n th s s le e p in g u n d e r a n in s e c tic id e
tr e a te d m o s q u ito n e t ( IT N : tr t'd < = 1 2 m )
( w o m a n q u e s t.)
IT N s * *
0
125
P e r c e n ta g e o f c h ild r e n 1 2 - 3 2 w h o
r e c e iv e d a t le a s t o n e h ig h d o s e v ita m in
A s u p p le m e n t w ith in th e la s t 6 m o n th s
17
0
10
323
2
P e r c e n ta g e o f c h ild r e n 6 - 5 9 w h o
r e c e iv e d a t le a s t o n e h ig h d o s e v ita m in
A s u p p le m e n t w ith in th e la s t 6 m o n th s
63
2
N o d a ta
66
P e r c e n ta g e o f c h ild r e n a g e d 1 2 -2 3
m o n th s w h o re c e iv e d 3 d o s e s o f D P T
v a c c in e
51
m is s in g
(% )
2001 D H S
%
P e r c e n ta g e o f c h ild r e n a g e d 1 2 -2 3
m o n th s w h o a r e im m u n iz e d a g a in s t H ib
66
P e r c e n ta g e o f c h ild r e n a g e d 1 2 -2 3
m o n th s w h o a r e im m u n iz e d a g a in s t
m e a s le s
In d ic a to r s *
E P I+
n
2 0 0 5 B e n in - U N IC E F
C o o p e r a t io n C o v e r a g e
S u rv e y
Table G1. EPI+ and ITN coverage indicators over time in PAK and DAA zones, Benin (weighted)
A56
IIP-JHU | Retrospective evaluation of ACSD in Benin
15 (33)
Percentage of pregnant women
sleeping under a mosquito net (woman
quest)
*among children 6-32 months of age
4 (186)
Percentage of children aged 0-59 months
sleeping under an insecticide treated
mosquito net (ITN treated <=6m) (women's
quest.)
4 (58)
50 (31)
Percentage of children aged 12-23 months
who received 3 doses of DPT vaccine
Percentage of children 6 - 59 who received
at least one high dose vitamin A supplement
within the last 6 months
41 (31)
Percentage of children aged 12-23 months
who are immunized against measles
12 (30)
6 (173)
28 (68)
77 (36)
59 (36)
2001 DHS %(n)
PAK
DAA
42 (145)
33 (1173)
47 (392)*
52 (229)
43 (224)
45 (91)
30 (782)
66 (235)*
49 (143)
50 (144)
2003 ACSD-CDC %(n)
PAK
DAA
n/a
31 (404)
79 (102)
n/a
n/a
n/a
23 (391)
87 (135)
n/a
n/a
2005 Coverage Survey
%(n)
PAK
DAA
Table G2. EPI+ and ITN coverage indicators over time stratified by PAK and DAA zones, Benin (weighted)
69 (426)
70 (258)
57 (254)
34 (139)
34 (189)
17 (1112) 18 (1236)
56 (394)
49 (235)
40 (229)
2006/7 DHS %(n)
PAK
DAA
IIP-JHU | Retrospective evaluation of ACSD in Benin
A57
151
43
30
Percentage of children aged 0-59
months with fever receiving
antimalarial drugs*
Percentage of children aged 0-59
months with suspected pneumonia
taken to an appropriate health
provider
Percentage of children aged 0-59
months with diarrhoea receiving
ORS, RHF or increased fluids and
continued feeding
*Any antimalarial medication
n*
IMCI case management
indicators
38
32
70
%
0
0
1
Missing
(%)
2001 DHS
0 - 77
20 - 44
55 - 85
95% CI
245
236
617
n*
46
31
70
%
2
1
2
Missing
(%)
2003 ACSD-CDC
141
38
241
n*
43
26
53
%
0
0
1
Missing
(%)
2005 Benin UNICEF
Cooperation
Table G3. Illness case management indicators from 2001 to 2006/7 in PAK and DAA zones, Benin (weighted)
176
156
633
n*
34
30
67
0.4
1
3
% Missing (%)
24 - 44
23 - 37
62 - 72
95% CI
2006/7 DHS &
Supplemental Survey
A58
IIP-JHU | Retrospective evaluation of ACSD in Benin
¥
¥
71 (89 )
DAA
4 5 (165 )
3 1 (153 )
6 5 (303 )
PAK
49 (80 )
31 (83 )
7 4 (314 )
DAA
2003 AC SD -CDC %(n)
¥ S am pl e size too sm all - >25 in e ither co mp ar ison **a ny antim al ari al trea tm ent in p re viou s 2 we eks
¥
¥
Perc entage of children aged 0- 59
m onths with suspected pneum onia taken
to an appropriate health provider
Perc entage of children aged 0- 59
m onths with diarr hoea r eceiv ing O RS,
RHF or incr eas ed fluids and c ontinued
feeding
70 (62 )
Perc entage of children aged 0- 59
m onths with fever receiving antimalarial
drugs**
PAK
2001 DHS %( n)
3 2 (92)
¥
4 9 (144 )
PAK
6 3 (49)
¥
5 9 (97)
DAA
2005 Cove rag e Survey
%(n)
Table G4. Illness case management indicators from 2001 to 2006/7, stratified by PAK and DAA zones, Benin (weighted)
3 0 (90)
3 9 (45)
64 (208)
3 7 (86)
26 (111)
69 (4 25)
DAA
200 6/7 DHS %( n)
PAK
IIP-JHU | Retrospective evaluation of ACSD in Benin
A59
88
45
27
16
Percentage of infants aged 0-5
months who are exclusively
breastfed
Percentage of infants aged 6-9
months who are breastfed and
receive complementary food
Percentage of children aged 20-23
months who are currently
breastfeeding
n
Percentage of newborns put to the
breast within one hour of birth
IMCI feeding behavior
indicator
n/a
80
40
45
%
0
0
0
0
n/a
54 100
23 57
29 62
Missing 95%
(%)
CI
2001 DHS
87
167
250
503
n
71
59
25
32
%
2
3
2
0
Missing
(%)
2003 ACSD-CDC
51
76
102
n
80
59
24
No data
%
0
0
0
Missing
(%)
2005 Benin UNICEF
Cooperation
Table G5. Feeding behaviour indicators from 2001 to 2006/7 in PAK and DAA zones, Benin (weighted)
157
215
278
608
n
66
78
27
47
%
0
0
0.3
0
56 75
70 87
21 34
41 54
Missing 95%
(%)
CI
2006/7 DHS &
Supplemental survey
A60
IIP-JHU | Retrospective evaluation of ACSD in Benin
46 (46)
¥
¥
¥
44 (42)
¥
¥
¥
¥ Sample size too small - >25 in either comparison
Percentage of newborns put to the
breast within one hour of birth
Percentage of infants aged 0-5
months who are exclusively
breastfed
Percentage of infants aged 6-9
months who are breastfed and
receive complementary food
Percentage of children aged 20-23
months who are currently
breastfeeding
2001 DHS %(n)
PAK
DAA
73 (59)
54 (96)
28 (163)
42 (311)
68 (28)
65 (71)
18 (87)
17 (192)
2003 ACSD-CDC %(n)
PAK
DAA
¥
52 (48)
4 (52)
n/a
¥
71 (28)
44 (50)
n/a
2005 Coverage Survey
%(n)
PAK
DAA
Table G6. Feeding behaviour indicators from 2001 to 2006/7, stratified by PAK and DAA zones, Benin (weighted)
69 (63)
83 (85)
22 (138)
43 (284)
63 (94)
76 (130)
32 (140)
51 (324)
2006/7 DHS %(n)
PAK
DAA
IIP-JHU | Retrospective evaluation of ACSD in Benin
A61
87
Percentage of pregnant women receiving 3
months of iron supplementation.
28
88
88
80
88
0
0
0
32
44
1.3
0
18-46
33-54
11-46
70-91
53-89
Missing
95%CI
(%)
No data
71
%
88
Percentage of newborns protected against
tetanus (2+ doses TT during pregnancy)
Percentage of pregnant women who report at
least 3 prenatal visits to a trained worker
(doctor, nurse or midwife)
Percentage of pregnant women who report at
least 3 prenatal visits to a trained worker
(doctor, nurse, midwife or auxiliary midwife)
Percentage of pregnant women receiving
intermittent preventative treatment for malaria
during pregnancy in previous year (any SP)
Percentage of pregnant women receiving
intermittent preventative treatment for malaria
during pregnancy in previous year (2+ doses
SP)
ANC indicators
Total
births*
2001 DHS
470
479
494
497
496
Total
births*
36
52
7
5
2
1
1
Missing
(%)
No data
18
79
74
%
2003 ACSD-CDC
177
177
No data
No data
1
0
594
597
606
606
606
No data
1
606
Total
births*
56
55
6
7
74
64
%
2
2
0.4
0.4
1
0.4
51-62
49-60
3-8
5-9
69-79
58-69
Missing
(%)
95%CI
2006/7 DHS &
Supplemental Survey
No data
0
Total
Missing
%
births*
(%)
UNICEF
Cooperation
Coverage Survey
Table G7: Antenatal care for women with a birth in the previous 12 months from 2001 to 2006/7 in PAK and DAA zones, Benin (weighted)
A62
IIP-JHU | Retrospective evaluation of ACSD in Benin
*weighted
88
88
88
88
5
0
0
0
0
1-10
87 100
84-100
59-92
Missing
95%CI
(%)
No data
94
92
76
%
70
%
491
21
No data
4
0
0
Missing
(%)
No data
503 80
503
Total
births*
2003 ACSD-CDC
**among women with live birth in 12 months previous to the survey
Percentage of women receiving vitamin A
supplementation within 2 months of birth
Percentage of newborns receiving a
postnatal visit by a trained worker (doctor,
nurse or midwife) within 3 days of delivery.
(NOassumption of postnatal care with
institutional delivery)
Percentage of newborns receiving a
postnatal visit by a trained worker (doctor,
nurse or midwife) within 3 days of delivery.
(women with institutional deliveries
assumed to have received postnatal care)
Percentage of births attended by skilled
health personnel (doctor, nurse, midwife or
auxiliary midwife)
Percentage of births attended by skilled
health personnel (doctor, nurse or midwife)
Assisted delivery and postnatal care Total
births*
indicators
2001 DHS
177
177
177
Total
births*
9
0
0
0
Missing (%)
No data
90
90
%
2005 Benin - UNICEF
Cooperation Coverage
Survey
Table G8. Assisted deliveries and postnatal care from 2001 to 2006/7 in PAK and DAA zones, Benin
608
606
607
607
607
Total
births*
38
68
84
85
74
%
0
0.4
0.2
0.2
0.2
Missing
(%)
33 - 44
62 - 73
79-89
79-90
68-79
95%CI
2006/7 DHS & Supplemental
survey
IIP-JHU | Retrospective evaluation of ACSD in Benin
A63
30 (41)
68 (42)
98 (42)
0 (42)
Percentage of pregnant women receiving 3
months of iron supplementation.
Percentage of births attended by skilled health
personnel (doctor, nurse or midwife)
Percentage of newborns receiving a postnatal
visit by a trained worker (doctor, nurse or
midwife) within 3 days of delivery**
Percentage of women receiving vitamin A
supplementation within 2 months of birth
10 (46)
90 (46)
82 (46)
34 (46)
64 (46)
3 (46)
80 (46)
13 (306)
n/a
61 (311)
39 (306)
43 (301)
4 (303)
65 (307)
34 (185)
n/a
85 (192)
31 (164)
67 (178)
39 (191)
88 (189)
2003 ACSD-CDC %(n)
PAK
DAA
9.0 (89)
n/a
83 (89)
n/a
n/a
0 (89)
n/a
9 (88)
n/a
97 (88)
n/a
n/a
2 (88)
n/a
2005 Coverage
Survey %(n)
PAK
DAA
¥ Sample size too small - < 25 in either comparison **Women with institutional deliveries are assumed to have had a post-natal visit
21 (42)
57 (42)
62 (42)
Percentage of newborns protected against
tetanus (2+ doses TT during pregnancy)
Percentage of pregnant women receiving
intermittent preventative treatment for malaria
during pregnancy in previous year (any SP)
Percentage of pregnant women who report at
least 3 prenatal visits to a trained worker
(doctor, nurse or midwife)
2001 DHS %(n)
PAK
DAA
28 (284)
73 (283)
63 (284)
57 (270)
38 (275)
1 (282)
52 (284)
47 (324)
94 (324)
83 (323)
56 (323)
69 (322)
12 (324)
74 (322)
2006/7 DHS %(n)
PAK
DAA
Table G9: Antenatal, delivery and postnatal care for women with a birth in the previous 12 months from 2001 to 2006/7, stratified by PAK and DAA
zones, Benin (weighted)
APPENDIX H
Tables presenting comparisons of priority indicators over time in ACSD
high-impact zones and the comparison area
A64
IIP-JHU | Retrospective evaluation of ACSD in Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A65
14
62
No data
6
10
359
323
63
51
%
31
579
No data
5
17
No data
67
54
3724
3283
760
748
%
0
4
5
1
2
Missing
(%)
Comparison area ¥
n
320
328
2349
2080
496
493
483
n
22
34
26
61
17
60
49
%
High impact zones
(PAK + DAA)
1684
1455
11785
10617
2579
2566
2520
n
17
42
28
63
14
64
60
%
0.4
13
6
5
0
1
2
Missing
(%)
Comparison area ¥
2006/7 EDS & Enquête Supplémentaire
*All vaccination indicators calculated based on MICS protocols (where distribution of children reported vaccination before 12m in card s applied to all children reported as
vaccinated).
** ITN = Insecticide treated net defined as treated within 12 months before the survey or long-lasting net.
¥ Comparison area is Benin – national level, excluding the HIZs and Cotonou
Percentage of pregnant women sleeping
under an insecticide treated mosquito net
(ITN) (household quest)
Percentage of pregnant women sleeping
under a mosquito net (women quest)
Percentage of children aged 0-59 months
sleeping under an insecticide treated
mosquito net (ITN) (woman quest.)
