Clearing the Global Health Fog

W O R L D

B A N K

W O R K I N G

P A P E R

N O .

1 6 6

Clearing the Global Health Fog A Systematic Review of the Evidence on Integration of Health Systems and Targeted Interventions

Rifat Atun Thyra de Jongh Federica V. Secci Kelechi Ohiri Olusoji Adeyi

S Y S C I E N C E V I P F A D I A S

H T E M A L T H I D E N E O P U L O G O N Y

P R I S N I N T E G H O R A C E T I L A T I O N A L

THE WORLD BANK

V O R I T I E R E R V E N T I L C I Z O N T A C L D A T A L P N P R G L O B A C O H

S

T O O L S E R V I C E U D I E S

I O N U T L R I S T I D O G M A N R A L T V E R A G E

W O R L D

B A N K

W O R K I N G

P A P E R

N O .

Clearing the Global Health Fog A Systematic Review of the Evidence on Integration of Health Systems and Targeted Interventions

Rifat Atun Thyra de Jongh Federica V. Secci Kelechi Ohiri Olusoji Adeyi

THE WORLD BANK Washington, D.C.

1 6 6

Copyright Š 2009 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First Printing: March 2009 printed on recycled paper 1 2 3 4 5 12 11 10 09 World Bank Working Papers are published to communicate the results of the Bank’s work to the development community with the least possible delay. The manuscript of this paper therefore has not been prepared in accordance with the procedures appropriate to formally-edited texts.

Some sources cited in this paper may be informal documents that are not readily available.

The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the International Bank for Reconstruction and Development/The World Bank and its affiliated organizations, or those of the Executive Directors of The World Bank or the governments they represent.

The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank of the legal status of any territory or the endorsement or acceptance of such boundaries.

The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grant permission promptly to reproduce portions of the work.

For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, Tel: 978-750-8400, Fax: 978-750-4470, www.copyright.com.

All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, Fax: 202-522-2422, email: pubrights@worldbank.org.

ISBN-13: 978-0-8213-7818-2 eISBN: 978-0-8213-7936-3 ISSN: 1726-5878

DOI: 10.1596/978-0-8213-7818-2

Library of Congress Cataloging-in-Publication Data has been requested.

Contents Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Acronyms and Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix 1.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2.

Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 The Conceptual Framework for Analysing Integration of Targeted Health Interventions into Health Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

3.

Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Outcomes Reported in the Studies Analyzed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 The Extent and Nature of Integration of Health Interventions into Critical Health Systems Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 The Extent of Intervention Success in the Studies Analyzed . . . . . . . . . . . . . . . . . . . 18 How the Context Influences the Extent and Nature of Integration . . . . . . . . . . . . . 19

4.

Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

5.

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Appendixes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 A. Search Strategy and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 B. Summary of the Included Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 C. Contextual Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

LIST OF TABLES 1. Critical Health Systems Functions and Elements of Integration . . . . . . . . . . . . . . . . . . 5

LIST OF FIGURES 1. Flow Chart Representing the Selection Process for Studies Included in the Review. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2. The Extent and Nature of Integration by Targeted Health Intervention and Intervention Success as Reported in the Study. . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 iii

Acknowledgments he work program underpinning this paper was initiated and funded by the World Bank and entitled “Integration of Health Systems and Priority Health, Nutrition and Population Interventions.� Preparation of the paper was co-funded in a collaborative exercise by the Center for Health Management at Imperial College London and the World Bank. In that context the authors acknowledge consultations with a number of colleagues during an international consultation hosted by the World Bank in May 2008, and comments from several reviewers who helped to improve the manuscript. The reviewers were Tom Novotny (Professor, University of California, San Francisco) and Richard Coker (Reader, London School of Hygiene & Tropical Medicine). Sylvia Robles reviewed the initial version of a Cochrane Review, which was an earlier step in this exercise. Logan Brenzel and Peter Berman commented on the conceptual framework on which the paper is based. Sonalini Khetrapal provided comments on an earlier version of the paper and Inas Ellaham helped with formatting. The authors alone are responsible for the contents of the paper.

About the authors:

T

Rifat Atun, MBBS, MBA, FRCGP, FFPH. Professor of International Health Management, Imperial College, London. E-mail: r.atun@imperial.ac.uk Thyra de Jongh, MSc, DIC, PhD. Researcher, Centre for Health Management, Imperial College London. E-mail: thyra.de-jongh06@imperial.ac.uk Federica V. Secci, MSc. Doctoral Researcher, Centre for Health Management, Imperial College London. E-mail: f.secci07@imperial.ac.uk Kelechi Ohiri, MD, MPH, MS. Health Specialist in the Human Development Network of the World Bank. E-mail: kohiri@worldbank.org Olusoji Adeyi, MD, DrPH, MBA. Coordinator of Public Health Programs in the Human Development Network of the World Bank. E-mail: oadeyi@worldbank.org Cover design: concept by Olusoji Adeyi, text by Olusoji Adeyi and Sonalini Khetrapal, graphics by Stuart K. Tucker.

v

Acronyms and Abbreviations AIDS BINP CWG EPI EPOC FP HIV ICDDR,B IDCS IMCI LHWP M&E MCH PHC PPI TB UNICEF USAID VCT WHO

Acquired Immune Deficiency Syndrome Bangladesh Integrated Nutrition Project Community Working Groups Expanded Program for Immunization Cochrane Effective Practice and Organisation of Care Group Family Planning Human Immunodeficiency Virus International Center for Diarrheal Disease Research, Bangladesh India Integrated Child Development Services Integrated Management of Childhood Illness Lady Health Worker Program Monitoring and Evaluation Maternal and Child Health Primary Health Care Pulse Polio Immunization Tuberculosis United Nations Children’s Fund United States Agency for International Development Voluntary Counselling and Testing World Health Organization

vii

Summary longstanding debate on health systems organization relates to benefits of integrating health programs that emphasize specific interventions into mainstream health systems to increase access and improve health outcomes. This debate has long been characterized by polarization of views and ideologies, with protagonists for and against integration arguing the relative merits of each approach. Recently, the debate has been rekindled due to substantial increases in externally funded programs for priority health, nutrition and population interventions and enhanced international efforts aimed at health systems strengthening. However, all too frequently these arguments have not been based on hard evidence.

In this paper we present findings of a systematic review that explores a broad range of evidence on: (i) the extent and nature of integration of targeted health programs that emphasize specific interventions into critical health systems functions (defined in the Methodology Section), (ii) how the integration or non-integration of health programs into critical health systems functions in different contexts have influenced program success, (iii) how contextual factors have affected the extent to which these programs were integrated into critical health systems functions. We use a new conceptual framework to guide the analysis. The review evaluates peer-reviewed studies that focus on health interventions, and which have been introduced on a regional or national scale.

The debate on health interventions has tended to narrowly focus on vertical or integrated descriptors. However, our analysis shows this to be a false dichotomy as few interventions are purely vertical (single-disease oriented) or horizontal (fully integrated into mainstream functions) health system.

As this review shows, in practice the nature of the problem, the interventions to address these and the adoption and assimilation of health interventions in health systems vary greatly in different contexts, as does success. The purpose, nature, speed and the extent of integration also vary—in part, dependent on the intervention complexity, the health system characteristics and the contextual factors.

There are few instances where there is full integration of a health intervention or where an intervention is completely non-integrated. Instead, there exists a highly heterogeneous picture both for the nature and also for the extent of integration.

The review suggests that the evidence base for integration versus targeted health programming is very limited. As success was measured in different ways (for example, in terms of varied programmatic goals related to efficiency, effectiveness, or equity), analysis and drawing lessons from this is further complicated. Perhaps the modesty of evidence creates the context for strong opinions for or against integration in global health. We have attempted to show that in practice the dichotomy between integrated and non-integrated (traditionally described as vertical and horizontal) is not rigid. Health systems combine both non-integrated and integrated interventions, but the balance of these interventions varies considerably. Our findings suggest that the purpose, nature and extent of integration vary enormously between interventions and in countries, creating a rich mosaic of local solutions to address emergent problems. We found rare instances of full integration but a wide range of instances where health interventions are integrated into one or more critical health system functions. Further, it is important to note that in many countries that ix

A

x

Summary

simultaneously have applied multiple health interventions, the nature of integration for these interventions varies and different degrees of integration for the same critical health system function co-exist.

The findings provide new synthesis of evidence to inform the debate on health systems and targeted interventions. Given the highly varied contexts and adoption systems that reflect local nuances, different health system capacities and the range of problems being addressed, it is not surprising that in practice a rich mix of solutions exist. While the discussion on the relative merits of integrating health interventions will no doubt continue, discussions should move away from the highly reductionist approach that has polarized this debate. Future efforts are best spent on generating and learning from useful evidence.