ITN **
Percentage of children 6 - 59 who received
at least one high dose vitamin A
supplement within the last 6 months
No data
66
Percentage of children aged 12-23 months
who received 3 doses of DPT vaccine
Percentage of children aged 12-23 months
who are immunized against Hib
66
n
Percentage of children aged 12-23 months
who are immunized against measles
EPI+ *
Indicators
High impact zones
(PAK + DAA)
2001 EDS
Table H1. EPI+ and ITN coverage indicators over time in high impact (PAK & DAA) and comparison zones, Benin (weighted)
A66
IIP-JHU | Retrospective evaluation of ACSD in Benin
63
70
%
30
43
38
32
Chloroquine
151
151
n*
545
472
1559
1559
n*
43
32
Chloroquine
62
66
%
2
0.4
6
5.8
Missing
(%)
Comparison zones¥
Comparison area is Benin – national level, excluding the HIZs and Cotonou
*Includes treatment with any antimalarial
**Includes treatment with first-line antimalarial recommended by national policy
Percentage of children aged 0-59
months with diarrhoea receiving
ORS, RHF or increased fluids
and continued feeding
Percentage of children aged 0-59
months with suspected
pneumonia taken to an
appropriate health provider
Benin antimalarial policy
Children 0-59m with fever in
previous 2 weeks, rec'd
appropriate antimalarial
treatment (effective)**
Percentage of children aged 0-59
months with fever receiving
antimalarial drugs (program)*
IMCI case management
indicators
High impact
zones
2001 DHS
176
156
633
633
n*
ACT
34
30
0
67
%
High impact
zones
1187
1198
3311
3311
n*
42
35
ACT
0.5
58
%
1
0.4
8
8
Missing
(%)
Comparison zones¥
2006/7 DHS & Supplemental survey
Table H2. Case management indicators over time in high impact (PAK & DAA) and comparison zones, Benin (weighted)
¥
IIP-JHU | Retrospective evaluation of ACSD in Benin
A67
4279
4615
243
294
426
921
n
72
67
67
66
38
47
%
8
0.2
1
1
0.2
0.1
Missing
(%)
Comparison zones¥
1405
1518
157
215
278
608
n
53*
49*
66
78
27
47
%
High impact
zones
12852
13683
689
986
1265
2838
n
58
55
59
73
45
54
%
9
3
0
1
1
0.2
Missing
(%)
Comparison zones¥
2006/7 DHS & Supplemental survey
¥ Comparison area is Benin – national level, excluding the HIZs and Cotonou
84
Percentage of households
consuming iodized salt
388
(>=15ppm)
* Salt not tested in 2007 supplemental survey
53
16
76
80
40
45
%
27
45
88
n
420
Percentage of households
consuming iodized salt
(>=15ppm)
Percentage of infants aged 6-9
months who are breastfed and
receive complementary food
Percentage of children aged 2023 months who are currently
breastfeeding
Percentage of infants aged 0-5
months who are exclusively
breastfed
Percentage of newborns put to
the breast within one hour of birth
IMCI feeding behavior
indicators
High impact
zones
2001 DHS
Table H3. Feeding behaviour indicators over time in high impact (PAK & DAA) and comparison zones, Benin
A68
IIP-JHU | Retrospective evaluation of ACSD in Benin
88
Percentage of births attended by skilled
health personnel (doctor, nurse or midwife)
88
5
94
76
32
44
28
71
%
921
919
920
870
914
915
907
n
17
78
63
39
47
6
65
%
0.1
0.3
0.2
6
1
1
2
Missing
(%)
Comparison zones¥
¥ Comparison area is Benin – national level, excluding the HIZs and Cotonou
* Women with institutional deliveries assumed to have appropriate postnatal care
Percentage of women receiving vitamin A
supplementation within 2 months of birth
88
87
Percentage of pregnant women receiving 3
months of iron supplementation.
Percentage of newborns receiving a
postnatal visit by a trained worker within 3
days of delivery*
88
88
Percentage of pregnant women receiving
intermittent preventative treatment for
malaria during pregnancy in previous year
(any SP)
Percentage of newborns protected against
tetanus (2+ doses TT during pregnancy)
88
n
Percentage of pregnant women who report
at least 3 prenatal visits to a trained worker
(doctor, nurse or midwife)
ANC, assisted delivery and
postnatal care indicators
High impact
zones
2001 DHS
608
607
607
594
597
606
606
n
38
84
74
56
55
7
64
%
High impact
zones
2836
2839
2844
2649
2806
2804
2800
n
40
80
75
53
51
7
67
%
0.3
0.2
0
7
1
1
2
Missing
(%)
Comparison zones¥
2006/7 DHS & Supplemental survey
Table H4: Antenatal, delivery and postnatal care indicators over time in high impact (PAK & DAA) and comparison zones, Benin
APPENDIX I
Tables presenting 2006-7 survey results for key indicators in the ACSD
high-impact zones by socio-demographic characteristics of the
population
IIP-JHU | Retrospective evaluation of ACSD in Benin
A69
A70
IIP-JHU | Retrospective evaluation of ACSD in Benin
56
DAA (Zou)
Total
Residence
30
70
2943
2051
892
507
17
Abom ey
Rural
530
18
Djidja
Urban
494
417
17
14
Pobe
606
21
Agbangnizoun
Kétou
390
13
1643
1300
Total Households
Adja-Ouere
Commune
44
High impact zones
PAK (Ouémé)
Percent
Table I1a: Households interviewed
Rural
Urban
Abomey
Djidja
Pobe
Kétou
Agbangnizoun
Adja-Ouere
1
Not sure
Total
Secondary school+
Prim ary School
Mother's education level
None
12
70
30
19
19
12
16
21
14
58
42
Yes
Currently pregnant
Residence
Commune
DAA (Zou)
High imp act zones
PAK (Ouém é)
Percent
2741
21
328
1924
817
508
506
337
442
573
375
1587
1154
Total Women
Table I1b: Eligible women with complete
interviews
Adja-Ouere
Rural
Urban
Abomey
Djidja
Pobe
Kétou
Agbangnizoun
T otal
Secondary school+
Primary School
M other's education level
None
Residence
Commune
DAA (Zou)
High impact zones
PAK (Ouémé)
73
27
15
19
14
18
20
14
54
46
Percent
2654
1947
706
389
504
376
487
537
360
1430
1223
Total Children
Table I1c: Under five Children: from Individual
interview
Table I1: Description of households, eligible women and children under five in 2006 DHS/2007 Supplemental survey in PAK and DAA, Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A71
78
S e c o n d a ry s c h o o l+
T o ta l
62
71
75
59
73
75
79
64
67
3
4
Leas t Po or
76
83
87
81
87
80
78
94
80
82
84
82
83
82
93
83
86
65
86
79
87
218
35
64
43
44
32
17
38
164
109
109
152
66
43
40
35
37
42
20
126
92
N um be r
w it h v a lid
date on
c a rd
229
49
65
57
55
45
29
80
50
47
52
46
47
53
63
483
77
106
92
97
111
24
81
378
226
257
343
140
84
79
83
49
102
87
48
254
52
29
58
48
57
40
A CS D
i n d ic a to r
(% )
N um be r
of
c h ild re n
12-23m *
63
74
78
68
59
40
81
75
59
68
59
60
71
79
61
55
51
76
52
72
53
R e c 'd
v a c c in e
(% )
93.1
96
98
99
93
95
93
1 00.0
95.7
95.4
96.9
94.6
94.8
98.2
1 00.0
91.7
1 00.0
91.2
1 00.0
83.9
97.6
233
36
68
49
46
34
16
43
174
120
113
165
68
48
40
34
43
51
17
140
94
Nu m be r
w ith v a lid
date on
c a rd
60
72
77
63
56
37
81
72
57
66
55
57
70
79
56
55
47
76
44
71
49
AC SD
i n d i c a to r
(% )
In n o c u la te d a g a in s t D P T
R e c'd
v a c c in e
b e f o re 1 2 m
(% )
49 3
80
10 7
94
98
11 4
24
82
38 7
23 0
26 3
34 7
14 6
87
83
80
10 3
87
52
25 8
23 5
Num ber
of
c h ild r e n
1 2 -2 3 m *
* n = C h ild re n 1 2 -2 3 m o n t h s o f a g e , s t ill a liv e w it h n o n -m is s in g d a ta f o r in d ic a t o r c a lc u la t io n : w e ig h te d
A C S D in d ic a to r: m u lt ip ly t h e p e rc e n t o f c h ild re n t h a t r e c e iv e d v a c c in a t io n b e fo re f irs t b irt h d a y , a c c o rd in g to im m u n iz a t io n c a rd , b y t h e t o t a l p e r c e n t a g e o f c h ild re n
56
46
2
37
85
62
57
62
57
57
65
68
62
57
44
67
60
66
52
R e c 'd
v a c c in e
(% )
P oores t
W e a l th i n d e x q u i n t il e s
66
66
P rim a r y S c h o o l
65
69
M a le
68
R u ra l
F e m a le
64
Urban
68
A bom ey
M o th e r 's e d u c a ti o n l e v e l
N one
Se x
R e s id e n c e
53
76
P obe
69
K etou
D jid ja
50
76
A d ja -O u e re
A g b a n g n iz o u n
74
D A A (Zou)
Com m une
59
H i g h im p a c t z o n e s
P A K (O u e m e )
% w ith
E P I c a rd S een
R e c 'd
v a c c in e
befor e 12m
(% )
I n n o c u la te d a g a in s t m e a s le s
C h il d r e n 1 2 - 2 3 m o n t h s o f a g e v a c c in a t e d a g a in s t m e a l s e s a n d D P T
Table I2: Vaccination rates by geographic and socio-demographic factors in PAK and DAA in 2006/7, Benin
Table I3: Vitamin A supplementation in previous 6 months by socio-demographic characteristics in
PAK and DAA zones, Benin 2006/7
Children 6-59 m onth s of age rece iving vitam in A
supplemen tation in the previou s 6 months
Vitam in A
supplem entation (% )
N umbe r of ch ildren
6-5 9 mon th s of
age*
High im pa ct zone
PAK (Ou eme)
54
97 4
D AA (Zou)
68
1 106
Com m une
Adja -Oue re
51
27 9
Ag bang nizo un
78
41 4
Ketou
58
40 0
Po be
50
29 4
D jidja
52
38 9
Abom ey
73
30 4
Urb an
63
55 8
Ru ral
61
1 522
Male
61
1 054
Fe male
62
1 026
6-11
63
30 9
12-23
63
48 8
24-35
63
41 8
36-47
59
45 4
48-59
58
41 1
Mothe r's educ ation leve l
No ne
58
1 645
Primary School
70
32 9
Second ary schoo l+
76
10 6
Res ide nce
Se x
Ag e in m onth s
We alth index quintiles
Total
Poorest
45
47 5
2
57
42 8
3
65
42 9
4
73
40 7
L east Po or
70
34 1
61
2 080
*n=Child ren 6 -59 mon ths o f ag e, still alive with no n-missin g data for
indicator calculation : w eigh te d
A72
IIP-JHU | Retrospective evaluation of ACSD in Benin
Table I4: Utilization of bednets by children under age five by socio-demographic characteristics in PAK
and DAA zones, Benin 2006/7
Children aged 0-59 m onths sleeping under a mosqu ito net, a trea ted net or an
IT N* the night preceding the s urvey
Pe rcentage of c hildre n who:
Slept und er
mo squito net last
nig ht (% )
Slep t un der a n eve rtre ated mo sq uito net
last n ight (% )
Slept und er an
IT N* la st night
(% )
N umb er of
childre n 0-5 9
mon th s of
ag e**
High im pa ct zone
PAK (Oue me)
36
33
25
1 167
D AA (Zo u)
38
32
25
1 303
Adja-Ouere
35
33
24
33 1
Agba ngn izou n
35
31
25
48 6
Ke to u
35
32
23
47 3
Pob e
37
34
27
36 3
D jid ja
37
28
21
44 9
Ab ome y
44
39
30
36 8
U rba n
41
37
28
68 5
Rural
36
31
24
1 785
Com m une
Res ide nce
Se x
Ma le
38
34
26
1 242
Fema le
36
32
24
1 228
Mothe r's educ ation leve l
Non e
Prim ary Scho ol
Se co ndary school+
Ag e in m onth s
0-1 1
41
36
28
58 0
12-2 3
43
37
28
50 5
24-3 5
38
33
25
45 6
36-4 7
32
28
22
48 8
48-5 9
We alth index quintiles
30
26
20
44 2
Total
Po ore st
27
21
14
56 9
2
31
27
22
48 8
3
35
30
23
51 7
4
42
39
30
48 5
Lea st poo r
55
51
41
41 2
37
33
25
2 470
*IT N=Mosquito ne t tre ated w ith in se cticide in the previous 12 mon th s, or a lo ng-la sting n et
**n=Total children und er five wh o sle pt in H H last night, with no n-missin g data for in dicator calculation:
we ig hted
As assessed through household questionnaire
IIP-JHU | Retrospective evaluation of ACSD in Benin
A73
Table I5: Utilization of bednets by currently pregnant women by socio-demographic characteristics in
PAK and DAA zones, Benin 2006/7
Percen tag e o f preg n an t wo me n w ho :
Sle pt un der
m os qu ito ne t las t
nig ht (% )
S lept u nder an ev er- Slep t u nd er an
treated m os qu ito
IT N *last n igh t
net las t nig ht (% )
(%)
N um b er of
preg nant
wo m en 15-4 9
ye ars o f age **
Hig h imp ac t z o n es
PA K (Oue m e)
33
28
23
137
D AA ( Zo u)
35
29
22
183
Co m mu n e
Adja -Oue re
38
30
27
50
Agb angni zou n
24
23
17
68
K eto u
31
28
20
55
31
Pob e
28
25
23
D jid ja
38
30
21
67
A bom ey
47
38
28
47
Re sid en ce
U rba n
40
33
25
78
R ur al
32
27
22
242
M oth er's e du ca tio n lev el
N on e
33
28
21
239
Prim a ry Sc ho ol
37
31
30
64
S ec onda ry s c hool+
42
35
12
16
W ealth in d ex q uin tile s
To tal
P oores t
24
19
12
74
2
26
20
19
53
3
34
29
21
64
4
40
33
25
77
L eas t po or
47
44
37
52
34
29
22
320
*IT N= M os quito net tr eated in the prev iou s 12 m onths
**n= To ta l p regna nt w om en w ho s lep t in H H las t night, w ith non -m iss in g data for in dic ator ca lc ulation:
we ighted
A74
IIP-JHU | Retrospective evaluation of ACSD in Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A75
342
38
Abomey
34
Secondary school+
Total
29
30
30
2
3
4
Least Poor
28
23
27
Poorest
Wealth index quintiles
26
31
27
Prim ary School
28
24
24-59
male
37
12-23
female
33
Mother's education level
None
Sex
20
0-5
6-11
2363
383
472
475
482
552
122
375
1867
1165
1199
1278
496
310
279
1723
640
442
17
Djidja
29
350
14
Pobe
23
456
19
Ketou
Rural
458
50
Urban
316
25
1242
1122
Adja-Ouere
35
19
Agbangnizoun
Ag e in month s
Residence
Commune
DAA (Zou)
High Impact zone
PAK (ouem e)
% with
fever
5
302
66
67
78
73
67
63
56
89
70
65
66
633
111
140
131
131
120
41
115
476
302
7
5
7
8
9
3
7
6
7
6
7
8
176
71
331
9
99
68
10
55
7
5
9
4
2
4
14
7
9
4
59
491
142
126
72
47
84
227
77
425
208
% with
suspected
pneumonia
30
26
29
37
29
25
71
33
156
28
36
35
45
13
9
22
126
8
10
6
7
6
8
4
11
7
7
69
26
8
87
4
13
15
4
8
7
9
5
8
6
8
11
7
8
% with
diarrhea
20
61
40
29
27
123
33
31
18
7
15
63
23
111
45
Number of
children 059m with
pneum onia*
37
30
36
26
24
28
36
24
39
47
66
23
19
26
39
% taken to
health
facility
Children 0-59 with suspected
pneumonia in previo us 2 weeks
78
66
72
68
69
61
73
69
55
69
64
Number of
Number of % given any children 0children 059m with
antimalarial
59m*
fever*
Children 0-59 with fever in previous 2 weeks
15
4
17
14
28
22
36
16
14
17
14
16
15
15
11
13
22
16
17
13
22
8
18
13
18
ORS
(%)
21
13
23
20
30
27
36
20
21
22
21
24
21
20
11
20
26
19
20
21
28
10
30
16
26
ORS +
RHF (%)
42
25
43
47
50
52
51
52
38
49
36
49
38
43
22
43
38
34
45
30
37
57
43
46
37
34
19
25
42
46
44
38
45
30
38
30
45
23
38
16
33
36
32
34
24
32
43
32
37
30
176
53
27
35
30
31
5
42
129
78
98
55
65
45
11
132
44
30
20
28
28
36
34
86
90
Number of
ORS/ RHF/
ORS/ RHF/ increased fluids children 0increased
59m with
with continued