CHAPTER 1

Introduction

longstanding debate on health systems organization relates to benefits of integrating health programs that emphasize specific interventions into mainstream health systems to increase access and improve health outcomes. This debate has long been characterized by polarization of views and ideologies, with protagonists for and against integration arguing the relative merits of each approach (Cueto 2004; Magnussen, Ehiri, and Jolly 2004; Newell 1988; Walsh and Warren 1979; Warren 1988; Wisner 1988).

Recently, the debate has been rekindled due to substantial increases in externally funded programs for priority health, nutrition and population interventions (such as those targeting communicable diseases, reproductive health interventions, and nutrition programmes—hereafter referred to as health interventions) and enhanced international efforts aimed at health systems strengthening (World Bank and World Health Organization 2006).

However, all too frequently these arguments have not been based on hard evidence (Atun, Bennett, and Duran 2008). The presence of both integrated and non-integrated programs in many countries suggest there may be benefits to either approach, but the relative merits of integration in various contexts and for different interventions have not been systematically analyzed and documented. Such an analysis is complicated as the term ‘integration’ is used to describe a variety of organizational arrangements in different settings (Atun, Bennett, and Duran 2008). Further, as the nature and extent of integration varies, there are methodological challenges to comparing various interventions.

In this paper we present findings of a systematic review that explores a broad range of evidence on three critical aspects of integration: (i) the extent and nature of integration of targeted health programs that emphasize specific interventions into critical health systems functions (defined in the Methodology Section), (ii) how the integration or non-integration

A

1

2

World Bank Working Paper

of health programs into critical health systems functions in different contexts have influenced program success, (iii) how contextual factors have affected the extent to which these programs were integrated into critical health systems functions. To guide our analysis, we have used a new conceptual framework, discussed in detail in a complementary paper (Atun and others 2009), to guide our analysis. Our review evaluates peer-reviewed studies that focus on health interventions, and which have been introduced on a regional or national scale.

This paper is organized in five chapters. This Introduction is followed by the methodology chapter, which includes a brief section on the conceptual framework used to analyze the studies retrieved and the programs presented within these to map the nature and extent of integration into critical health system functions. The Results chapter includes: a summary of the outcomes for each study grouped by the disease area or the clinical problem the intervention seeks to address, including the reported success; for each program, analysis and mapping of the nature and extent of integration into critical health system functions; and an analysis of how contextual factors either created opportunities for introducing or integrating a program or influenced the desirability or feasibility of program integration.

The Discussion chapter provides an overview of the implication of findings for policy makers, practitioners and researchers. The final chapter draws conclusions.

CHAPTER 2

Methodology

e developed a search strategy based on the use of exploded-MeSH terms, supplemented with a broad search for keywords in the titles or abstracts for which no appropriate MeSH terms exist. In all cases, MeSH terms were chosen to represent the highest order of relevance within each MeSH-tree. A systematic review by Briggs and Garner, which used Cochrane Systematic Review methodology, served as the basis for the development of the search strategy (Briggs and Garner 2006). The strategy combined two parts: the first was designed to identify articles related to organizational arrangements for health care delivery, and the second was designed to limit the search to specific areas identified as relevant to the World Bank health programs. The search strategy is shown in detail in Appendix A. We reviewed 8,274 potential articles, which yielded 55 articles that were included in this review (Figure 1). These papers are summarized in Appendix B.

W

The Conceptual Framework for Analysing Integration of Targeted Health Interventions into Health Systems A separate paper (Atun and others 2009) provides details of the analytic framework1 on which this systematic review is based. Here, we provide a summary of the key features of the framework.

1. A longer version of the conceptual framework is available as a World Bank Health Nutrition and Population Discussion Paper. (Atun, R., Ohiri, K., Adeyi, O. Integration of Health Systems and Priority Health, Nutrition and Population Interventions: A Framework for Analysis and Policy Choices. August 2008. Available online at: http://siteresources.worldbank.org/HEALTHNUTRITIONANDPOPULATION/ Resources/281627-1095698140167/IntegrationHealthSystemsandInterventionsFramework2008.pdf

3

4

World Bank Working Paper

Figure 1. Flow Chart Representing the Selection Process for Studies Included in the Review Database search 8,274 6,723 excluded

I. Title scanning 1,551

1,046 excluded

II. Abstract scanning

200 III. Full text scanning

28 not available 172 117 excluded

55 Included studies

In this framework, we define integration as the extent, pattern, and rate of adoption and eventual assimilation of health interventions into each of the critical health system functions, which include inter alia: (i) stewardship and governance, (ii) financing, (iii) planning, (iv) service delivery, (v) monitoring and evaluation (M&E), and (vi) demand generation.

An “intervention” in this context refers to combinations of technologies (for example, vaccines, drugs), inputs into service delivery, organizational changes and modifications in processes related to decision making, planning, and service delivery.

We view a health intervention using a “diffusion of innovation” lens, and consider this as an idea, practice, or object that is perceived as new by an individual or a unit of adoption, while recognising that in some cases the interventions that have previously been implemented on a small scale are scaled up and increased in intensity. In such instances, the “newness” relates less to the technical element of the intervention itself but the organizational changes, new financing schemes and novel processes that accompany scaling up, intensification, integration and eventual institutionalization of the intervention.

Drawing on relevant empirical evidence and theory (Atun and others 2009), we propose that the adoption and diffusion of new health interventions and the extent to which they are “assimilated” or “integrated” into the “general” health system will be influenced by five factors. These include the nature of the problem being addressed, the intervention, the adoption system which adopts and assimilates the intervention (key actors and institutions), the health system characteristics such as the absorptive capacity, and the broader context of the health system setting.

Clearing the Global Health Fog

5

Table 1. Critical Health Systems Functions and Elements of Integration Critical Health System Function

Elements

Stewardship and governance

– Accountability function – Reporting – Performance management – Pooling of funds – Provider payment methods – Needs assessment – Priority setting – Resource allocation – Structural – Human resources – Infrastructure – Operational integration – Referral and counter-referral systems – Guidelines or care pathways – Procurement

Financing Planning

Service Delivery

Monitoring and Evaluation Demand Generation

– Supply chain management – Information management and technology – Data collection and analysis – Incentives—financial (e.g. conditional cash transfers, health or social insurance) and nonfinancial (peer recognition, non-monetary awards) – Population interventions—e.g. education and promotion

The conceptual framework provides a basis for evaluating these five constituents with respect to the purpose, extent and nature of integration of the health intervention(s) under study into critical health system functions.

We consider integration of elements of a health intervention or program with critical health system functions, to include, inter alia, stewardship and governance, financing, planning, service delivery, M&E, and demand generation (Table 1). With respect to extent, we identify whether the integration is full, partial or non-existent, and by level we refer to integration of these functions at local (provider unit), district, regional or national tiers of the health system.

Given the broad mix of outcome measures used in different studies it was difficult to directly compare the relative success of programs and interventions. For this analysis, we attempted to ascertain the “success” for each program by analyzing the reported achievements against the outcome measures and indicators predefined at the start of the study. We assigned a “high success” rating to a program if all or most of the explicitly stated objectives were achieved in the study, even if the number of reported outcome measures was limited.

CHAPTER 3

Results

Outcomes Reported in the Studies Analyzed In this section we present all interventions grouped by the disease area or outcomes they seek to address. Data underpinning the assertions and narrative in this section are presented in Appendix B, which includes a summary of the studies evaluated.

Neglected Tropical Diseases The control of neglected diseases in developing countries has been predominantly carried out through specific interventions that have targeted one or more of these diseases with limited integration into mainstream health system functions.

Dengue. To build a sustainable national program for dengue control, areas of Cuba integrated community working groups (CWG), which included Primary Health Care (PHC) workers, into the existing program that had limited integration into the mainstream health system (Toledo Romani and others 2007). Following integration, areas that had CWG involvement reported greater improvements in entomological indices and were able to maintain these effects with community level support for ongoing control though no data was provided on incidence.

Malaria. Certain areas in Colombia strengthened community participation and networks to improve malaria control (Rojas, Botero, and Garcia 2001). The program, which was managed by a Central Coordinating Committee and which was centered on education, provision of diagnosis and treatment, as well as vector control at the local level, succeeded in decreasing malaria incidence by 45 percent. However, there were no control sites 7

8

World Bank Working Paper

for comparison or figures on changes in malaria incidence in the periods that preceded the introduction of the program.