feed ing (%)
fluids (%)
diarrhoea*
Children 0-59 with diarrhoea in previou s 2 weeks
Care management of fever, suspected pneumonia, and diarrhea for children 0-59 months with illness in the previous 2 weeks
Table I6: Illness case management by socio-demographic characteristics in PAK and DAA zones, Benin 2006/7
A76
IIP-JHU | Retrospective evaluation of ACSD in Benin
1
1
National comparison
0
0
0
5
1
0
0
1
0
67
0
10
1
63
0
59
Appropriate
AM**
4
Quinine
ACT/
CoArtem
N/A
54
50
N/A
N/A
Any AM w/in
24h
***n=Children under five with non-missing data for indicator calculation: weighted
**Appropriate antimalarial treatment defined as Benin policy for first line malaria treatment (CQ in 2001 & 2003; ACT in 2006/7)
* See CDC Data quality issues document
Anti-malarial treatment columns are not mutually exclusive
50
62
67
5
High Impact zones
ACSD-DHS 2006/07
High Impact zones
1
63
2
High Impact zones
ACSD-CDC 2003*
0
2
59
Amodiaquine
1
Chloroquine
National comparison
DHS 2001
SP/
Fansidar
Children with a fever in the last two weeks who were treated with:
58
67
70
70
66
Any AM
treatment
3311
633
617
151
1559
No. of children with
fever in last two
weeks***
Table I7: Reported antimalarials given for fever in the 2 weeks preceding the survey in PAK and DAA zones and comparison areas in 2001,
2003 and 2006/7, Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A77
24
National comparison
21
42
54
60
57
53
ORS, ORT and/or
Increased fluids
**n=Children under five with non-missing data for indicator calculation: weighted
High Impact zones
15
* See CDC Data quality issues document
32
47
15
High Impact zones
ACSD-DHS 2006/07
High Impact zones
ACSD-CDC 2003*
32
49
23
42
ORS and/or ORT
National comparison
DHS 2001
ORS
77
78
71
76
73
Continued
feeding
34
42
46
43
38
176
1187
245
545
30
ORS, ORT and/or
Increased fluids Number of children aged
AND continued
0-59 months with
feeding
diarrhoea**
Children with diarrhoea in the last two weeks who were given:
Table I8: Reported case management for diarrhoea in the 2 weeks preceding the survey in PAK and DAA zones and comparison areas in
2001, 2003 and 2006/7, Benin
A78
IIP-JHU | Retrospective evaluation of ACSD in Benin
55
26
36
41
High Impact zones
ACSD-DHS 2006/07
National comparison
8
9
8
3
6
Private Health center
/ facility
**n=Children under five with non-missing data for indicator calculation: weighted
21
27
54
High Impact zones
ACSD-CDC 2003 *
High Impact zones
42
* See CDC Data quality issues document
24
50
Public health
center / facility
National comparison
DHS 2001
Not treated /
treated at home
/ neighbors
22
13
2
0
12
Private sector drug
vendor
7
13
0
16
8
Other
Children with suspected pneumonia in the last two weeks who were taken to:
156
1198
236
43
472
Number of
children aged 059 months with
pneumonia**
Table I9: Locations where care was sought for suspected pneumonia in the 2 weeks preceding the survey in PAK and DAA zones and
comparison areas in 2001, 2003 and 2006/7, Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A79
A dja- O uer e
fem a le
m a le
3- 5
0- 2
R ur al
U rba n
A b om e y
D jid ja
P ob e
K e to u
A gba ngn izou n
47
53
45
46
49
42
49
47
34
48
47
35
44
44
49
54
58
43
51
6 08
1 45
1 28
1 20
1 14
1 02
4 56
1 04
48
-
-
1 64
4 44
80
1 36
1 03
1 01
1 15
73
2 84
3 24
27
26
29
27
28
27
28
19
37
29
26
39
16
26
28
27
26
20
22
37
32
22
32
B irt h w ithin
previ ous
E xcl u sive ly
1 2m *
b reas tfe ed
278
70
51
50
64
42
214
47
18
133
145
136
142
82
196
34
45
49
55
56
39
138
140
N um ber o f
c hildren 05m **
21 5
42
62
47
24
40
16 0
32
23
10 0
11 5
-
49
16 6
32
62
27
27
44
23
85
13 0
Nu m be r of
c hildren 69m **
** n= T otal ch ild, s till liv ing , liv ing w ith mo th er an d m os t re ce ntly bo rn w ith non- m iss ing d ata fo r indic ator ca lc ula tio n: w e ig hted
78
84
75
76
60
91
75
85
96
82
75
75
79
87
80
75
85
73
70
83
76
Co m p lem e n t
ary fe ed in g
* n =W om en w ith a liv e birth in pr ev ious 12 m on th s w ith n on- m is s ing da ta fo r indic ator a naly s is: w e ighted
T otal
Le as t P oo r
4
3
2
P o ore st
W e alth in d ex q u in t iles
S e co ndar y s ch ool+
P rim ary Sc ho ol
M o th e r's ed u c atio n leve l
N on e
Se x
Ag e in m o n th s
R es id e n ce
Com m une
D AA (Z o u)
H ig h Im p a ct z o n es
P A K ( oue m e)
T i m ely
in itiati o n o f
b reas tfeed in g
66
57
73
81
62
54
71
47
55
69
62
42
76
72
86
65
72
66
30
69
63
C o n tin u ed
b reas tfeed in g
15 7
25
36
28
38
30
11 8
27
12
75
82
-
49
10 8
12
37
27
24
29
28
63
94
Nu m be r of
c hildren 2023 m **
T i m e l y i n it i a ti o n o f b re a s t fe e d i n g , e x c l u s iv e b r e a s tf e e d i n g a m o n g c h il d re n 0 - 5 m o n th s , c o m p l e m e n ta r y f e e d in g a m o n g
c h ild r e n 6 - 9 m o n t h s a n d c o n t in u e d b re a s tf e e d in g ra te s a m o n g c h ild r e n 2 0 -2 3 m o n t h s
Table I10: Infant feeding behaviours as reported by mothers by socio-demographic characteristics in PAK and DAA zones, Benin 2006/7
A80
IIP-JHU | Retrospective evaluation of ACSD in Benin
91
Primary School
Secondary school+
Least Poor
75
96
4
71
95
77
69
75
49
91
81
67
74
68
70
75
95
76
55
69
79
55
81
60
2+
103
67
42
53
36
606
64
43
34
40
18
48
48
32
39
34
34
42
67
100
114
120
128
145
48
103
455
320
286
443
163
70
115
21
30
43
14
48
22
Blood test
taken
90
67
65
66
40
88
72
59
68
58
63
64
90
101
63
44
136
80
322
284
Birth within
previous
12m**
72
47
74
52
3+
606
102
114
120
128
143
48
104
453
319
287
442
164
73
115
101
103
136
77
324
281
Birth within
previous
12m**
** n=Women with a live birth in previous 12 months with non-missing data for indicator analysis: weighted
* Trained health care worker: doctor or nurse/midwife only
Total
78
2
3
56
72
84
Poorest
Wealth index quintiles
72
84
Mother's education level
None
73
77
0-5
6-11
80
74
Djidja
Abomey
Rural
80
97
Pobe
Urban
72
59
Ketou
59
85
Adja-Ouere
86
64
1+
Agbangnizoun
Months since birth
Residence
Commune
DAA (Zou)
High Impact Zone
PAK (oueme)
Number of ANC visits* (%)
75
96
76
83
71
54
91
84
71
76
73
73
79
95
78
58
71
85
58
85
63
608
102
114
120
128
145
48
104
456
321
287
444
164
73
115
101
103
136
80
324
284
71
91
77
77
67
50
84
82
67
74
67
68
77
95
75
52
67
77
59
80
60
608
102
114
120
128
145
48
104
456
321
287
444
164
73
115
101
103
136
80
324
284
Percent of pregnant women who had:* (%)
Birth within
Urine
Birth within
Blood
previous
specimen
previous
pressure
12m**
taken
12m**
measured
Characteristics of antenatal care among women who have given birth in the previous 12 months
75
96
77
84
71
54
91
82
72
77
73
73
80
97
79
58
71
85
59
85
63
Weight
measured
608
102
114
120
128
145
48
104
456
321
287
444
164
73
115
101
103
136
80
324
284
Birth within
previous
12m**
Table I11: Antenatal care among women giving birth in the previous 12 months by socio-demographic characteristics in PAK and DAA zones,
Benin 2006/7
Table I12: Antenatal interventions among women giving birth in the previous 12 months by sociodemographic characteristics in PAK and DAA zones, Benin 2006/7
Antenatal care (including IPT, TT, Fe) among women who have given birth in the previous 12 months
IPT for
pregnant
women (2
doses SP)
IPT during
pregnancy
(any dose SP)
Birth within
previous
12m*
Neonatal
tetanus
protection**
Birth within
previous
12m*
Iron supplementation for at least 3
months.
Birth within
previous
12m*
High Impact Zones
PAK (oueme)
1
1
282
38
275
57
270
DAA (Zou)
10
12
324
69
322
56
323
Commune
Adja-Ouere
3
3
78
32
75
56
75
Agbangnizoun
10
11
135
77
134
42
136
Ketou
0
0
103
45
102
66
103
Pobe
1
1
101
36
99
49
92
Djidja
6
7
115
52
115
60
115
Abomey
15
20
73
79
73
77
72
Urban
6
8
164
46
163
64
159
Rural
5
6
442
58
434
54
434
0-5
5
6
286
48
282
54
281
6-11
6
8
320
61
315
59
312
Mother's education level
None
4
4
454
50
445
53
445
Primary School
6
10
104
63
104
58
101
Secondary school+
23
26
48
76
48
88
47
Poorest
1
2
145
34
139
45
137
2
7
7
128
60
124
51
127
3
3
3
119
63
120
53
120
4
3
5
113
55
113
66
112
Least Poor
16
18
102
65
102
72
97
6
7
606
55
597
56
594
Residence
Months since birth
Wealth index quintiles
Total
* n=Women with a live birth in previous 12 months with non-missing data for indicator analysis: weighted
** n=At least 2 doses of TT during the pregnancy
IIP-JHU | Retrospective evaluation of ACSD in Benin
A81
Table I13: Assisted delivery and post-natal care among women giving birth in the previous 12 months
by socio-demographic characteristics in PAK and DAA zones, Benin 2006/7
Delivery and postnatal care indicators among women who have given birth in the previous 12 months
Skilled birth
attendant*
High Impact Zones
PAK (oueme)
63
DAA (Zou)
83
Birth within Postnatal care within 3 Birth within
days of delivery by
previous
previous
12m**
12m**
trained health worker*
284
73
323
94
Postnatal
supplementation
with Vitamin A*
Birth within
previous
12m**
283
28
284
324
47
324
Commune
Adja-Ouere
56
80
69
80
17
80
Agbangnizoun
85
136
97
136
59
136
Ketou
76
103
85
103
34
103
Pobe
55
101
63
100
30
101
Djidja
74
114
89
115
31
115
Abomey
93
73
96
73
51
73
Urban
85
164
90
163
38
164
Rural
69
443
82
444
38
444
0-5
72
286
81
286
34
287
Residence
Months since birth
6-11
75
321
87
321
42
321
Mother's education level
None
69
455
80
456
34
456
Primary School
84
104
97
103
55
104
Secondary school+
97
48
100
48
41
48
Poorest
48
145
59
145
21
145
2
72
128
86
128
40
128
3
79
120
91
120
48
120
4
82
112
96
114
43
114
Least Poor
96
102
98
101
44
102
74
607
84
607
38
608
Wealth index quintiles
Total
* Trained health care worker: doctor or nurse/midwife only
**n= Women with a live birth in previous 12 months with non-missing data for indicator analysis: weighted
A82
IIP-JHU | Retrospective evaluation of ACSD in Benin
APPENDIX J
Summary of contextual factors
This section is comprised of narrative and tables concerning contextual factors that may be associated
with ACSD coverage and impact outcomes. The examination of these factors contributes to the
plausibility analysis—i.e. to determine if observed changes can be attributed to the ACSD and partner
activities. The contextual factors considered in the evaluation were based on previous work by Victora
et al.65 We examine other activities taking place in the ACSD high impact zones, changes in health
care provision, changes in national policies. Other factors, such as changes in national policies and
pricing, which may influence intervention coverage, are also considered. Socio-economic and
demographic factors in 2001 and 2006 are presented in the body of the report.
The ACSD strategy emphasized working with national and local partners. Table J1a shows activities
carried out by UNICEF’s national and local partners in the period between 2002 and present. ACSD
worked most closely with activities supported by the Benin national ministries, although in the DAA
health zone, Population Services International, an international NGO specializing in social marketing,
promoted and distributed ITNs.
Table J1a: Child health and nutrition activities implemented by UNICEF partners in DAA and PAK
health zones in 2002-present, Benin
DEVELOPMENT
ACTIVITIES
LEAD AGENCY
GEOGRAPHICAL
COVERAGE &
TIMING
DESCRIPTION OF
COLLABORATIONS
WITH UNICEF
Maternal
and
child
health; nutrition; family
planning; youth and
AIDS
MOH
Family
Health
Department (Direction de
la Santé Familiale)
National-level;
DAA: on-going
PAK
&
Prevention
and
treatment of malaria;
Distribution of ITNs as
part of Roll Back
Malaria Initiative
National
Programme
against Malaria (PNLP)
National-level;
DAA: on-going
PAK
&
IMCI
training
&
supervision; monitoring
systems, including PBT;
BCC
for
malaria
treatment & prevention;
C-IMCI
&
mother's
groups
PROPLIPO
(MOH-led
malaria control project in
Ouémé-Plateau region in
Benin, implemented by
CDC with support from
USAID)
Oeume-Plateau:
2000-2005
Prevention
and
treatment of HIV/AIDS
including PMTCT
National Plan against
HIV/AIDS (PNLS), with
local NGOs
National level; PAK &
DAA; on-going
Collaborate with UNICEF
in PMTCT
Distribution of ITNs
Population
Services
International (PSI)
Zou-Collines; DAA: 2002present
Funded by UNICEF/ACSD
to distribute & promote
ITNs
IIP-JHU | Retrospective evaluation of ACSD in Benin
PAK;
Strong collaboration with
UNICEF on all aspects of
ACSD, also collaboration
with UNFPA
Collaborate with UNICEF
on ITN distribution
Collaborate with UNICEF
on malaria BCC, ITN
distribution & promotion,
mother's groups,
A83
Many of the other project activities taking place in 2002 to the present in the HIZs focused on clinical
improvements, especially in maternal and neonatal health (table J1b). Many of these smaller projects
focused on support and care for AIDS orphans and vulnerable children, as well as prevention of
mother-to-child transmission of HIV (PMTCT). Local and international NGOs in DAA also carried out
nutritional rehabilitation and education in selected communes. Table J1b presents other development
project activities taking place in the “high impact” zones.
Table J2 presents other contextual factors possibly associated with levels of coverage in the HIZs and
the rest of Benin. To our knowledge, there were no natural disasters, famines or other emergencies in
the HIZs or comparison areas from 2000 to present. In 2005, the regions of Aribori and Atakora in the
66
north of Benin did experience food insecurity, associated with the famine in neighboring Niger.