Schistosomiasis. Various strategies have been implemented to control schistosomiasis. For example Brazil, Burundi, Cameroon, and Saudi Arabia have integrated some targeted interventions into primary care structures or community centers (Coura Filho and others 1992; Engels, Ndoricimpa, and Gryseels 1993; Engels, Sindayigaya, Gryseels 1995; Bausch and Cline 1995; Cline and Hewlett 1996; Ageel and Amin 1997; Jarallah and others 1993). In contrast, Uganda and China have adopted essentially vertical structures (Kabatereine and others 2006; Sleigh and others 1998a, 1998b, 1998c).

In the case of Brazil (Coura Filho and others 1992), integrated strategies helped to successfully reduce the incidence and prevalence of schistosomiasis infections at rates similar to those reported in targeted interventions without health system integration.

In Saudi Arabia (Ageel and Amin 1997; Jarallah and others 1993) integration of targeted interventions into PHC led to a greater decline in the incidence, prevalence and intensity of infections than that achieved by targeted interventions with no integration.

In Cameroon population knowledge about schistosomiasis and the utilization of health centers increased after introduction of a program that integrated targeted interventions into PHC (Bausch and Cline 1995; Cline and Hewlett 1996).

In Burundi, an integrated model of control resulted in an annual yield of 60 percent of cases detected by the PHC services. The participation rate in the integrated model (60 percent) was lower than in the targeted interventions (80 percent), with no integration aimed at selective population treatment. Under the integrated approach, the proportion of cases treated declined from 16 to 10 percent. However, these outcomes were achieved at an eighth of the cost of targeted interventions with no integration (Engels, Ndoricimpa, and Gryseels 1993).

In China, programs that were predominantly non-integrated in nature but which enjoyed strong community participation were successful in eradicating schistosomiasis in all endemic areas at an annual cost of less than US$ 0.50 per capita for the protected population (Sleigh and others 1998a, 1998b, 1998c).

In Uganda (Kabatereine and others 2006) a large pilot project demonstrated the viability of using schools as channels for drug distribution and the feasibility of using community-directed treatment for mass drug distribution to reach poor communities in remote areas.

Onchocerciasis. A study in Uganda demonstrated that provision of additional PHC tasks by community health workers, such as malaria or tuberculosis (TB) control, family planning, or immunization, was positively correlated with an increase in the quality of treatment for onchocerciasis (Katabarwa and others 2005).

Leprosy. In many countries, leprosy services have historically been delivered as targeted interventions with no integration. An independent evaluation (data not presented) is reported to have found that in Tamil Nadu, India, following integration, the number of detected cases increased slightly, but also that, as follow-up and treatment completion were no longer monitored, completion rates declined (Rao and others 2002).

Clearing the Global Health Fog

9

A pilot project in Andhra Pradesh, India showed negligible differences in leprosy prevalence and newly detected cases before and after integration. After integration, there was an increase in detection of previously hidden cases, but the rate of case-holding (adherence to treatment in accordance with program guidelines) declined (Parkash and Rao 2003).

Whereas, a comparison of leprosy programs in the area of Jamkhed (integrated) and in the neighboring district of Osmanabad (targeted without integration) in Maharashtra, India showed that, in communities with an integrated approach to leprosy care, social stigma experienced by leprosy patients (reduction of which was a program objective) was greatly diminished (Arole and others. 2002).

In Sri Lanka, integration of targeted interventions for leprosy into PHC was associated with higher case detection reflected by a 36 percent increase in prevalence and 41 percent increase in new cases (Kasturiaratchi, Settinayake, and Grewal 2002).

Nutrition The nutrition programs studied included general programs to address undernutrition and programs providing micronutrient supplementation.

The Bangladesh Integrated Nutrition Project (BINP) was designed to provide education, basic care and nutritional services to children and mothers. An evaluation concluded that the BINP had failed to achieve its objectives (reducing prevalence of severe undernutrition by 40 percent and moderate undernutrition by 25 percent). Although the mothers in the intervention area reported better care practices than those in the control area, there was no substantial difference in the rate of malnutrition between the two groups of children (Hossain, Duffield, and Taylor 2005).

Multi-country studies on micronutrient supplementation (including vitamin A, iodine and iron) demonstrate that in many countries these projects were successfully launched, implemented and sustained, but that campaign-based interventions (e.g. vitamin A supplementation during immunization days) were more successful in achieving their objectives than facility-based models (e.g. iron distribution at antenatal care clinics) (Deitchler and others 2004).

In Peru, compared to a non-integrated food-distribution program, a weekly multimicronutrient supplementation program with an integral communication campaign and community involvement in education and distribution targeting children, women and adolescent girls of childbearing age, improved coverage and led to an increase in the knowledge on the beneficial effects of supplementation (Gross, Mamani Diaz, and Valle 2006a; Gross, Valle, Mamani Diaz 2006b; Lechtig and others 2006a, 2006b; Lopez de Romana and others 2006). In addition, it had a protective effect on the haemoglobin levels of the three target groups, and was more cost-effective and better targeted to the population needs.

Immunization Evaluation of mass immunization campaigns (targeted interventions with no integration) in three different contexts revealed contrasting results. In rural India, the Pulse Polio Immunization (PPI) campaign, which was launched in 1995, aimed at promoting social mobilization and immunizing the population in remote areas of the country during National Immunization Days (NID). This program led to a significant increase in immunization

10

World Bank Working Paper

coverage for the first dose of polio vaccine. However, the coverage declined between the first and the third dose, and there was no change in inequities related to gender, caste, wealth status, religion, and geography for polio immunization or for other non-polio Expanded Program for Immunization (EPI) immunizations before and after the PPI campaigns (Bonu, Rani, and Baker 2003).

In Sierra Leone, mass immunization campaigns led to high coverage rates for those vaccines requiring only one dose (90 percent BCG, 62 percent measles), with lower coverage rates for the vaccinations requiring three doses (59 percent DPT, 58 percent polio, 76 percent tetanus). With each full dose of vaccines there was a significant decline in infant and child mortality across all the socio-economic groups (Amin 1996).

Two mass immunization campaigns in seven refugee camps in the Former Yugoslav Republic of Macedonia during the military conflict in 1998, achieved coverage rates of around 90 percent of children, but no other outcomes were reported (Koop, Jackson, and Nestel 2001).

Child Health and Development Integrated Management of Childhood Illness. Studies that compared the effectiveness of the Integrated Management of Childhood Illness (IMCI) strategy with routine care comprising a series of non-integrated programs, have reported positive impacts on the quality of care delivered in a range of settings.

In Bangladesh, the quality of care (i.e. adherence to IMCI guidelines), care-seeking behavior and utilization of government health facilities significantly improved in facilities providing IMCI when compared with 20 paired facilities that had not implemented IMCI guidelines (El Arifeen and others 2004).

In South Africa, introduction of IMCI in selected districts led to improved process and health outcomes (as measured by a quality index based on the quality indicators developed by WHO) (Chopra and others 2005), while in Tanzania it led to significantly improved case management and prescribing and a reduction of under-5 mortality rates as compared with those in districts where IMCI was not introduced (Armstrong Schellenberg and others 2004; Masanja and others 2005)—as well as improvements in equity, immunization coverage, caretaker knowledge and behavior, nutritional status and morbidity (Masanja and others 2005). Additional studies, which examined the total and incremental start up and implementation costs of IMCI (Adam and others 2005) and the costs borne by various health system levels in relation to a composite measure for correct management of childhood illness (Bryce and others 2005a), demonstrated that implementation of IMCI was not associated with higher costs than routine care, but led to significant improvement in case management compared to controls that used routine training approaches.

In Uganda introduction of IMCI led to improvements in the performance of trained health workers and the quality of care delivered to children aged less than five years (Pariyo and others 2005), while in China, implementation of IMCI helped to strengthen the local health system and improved communications between physicians and patients (Zhang, Dai, and Zhang 2007). Similarly, in Brazil (Amaral and others 2004) and Morroco (Naimoli and others 2006), introduction of IMCI led to significant improvements in quality of care (e.g. weight checked against growth chart; checking for at least three danger signs; and evaluation of feeding practices in Brazil and adherence to guidelines and correct prescription

Clearing the Global Health Fog

11

of antibiotics in Morroco ) as compared with routine care provided in facilities where IMCI was not introduced. By contrast, in Peru, implementation of IMCI did not positively affect the quality of care and utilization of services in facilities where it was implemented (Huicho and others 2005a, 2005b).