There were a number of important national policy changes during this period, which are also reviewed
in the table, notably:
• Introduction of pentavalent vaccination in June 2005
• Change in first-line antimalarial policy from chloroquine to ACTs, with implementation not
generalized
• Importation of polio from Nigeria to Benin, with 2 cases in late 2003,and 6 cases in early 20045
necessitating the organization of national immunization days (2 campaigns/year in 2004, 4
campaigns/year in 2005 and 2006, and 1 campaign in April 2007)
• Distribution of Vitamin A supplementation twice a year, coupled with polio campaigns when they
were organized, starting in 2002
• Changes in policies concerning the price of ITNs
Available information about changes in health services over the study period is presented in further
details in table J3. Finally, table J4 describes the evolution of antimalarial resistance, policies and
availability in Benin.
A84
IIP-JHU | Retrospective evaluation of ACSD in Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A85
ITNs, Case
management
& feeding
practices
IMCI+
ACSD
ELEMENT
Financial and nutritional care and
support for children & advanced
strategy plan
Care, support, and consultation for
orphans and vulnerable children.
IMCI centre and curative care.
Care and support to pedetric centre
in Abomey, especially children with
severe malnutrition. Built Sedogoho
centre and trained personel.
Care and support for orphans and
vulnerable children; malaria,
HIV/AIDS, and/or TB patients; and
children with a Buruli ulcer.
Nutritional, pschycological, and
medical care and support for 360
Orphans and Vulnerable Children
due to HIV/AIDS
Maternal & child health through MOH
support; ITNs & antimalarial
treatment
Orasel (ORS) advocacy;
Nutritional rehabilitation and home
training
Quality assurance through
accompanied strengthening of
medical capacities by Tutorat*
method.
OTHER DEVELOPMENT
ACTIVITIES WITH POSSIBLE
INFLUENCE
Abomey
Abomey; 2000 to
present
Abomey:
present
Abomey; 2001 to
present
Centre de Santé St Enfants
Jesus
Terres des Hommes
Centre de Sante de Davougon
Catholic Relief Services (CRS)
1980s
Agbangnizoun; 20052006
DAA; late 2006-2007
DAA; 1993 - Present
PAK; 2005
PAK; 2004-2005
GEOGRAPHICAL
COVERAGE &
TIMING
Bornn Fonden (German NGO)
Integrated Family Health
Project (PISAF)
Ministère de la Famille et de
l'Enfant (Child and Family):
Centre de Promotion Sociale
in collaboration with CRS
Population Services
International (PSI)
USAID/URC (University
Research Corporation)
LEAD AGENCY
Table J1b: Health and development project activities in PAK and DAA zones, Benin
-
Collaboration with the Child Protection
Division of UNICEF
Inaugurated in mid-2006, only ITN
distribution in Feb 2007; Collaborated with
UNICEF on ITNs
In Tutoral, a team integrates into an existing
team and helps them to strenghten their
quality of care and services.
NOTES
A86
IIP-JHU | Retrospective evaluation of ACSD in Benin
Other
activities
(without
likely shortterm
influence on
coverage)
Antenatal,
delivery and
postnatal
care
ACSD
ELEMENT
WFP
MoH: PADSEA (Program
d'Appui au Developpement du
Secteur Eau et
Assainissement) with
support/funding from DANIDA
Hygiene and sanitation: latrine
building and increase provision of
potable water; promoting proper
hygiene and sanitation methods (ie.
washing-hands)
Population Service
International (PSI)
GADMIR/Action Social
UNFPA
ABPF (Association Béninoise
de Planning Familial) - Family
Planning
USAID
LEAD AGENCY
Promotion of formal and non-formal
education for children and youth
(cantines scolaires)
Care and support for OVC (Orphans
and Vulnerable Children) due to
HIV/AIDS; PMTCT follow-up
HIV awareness for sexually-active
adults, young adults, women of
reproductive age, pregnant women
and mothers with children under five
years of age; Ditribution of
contraceptives incl. condoms.
Maternal and child health; Family
Planning consultations
Maternal and neonatal mortality
reduction: Technical strenghtening
on: (1) Medico-technical equipment;
(2) information technology; (3)
provision of emergency obstetrical
kits, and contraceptives; (4) vehicle
and motorcycle provision.
Training of quality maternity health
providers in postpartum hemorrage
prevention
OTHER DEVELOPMENT
ACTIVITIES WITH POSSIBLE
INFLUENCE
DAA (Project is
applied to the dept);
1993-2009
DAA; Present
DAA; 1993 - Present
PAK; 2000-present
Abomey; on-going
PAK: 1988-present
PAK & DAA; 2004
GEOGRAPHICAL
COVERAGE &
TIMING
NOTES
IIP-JHU | Retrospective evaluation of ACSD in Benin
A87
ACSD
ELEMENT
Water and sanitation
Agriculture and health
Maternal mortality reduction:
Personnel training on:(1) organizing
audits of maternal deaths; (2) Family
planning and STI/HIV/AIDS
integrated care and management
protocols.
OTHER DEVELOPMENT
ACTIVITIES WITH POSSIBLE
INFLUENCE
Prevention and treatment of
HIV/AIDS including Prevention of
Mother-to-Child Transmission. Have
CIPEC (Centre d'Information de
Prospective d'Education et de
Conseils pour le SIDA)
PAK
UNFPA
Ministère de l'Eau et des
Ressources Hydrologiques:
Direction Hydraulique
Villageoise in collaboration
with GTZ
PAK
Djidja: present
DAA & PAK; late
1980s-present
CIRAPIP
LEAD AGENCY
PNLS National Plan against
HIV/AIDS in collaboration with
local NGOs. In Abomey,
particularly with Peace Benin.
And CNLS (National AIDS
Committee)
GEOGRAPHICAL
COVERAGE &
TIMING
NOTES
Table J2: Contextual factors and possible associations with ACSD implementation
packages, Benin.
Crosscutting
EPI+
CONTEXTUAL FACTOR
TIMING
Introduction & set up of
additional health centers
(HC) in DAA
2003- Outo HC in DAA;
2004-Monsourou
28
HC
in
DAA;
2005- Lobeta HC & Sahè
maternity in DAA;34, 40
2006- Honhou HC in DAA
Possible
links
to
interventions
delivered
through
outreach
(vaccination) and through
facilities (IMCI & ANC)
Road
construction
Ouémé-Plateau (PAK)
Periodically
Possible
increased
access to health services
& economic activity
Introduction of Hepatitis B
vaccine
2003
?
Introduction of de-worming
with campaigns
2003- Ouémé/Plateau and
Zou/Collines including DAA
and PAK
2005- national level
Possible
better
attendance at campaigns
due to de-worming
Introduction of pentavalent
vaccine into national policy
& EPI
June 2005
Hib vaccination
Wild poliovirus
Benin5
2003 and 2004
found
in
in
A88
Sizable investments
time & resources
combat polio
Increase coverage
measles and decrease
measles cases
of
to
in
in
Measles campaigns
2003 and 2005
Polio campaigns
on-going
Vitamin A distributed
during campaigns
Pricing changes in ITNs
2002 - CFA 3500 (~7 USD);
2003 - official price reduced to
cfa1500 (~ 3 USD) for
pregnant women and children
under 5 years of age.
2005 -- CFA 500 (~ 1 USD)
for targeted populations
Possible
changes
in
demand (& supply) due to
pricing
ITN stock-outs
late 2005 to mid-2007
Few ITNs available in
HIZs
ITNs
IMCI+ Case
management
&
feeding
practices
POSSIBLE
ASSOCIATION WITH
COVERAGE
OUTCOME(S)
Increasing
levels
chloroquine resistance
of
ACTs (CoArtem) as national
1st line antimalarial policy
Since
2001-present
appendix table J4)
2004
(see
Decrease in
chloroquine
use
of
No availability of 1st line
antimalarial in most of
Benin at present
IIP-JHU | Retrospective evaluation of ACSD in Benin
Table J3: Evolution of health service factors in PAK, DAA and comparison areas, Benin
Health services factors over time (source: Annuaire
statistique 2003, 2005)
Number of health zones
PAK
DAA
Benin, exclude Cotonou & HIZ
2003
2005
1
1
28
1
1
28
1/1
0/1
23/28
1/1
0/1
24/28
3
3
2.2
3
3
2.3
13
17
14
14
21
19
94%
69%
87%
100%
72%
91%
17,056
11,224
13,406
17,116
9,902
14,217
19,313
27,288
28,696
23,266
29,813
23,679
4,383
13,528
6,106
3,071
5,024
3,770
Functional Hospital per zone
PAK
DAA
Benin, exclude Cotonou & HIZ
Commune health center per zone (2ndary)
PAK
DAA
Benin, exclude Cotonou & HIZ
Arrondissement health center per zone (1ery)
PAK
DAA
Benin, exclude Cotonou & HIZ
% arrondisements covered in CSA or CSC
PAK
DAA
Benin, exclude Cotonou & HIZ
Population per public health center
PAK
DAA
Benin, exclude Cotonou & HIZ
Population per Public Doctor
Zou (DAA+COZ+BZZ)
Plateau (PAK + SI)
Benin, exclude Cotonou, Zou, Plateau
Population per Public Nurse
Zou (DAA+COZ+BZZ)
Plateau (PAK + SI)
Benin, exclude Cotonou, Zou, Plateau
IIP-JHU | Retrospective evaluation of ACSD in Benin
A89
A90
IIP-JHU | Retrospective evaluation of ACSD in Benin
2000
2001
PNLP
PNLP
67
Lokoss
a-Mono
SèmèPodji –
Ouémé
67
42
21%
42
14%
PNLP
PNLP
67
AdjarraOuémé
Parako
uBorgou
67
40
18%
43
14%
, 69
69
69
PNLP,
World
Bank68,
PNLP,
World
Bank68,
PNLP,
World
Bank68
Aubouy,
et al70
Ouidah
Atlantique
“National
reported
by PNLP
Lokoss
aMono
DassaZouméCollines
Abome
y-Zou
PNLP,
World
68
Bank
14
86%
315
35%
2005
62
61%
2004
Source
No data
2006
No data
2007
Ltd availability in Mono-Couffo*
65
63
39%
2003
No data
No data
*Starting in 2005, ACTs were available at a very limited scale in Mono-Couffo departments through Global Fund support.
NOTE: Chloroquine resistance data adapted from review prepared and shared by Alex Rowe of CDC-Atlanta
Area of study
N of patients
% treatment
failure
28%
2002
Chloroquine Resistance in South & Central Benin
Year
ACTs
Chloroquine
Availability of:
Chloroquine
ACTs
1st line antimalarial policy
Table J4: Cholorquine resistance patterns in southern and central Benin and first line treatment availability and policies in Benin
2000 to 2007
APPENDIX K
Description of methodological challenges
This section discusses the methodological challenges faced by the evaluation team. These are related to
the retrospective nature of the evaluation, which necessitates relying on existing—even if imperfect—data
and information. The drawbacks of retrospective evaluations have been explained elsewhere.71 This
section first discusses general methodological considerations, and then describes challenges in
measuring levels of coverage for each ACSD implementation package. Complementing this section,
appendix F provides descriptions of surveys included in the evaluation, appendix E provides a list of the
questions utilized for indicator calculation from each survey, and appendix Q compiles the questions from
each survey.
General methodological challenges.
The principal methodological weakness in the retrospective evaluation is the limited sample size available
in the 2001 DHS survey for calculation of baseline coverage indicators, especially those indicators
measured among limited subgroups, such as exclusive breastfeeding among infants less than 6 months
or complementary feeding among children 6-9 months of age. Although these small sample sizes are still
representative of the HIZs and do not introduce a bias into the estimations, they are much less precise
than later estimates based on larger sample sizes. Confidence limits for the point estimates are
presented in appendices G and H. These small sample sizes also affect the statistical power to detect
small differences over time.
Appendix F provides a full description of the methodology and conduct of surveys utilized in the analysis.
The 2001 and 2006/7 DHS used in both the adequacy and plausibility comparisons in the evaluation were
very similar in methodology and conduct, including; the sampling strategy, technical assistance provided,
field agents recruited, and data processing procedures. Some differences, however, are worth noting.
The 2006 DHS was conducted in conjunction with an economic survey that was representative at the
commune level; thus its sample size was three times larger than in 2001. Interviewer training, especially
for the vaccination and bed-net modules, was also more intensive in 2006. It is unlikely that these
differences would have greatly biased the measured levels of coverage or the differences between
districts.
The data from the supplemental surveys carried out in the HIZs in May of 2007 were merged with the
data from households surveyed during the DHS 2006. These surveys were methodologically similar in
almost all aspects. If anything the quality of the 2007 supplemental survey may have been superior due
to: 1) interviewers were already experienced in the DHS questionnaire and received additional training,
based partially on common errors seen in the 2006 DHS; 2) shorter questionnaires in 2007; and 3) more
intensive supervision. In the DHS 2006, supervision occurred once every two weeks for three months,
while in the supplemental survey; supervisors remained in the field for the entire month of data collection.
The period of data collection in the 2001 and 2006 DHS surveys was between August and September,
spanning the end of the rainy season and the beginning of the dry season. The supplemental survey was
carried out in May of 2007, which was the start of the rainy season. This difference in seasonality did not
appear to influence estimates of ITN use or malarial treatment, which were not statistically different
between the 2006 and 2007 surveys (table K1).
IIP-JHU | Retrospective evaluation of ACSD in Benin
A91
Table K1: Comparison of coverage indicator estimates from the DHS 2006 and Supplemental DHS 2007
in PAK and DAA zones, Benin
ACSD Indicator
Indicator (%)
2006
2007
Chi2 (p)
Combined
Estimate
EPI+
Any measles vaccination (12-23m)
Any 3x DPT vaccination (12-23m)
Any 3x HIB vaccination (12-23m)
Percentage of children 6 - 59 who received at least one high
dose vitamin A supplement within the last 6 months
Percentage of children aged 0-59 months sleeping under an
insecticide treated mosquito net (ITN: trt'd <=12m) (woman
quest.)
Percentage of pregnant women sleeping under an insecticide
treated mosquito net (ITN: <=12) (household quest)
60
64
8
59
62
31
NS
NS
<0.01
59
63
18
56
67
0.01
61
28
24
NS
26
22
23
NS
22
26
36
NS
30
68
66
NS
67
38
31
NS
34
47
48
NS
47
31
22
NS
27
70
91
<0.01
78
67
64
NS
66
64
63
NS
64
89
78
0.03
84
59
49
0.07
55
51
64
0.01
56
78
68
0.06
74
35
43
NS
38
4
11
0.04
7
IMCI+
Percentage of children aged 0-59 months with suspected
pneumonia taken to an appropriate health provider
Percentage of children aged 0-59 months with fever receiving
antimalarial drugs*
Percentage of children aged 0-59 months with diarrhoea
receiving ORS, RHF or increased fluids and continued feeding
Percentage of newborns put to the breast within one hour of
birth
Percentage of infants aged 0-5 months who are exclusively
breastfed
Percentage of infants aged 6-9 months who are breastfed and
receive complementary food
Percentage of children aged 20-23 months who are currently
breastfeeding
ANC+
Percentage of pregnant women who report at least 3 prenatal
visits to a trained worker (doctor, nurse or midwife)
Percentage of newborns receiving a postnatal visit by a trained
worker (doctor, nurse or midwife) within 3 days of delivery.
(women with institutional deliveries assumed to have received
postnatal care)
Percentage of newborns protected against tetanus (2+ doses
TT during pregnancy)
Percentage of pregnant women receiving 3 months of iron
supplementation.