Integrated Child Development Services. The Indian Integrated Child Development Services (ICDS) program (which comprises a package of education services, nutritional support for pregnant and nursing women and children, and links to other primary care based health services for children) has been evaluated in a number of studies. Districts that adopted the ICDS scheme demonstrated improvements in the quality and quantity of services offered, improved coverage of immunization and a marked reduction in malnutrition in children aged under six years (Lal 1980), as well as higher uptake of healthcare services such as immunization and antenatal care, and significant improvement in feeding practices and nutritional and health status of children (Gupta and others 1984).

Addition of therapeutic food supplementation and nutritional support to ICDS led to declines in malnutrition levels in severely malnourished children (Kapil and others 1999), improvements in weight gain for pregnant women and reductions in the number of pre-term and low birth weight deliveries in ICDS districts compared to non-intervention sites (Agarwal and others 2000). ICDS services, when delivered in a coordinated manner with EPI, led to significant increases in overall immunization coverage with greater improvement in knowledge, attitudes and practices of carers with regard to immunization in intervention districts compared with districts where ICDS and EPI were delivered separately (Tandon and others 1992; Tandon and Sahai 1988). One study demonstrated a strong correlation between the degree of utilization of all ICDS services and nutritional status and morbidity of children between 0–36 months of age (Saiyed and Seshadri 2000). In contrast, one study found no statistically significant impact of ICDS on malnutrition levels in ICDS interventions sites but observed higher rates of coverage for DPT and measles immunization in non-ICDS districts (Trivedi, Chapararwal, and Thora 1995).

Reaching intended target groups for ICDS has proved difficult. In response, the Dular program developed by UNICEF-India has built on the existing ICDS infrastructure with an enhanced package of education, training of health workers and active outreach to the community. This program was successful in enhancing the effectiveness of ICDS and led to statistically significant improvements in all measured outcomes (for example, iodized salt use, prenatal care, colostrum feeding, and use of a delivery kit for childbirth) between Dular and non-Dular villages (Dubowitz and others 2007).

Family Planning Services Matlab Family Planning (FP) and Health Services Project. The “Matlab FP and Health Services Project� (FPHSP), initiated by the International Center for Diarrhoeal Disease Research, Bangladesh (ICDDR,B), which combined FP services with a basic package of Maternal and Child Healthcare (MCH) was more cost-effective in stimulating uptake of modern contraceptive services than the pre-existing governmental program which provided these services in a non-integrated manner (Simmons, Balk, and Faiz 1991). However, extension of the original program with provision of a wider range of MCH services

12

World Bank Working Paper

did not increase uptake of contraceptive services when compared with the original program (DeGraff and others 1986; Phillips and others 1984).

Lady Health Worker Program. The Pakistan National Program of Family Planning and Primary Health Care, more commonly known as the Lady Health Worker Program (LHWP), delivers a bundle of integrated services related to MCH and FP door-to-door in rural areas. It has achieved significantly higher use of reversible, modern contraceptive methods in areas served compared to routine practice where these services are provided in primary care centers (Douthwaite and Ward 2005).

Combined Family Planning and MCH Programs with Basic Health Services. A study in Nepal compared the effectiveness of two FP/MCH programs; one providing only FP/MCH services and the other (the ‘integrated’ program) combining the range of FP/MCH services with other basic health services that included immunizations and case finding and treatment for TB and leprosy. While the non-integrated FP/MCH services achieved greater impact on knowledge of FP, intention to use FP services and infant mortality, the integrated program had a relatively greater impact on child survival (Tuladhar and Stoeckel 1982).

HIV/AIDS The GHESKIO clinic in Haiti was set up to provide Voluntary Testing and Counselling (VCT) and care for patients with HIV/AIDS (Peck and others 2003). It has gradually added other primary care services for communicable diseases and reproductive health. The integration of on-site primary care services was accompanied by an increased demand for VCT. Patients attending the clinic could receive coordinated care for co-morbidities.

The Extent and Nature of Integration of Health Interventions into Critical Health Systems Functions We present below an analysis that describes, for each intervention included in the study, how the intervention elements are integrated into the six critical health system functions.

Further, using published data from the studies, we map the nature and extent of integration of intervention elements into critical health system functions. These are further discussed below and summarized in Figure 2.

Our analysis draws on the data available in the published papers. The mapping of the extent of integration reflects the actual situation as reported in the published paper in the period under concern rather than what was intended.

Stewardship and Governance This dimension captures aspects related to distribution of governance responsibilities, accountability and performance management for an intervention. We considered full integration to have occurred when the governance arrangements for the targeted health intervention were the same as those for the general health services or the local/national administrative structures. For example, in a number of settings interventions for schistosomiasis control

Clearing the Global Health Fog

Figure 2. The Extent and Nature of Integration by Targeted Health Intervention and Intervention Success as Reported in the Study

Most to all outcomes

Partially integrated

Mixed outcomes

Not integrated

Few to no outcomes

Unknown

Unknown

Dengue Cuba (Toledo Romani 2007) Malaria Colombia (Rojas 2001) Schistosomiasis control Brazil (Filho 1992) Burundi (Engels 1993, 1995) Cameroon (Clino 1996) China (Sleigh 1998) Saudi Arabia (Ageel 1997, Jarallah 1993) Uganda (Kabatoroine 2006) Leprosy India (Rao 2002, Thakar 2003) Sri Lanka (Kasturiaratchi 2002) Nutrition Peru (Gross2006ab, Lachtig2006ab, Lopez de Romana 2006) Bangladesh (Hossain 2005) Various (Doitchler 2004) Immunisation Sierra Loone (Amin 1996) India (Bonu 2003) Macedonia (Koop 2001) Child health & development IMCI* ICDS (Agarwal 2000, Gupta 1984, Kapil 1999, Lal 1980, Trivedi 1995, Saiyed 2000) ICDS + Dolar (Dubowitz 2007) Family planning services Bangladesh - FPHSP (Philips 1984, de Graff 1986) Pakistan - LHWP (Douthwaite 2005) Nepal (Tuladhar 1982) HIV/AIDS & STD services Haiti (Pack 2003)

St

Fully integrated

ew ar d na sh nc ip Pl in & G an g ov n er Se ing na rv nc ic e e M d on e ito liv e D em ring ry an & E d ge val ne ua Su ra tio cc tio n es n s

Degree of Success

Fi

Level of Integration

13

14

World Bank Working Paper

adopted a fully integrated governance structure—such as that in Brazil, where the intervention was managed by the municipality (Coura Filho and others 1992); in Cameroon, where the PHC service was accountable for schistosomiasis control (Cline and Hewlett 1996) and Saudi Arabia, where the PHC team cooperated and shared responsibility with the district supervisor and regional health authorities (Ageel and Amin 1997; Jarallah and others 1993). In India the governance structures for interventions aimed at leprosy control were successfully integrated with the PHC services (Rao and others 2002), while in Nepal governance structures for the FP/MCH intervention were integrated with services provided in the district health offices (Tuladhar and Stoeckel 1982).

Partial integration occurs where responsibility is shared by the existing general healthcare system and a specific structure created purposely for the intervention. For instance, in Cuba composite entities comprising PHC level and vector control team workers jointly managed the interventions for dengue control (Toledo Romani and others 2007), whereas in Colombia a Central Coordinating Committee consisting of representatives of the regional health services, NGOs, scientific centers, and local politicians was formed to manage and oversee the interventions for malaria control (Rojas, Botero, and Garcia 2001).

In Sri Lanka, the responsibility for managing interventions for leprosy control moved between the local health authorities and the Central Leprosy Clinic personnel (Kasturiaratchi, Settinayake, and Grewal 2002). In India anganwadi centers, created as part of a self-managed system for child health and development, were also responsible for organization of a wide range of primary care services (Agarwal and others 2000; Gupta and others 1984; Kapil and others 1999; Lal 1980; Saiyed and Seshadri 2000; Tandon and others 1992; Tandon and Sahai 1988; Trivedi, Chaparwal, and Thora 1995).

This review considers the governance role to be non-integrated when accountability remains exclusively with dedicated specialist entities that are charged with implementation and management of the targeted health interventions, without involvement of the general healthcare system—for example, interventions directly managed by dedicated units within national or regional governments without integration into main health system functions as with the interventions for schistosomiasis control in China (Sleigh and others 1998a, 1998b, 1998c) and Uganda (Kabatereine and others 2006), nutrition-related campaigns in Peru (Gross, Mamani Diaz, and Valle 2006a; Gross, Valle, and Mamani Diaz 2006b; Lechtig and others 2006a, 2006b; Lopez de Romana and others 2006) and various Asian countries (Deitchler and others 2004; Hossain, Duffield, and Taylor 2005); the LHWP in Pakistan (Douthwaite and Ward 2005); and the FPHSP in Bangladesh (DeGraff and others 1986; Phillips and others 1984).

While non-governmental organizations (NGOs) and external donors work with regional or national health services, they often retain direct governance roles beyond financing or service delivery, as with the stand-alone HIV/AIDS clinic in Haiti offering voluntary counseling and testing with additional services for HIV/AIDS STI and reproductive health services (Peck and others 2003); the nutrition interventions in Peru (Gross, Mamani Diaz, and Valle 2006a; Gross, Valle, and Mamani Diaz 2006b; Lechtig and others 2006a, 2006b; Lopez de Romana and others 2006) (PISA); the mass immunization campaigns in Sierra Leone implemented by UNICEF (Amin 1996) and in Macedonia implemented by International Medical Corps (Koop, Jackson, and Nestel 2001); or the Dular program in India managed by UNICEF (Dubowitz and others 2007).

Clearing the Global Health Fog

15

Financing For the purposes of this review, financing refers to both the pooling of financial resources and the provider-payment methods used to allocate these. Revenue generation, a critical function of financing, is beyond the scope of the review.

We considered an intervention to be fully integrated into the financing function if it was funded entirely through the national or regional general healthcare budget. Examples of full integration of financing include interventions for schistosomiasis control in Brazil (Coura Filho and others 1992) and interventions for leprosy control in India (Rao and others 2002) and Sri Lanka (Kasturiaratchi, Settinayake, and Grewal 2002).

Partial integration into the financing function was achieved with interventions for schistosomiasis control in Cameroon, where earmarked funding was provided by the United States Agency for International Development (USAID) but channeled through the PHC system (Cline and Hewlett 1996). Under the ICDS scheme in India resources were provided by the national government directly to the intervention which comprised a range of essential health services to supplement other general services locally provided and funded by local and national governments (Agarwal and others 2000; Gupta and others 1984; Kapil and others 1999; Lal 1980; Saiyed and Seshadri 2000; Tandon and others 1992; Tandon and Sahai 1988; Trivedi, Chaparwal, and Thora 1995).

When financing was provided directly to an intervention and addressed only a particular disease or problem, the function was considered to be non-integrated. In some instances interventions were directly funded by the government, either at local level, for example dengue control in Cuba (Toledo Romani and others 2007), or at national level, for example malaria control in Colombia (Rojas, Botero, and Garcia 2001) and the LWHP in Pakistan (Douthwaite and Ward 2005).

Many interventions were directly funded by external donors: such as the BINP funded by the International Development Association of the World Bank (Hossain, Duffield, and Taylor 2005); the Ugandan National Schistosomiasis Control Program funded from the Bill & Melinda Gates Foundation through its collaboration with the international Schistosomiasis Control Initiative (Kabatereine and others 2006); the Dular ICDS financed

by UNICEF-India; the micronutrient supplementation programs in many Asian countries supported by the Canadian International Development Agency and UNICEF (Deitchler and others 2004), and in Peru by the NGO PROMESA (Gross, Mamani Diaz, and Valle 2006a; Gross, Valle, and Mamani Diaz 2006b; Lechtig and others 2006a, 2006b; Lopez de Romana and others 2006); and the immunization program in Sierra Leone funded by a range of external donors (Amin 1996).

Planning In our framework, the planning function includes activities, processes and systems for needs assessment, priority setting, and resource allocation. We consider planning for an intervention to be fully integrated if the decision-making in relation to the above three areas is undertaken by institutions/stakeholders who are involved in the same tasks for the general health system. Examples of full integration include community leadership in schistosomiasis control in Cameroon (Cline and Hewlett 1996) and in India integration of decision making for leprosy control with the PHC system (Rao and others 2002).

16

World Bank Working Paper

Partial integration occurs when the decision-making responsibility for planning is retained by those managing the health intervention but involve a range of stakeholders (such as civil society representatives, PHC level, or local/regional/national government) through inclusive groupings as illustrated by the Community Working Groups for dengue control in Cuba (Toledo Romani and others 2007), the Central Coordinating Committee for malaria control in Colombia (Rojas, Botero, and Garcia 2001) and interventions for leprosy control in Sri Lanka where the responsibility for planning rested with the PHC services but relied heavily on the workers from the program (Kasturiaratchi, Settinayake, and Grewal 2002).

Planning was considered to be non-integrated when the decision-making focused solely on the intervention without consideration of general healthcare activities. This may include specific national government units at national level, as in the schistosomiasis control project in China (Sleigh and others 1998a, 1998b, 1998c) or the LHWP in Pakistan (Douthwaite and Ward 2005); dedicated units at national and local levels, as with the micronutrient supplementation programs in several Asian countries (Deitchler and others 2004); or NGOs and external donors—for example, PISA in the micronutrient supplementation interventions in Peru, UNICEF in the Dular program in India (Dubowitz and others 2007) and the International Medical Corps in Macedonia (Koop, Jackson, and Nestel 2001).

Service Delivery Service delivery relates to structural and organizational dimensions of a particular intervention, either at or close to the interface with the customer. In our analysis services within a health intervention are considered to be fully integrated if their provision is the responsibility of general or multi-purpose health workers, as with the schistosomiasis control in Burundi and Cameroon (Cline and Hewlett 1996; Engels, Ndoricimpa, and Gryseels 1993), leprosy care in Sri Lanka (Kasturiaratchi, Settinayake, and Grewal 2002) and ICDS services in India (Agarwal and others 2000; Gupta and others 1984; Kapil and others 1999; Lal 1980; Saiyed and Seshadri 2000; Tandon and others 1992; Tandon and Sahai 1988; Trivedi, Chaparwal, and Thora 1995).

Partial integration refers to instances where there is shared responsibility for the provision of services between general health workers and the staff of the targeted health intervention; as with interventions for dengue control in Cuba (Toledo Romani and others 2007), schistosomiasis control in Brazil (Coura Filho and others 1992) and Saudi Arabia (Ageel and Amin 1997; Jarallah and others 1993), and the interventions for leprosy control in India which after integration with PHC services were provided by the former staff of leprosy centers as the general purpose PHC workers were insufficiently trained to take over this responsibility (Rao and others 2002). Partial integration was also achieved in the ICDS intervention in India through collaboration of anganwadi workers and purposetrained volunteers (Agarwal and others 2000; Kapil and others 1999; Tandon and others 1992; Tandon and Sahai 1988). Partial integration also occurred when service delivery for a number of interventions were linked; for example family planning and maternal and child health services (DeGraff and others 1986; Phillips and others 1984; Tuladhar and Stoeckel 1982) or integration of general health services with HIV/AIDS VCT services in Haiti (Peck and others 2003). In Uganda, National Immunization Days were used as a

Clearing the Global Health Fog

17

vehicle to deliver interventions for schistosomiasis control (Kabatereine and others 2006).

The IMCI strategy is a good example of partial integration where interventions for management of a number of childhood illnesses are bundled together (Bryce and others 2005b), but can be administered either separately through community workers trained exclusively in IMCI, or through PHC facilities using general staff.

A number of interventions rely solely on single purpose workers and have no integration with other interventions or general health services; such as the interventions for malaria control in Colombia (Rojas, Botero, and Garcia 2001), interventions for schistosomiasis control in China (Sleigh and others 1998a, 1998b, 1998c), or the immunization for polio in India (Bonu, Rani, and Baker 2003) and other childhood illnesses in Macedonia (Koop, Jackson, and Nestel 2001).

Monitoring & Evaluation The M&E function of a targeted health intervention was considered to be fully integrated if the responsibility for this rested with institutions that retained overall responsibility for M&E in the health system—for example, with interventions for schistosomiasis control in Cameroon and Saudi Arabia, monitored respectively by the Ministry of Public Health and the district supervisor of the local PHC services (Ageel and Amin 1997; Cline and Hewlett 1996; Jarallah and others 1993), or the interventions for leprosy control in Sri Lanka monitored by the regional PHC services (Kasturiaratchi, Settinayake, and Grewal 2002). In contrast, M&E of interventions for malaria control in Colombia was partially integrated as it was undertaken jointly by staff from the regional health services and the malaria control program (Rojas, Botero, and Garcia 2001).

For many health interventions M&E is not integrated with mainstream M&E systems.

The interventions for schistosomiasis control in China used a dedicated parallel system; the international sponsor for schistosomiasis control in Uganda employed its own systems for M&E (Sleigh and others 1998a, 1998b, 1998c); the nutrition interventions in Peru managed by an international organization, PISA undertook its own M&E with the assistance of independent institutions (Gross, Mamani Diaz, and Valle 2006a; Gross, Valle, and Mamani Diaz 2006b; Lechtig and others 2006a, 2006b; Lopez de Romana and others 2006); and the immunization program in Macedonia was monitored directly by volunteers of the implementing organization (Koop, Jackson, and Nestel 2001).