Percentage of births attended by skilled health personnel
(doctor, nurse or midwife)
Percentage of women receiving vitamin A supplementation
within 2 months of birth
Percentage of pregnant women receiving intermittent
preventative treatment for malaria during pregnancy in previous
year (any SP)
A92
IIP-JHU | Retrospective evaluation of ACSD in Benin
Challenges in measuring EPI+ coverage.
Because of the recent introduction of the pentavelent vaccine, there have also been changes in the way
that vaccination cards are filled in, which might influence the vaccination coverage estimate. In 2006/7,
interviewers received more intensive training on how to extract vaccination dates from the vaccination
cards than in 2001. Vaccinations were recorded separately if the child was given the pentavalent
vaccine, because some children who got the DTP without the other 2 antigens (Hib & Hep B). Thus,
there were possible differences in the way vaccination responses were collected in 2001 versus 2006/7.
Samples of vitamin A were given to interviewers at the start of the 2001 DHS and the 2006 DHS;
however, samples were only available for the second part of the supplementary DHS survey, which could
have led to a possible reduction in coverage in the 2007 supplemental sample relative to the main 2006
survey. This is assuming that mothers in 2006 may have reported different medications as being vitamin
A, and by showing them the capsule in 2007 there would be fewer false positives. On the other hand,
vitamin A mass distribution, coupled with the national polio campaign, took place in April 2007, just before
the supplemental survey. We would expect better recall of more recent vitamin A supplementation. The
finding that vitamin A coverage estimates from the supplemental survey in 2007 were higher than those
from the original 2006 survey suggests an effect of the campaign (table K1).
Challenges in measuring ITN coverage.
Technical staff at the Benin national statistical offices reported few differences in the collection of ITN
information between 2001 and 2006. Training in 2006/7 was more detailed than in 2001, with
demonstrations of types of bed-nets and retreatment kits. In 2001 bed-net use was reported by women in
the individual questionnaire, while in 2006/7, bed-net use was collected in both the household and
women’s questionnaire. The respondent for the household questionnaire is often the head of the
household, and it was found that the information about use of ITNs was statistically lower if assessed
through the household versus women’s questionnaire. In order to maintain comparability of indicators
between 2001 and 2006/7, we utilized the information collected in the women’s question for ITN indicator
calculation, except ITNs among pregnant women where treatment status had not been assessed in the
women’s questionnaire.
In the 2001 DHS survey, the questions pertaining to bed-net use among pregnant women did not assess
the treatment status of the bed-net. Thus it was not possible to calculate ITN use among pregnant
women in 2001, and bed-net use—irrespective of treatment status—is utilized as a proxy indicator.
Challenges in measuring case management and feeding practices.
The preferred indicator for treatment of fever is “treatment with an appropriate antimalarial within 24 hours
of the onset of fever.” However, the 2001 DHS survey did not contain any information about the timing of
antimalarial, and therefore we have presented treatment of fever within the last 2 weeks for all
comparisons.
Challenges in measuring ANC+
The measurement of IPT with SP among pregnant women presented many challenges. The DHS 2001
contained no information on the number of doses of SP. Also, IPT with SP during pregnancy was not
available or implemented in 2001, yet over one-fourth of women reported receiving SP during pregnancy.
The question concerning SP use during pregnancy asks about treatments taken to avoid (eviter) malaria
in pregnancy and it is possible that women interpreted this as malaria treatment. There is also a possible
nd
bias in this indicator because there were no antimalarial samples until the 2 week of the supplemental
survey in 2007. Pregnant women that had received SP for IPT, may have reported use of chloroquine—
another white, commonly known antimalarial tablet—if the health agent did not tell women the name of
the tablet.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A93
In 2001, only women who had delivered outside a facility were questioned about postnatal consultation,
and it was assumed that women who had delivered in a facility received a postnatal visit. In the 2006
DHS and 2007 supplemental survey, all women were questioned about postnatal consultation, regardless
of place of delivery. In order to maintain comparability of indicators over time, we utilized the 2001
assumption that women delivering in a facility received a postnatal visit.
Challenges in measuring mortality
The aim of this section is to provide more detail on child mortality data in Benin “high-impact” zones
(HIZs), particularly as to the data quality and its likely impact on the estimates documented in the main
report.
As explained in the main report, the focus in this annex is on under-five mortality rate (U5MR) data from a
single survey that collected data in 2006 and 2007. Figure K1 shows mortality decline by year for the
HIZs and national comparison areas. There is an apparent decline in U5MR over the ten years displayed
for both areas. However, with the large 95% confidence limits around these yearly estimates, particularly
for the HIZs, little else is clear, including differences between the mortality decline in high-impact areas
versus decline in the national comparison area. Hence, it is necessary to consider other measures that
can provide more specifics on the likely survey data quality.
Figure K1: Annual rates of under-five mortality in “high-impact” zones and comparison areas as
measured by the 2006-7 DHS, Benin
U5MR (deaths per 1000 births)
250
200
150
100
50
0
1997
1998
High-impact
95% confidence bounds
National comparison
95% confidence bounds
1999
2000
2001
2002
2003
2004
2005
2006
2007
Year
Mortality data - overall quality assessment
A first step in the data quality assessment is to focus on the elements included in table K2. This table is
used extensively in the DHS final reports to provide an assessment of data quality (see for example the
Benin DHS 2006 report, page 351). The table naturally divides into three parts.
The first part, on number of births, is used to identify any unexpected peaks or dips in the number of
living, dead or total births, and the right-most set of three columns in the table, headed Calendar year
ratio helps more easily identify these variations. If the number of births changed in the same direction by
the same amount each year, the value in these last three columns would be 100. The wider the
difference from this smooth change in the number of births, the larger the divergence from 100. Table K2
shows a wide variation around 100 – from 68 to 137.
A94
IIP-JHU | Retrospective evaluation of ACSD in Benin
Despite the detail provided by these data, a chart can provide a clearer picture. Figure K2 shows the
number of births by year from table K2 and highlights an issue that has become a common occurrence in
DHS – the shift of births from the 5-year period immediately before the survey data collection, to the
previous 5-year period.
The primary cause of this shift of births has been ascribed to interviewers pushing births outside a period
where they have to ask many detailed questions about a child. For the DHS 2006, this period applied to
any child born after 1 January 2001, and for the 2007 survey component the period was for any child born
after 1 January 2002. The dips in births for 2001 and 2002 are evident in figure K2, as is the peak in
2000. The result of this can be a shift in mortality between the two 5-year periods generally used for
reporting U5MR by DHS. In general, this appears to lead to a decrease in mortality for the 5-year period
immediately before data collection, and an increase in mortality for the preceding 5-year period – leading
to an estimated faster decline in mortality than is actually occurring.
However, the comparison periods used in this ACSD evaluation, shown in the boxes at the base of figure
K2, result in a reduction of the impact of this shift of births on mortality estimation. This is due to the
averaging of births over each of the comparison periods and, in particular, the baseline period includes
both the major dips and peak.
Figure K2: Births and deaths by calendar year for “high-impact” zones, combined 2006 and 2007
DHS data, Benin
700
Living
Dead
Total
600
Births
500
400
300
200
100
Jul 1999 - Jun 2002
0
1997
1998
1999
2000
2001
Jan 2004 - Dec 2006
2002
2003
2004
2005
2006
2007
2008
Year
IIP-JHU | Retrospective evaluation of ACSD in Benin
A95
A96
IIP-JHU | Retrospective evaluation of ACSD in Benin
3
2
105
539
563
522
538
449
411
584
480
478
2267
2403
1854
1277
1209
9010
Total
98.0
96.8
94.3
93.8
94.7
87.4
83.7
83.8
80.1
90.9
85.7
74.8
74.0
71.6
81.7
Living
95.7
93.1
95.2
85.0
85.0
77.1
72.6
62.1
67.6
91.6
72.5
70.9
66.4
65.0
72.1
Dead
97.8
96.6
94.5
92.6
93.4
85.7
81.9
80.7
78.0
91.0
83.6
74.2
72.4
69.9
80.1
Total
Percentage with complete birth
[2Bx/(Bx-1+Bx+1)]x100, where Bx is the number of births in calendar year x
(Bm/Bf)x100, where Bm and Bf are the numbers of male and female births, respectively
Both year and month of birth given
0
38
35
65
67
62
68
94
69
80
204
373
314
273
316
1481
105
503
528
457
470
387
342
490
411
399
2063
2029
1540
1004
893
7529
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
2003-2007
1998-2002
1993-1997
1988-1992
<= 1991
All
1
Dead
Living
Calendar
year
Number of births
101.8
101.2
116.9
93.8
96.1
87.3
92.3
105.1
101.1
102.7
96.3
102.4
108.8
97.8
101.1
Living
115.9
119.3
125.3
110.3
105.2
77.3
100.2
147.0
110.9
117.4
105.5
123.1
127.9
115.7
116.9
Dead
Sex ratio at birth
Table K2: Births by calendar year in “high-impact” zones for combined 2006 and 2007 DHS, Benin
102.7
102.3
117.9
95.7
97.3
85.4
93.5
110.2
102.7
104.0
97.7
105.6
112.6
102.2
103.5
Total
68.0
127.5
105.5
91.9
87.2
137.2
79.3
-
-
Dead
110.0
91.6
111.4
95.3
78.0
130.1
92.5
Living
Calendar year ratio
-
106.1
94.8
110.8
94.6
79.6
131.1
90.4
Total
The second part of table K2 is the three sets of columns headed Percentage with complete birth date.
This shows that births with a complete birth date vary from 98% down to 62% over the ten-year period
from 1998 to 2007. Not having a complete birth date (month and year) increases the uncertainty of
the mortality estimates and hence one would like to have close to 100% of births with complete birth
dates. However, respondents in Benin and other countries in West Africa have difficulty in providing
complete birth dates, as can be see from a review of the comparable table in DHS reports in Mali and
Senegal. Indeed, the previous DHS in Benin in 2001 had incomplete birth dates ranging from 99%
down to 26% in the ten-year period before the survey.
However, month is the major missing part of the birth date. For example, the 2006 DHS had 13.5% of
birth dates with missing month and only 0.1% with missing year. The 2001 DHS had 38.9% missing
month and 0.1% missing year. The implication is that mortality estimates for multiple year periods
should reduce the impact of missing month.
The third part of table K2 is the three columns headed Sex ratio at birth. These ratios are used to
check for the last row of table that the sex ratio of total births is around 105, as generally more males
than females are born. The sex ratio for those that have died should also be larger than sex ratio for
total births since in general more males die than females. In addition, the table is used to assess
variability by year. In the latter case, there is a noticeable dip around 2001, to 77, and a peak around
1999 to147. This suggests that the shift in births noted in figure K2 may also be associated with a
differential shift with respect to sex, and particularly in terms of deaths.
However, table K3 shows that the periods used for calculating mortality (as delineated in figure A2)
provide an averaging of births and deaths data across the low and high sex-ratios. Hence the periods
used in the ACSD evaluation in Benin for estimating endline and, particularly, baseline mortality,
reduce the impact of these sex-ratio variations.
Table K3: Sex ratio at birth by calendar year in the “high-impact” zones for combined 2006 and
2007 DHS, Benin
Sex ratio at birth*
Living
Dead
Total
2007
2006
101.8
115.9
102.7
Multi-year sex ratio
Living
Dead
Total
2005
2004
2003
2002
2001
2000
1999
1998
All
101.2
116.9
93.8
96.1
87.3
92.3
105.1
101.1
101.1
119.3
125.3
110.3
105.2
77.3
100.2
147.0
110.9
116.9
102.3
117.9
95.7
97.3
85.4
93.5
110.2
102.7
103.5
106.0
121.1
107.2
95.2
104.2
96.5
Calendar year
* (Bm/Bf)x100, where Bm and Bf are the numbers of male and female
births, respectively
A conclusion from the above is that there are quality concerns with the mortality data from the HIZs,
but that they are reduced by the selection of baseline and endline periods for calculation of U5MR.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A97
Mortality data – quality assessment by sub-samples
The following focuses on disaggregation of the mortality estimates in terms of male-female, HIZs and
survey data collection period.
First, the HIZs (PAK and DAA).are located in two different regions with somewhat different situations.
Additionally, there were two survey data collection periods, with the first in late 2006 as part of the
DHS 2006. However, this produced too small a sample for assessing mortality for the ACSD project,
so an additional sample was selected and interviewed in early 2007 in the HIZs only. The
questionnaires for both survey periods used the same full birth history set of questions for estimating
child mortality, and included many of the same questionnaire modules as the DHS 2006.
Table K4 presents these disaggregated by survey year (2006 versus 2007), child’s sex and health
zone. Most striking is the difference between the 2006 survey in HIZs versus the 2007 survey. These
two sets of estimates should be the same, within sampling error, since they use the same questions
and were sampled from the same population. However, when one calculates the difference between
the two baseline estimates (for period July 1999 – June 2002), which is 60, and then the standard
error (SE) of this difference, one obtains 95% confidence limits for the difference of 23 to 97; even the
99% confidence limits do not include zero. Hence, while there is a chance that this difference is due to
sampling error, the likelihood of this is very small, at less than 1 in 100.
Table K4: Disaggregated under-five mortality rates in “high-impact” zones and comparison area
zone as measured in the 2006-7 DHS, Benin
Jan 2004 - Dec 2006
U5MR
SE
July 1999 - Jun 2002
U5MR
SE
123
8.0
141
9.7
% reduction
from baseline
13
High impact 2007
High impact 2006
120
125
10.0
12.4
174
114
14.2
11.8
31
-10
Male
Female
122
123
12.7
10.0
159
123
14.4
13.6
23
0
PAK
DAA
101
142
10.0
10.8
133
148
12.2
14.9
24
4
109
4.1
145
4.9
25
110
108
5.5
5.4
154
136
6.2
6.3
29
21
Area
High impact (2006+2007)
National comparison
Male
Female
What are the implications from such a finding? If one could identify data from one of the survey
components to be considerably better than the other, then it would be appropriate to weight the survey
estimates so as to favor the better quality data. However, given the wide differences between the
baseline estimates of mortality from the two survey components, this can have a very major impact,
varying from a 10% increase in mortality over the ACSD project period (from the 2006 component) to a
31% decrease in mortality (from the 2007 component).
Table K5 provides a basis for assessing data quality between the two surveys. For the most recent
ten years the calendar year ratio has 2007 survey extremes of 65 to 126, whereas the 2006 survey
extremes are 60 to 156; for the sex ratio at birth the 2007 extremes are 87 to 136, whereas the 2006
extremes are 40 to 163. However, the percentage with complete birth date has extremes of 54 to 100
A98
IIP-JHU | Retrospective evaluation of ACSD in Benin
for 2007 and a lower range of 73 to 100 for 2006. A conclusion from this is that the 2007 data may be
better than the 2006 data, but the evidence is not strong.
Table K5: Births by calendar year for living and dead children by survey sample year in the “highimpact” zones, Benin
Number of births
Calendar
year
Living
Dead
Percentage with complete birth
date¹
Total
Living
Dead
Total
Sex ratio at birth²
Living
Dead
Total
Living
Dead
Total
87
134
188
108
88
108
101
136
93
125
105
150
128
132
125
99
110
116
119
91
91
103
110
109
86
108
100
99
114
92
103
101
102
105
86
100
112
98
90
-
79
126
95
65
118
126
84
99
-
99
104
104
83
103
115
95
92
-
195
103
95
112
117
40
99
163
134
112
106
103
128
107
111
96
92
117
100
103
67
83
111
117
100
96
111
112
109
104
119
83
117
103
62
147
89
117
-
56
130
112
117
60
156
74
114
-
113
88
117
105
62
148
87
116
-
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
2003-2007
1998-2002
1993-1997
1988-1992
<= 1991
All
105
244
237
226
207
169
184
201
175
155
1,019
884
652
395
310
3,260
0
24
20
27
23
22
44
52
39
41
94
198
150
99
121
663
105
267
257
253
231
191
228
253
214
196
1,113
1,083
802
494
431
3,923
97
96
93
93
96
79
81
78
74
85
82
65
60
52
74
HIGH IMPACT AREAS (2007)
99
93
97
113
93
96
115
88
93
112
71
91
89
69
93
91
65
76
102
57
76
112
54
73
104
59
71
84
86
85
106
60
78
99
48
62
90
40
56
111
38
48
80
54
71
99
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
2003-2007
1998-2002
1993-1997
1988-1992
<= 1991
All
259
291
231
263
218
158
289
236
244
1,044
1,145
888
609
583
4,269
14
15
38
44
40
24
42
30
39
110
175
164
174
195
818
272
306
269
307
258
183
331
266
282
1,154
1,320
1,052
783
778
5,087
99
98
96
94
94
97
86
88
84
97
89
82
83
82
88
HIGH IMPACT AREAS (2006)
100
99
93
93
98
91
100
96
122
92
94
98
94
94
100
100
97
73
92
86
81
73
87
106
76
83
114
96
97
99
87
89
94
92
84
112
82
83
108
82
82
109
87
88
103
1
Both year and month of birth given
2
(Bm/Bf)x100, where Bm and Bf are the numbers of male and female births, respectively
3
[2Bx/(Bx-1+Bx+1)]x100, where Bx is the number of births in calendar year x
Calendar year ratio³
Further disaggregation of the data in table K4 may also be informative and specifics are shown in table
K6. As can be clearly seen, the difference between the 2006 and 2007 surveys continues even when
disaggregated by male and female and by high-impact area (PAK and DAA). The largest difference is
for DAA where the 2007 survey estimates a 28% reduction in mortality, whereas the 2006 estimates a
25% increase over the period of the ACSD project. While it is difficult to accept a 25% increase in
U5MR in the face of a widespread decline in mortality in Benin, this increase is not statistically
significant.
In summary, while the difference between the mortality estimates from the 2006 and 2007 surveys has
only a small chance of being due to sampling error, it is difficult to attribute this difference to one
survey being of better quality than another. At the same time an increase in mortality over the period
of the ACSD project goes against the overall trend of a reduction in mortality in Benin. Hence, while
IIP-JHU | Retrospective evaluation of ACSD in Benin
A99
one is tempted to give greater weight to the 2007 survey data, there is insufficient strength of evidence
to support a change in weight. The conclusion from this is that the reduction of 13% in U5MR reported
in the main section should stay as is, but with a strong caveat that the quality of the mortality data is
problematic, due to evidence of sizeable non-sampling errors in addition to sampling errors.
Table K6: Under-five mortality rates disaggregated by survey sample year, child sex, and health
zone in the “high-impact” zones, Benin
U5MR
Jan 2004 - July 1999 Dec 2006
Jun 2002
Area
Male
% reduction
from baseline
2007
2006
111
132
199
127
44
-4
2007
2006
128
118
148
103
14
-15
2007
2006
98
104
149
118
34
12
2007
2006
146
140
204
112
28
-25
Female
PAK
DAA
A100
IIP-JHU | Retrospective evaluation of ACSD in Benin
APPENDIX L
Tables presenting additional nutritional analyses
Figure L1 : Protocol for inclusion and exclusion of cases for nutrition analyses in “high-impact” zones
and national comparison as measured in 2001 DHS, Benin
Ex clu d ed:
N ation al
co mp ariso n
2001
T otal Number of children
under fiv e
n =4435
216 (5%)
408 (10%)
28 (1% )
Exclud ed :
H igh imp act
zo n es
n= 391
Did not s leep in
household las t
night
Non- biological
c hildren
Inc omplete
res ult
12 (3% )
17 (5% )
0 (0%)
Wt/H t
3% outliers
Wt /H t
3% outliers
n =3677
n= 352
531 (14%)
Unknown DOB
20 (6% )
H t/ag e
4% outliers
H t/ag e
3% ou tlie rs
n =3139
n =331
W t/age
1% outliers
Wt /ag e
0% ou tlie rs
n =3230
n =342
IIP-JHU | Retrospective evaluation of ACSD in Benin
A101
Figure L2 : Protocol for inclusion and exclusion of cases for nutrition analyses in “high-impact” zones
and national comparison as measured in 2006-7 DHS, Benin
Excluded:
National
comparison
2006/2007
T otal Number of children
unde r five
n=14061
388 (3%)
1560 (11%)
Did not sleep in
household la st
nigh t
N on-biological
Excluded: High
impact zones
2006
2007
n=1376
n=1128
20 (2%)
8 (1%)
122 (9%)
58 (5%)
14 (1%)
8 (1%)
ch ildren
268 (2%)
Wt/Ht
6% outliers
Wt/Ht
8% outliers
Wt/Ht
2% outliers
n=10726
n=1120
n=1035
46 (4%)
46 (4%)
Ht/age
9% outliers
Ht/age
8% outliers
Ht/age
5% outliers
n=10283
n=1078
n=963
Wt/age
2% outliers
Wt/age
1% outliers
Wt/age
1% outliers
n=11172
n=1166
n=1000
93 (1%)
A102
Incomplete
result
U nkno wn DOB
IIP-JHU | Retrospective evaluation of ACSD in Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A103
43%
45%
29%
Mother's education level
None
Any formal education
12%
44%
46%
63%
47%
52%
45%
40%
49%
21%
0-11
12-23
24-35
36-47
48-59
Wealth index quintiles
Poorest
2
3
4
Least Poor
Age
39%
43%
Rural
Male
30%
Urban
Female
Sex
Residence
% stunted (<
-2 SD)
5%
22%
16%
23%
22%
24%
20%
20%
19%
6%
10%
20%
19%
16%
20%
8%
% severely
stunted
(< -3 SD)
73
80
64
79
35
65
56
67
65
78
92
239
148
183
259
72
n
HIGH IMPACT ZONES
27%
37%
39%
43%
41%
44%
45%
49%
44%
14%
30%
40%
36%
39%
40%
32%
8%
17%
15%
19%
17%
16%
18%
23%
17%
5%
10%
17%
14%
16%
16%
12%
%
severely
% stunted stunted (<
(< -2 SD)
-3 SD)
570
636
647
639
648
477
533
637
688
805
748
2392
1583
1556
2320
820
n
COMPARISON AREA
2001 DHS
26%
41%
50%
47%
53%
53%
52%
54%
46%
27%
49%
30%
41%
49%
48%
37%
% stunted
(< -2 SD)
10%
16%
24%
25%
26%
26%
25%
26%
20%
12%
23%
12%
19%
23%
23%
16%
% severely
stunted
(< -3 SD)
310
406
419
412
494
337
374
367
444
520
1588
453
1017
1024
1487
554
n
HIGH IMPACT ZONES
34%
44%
49%
49%
51%
46%
52%
53%
47%
31%
48%
36%
42%
49%
48%
39%
% stunted
(< -2 SD)
17%
23%
25%
28%
28%
21%
28%
29%
26%
18%
26%
18%
22%
27%
26%
20%
% severely
stunted
(< -3 SD)
1923
2055
2038
2117
2147
1802
2079
1932
2167
2300
7958
2322
5099
5180
7118
3162
n
COMPARISON AREA
2006/2007 DHS
Table L1: Prevalence of stunting among children 0-59 months of age by sub-groups of the population in the “high impact” zones and comparison
area as measured by the 2001 and 2006-7 DHS, Benin
A104
IIP-JHU | Retrospective evaluation of ACSD in Benin
11%
13%
Mother's education level
None
Any formal education
23%
15%
7%
6%
5%
19%
11%
13%
7%
11%
0-11
12-23
24-35
36-47
48-59
Wealth index quintiles
Poorest
2
3
4
Least Poor
Age
11%
Female
11%
Rural
12%
12%
Urban
Male
Sex
Residence
5%
1%
1%
5%
3%
2%
0%
3%
5%
5%
3%
3%
3%
3%
3%
3%
% severely
% wasting (<
wasting
-2 SD)
(< -3 SD)
74
82
69
85
42
75
59
69
66
82
93
259
159
192
273
78
n
HIGH IMPACT ZONES
5%
8%
9%
11%
11%
3%
4%
5%
14%
18%
7%
10%
8%
10%
9%
9%
0%
2%
3%
3%
4%
1%
1%
1%
4%
5%
2%
3%
2%
3%
2%
3%
% severely
% wasting wasting
(<
(< -2 SD)
-3 SD)
588
701
781
772
835
652
702
741
740
842
792
2885
1848
1829
2734
944
n
COMPARISON AREA
2001 DHS
5%
7%
8%
10%
9%
3%
5%
3%
11%
15%
9%
6%
7%
9%
8%
8%
2%
2%
1%
2%
2%
0%
1%
0%
4%
4%
2%
2%
3%
1%
2%
2%
% severely
% wasting wasting
(< (< -2 SD)
3 SD)
327
414
432
441
542
368
404
386
457
540
1692
464
1078
1078
1570
586
n
HIGH IMPACT ZONES
6%
9%
8%
9%
9%
5%
5%
8%
10%
12%
9%
7%
8%
9%
8%
8%
2%
3%
2%
3%
3%
1%
1%
3%
3%
4%
3%
2%
2%
3%
3%
2%
% severely
% wasting wasting
(< (< -2 SD)
3 SD)
1991
2122
2140
2217
2252
1847
2162
2019
2299
2395
2406
8316
5320
5402
7426
3296
n
COMPARISON AREA
2006/2007 DHS
Table L2: Prevalence of wasting among children 0-59 months of age by sub-groups of the population in the “high impact” zones and comparison
area as measured by the 2001 and 2006-7 DHS, Benin
IIP-JHU | Retrospective evaluation of ACSD in Benin
A105
31%
25%
32%
32%
28%
24%
18%
36-47
48-59
Wealth index quintiles
Poorest
2
3
4
Least Poor
30%
26%
12-23
24-35
22%
16%
Any formal education
0-11
30%
Mother's education level
None
Age
29%
23%
28%
21%
Female
Rural
Urban
Male
Sex
Residence
6%
9%
3%
14%
12%
5%
6%
8%
14%
10%
6%
10%
8%
9%
9%
8%
% severely
% underweight underweight
(< -2 SD)
(< -3 SD)
76
81
66
82
36
65
56
69
66
87
93
249
152
190
268
74
n
HIGH IMPACT ZONES
9%
17%
23%
26%
24%
21%
15%
20%
25%
19%
13%
22%
18%
22%
21%
16%
% underweight
(< -2 SD)
1%
4%
9%
7%
8%
4%
3%
6%
9%
6%
4%
7%
5%
7%
6%
6%
% severely
underweight
(< -3 SD)
n
583
657
669
655
665
479
535
647
712
858
765
2465
1631
1599
2376
854
COMPARISON AREA
2001 DHS
11%
18%
23%
23%
23%
24%
19%
22%
22%
16%
22%
14%
17%
23%
21%
18%
%
underweight
(< -2 SD)
2%
2%
4%
6%
7%
5%
3%
4%
5%
5%
5%
2%
5%
4%
5%
3%
% severely
underweight
(< -3 SD)
321.9
429.5
442.8
440.9
530.1
347.3
397.9
384.7
468.4
566.9
1695
470.3
1071
1094
1585.0
580.6
n
HIGH IMPACT ZONES
11%
18%
20%
25%
26%
20%
22%
22%
21%
17%
22%
14%
18%
22%
22%
16%
% underweight
(< -2 SD)
3%
5%
6%
8%
9%
5%
7%
6%
7%
6%
7%
4%
5%
7%
7%
5%
% severely
underweight
(< -3 SD)
n
2074
2207
2246
2298
2343
1884
2231
2126
2387
2540
8674
2495
5531
5638
7779
3389
COMPARISON AREA
2006/2007 DHS
Table L3: Prevalence of underweight among children 0-59 months of age by sub-groups of the population in the “high impact” zones and
comparison area as measured by the 2001 and 2006-7 DHS, Benin
Table L4: Prevalence of stunting, wasting and underweight among children 0-59 months of age in
2006-7 in the “high impact” zones and regions in the comparison area as measured by the 2006-7
DHS, Benin
Height/age
High Impact Zones
Total National Comparison area
Weight/height
Weight/age
% stunted
(< -2 SD)
% severely
stunted
(< -3 SD)
% severely
wasted
% underweight
(< -3 SD)
(< -2 SD)
% wasted
(< -2 SD)
% severely
underweight
(< -3 SD)
44.7
45.5
20.9
24.3
8.2
8.2
2.1
2.5
20.2
20.4
4.6
6.3
32.3
11.8
3.9
33.0
12.0
24.4
9.5
3.3
19.4
6.1
21.8
p<0.01
6.3
p<0.01
1.6
p<0.01
16.4
p<0.01
4.4
p<0.01
43.8
20.4
19.7
26.3
23.7
14.2
20.8
25.9
17.9
21.2
37.5
10.0
14.0
9.8
9.5
6.4
3.7
15.2
5.0
6.3
7.7
3.7
3.3
4.6
2.8
3.2
2.3
1.0
5.3
0.4
1.7
1.7
0.8
35.4
30.2
19.3
19.8
17.3
15.7
22.2
15.1
16.3
19.6
12.9
14.0
9.6
5.0
5.6
6.1
4.0
7.4
3.3
4.6
7.8
3.7
Sub-groups of NC (exclude
Cotonou & HIDs)
Far North Regions (Alibori, Atakora) 51.3
Central Regions (Borgou, Donga, &
Collines) 46.6
Southern Regions (Couffo, Mono,
Atlantique, Plateau, Oeume, Zou) 43.1
p-value p<0.01
Regions
Alibori 62.1
Atacora 40.2
Atlantiq 41.7
Borgou 48.9
Collines 45.0
Couffo 37.3
Donga 43.8
Mono 46.8
Quémé 38.0
Plateau 47.8
Zou 57.1
A106
IIP-JHU | Retrospective evaluation of ACSD in Benin
APPENDIX M
Tables presenting additional equity analyses
Table M1: Selected coverage indicators by wealth quintile, and concentration indices, in the “high
impact” zones and comparison area as measured by 2006-7 DHS, Benin.