Demand Generation Demand generation is increasingly recognized as an important health system function. It relies on a number of interventions such as the use of appropriate financial incentives and monetary support such as insurance, or of information, education and communication (IEC) activities designed to change behavior.

Demand generation was considered to be fully integrated if mechanisms used to create financial incentives or IEC activities were provided jointly with the general services or were delivered by PHC workers. For example, in Burundi, Cameroon and Saudi Arabia, IEC for schistosomiasis control was the responsibility of staff of the public health centers (Ageel and Amin 1997; Cline and Hewlett 1996; Engels, Ndoricimpa, and Gryseels 1993; Jarallah and others 1993) and in Sri Lanka IEC activities were considered a fundamental

18

World Bank Working Paper

part of the integration strategy (Kasturiaratchi, Settinayake, and Grewal 2002). A particularly interesting example of integration of demand generation was in Haiti, where VCT activities for HIV/AIDS at a clinic were used to generate demand for related services for STIs, reproductive health and basic PHC (Peck and others 2003).

In Colombia, education on the prevention, diagnosis and treatment of malaria was provided both by community volunteers working in malaria control and staff from the general health services (Rojas, Botero, and Garcia 2001). In Peru and Bangladesh health education for nutrition interventions was jointly provided by staff working for the targeted intervention and regional health workers (Gross, Mamani Diaz, and Valle 2006a; Gross, Valle, and Mamani Diaz 2006b; Hossain, Duffield, and Taylor 2005; Lechtig and others 2006a, 2006b; Lopez de Romana and others 2006). In China, information and education related to schistosomiasis control was provided through a partnership between staff working on schistosomiasis control, general health workers, schoolteachers and community health workers (Sleigh and others 1998a, 1998b, 1998c). Similarly, staff involved in the LHWP in Pakistan and the ICDS in India provided a wide range of IEC activities on general health, women’s health, family planning, maternal care and child health in addition to those related to their specific interventions (Agarwal and others 2000; Douthwaite and Ward 2005; Gupta and others 1984; Kapil and others 1999; Lal 1980; Saiyed and Seshadri 2000; Tandon and others 1992; Tandon and Sahai 1988; Trivedi, Chaparwal, and Thora 1995). In contrast, in a number of countries, information campaigns related to health interventions tended to be stand-alone activities, focusing solely on a single problem or disease, and delivered by single-purpose health workers or volunteers: for example for dengue control in Cuba (Rojas, Botero, and Garcia 2001), schistosomiasis control in the Ugandan National Control Program (Kabatereine and others 2006) and for childhood immunization in Macedonian refugee camps (Koop, Jackson, and Nestel 2001).

The Extent of Intervention Success in the Studies Analyzed We have summarized in Figure 2 the extent of outcome success for each study included in the review, and relate this to the nature and level of integration.

We assigned a “high success” rating to a program if all or most of the explicitly stated objectives were achieved in the study, even if the number of reported outcome measures was limited.

A number of programs was given a ‘high success’ rating, namely the Cuban dengue control program (Toledo Romani and others 2007); the malaria control program in Colombia (Rojas, Botero, and Garcia 2001), the schistosomiasis programs in Brazil (Coura Filho and others 1992), Cameroon (Cline and Hewlett 1996), China (Sleigh and others 1998a, 1998b, 1998c), Saudi Arabia (Ageel and Amin 1997; Jarallah and others 1993) and Uganda (Kabatereine and others 2006); the immunization campaigns in Sierra Leone (Amin 1996) and Macedonia (Koop, Jackson, and Nestel 2001); the Indian ICDS and Dular programs (Agarwal and others 2000; Dubowitz and others 2007; Gupta and others 1984; Kapil and others 1999; Lal 1980; Saiyed and Seshadri 2000); the LHWP in Pakistan (Douthwaite and Ward 2005) and the HIV/AIDS VCT clinic in Haiti (Peck and others 2003).

We assigned a “moderate success” rating for programs that reported success against a number of outcome measures identified at the outset of the study but failure against others.

Clearing the Global Health Fog

19

Examples of these are the schistosomiasis program in Burundi (Engels, Ndoricimpa, and Gryseels 1993), the integrated leprosy services in Sri Lanka (Kasturiaratchi, Settinayake, and Grewal 2002) and India (Rao and others 2002), the immunization program in India (Bonu, Rani, and Baker 2003) and the Peruvian nutrition program (Gross, Mamani Diaz, and Valle 2006a; Gross, Valle, and Mamani Diaz 2006b; Lechtig and others 2006a, 2006b; Lopez de Romana and others 2006). The nutrition programs in a number of Asian countries and South Africa were not individually reported and were collectively assigned a “moderate success” rating to reflect varying levels of success achieved in different settings (Deitchler and others 2004).

Programs that failed to achieve most or all of the intended outcomes were assigned a “low success” rating. These included the Bangladesh Integrated Nutrition Project (Hossain, Duffield, and Taylor 2005), the extended form of the—otherwise successful—FPHSP in Bangladesh (DeGraff and others 1986; Phillips and others 1984) and the integrated strategy for FP/MCH services in Nepal (Tuladhar and Stoeckel 1982).

How the Context Influences the Extent and Nature of Integration To better understand which factors have contributed to success or failure of health interventions in a number of settings, we analyzed the context in which these interventions were implemented, in particular factors pertaining to the adoption system, the health system and those relating to the wider politico-economic and socio-cultural context. Collectively, these factors interact to either produce opportunities, or create barriers, or generate receptive contexts thereby influencing the desirability and sustainability of an intervention.

These are summarized in Appendix C.

Sustainability A changing socio-demographic and economic context impacts on population needs and affects sustainability of programs. For example, in Peru rapid urbanization and arrival of large numbers of indigent slum-dwellers significantly altered the morbidity profile of the local population with a substantial rise in micronutrient deficiencies. This created a significant new burden that had to be addressed by targeted interventions (Gross, Mamani Diaz, and Valle 2006a; Gross, Valle, and Mamani Diaz 2006b; Lechtig and others 2006a, 2006b; Lopez de Romana and others 2006). In contrast, a rise in living standards in Saudi Arabia which led to improvements in general hygiene and sanitation helped reduce the burden of schistosomiasis (Ageel and Amin 1997; Jarallah and others 1993).

Understanding local socio-cultural context is critical to the design of an appropriate intervention as illustrated by the schistosomiasis control in Uganda. This intervention initially targeted school children, but, as up to 30 percent or more of children were not enrolled in schools in certain endemic areas, the intervention failed to reach large population groups (Kabatereine and others 2006). The intervention was then modified to use a wider range of opportunities, such as National Immunization Days, for deworming activities.

While the disease-specific focus of interventions for the control and eradication of leprosy (India, Sri Lanka) or schistosomiasis (such as in Burundi and Saudi Arabia) was

20

World Bank Working Paper

appropriate when these diseases were highly endemic, success of these interventions in reducing the disease burden meant that strategies which were previously cost-effective became financially unsustainable in the new epidemiological context. Given the low endemicity and prevalence, the more financially sustainable option was to integrate these interventions into mainstream PHC services.

In Brazil, integration of the intervention for schistosomiasis control into PHC was possible as municipalities possessed enough absorptive capacity to assume responsibility for planning, financing and delivery of the intervention when external funding ceased (Coura Filho and others 1992). Similarly, the presence of a well-developed PHC infrastructure with high population coverage and utilization rates enabled Saudi Arabia to embed schistosomiasis control into PHC (Ageel and Amin 1997; Jarallah and others 1993), and allowed integration of IMCI into the general health services in Peru (Huicho and others 2005a; Huicho and others 2005b), South Africa (Chopra and others 2005) and Tanzania (Armstrong Schellenberg and others 2004; Masanja and others 2005).

The FPHSP intervention in Bangladesh was able to provide efficient family planning services for many years thanks to its dedicated human and physical resources (Simmons, Balk, and Faiz 1991). On the other hand, the LHWP program in Pakistan struggled to find a place in a fragile health system which was already overstretched and experienced significant funding shortfalls (Douthwaite and Ward 2005). In contrast, the Dular strategy for health education and child health services was successfully sustained as it managed to build on the infrastructure of the ICDS at low cost (Dubowitz and others 2007).

Opportunity and Necessity Critical events create windows of opportunity or a necessity for action, by mobilizing civil society and other key actors (such as health professionals and policy makers) to introduce new systems and designs for finance and delivery of health interventions. For example, an outbreak of dengue in a non-endemic area of Cuba motivated the local community to strengthen existing interventions for dengue control (Toledo Romani and others 2007).

In Cameroon, a number of development projects raised concerns about the possible expansion of snail habitats with a concomitant increase in schistosomiasis infection, prompting the government to establish a dedicated intervention for schistosomiasis control with strong community involvement and integration into PHC. This intervention was in line with the Government’s commitment to strengthen the national PHC system (Cline and Hewlett 1996).

Confronted with unsustainably high rates of population growth, the Government of Bangladesh introduced the FPHSP program to provide family planning education and contraceptive services (DeGraff and others 1986; Simmons, Balk, and Faiz 1991). Overcrowding in refugee camps that followed the military conflict in Macedonia increased the risk of rapid transmission of vaccine preventable diseases, necessitating urgent implementation of an EPI for children (Koop, Jackson, and Nestel 2001). The Peruvian nutrition supplementation program was set up in response to evidence which showed a high prevalence of iron-deficiency in the newly urbanized population (Gross, Mamani Diaz, and Valle 2006a; Gross, Valle, and Mamani Diaz 2006b; Lechtig and others 2006a, 2006b; Lopez de Romana and others 2006).

Clearing the Global Health Fog

21

Synergies between targeted health interventions can create opportunities for integration and can positively influence the ultimate success of these interventions. For example, positive synergies between FP and MCH services in Bangladesh (DeGraff and others 1986; Phillips and others 1984), HIV/AIDS VCT, STI and reproductive health services in Haiti (Peck and others 2003), and the ICDS and EPI in India helped to mutually reinforce effectiveness of each intervention with improved results. In contrast, poor coordination between related interventions can be detrimental to effectiveness and sustainability. For example, in Morocco multiple and contrasting guidelines for managing child health acted as a barrier to effective implementation of IMCI (Naimoli and others 2006). Similarly, in Peru the presence of a number of child health programs with overlapping remits led to inefficiencies in delivery and destructive competition for resources for IMCI (Huicho and others 2005a; Huicho and others 2005b). In contrast, successful introduction of IMCI in Tanzania coincided with other measures to improve the management of the district health systems, creating an enabling environment for the implementation of IMCI strategy (Armstrong Schellenberg and others 2004; Masanja and others 2005).

Desirability A desire to implement and sustain an intervention could emanate not just from the leadership but also from the health workers. Government commitment was critical to the success of the interventions aimed at schistosomiasis control in China (Sleigh and others 1998a, 1998b, 1998c) and successful integration of leprosy services with PHC in Sri Lanka (Kasturiaratchi, Settinayake, and Grewal 2002). In contrast, a lack of government commitment contributed directly to the poor IMCI results in Peru (Huicho and others 2005a, 2005b). In India, strong opposition from the staff involved in delivering the National Leprosy Program hindered for a considerable time integration of the program with PHC (Rao and others 2002). Similarly, in Cuba, reluctance of health workers to relinquish responsibilities delayed integration of interventions for malaria control into mainstream health services (Rojas, Botero, and Garcia 2001).

CHAPTER 4

Discussion

he debate on health systems and targeted interventions has tended to narrowly focus on vertical or integrated descriptors. However, our analysis shows this to be a false dichotomy as few interventions are purely vertical (single-disease oriented) or horizontal (fully integrated into mainstream health system functions). As this review shows, in practice the nature of the problem, the interventions to address these and the adoption and assimilation of health interventions in health systems vary greatly in different contexts, as does success. The purpose, nature, speed, and the extent of integration also vary—in part, dependent on the intervention complexity, the health system characteristics and the contextual factors.

There are few instances where there is full integration of a health intervention or where an intervention is completely non-integrated. Instead, there exists a highly heterogeneous picture both for the nature and also for the extent of integration.

The review suggests that the evidence base for integration versus specific health programming is very limited. As success was measured in different ways (for example, in terms of varied programmatic goals related to efficiency, effectiveness, or equity), analysis and drawing lessons from this is further complicated. Perhaps the modesty of evidence creates the context for strong opinions for or against integration in global health. We have attempted to show that in practice the dichotomy between integrated and non-integrated (traditionally described as vertical and horizontal) is not rigid. Health systems combine both non-integrated and integrated interventions, but the balance of these interventions varies considerably. Our findings suggest that the purpose, nature and extent of integration vary enormously between interventions and in countries, creating a rich mosaic of local solutions to address emergent problems. We found rare instances of full integration of health intervention into all critical health system functions, but a wide range of instances

T

23

24

World Bank Working Paper

where health interventions are integrated into one or more critical health system functions.

Further, it is important to note that in many countries that simultaneously have multiple health interventions, the nature of integration for these interventions varies and different degrees of integration for the same critical health system function co-exist.

This study has a number of limitations. Firstly, program evaluations reported in the peer-reviewed literature used in this study provide limited and varied detail on the organizational structures surrounding the intervention and the health systems within which these interventions were implemented. However, the framework presented (Atun and others 2009) has enabled us to consistently analyze the data on health system elements and contextual factors. Secondly, our assessment of the ‘success’ of interventions was based on that reported by the authors against a range of measures specified in the studies. This assessment could potentially overlook other important but unreported measures of success or failure. Thirdly, in our review, we selected predominantly “program evaluations”— not a well-defined study design. Unlike the Cochrane criteria for appraising quality of randomized controlled studies there is no consensus on the criteria that could be used to appraise the quality of program evaluations. Hence, we selected these studies on the basis of their relevance to our research question but their quality was not rated. Finally, the inherent heterogeneity of the included studies, both in their clinical focus and in the setting, makes it difficult to generalize findings without a better understanding of how contextual factors have shaped the organizational structures of health interventions. We intend to address this issue through a more comprehensive and in-depth approach, using a multicountry case study design.

CHAPTER 5

Conclusions

n spite of the study limitations that we tried to address, the findings provide new synthesis of evidence to further the debate on health systems and targeted interventions which has ossified in a binary mode. Given the highly varied contexts and adoption systems that reflect local nuances, different health system capacities and the range of problems being addressed, it is not surprising that in practice a rich mix of solutions exist.

While the discussion on the relative merits of integrating health interventions will no doubt continue, discussion should move away from the highly reductionist approach that has polarized this debate.

Future efforts are best spent on generating and learning from useful evidence. What is needed is a series of robust country case studies of health interventions that use replication logic with common methodology and shared methods, informed by appropriate theoretical frameworks (such as the one used in this study). The case studies would allow comparisons among countries and programs, generating evidence that has relevance beyond a country case.

I

25

Appendixes

27

APPENDIX A

Search Strategy and Methods

1. ((vertical[Tiab] OR horizontal[Tiab] OR integrat*[Tiab] OR coordinat*[Tiab] OR co-ordinat*[Tiab] OR link*[Tiab]) AND (program*[Tiab] OR care[Tiab] OR service*[Tiab]) OR “Delivery of Health Care, Integrated”[MeSH]) AND 2. ((“child health services”[MeSH] NOT “early intervention (education)”[MeSH]) OR “immunization programs” [MeSH] OR “family planning services”[MeSH] OR (“maternal health services”[MeSH] NOT “preconception care”[MeSH]) OR “maternal-child health centers”[MeSH] OR “community health centers”[MeSH] OR immunization[MeSH] OR “reproductive medicine”[MeSH] OR “adolescent health services”[MeSH] OR cholera[MeSH] OR dengue[MeSH] OR “Dengue Hemorrhagic Fever”[MeSH] OR Fascioliasis[MeSH] OR Trypanosomiasis[MeSH] OR Leishmaniasis[MeSH] OR Elephantiasis[MeSH] OR yaws[Mesh] OR “buruli ulcer”[MeSH] OR dracunculiasis[MeSH] OR leprosy[MeSH] OR schistosomiasis[MeSH] OR helminths[MeSH] OR “chagas disease”[MeSH] OR onchocerciasis[MeSH] OR “alcohol-related disorders”[MeSH] OR “cocaine-related disorders”[MeSH] OR “opioid-related disorders”[MeSH] OR “substance abuse, intravenous”[MeSH] OR “anxiety disorders”[MeSH] OR “depressive disorder”[MeSH] OR schizophrenia[MeSH] OR “eating disorders”[MeSH] OR dementia[MeSH] OR HIV/AIDS[MeSH] OR tuberculosis[MeSH] OR malaria[MeSH] OR “sexually transmitted diseases”[MeSH] OR “mental health”[MeSH] OR “Anemia, Iron-Deficiency”[MeSH] OR “Vitamin A Deficiency”[MeSH] OR “Food, Fortified”[MeSH] OR micronutrients[MeSH] OR zinc[MeSH]) Search results were limited using the PubMed filters “English” and “humans”.