2 0 0 6 /7 D H S
A C S D c o v er ag e in d ica t o r
H ig h I m p a c t
Zo nes
n
%
A n y m e a s le s In n o c u l a t io n ( 1 2 23m )
G eo g ra p h ic
c o m p a ris o n a re a **
n
%
59
p
68
W e a l th In d e x Q u i n ti le s
P o o re s t
2
3
4
L e as t P o o r
122
92
103
3 9 .4
5 5 .2
5 9 .2
503
501
513
59 .9
61 .9
65 .3
90
76
7 6 .5
7 6 .1
0 .1 3 2
26
502
499
70 .3
80 .7
0 .0 5 9
28
1 5 .3
2 0 .4
2 6 .5
3 2 .9
4 5 .6
0 .2 0 7
2544
2414
2353
2302
2167
C o n c e n tr a t io n In d e x
IT N u s e f o r u n d e r f iv e c h il d r e n
0 .13
W e a l th In d e x Q u i n ti le s
P o o re s t
2
3
4
L e as t P o o r
590
486
487
450
337
C o n c e n tr a t io n In d e x
V i t a m in A s u p p le m e n t a t i o n o f
c h i ld r e n ( 6 - 5 9 m )
13 .4
20 .3
25 .8
34 .1
50 .2
0 .2 4 5
0 .61
63
61
W e a l th In d e x Q u i n ti le s
P o o re s t
2
3
4
L e as t P o o r
512
433
437
392
305
C o n c e n tr a t io n In d e x
O R T f o r d ia r rh e a
4 6 .7
5 6 .3
6 5 .2
7 5 .7
6 8 .0
0 .0 8 8
33
2263
2174
2147
2071
1957
1 9 .9
2 7 .4
3 4 .3
5 5 .8
4 2 .7
0 .1 7 6
73
304
284
264
196
139
5 1 .2
7 1 .2
7 8 .4
8 3 .4
9 4 .9
0 .1 1 0
63
596
566
588
572
522
4 1 .6
6 3 .5
6 6 .8
6 8 .7
8 9 .4
0 .1 2 5
589
559
579
560
512
52 .8
58 .4
63 .2
67 .8
76 .6
0 .0 7 1
42
0 .60
W e a l th In d e x Q u i n ti le s
P o o re s t
2
3
4
L e as t P o o r
46
36
41
24
29
C o n c e n tr a t io n In d e x
S k i lle d b ir t h a t t e n d a n t
42 .0
42 .8
41 .4
39 .3
44 .5
-0 . 00 2
75
0 .03
W e a l th In d e x Q u i n ti le s
P o o re s t
2
3
4
L e as t P o o r
150
137
115
115
90
C o n c e n tr a t io n In d e x
3 + v is it s A N C c a r e
53 .8
65 .4
74 .4
87 .0
95 .8
0 .1 1 2
67
0 .96
W e a l th In d e x Q u i n ti le s
P o o re s t
2
3
4
L e as t P o o r
150
137
115
116
88
C o n c e n tr a t io n In d e x
IIP-JHU | Retrospective evaluation of ACSD in Benin
46 .2
57 .2
67 .3
76 .3
92 .7
0 .1 3 1
0 .91
A107
A108
IIP-JHU | Retrospective evaluation of ACSD in Benin
Poorest
2
3
4
Least Poor
Poorest
2
3
4
Least Poor
359
300
291
226
160
births
155
142
182
88
122
-0.058046
U5MR
1705
1801
1687
1495
1196
births
157
161
155
139
101
-0.064
U5MR
Geographic
comparison area¥
n
%
¥ Comparison area is Benin – national level, excluding the HIZs and Cotonou
Concentration Index
Wealth Index Quintiles
Concentration Index
Under-five mortality
Wealth Index Quintiles
Moderate stunting (24-59m)
High Impact
Zones
n
%
Baseline
0.94
p
407.4
332.2
312.7
263.2
165.4
births
255
211
230
224
158
1078
120.6
133.2
154.0
98.2
100.9
-0.028118
U5MR
62.7
58.8
60.8
46.7
25.9
-0.114695
High Impact
Zones
n
%
1613
1576
1598
1559
1455
births
1246
1219
1145
1144
1058
5812
113.9
125.5
122.8
107.4
68.6
-0.077
U5MR
57.5
55.8
52.9
47.7
36.9
-0.075
Geographic
comparison area¥
n
%
Endline
0.53
0.62
p
Table M2: Selected impact indicators by wealth quintile, and concentration indices, in the “high impact” zones and comparison area as
measured by 2006-7 DHS, Benin.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A109
male
female
Urban
Rural
Adja
Fon
Yoruba
Other
male
female
Adja
Fon
Yoruba
Other
male
female
Adja
Fon
Yoruba
Other
male
female
0.0
38
64
n=30; too small
13.0
11.2
9.5
10.4
10.4
259
177
146
69
254
5.0
67
10.4
6.7
7.0
5.6
18.2
3.2
292
198
162
78
282
46.0
14
6.4
78.4
69.3
76.6
78.7
69.0
52
47
19
16
50
71.4
0.37
1.00
0.59
0.59
0.02
0.69
0.09
0.65
0.57
High Impact Zones
n
%
p
73
160
29
282
274
271
125
420
476
1301
312
1194
1634
1649
834
2450
550
1443
347
1383
1865
1859
938
2786
112
279
73
284
359
389
204
544
27.5
42.9
33.1
48.1
49.3
36.7
51.8
40.4
15.8
17.6
16.3
17.7
43.0
17.4
17.1
25.6
14.4
17.3
5.7
4.9
6.7
3.6
4.4
5.1
8.3
3.5
4.7
63.7
70.5
75.2
62.5
68.7
65.2
71.4
65.2
66.9
0.05
0.10
0.01
0.89
0.00
0.83
0.17
0.00
0.44
0.24
0.20
0.23
¥ Comparison area is Benin – national level, excluding the HIZs and Cotonou
Urban
Rural
Ethnicity
Gender
ORT for diarrhea
Urban
Rural
Ethnicity
Gender
Vitamin A
supplementation of
children (6-59m)
Urban
Rural
Ethnicity
Gender
Adja
Fon
Yoruba
Other
ITN use for under five
children
Ethnicity
Gender
Any measles
Innoculation (12-23m)
ACSD coverage
indicator
Geographic
comparison area¥
n
%
p
2001 DHS
3
84
80
6
98
78
44
132
59
1177
784
24
1054
1026
558
1522
55
1330
893
27
1187
1163
625
1724
19
270
179
5
257
228
140
343
20.9
35.9
30.7
61.3
30.0
38.0
36.0
33.0
69.9
67.3
52.7
63.7
33.5
61.0
62.0
63.0
61.0
61.2
40.1
27.1
24.5
36.2
27.4
25.3
30.1
25.0
26.4
84.7
65.4
49.5
45.3
57.0
62.0
64.0
57.1
59.3
0.57
0.70
0.39
0.00
0.58
0.81
0.42
0.16
0.28
0.01
0.23
0.28
High Impact Zones
n
%
p
140
387
108
510
631
556
311
876
1873
3921
903
3571
5345
5272
3257
7360
2083
4461
980
3860
5925
5860
3587
8197
468
1005
214
755
1283
1237
787
1733
37.9
33.8
48.8
47.7
45.3
38.0
44.0
41.2
66.1
63.4
80.4
57.3
41.9
64.0
62.7
70.2
60.3
63.0
31.4
30.9
36.9
20.8
28.6
27.6
36.2
24.5
28.0
68.9
68.5
77.4
62.7
66.4
68.8
75.7
63.9
67.6
<0.001
0.41
0.01
<0.001
<0.001
0.22
<0.001
<0.001
0.25
0.01
<0.001
0.23
Geographic comparison
area¥
n
%
p
2006/7 DHS
-
-
-
-
0.47
0.66
-
0.156
0.528
-
0.85
0.41
0.597
0.926
0.123
0.01
0.10
0.50
0.814
0.133
0.574
0.01
0.40
0.59
NC v. HIZ NC v. HIZ
2001
2006/7
Table M3: Selected child health coverage indicators by gender, residence and ethnicity, in the “high impact” zones and comparison area as
measured by 2001 and 2006-7 DHS, Benin.
A110
IIP-JHU | Retrospective evaluation of ACSD in Benin
Adja
Fon
Yoruba
Other
71.0
92.4
66.5
68.0
85.6
73
15
0.21
0.13
64.8
15
15
73
0.34
77.8
0.56
73
15
73
75.6
84.9
73.6
HighImpact Zones
n
%
p
122
386
73
326
223
684
124
389
73
334
227
693
67.0
74.8
78.9
48.5
64.6
73.4
61.8
64.0
74.1
79.8
46.5
63.1
73.4
59.8
<0.001
0.012
<0.001
0.037
2006/7 DHS
17
346
227
5
163
443
17
347
227
5
164
443
79.4
73.4
47.8
85.3
63.5
64.0
63.0
83.1
83.4
58.5
100.0
73.5
85.0
69.0
0.00
0.91
0.00
0.01
472
1063
214
940
851
1948
481
1073
217
961
862
1983
72.6
79.1
78.7
49.2
67.3
75.8
63.6
82.2
90.4
81.4
52.5
74.8
83.2
71.1
<0.001
<0.001
<0.001
<0.001
-
0.227
-
0.95
0.057
0.049
0.273
0.602
Geographic comparison
NC v. HIZ NC v. HIZ
HighImpact Zones
area¥
2001
2006/7
n
%
p
n
%
p
¥ Comparison area is Benin – national level, excluding the HIZs and Cotonou
Urban
Rural
Ethnicity
Adja
Fon
Yoruba
Other
3+ visits ANC care
Urban
Rural
Ethnicity
Skilled birth attendant
ACSDcoverage
indicator
Geographiccomparison
area¥
n
%
p
2001 DHS
Table M4: Selected ANC+ coverage indicators by gender, residence and ethnicity, in the “high impact” zones and comparison area as
measured by 2001 and 2006-7 DHS, Benin.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A111
male
female
95
93
40
148
male
female
0.66
0.10
0.22
births
236
724
186
492
824
823
441
1206
U5MR
49.6
45.1
35.5
50.5
47.7
44.9
36.0
50.0
46.3
0.02
<0.001
0.23
births
20
637
393
14
552
526
285
793
122
123
106
129
U5MR
42.5
53.7
52.4
46.6
56.1
49.4
44.0
56.0
52.8
0.75
0.00
0.01
3937
3864
2378
5423
births
1008
2197
498
1923
2964
2849
1776
4038
110
108
95
115
U5MR
45.4
52.0
43.4
53.9
52.7
48.4
42.2
54.2
50.6
<0.001
<0.001
0.001
-
0.69
0.10
0.35
0.97
0.37
Geographic comparison
NC v. HIZ NC v. HIZ
area¥
2001
2006/7
n
%
p
Endline
High Impact Zones
n
%
p
653
4067
756
159
154
683
123
3818
136
725
Urban
299
153
2245
126
343
Rural
1036
137
5640
153
1138
¥ Comparison area is Benin – national level, excluding the HIZs and Cotonou
Gender
U5MR
56
50.0
46.9
55.5
42.2
53.5
51.1
High Impact Zones
n
%
p
Adja
Fon 154.0
Yoruba
Other 35
Under-five mortality:
births
Ethnicity
Urban
Rural
Gender
Moderate stunting (2459m)
ACSD Impact
measure
Geographic comparison
area¥
n
%
p
Baseline
Table M5: Selected impact indicators by gender, residence and ethnicity, in the “high impact” zones and comparison area as measured by
2001 and 2006-7 DHS, Benin.
APPENDIX N
References for Appendices
1.
République_du_Bénin. Troisième recensement général de la population et l’habitation
(RGPH-3), 2002. Cotonou, Benin, Institut National de la Statistique et de l’Analyse Economique,
Ministère Chargé du Plan, de La Prospective et du Développement, 2004.
2.
INSAE & MEASURE_DHS. Enquête Démographique et de Santé, Benin 2006 - Rapport
Finale. Cotonou, Benin et Calverton, MD, Institut National de la Statistique et de l’Analyse
Économique (INSAE) et ORC Macro, 2008.
3.
Bryce, J., Requejo, J. & 2008_Countdown_Working_Group. Tracking progress in maternal,
newborn, and child survival: the 2008 report. (avilable at http://www.countdown2015mnch.org).
New York, UNICEF, 2008.
4.
MARA_ARMA. Benin: Duration of the Malaria Transmission Season, available at:
http://www.mara.org.za/pdfmaps/BenMonthsRisk.PDF, 2000.
5.
Wild poliovirus importations--West and Central Africa, January 2003-March 2004. MMWR
Morb Mortal Wkly Rep. 53 (20): 433-5 (2004).
6.
UNICEF. Accelerated Child Survival and Development in Benin. Cotonou, Benin, Unpublished
report, 2006.
7.
Agbo-Ola, L. Rapport de mise en œuvre de la SASDE dans la zone sanitaire Pobè-Kétou
Adja-Ouèrè, Benin Cotonou, Benin, UNICEF, 2004.
8.
Information extracted from available UNICEF-Benin expediture data 2002-2004.
9.
Agbo-Ola, L. Rapport trimestriel d'activités de mise en œuvre de la SASDE dans l'OuéméPlateau pour la période de janvier à mars 2004. Cotonou, Benin, UNICEF, 2004.
10.
Ahlan, J. Rapport de supervision intégrée sur les activités CPN, vaccination, soins curatifs et
financement communautaire dans les formations sanitaires de Kpoto, Tanve, Mougnon, Djidja,
Dan, Oumbega, et Djegbe. Abomey, Benin, Ministère de la Santé Publique, Direction
Departementale de la Santé Publique du Zou et des Collines, Zone Sanitaire Djidja-AbomeyAgbangniizoun & UNICEF, 2004.
11.
Agbo-Ola, L. Rapport trimestriel d'activités de mise en œuvre de la SASDE dans l'OuéméPlateau pour la période de avril à juin 2004. Cotonou, Benin, UNICEF, 2004.
12.
Abada, R. Rapport des campagnes de re-imprégnation des moustiquaires et de distribution de
la vitamine A. Porto-Novo, Benin, Ministère de la Santé Publique, Direction Départementale de la
Santé Publique de l'Ouémé et du Plateau, 2004.
13.
Aglin, M. Rapport du ratissage dans les arrondissements à faible taux de couverture
vaccinale: Monsourou, Agouna et Houto. Djidja, Benin, Ministère de la Santé Publique, Zone
Sanitaire Djidja-Abomey-Agbangniizoun, Centre de Santé Commune Djidja, 2004.
14.
Agbo-Ola, L. Rapport trimestriel d'activités de mise en œuvre de la SASDE dans l'OuéméPlateau pour la période de juillet à septembre. Cotonou, Benin, UNICEF, 2004.
15.
A112
Rapport général de l'atelier de formation/recyclage sur la gestion du PEV (Zone sanitaire
IIP-JHU | Retrospective evaluation of ACSD in Benin
DAA, Bénin). Abomey, Benin, Ministère de la Santé Publique, Zone Sanitaire Djidja-AbomeyAgbangniizoun, 2004.
16.
Atelier de formation/recyclage des agents de santé du secteur publique et privé de la zone
sanitaire Djidja-Abomey-Agbangniizoun (DAA) sur la gestion du PEV à Sedo-Goho les 4, 5 et 6
janvier 2005. Abomey, Benin, Ministère de la Santé Publique, Zone Sanitaire Djidja-AbomeyAgbangniizoun, 2005.
17.
Rapport d'activités de fin d'année 2005: PEV (Zone sanitaire DAA, Bénin). Abomey, Bénin,
Ministère de la Santé Publique, Zone Sanitaire Djidja-Abomey-Agbangniizoun, 2005.
18.
Rapport d'investigation d'epidemie de rougeole a Djidja (Zone sanitaire DAA, Benin) Djidja,
Benin, Ministère de la Santé Publique, Zone Sanitaire Djidja-Abomey-Agbangniizoun, Centre de
Santé Commune Djidja, 2005.
19.
Azonhou, V. Rapport de la formation des aides-soignants sur la logistique du PEV de la zone
sanitaire Djidja-Abomey-Agbangniizoun Abomey, Benin, Ministère de la Santé Publique, Zone
Sanitaire Djidja-Abomey-Agbangniizoun, 2006.
20.
Zinsou, J., Sondjinou, V., Lawale, T., Adjai, O., Aiko, S., Dansou, L. et al. Resultats du
monitorage du 2ème Semestre 2005 dans les centres de sante de la zone sanitaire PAK. Pobe,
Benin, République de Benin, Ministère de Sante Publique, Direction Départementale de la Sante
Publique de L'Ouémé et du Plateau, Zone Sanitaire Pobè-Kétou Adja-Ouèrè, 2006.
21.
Edjitche, M. Rapport du deroulement de la campagne d'imprégnation des moustiquaires dans
la zone sanitaire Pobè-Kétou-Adja-Ouèrè, Benin Pobe, Benin, Ministère de la Santé Publique,
Zone Sanitaire Pobè-Kétou Adja-Ouèrè, 2003.
22.
Synthèse de l'opération d'imprégnation des MI et distribution d'albendazole: période de la
campagne: octobre-novembre 2005. Pobe, Benin, Zone Sanitaire Pobè-Kétou Adja-Ouèrè, 2005.
23.
Edjitche, M. Rapport de l'atelier d'élaboration et d'adaptation des fiches pédagogiques de
formations relais communautaires et chefs d'arrondissement des zones sanitaires PAK et DAA,
Benin 16-18 Juin, 2003. Porto-Novo, Benin, Ministère de la Santé Publique, Benin & UNICEF,
2003.
24.
Edjitche, M. Rapport de l'atelier de formation des formateurs des ASBC et chefs
d'arrondissement de la zones sanitaire PAK, Benin 23-24 Juin, 2003. Pobe, Benin, Ministère de la
Santé Publique, Zone Sanitaire Pobè-Kétou Adja-Ouèrè, 2003.
25.
Rapport de formation des agents de services à base communautaire de la zone sanitaire de
Djidja-Abomey-Agbangnizoun Abomey, Benin, Ministère de la Santé Publique, Zone Sanitaire
Djidja-Abomey-Agbangniizoun, 2003.
26.
Rapport de la supervision sur sites des agents des services a base communautaire et
groupements feminins dans la zone sanitaire Pobè-Kétou Adja-Ouèrè, Benin. Pobe, Benin,
Ministère de la Santé Publique, Zone Sanitaire Pobè-Kétou Adja-Ouèrè, 2005.
27.
Rapport de synthèse 1ère, 2ème, 3ème, et 4ème sessions de formation des prestataires de
soins, sage femmes, et infirmiers sur le paquet minimum d'activités de nutrition (PMA/Nut).
Abomey, Benin, Ministère de la Santé Publique, Zone Sanitaire Djidja-Abomey-Agbangniizoun,
2004.
28.
Notes sur les réunions mensuelles de l’équipe d'encadrement de la zone sanitaire (EEZS)
IIP-JHU | Retrospective evaluation of ACSD in Benin
A113
2004, Djidja-Abomey-Agbangnizoun. Abomey, Benin, Ministère de la Santé Publique, Zone
Sanitaire Djidja-Abomey-Agbangniizoun, 2004.
29.
Rapport de la supervision groupée (réunion de concertation) des agents des services a base
communautaire (ASBC) dans la zone sanitaire Pobè-Kétou Adja-Ouèrè - DRAFT, Benin Pobe,
Benin, Ministère de la Santé Publique, Zone Sanitaire Pobè-Kétou Adja-Ouèrè, 2004.
30.
Odounfa, V. Atelier de finalisation des boites à images de la PCIME-communautaire dans les
zones DAA, PAK et BADA 12-15 juillet 2004 : Rapport général de l’atelier Porto-Novo, Benin,
Ministère de la Santé Publique & UNICEF, 2004.
31.
Rapport de l'atelier de pre-test et de test des images et messages sur le paludisme, la
diarrhee, et les IRA, 20-25 septembre 2004. Porto-Novo, Benin, Ministère de la Santé Publique &
UNICEF, 2004.
32.
Ahoyo, H. Rapport générale de la synthese issue de la supervision des ASBC dans la zone
sanitaire DAA. Abomey, Benin, Ministère de la Santé Publique, Zone Sanitaire Djidja-AbomeyAgbangniizoun, 2004.
33.
Rapport de la 2eme serie de formation des agents de services à base communautaire (ASBC)
sur la prise en charge des infections respiratoires aigue (IRA) chez les enfants moins de 5 ans; 2326 mars 2005, Agbokpa, Benin. Abomey, Benin, Ministère de la Santé Publique, Direction
Departementale de la Santé Publique du Zou et des Collines, Zone Sanitaire Djidja-AbomeyAgbangniizoun & UNICEF, 2005.
34.
Notes sur les réunions mensuelles de l’équipe d'encadrement de la zone sanitaire (EEZS)
2005, Djidja-Abomey-Agbangnizoun. Abomey, Benin, Ministère de la Santé Publique, Zone
Sanitaire Djidja-Abomey-Agbangniizoun, 2005.
35.
Rapport de la 3eme serie de formation des agents de services à base communautaire (ASBC)
sur la prise en charge des infections respiratoires aigue (IRA) chez les enfants moins de 5 ans; 4-7
mai 2005, Agbokpa, Benin. Abomey, Benin, Ministère de la Santé Publique, Direction
Departementale de la Santé Publique du Zou et des Collines, Zone Sanitaire Djidja-AbomeyAgbangniizoun & UNICEF, 2005.
36.
Guedou, R. & Assogba, T. Rapport de formation des formateurs des agents de services à
base communautaire (ASBC) sur la utilisation des boites a images, 16-18 juin 2005 Abomey,
Benin, Ministère de la Santé Publique, Direction Departementale de la Santé Publique du Zou et
des Collines, Zone Sanitaire Djidja-Abomey-Agbangniizoun, 2005.
37.
Rapport de la 4eme serie de formation des agents de services à base communautaire (ASBC)
sur la prise en charge des infections respiratoires aigue (IRA) chez les enfants moins de 5 ans; 3-6
August 2005, Agouna, Benin. Abomey, Benin, Ministère de la Santé Publique, Direction
Departementale de la Santé Publique du Zou et des Collines, Zone Sanitaire Djidja-AbomeyAgbangniizoun & UNICEF, 2005.
38.
Degue, M. Rapport de briefing des agents de santé sur le suivi des agents de services à base
communautaire (ASBC) formes sur la prise en charge des infections respiratoires aigue (IRA) au
niveau communautaire. Abomey, Benin, Ministère de la Santé Publique, Direction Departementale
de la Santé Publique du Zou et des Collines, Zone Sanitaire Djidja-Abomey-Agbangniizoun &
UNICEF, 2005.
39.
Ahoyo, H. & Jocoue, C. Rapport de la supervision générale des ASBC de la commune de
Djidja. Djidja, Benin, Ministère de la Santé Publique, Zone Sanitaire Djidja-Abomey-Agbangniizoun,
A114
IIP-JHU | Retrospective evaluation of ACSD in Benin
Centre de Santé de Commune Djidja, 2005.
40.
Notes sur les réunions mensuelles de l’équipe d'encadrement de la zone sanitaire (EEZS)
2006, Djidja-Abomey-Agbangnizoun. Abomey, Benin, Ministère de la Santé Publique, Zone
Sanitaire Djidja-Abomey-Agbangniizoun, 2006.
41.
Houssou, S. & Balogoun, H. Rapport de la formation des groupes de soutien à l'allaitement
maternel exclusif dans la commune d'Abomey. Abomey, Benin, Ministère de la Santé Publique,
Zone Sanitaire Djidja-Abomey-Agbangniizoun, 2006.
42.
Rapport général de l'atelier du 6-8 avril 2004 sur la prise en charge du nouveau-né à la
maternité. Abomey, Benin, Ministère de la Santé Publique, Zone Sanitaire Djidja-AbomeyAgbangniizoun, 2004.
43.
Rapport de l'enquête TPI (Traitement Préventif Intermittent a la Sulfadoxine Pyriméthamine)
dans les quatre zones sanitaires DAA, ZOBOZA, PAK, SAKIF. Benin, Ministère de la Santé
Publique, 2004.
44.
Balogoun, H. Rapport de la supervision sur la CPN recentrée. Abomey, Benin, Ministère de la
Santé Publique, Zone Sanitaire Djidja-Abomey-Agbangniizoun, 2004.
45.
Rapport général de la formation sur la prévention de la transmission mère enfant du VIH/SIDA
(PTME) dans la Zone Sanitaire Djidja-Abomey-Agbangniizoun du 18 april au 22 avril 2005.
Abomey, Benin, Ministère de la Santé Publique, Zone Sanitaire Djidja-Abomey-Agbangniizoun,
2005.
46.
Balogoun, H. & Adegoute, S. Rapport général de l'atelier de formation sur la prévention de la
transmission mère enfant du VIH/SIDA (PTME) dans la Zone Sanitaire Djidja-AbomeyAgbangniizoun. Abomey, Benin, Ministère de la Santé Publique, Zone Sanitaire Djidja-AbomeyAgbangniizoun, 2005.
47.
Deguenon, G., Vinalide, V. & AZadji, E. Gestion active de la troisième de travail
d’accouchement GAPTA 2eme session de formation des prestataires de soins en santé familiale;
Sèdo, Goho, Abomey – 23-28/06/2005. Abomey, Benin, Ministère de la Santé Publique, Direction
départementale de la santé du Zou et des Collines, Zone Sanitaire Djidja-Abomey-Agbangniizoun,
2005.
48.
Rapport de l’atelier de concertations et de validation des données de référence et de contreréférence des urgences obstétricales et néonatales entre les prestataires des CSC/CSA et ceux de
l’hôpital de référence CHD (Maternité-Pédiatrie-SAMU) Sedo-Goho, Benin, Ministère de la Santé
Publique, Direction départementale de la santé du Zou et des Collines 2005.
49.
Deguenon, M. Rapport d'atelier d'adaptation des messages de la Zone Sanitaire Pobè-AdjaOuèrè-Kétou à la Zone Sanitaire DAA sur la prévention de la transmission mère enfant. Abomey,
Benin, Ministère de la Santé Publique, Zone Sanitaire Djidja-Abomey-Agbangniizoun, 2005.
50.
Noutangni, J. Rapport d'atelier de validation du (PIC) Plan Intégré de Communication de la
PTME (prévention de la transmission mère enfant) du VIH/SIDA de la Zone Sanitaire DjidjaAbomey-Agbangniizoun. Abomey, Benin, Ministère de la Santé Publique, Zone Sanitaire DjidjaAbomey-Agbangniizoun, 2005.
51.
Noutangni, J. Rapport de l’atelier d'élaboration de contrats de partenariat avec les médias
pour la promotion de la PTME dans de la zone sanitaire DAA. Abomey, Benin, Ministère de la
Santé Publique, Zone Sanitaire Djidja-Abomey-Agbangniizoun, 2005.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A115
52.
Tournée d’information, de sensibilisation et de plaidoyer auprès des palais royaux et couverts
de la Zone Sanitaire Djidja-Abomey-Agbangniizoun pour leur implication dans l’identification de
stratégies d’élaboration et de mise en œuvre des PIC/PTME. Abomey, Benin, Ministère de la
Santé Publique, Zone Sanitaire Djidja-Abomey-Agbangniizoun, 2005.
53.
WHO. WHO Child Growth Standards: Methods and development: Length/height-for-age,
weight-for-age, weight-for-length, weight-for-height and body mass index-for-age. Geneva, World
Health Organization, 2006.
54.
UNICEF. Revue des études et évaluations 2001-2006: Volume 5. Cotonou, Benin,
Programme Suivi Evaluation Benin-UNICEF, 2006.
55.
INSAE & ORC_Macro. Enquête Démographique et de Santé au Benin 2001. Cotonou, Benin
& Calverton, MD, Institut National de la Statistique et de l’Analyse Économique (INSAE) & ORC
Macro, 2002.
56.
CERTI. Enquêtes de couverture de la Stratégie Accélérée pour la Survie et le Développement
de Enfant [Benin] 2003. Cotonou, Benin, CERTI, CDC & UNICEF, 2003.
57.
CEFORP. Etude de base dans les zones d'intervention du programme de coopération BeninUNICEF 2004-2008. Cotonou, Benin, Centre d'Etudes, de Formation et de Recherches sur la
Population (CEFORP), République du Benin & UNICEF, 2005.
58.
INSAE & MEASURE_DHS. Enquête Démographique et de Santé, Benin 2006 - Rapport
Préliminaire. Cotonou, Benin, Institut National de la Statistique et de l’Analyse Économique
(INSAE), 2007.
59.
Jekinnou, P. Enquête de couverture vaccinale par grappes couplee avec des investigations
sur l'utilisation des moustiquaires impregnees et la supplementations en vitamine A chez les
enfants moins de 5 ans. Cotonou, Benin, Republique du Benin & UNICEF, 2003.
60.
Adjimon, R., Sessou, C., Tollegbe, A. & Gigigaye, M. Enquête mi-parcours au Zou-Collines
sur les comportements MIIs. Cotonou, Benin, PSI-Benin, 2006.
61.
MOH_Benin. Annuaires des statistiques sanitaires 2003. Cotonou, Benin, Ministère de la
Santé Publique, République du Bénin, 2003.
62.
MOH_Benin. Annuaires des statistiques sanitaires 2004. Cotonou, Benin, Ministère de la
Santé Publique, République du Bénin, 2004.
63.
MOH_Benin. Annuaires des statistiques sanitaires 2005. Cotonou, Benin, Ministère de la
Santé Publique, République du Bénin, 2005.
64.
MOH_Benin. Annuaires des statistiques sanitaires 2006. Cotonou, Benin, Ministère de la
Santé Publique, République du Bénin, 2006.
65.
Victora, C.G., Schellenberg, J.A., Huicho, L., Amaral, J., El Arifeen, S., Pariyo, G. et al.
Context matters: interpreting impact findings in child survival evaluations. Health Policy Plan. 20
Suppl 1: i18-i31 (2005).
66.
World_Food_Program. Where we work - Benin. Information obtained from:
http://www.wfp.org/country_brief/indexcountry.asp?country=204 accessed 15 February 2008.
67.
A116
PNLP. Situation de la chloroquine-resistance au Benin, presentation at the ProLIPO Partners
IIP-JHU | Retrospective evaluation of ACSD in Benin
Meeting in January 2002 by Dr. Dorothée KINDE-GAZARD (Coordinator of the National Malaria
Control Program in Benin). 2002.
68.
World_Bank. Project Appraisal Document for the Booster Program in Benin, 2006.
69.
PNLP. Politique et stratégies nationales de lutte contre le paludisme, presentation at CIEVRA
in July 2006 made by Dr. Hortense KOSSOU (Coordinator of the National Malaria Control Program
in Benin). 2006.
70.
Aubouy, A., Fievet, N., Bertin, G., Sagbo, J.C., Kossou, H., Kinde-Gazard, D. et al.
Dramatically decreased therapeutic efficacy of chloroquine and sulfadoxine-pyrimethamine, but not
mefloquine, in southern Benin. Trop Med Int Health. 12 (7): 886-94 (2007).
71.
Bryce, J., Gilroy, K., Black, R.E., Jones, G. & Victora, C.G. A Retrospective Evaluation of the
Accelerated Child Survival and Development Project in West Africa; Inception Report. Baltimore,
MD, Johns Hopkins University Institute for International Programs, 2007.
IIP-JHU | Retrospective evaluation of ACSD in Benin
A117
A118
IIP-JHU | Retrospective evaluation of ACSD in Benin