We searched the following electronic databases: PubMed; the Cochrane Central Register of Controlled Trials (CENTRAL); and the Cochrane EPOC specialized register and Database of Abstracts of Reviews of Effectiveness. Supplementary searches were conducted through reference and citation tracking of the key articles retrieved during the search.

Review Methods The search strategy retrieved 8,274 potentially relevant articles from PubMed. This was followed by selection of articles deemed relevant for further analysis by two independent reviewers on the (continued)

29

30

World Bank Working Paper

basis of the titles retrieved. In order to establish a common set of inclusion criteria the first 100 titles were assessed jointly. The criteria for inclusion at the first stage were based on the potential relevance of each article to the research question for the review, independent of study design.

Once a common approach and understanding was developed each reviewer independently assessed one half of the remaining 8,274 titles (i.e. 4,137 each). In addition, to verify sufficient inter-rater agreement each reviewer evaluated 20 percent of titles assessed and rated by the other reviewer.

Each study deemed relevant by either of the reviewers was further considered for inclusion in the second stage of analysis. This yielded a total of 1,551 titles, which were retained for the second stage of selection. In this stage, the two reviewers independently assessed the abstracts of each of the 1,551 remaining studies for relevance to the review. This exercise showed the inter-rater agreement (Cohen’s κ coefficient) to be equal to 0.78. The studies selected by each of the reviewers were then compared. Where there was concordance to include or exclude a study they were retained or discarded. Where there was disagreement, the article in question was retained for full text analysis.

A total of 200 studies were retained for the third stage of the review at which the full-text was analyzed. The full-text of 172 of these identified papers was retrieved, whereas the remaining 28 studies could not be accessed. Studies were then evaluated based on study design as well as relevance.

The considered studies were primarily program evaluations, as well as studies that assessed the relative performance of different models of care after a change in organizational structure. For inclusion, the study had to present data on outcome measures such as health outcomes, quality of care, access to care and service utilization, patient satisfaction and cost or cost-effectiveness.

Papers that merely described the process of development of a care model, without discussion of any of the relevant process or outcome measures, were therefore excluded. Programs, furthermore, needed to have been implemented on a regional or national scale (i.e. beyond a trial setting) to be eligible for inclusion.

Data Extraction and Management Data extraction was done independently by the two reviewers, each using a common checklist purposefully developed for this review and based on the analytical framework outlined in detail below. Extracted data included: (i) General information: title, authors, and year of publication; (ii) Setting: clinical area, country, setting/care delivery system, and target groups; (iii) Intervention: description, duration, comparisons, and co-interventions; (iv) Study characteristics: study design, duration; (v) Participants: unit of analysis, number of participants in intervention and comparison groups; (vi) Outcomes: outcome measures, results of the intervention; (vii) Extent and nature of integration: stewardship & governance, financing, planning, service delivery, monitoring & evaluation, demand generation; (viii) Contextual factors: sustainability, opportunities, desirability

APPENDIX B

Summary of the Included Studies

31

Name

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

Cuba

2001–2004

CWG integrated into vertical program

Strengthened vertical program

General; 3 health zones within a district

General; 3 health zones within the same district

Maintenance of effects at 2 years after the end of the intervention; institutionalization; capacity building in the community

Correct water storage cover in 87.5% of the surveyed houses in intervention area and 21.5% in control area; main breeding sites: communal areas in the intervention area (75.8%) and domestic environment in the control area (86.4%)

Colombia

not clear

Education, diagnosis and treatment within communities, vector control

n/a

General; 23 communities in 2 areas

n/a

Incidence; length of sick leave

Incidence of malaria decreased by 45.4%; length of sick leave decreased from 7.5 to 3.7 days

Saudi Arabia

1990–1996

PHC provided education, detection and treatment.

Schoolchildren helped in diagnosis within the community

n/a

Population of high-risk villages, who attended the health centers

n/a

Coverage; prevalence of infection

Better coverage of the population (60%–90%) than the vertical program (6%–50%); overall prevalence of infection maintained at <1% for >5 years

Results

Dengue ToledoRomani 2007

32 Malaria Rojas 2001

Schistosomiasis Ageel 1997

Cameroon

n/a

Local health centers and elementary teachers performed health education, screening and drug therapy

No schistosomiasis control activities

All Takelale quarters

3 nonTakelale quarters accidentally excluded from the control activities

Prevalence; disease-specific knowledge

Major reduction in prevalence (7% vs.

71%) and intensity (1% vs. 26%) of schistosomiasis infection; improved knowledge about the disease (sources, risk factors)

Cline 1996

Cameroon

1991â&#x20AC;&#x201C;1993

Local health centers and teachers performed health education, treatment, snail control

Same as the intervention, but without health education

4 selected villages within Kaele subdivision

1 village in another subdivision

Prevalence of infection; disease-specific knowledge; health service utilization

Reduction in prevalence (from 21% to 7%) and in the number of heavy infections (from 23 to 8); increased student knowledge on the role of snails and water contact in infection (respectively, from 16% to 76% and from 11% to 72%); increased utilization of health centers (20%)

Coura Filho 1992

Brazil

1974â&#x20AC;&#x201C;1987

CPqRR and, later, local staff performed demographic census, examination and treatment

n/a

Peri Peri district population

n/a

Prevalence; incidence

Prevalence decreased from 43.5% to 15.2% under CPqRR and further to 4.4% under municipal management; incidence decreased from 19% to 10.9% and further to 2.9%

33

Bausch 1995

(continued )

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

Engels 1993

Burundi

1985–1990

Selective population based chemotherapy by mobile control teams, later integrated into PHC

n/a

General; population

n/a

Number of detected cases

10% of all cases received treatment every year; annual yield of cases detected by basic health services was 60% at an eighth of the cost per inhabitant

Engels 1995

Burundi

1993–1994

Case detection based on primary care

n/a

General; population

n/a

Number of detected cases

15 months after the civil unrest, the service output has recovered in most endemic provinces and the program output has continued to increase after an initial depression in the least affected provinces

Jarallah 1993

Saudi Arabia

1984–1989

Integration of the existing schistosomiasis vertical program into primary care

n/a

Riyadh population and areas

n/a

Population coverage; prevalence; treatment and follow-up; snail control

Six-fold increment in population coverage (up to more than 95% in some health centers); significant reduction in prevalence (from 13.2% to 0.2%); reduction in dropout rate (from 54.4% to 22%); significant reduction in the number of surveyed water sources that resulted positive for snails (from 1.8% to 0.1%)

34

Name

Results

Uganda

n/a

Health education, drug distribution, and deworming by school teachers and community drug distributors

n/a

Communities at high risk in schistosomiasisendemic areas

n/a

Coverage; costs

Coverage was 91.4% in schools and 64.7% in communities; estimated financial costs per person equal to US$ 0.34

Sleigh 1998a

China

1953â&#x20AC;&#x201C;1992

Schistosomiasis Control Program: case detection, treatment, snail eradication and health education

n/a

Guangxi population and areas (>11 million, of which 1.1 million at risk)

n/a

Snail distribution and control; water supplies and sanitation; detection, control and treatment; benefits of eradication

Eradication of schistosomiasis in the 19 endemic countiesâ&#x20AC;&#x201C;reduction in the number of infected snail foci from 4,716 in 1958 to 0 after 1992; people and animals found positive at any test were treated (96,537 people received drug therapy)

Sleigh 1998b

China

1953â&#x20AC;&#x201C;1992

Schistosomiasis Control Program: case detection, treatment, snail eradication and health education

n/a

Guangxi population and areas

n/a

Program costs by category and period, relating time trends to political, economic and policy factors

Annual variable costs of <US$ 0.50 per capita for the protected population

35

Kabatereine 2006

(continued )

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

Sleigh 1998c

China

1993–1996

Schistosomiasis Control Program– maintenance of schistosomiasis eradication

n/a

Guangxi population and areas

n/a

Experience, knowledge, attitude about schistosomiasis and its control

The less educated, including women, knew less about schistosomiasis and were at higher risk; 33% of population surveyed had never heard of schistosomiasis– the older educated knew more than the younger less educated; 79% of the people who knew about the disease were willing to help the control program

Uganda

n/a

Communitydirected health workers (CDHW) and supervisors (CDHS) distributed ivermectin and some provided care

n/a

625 CDHWs, 864 CDHSs and 250 households randomly selected

n/a

CDTI coverage

Enhanced performance of CDHWs/CDHS with additional responsibilities, maintaining 90% treatment coverage of the ultimate treatment goal; no increase in attrition rate of <2%; no change in community support

36

Name

Results

Onchocerciasis Katabarwa 2005

Leprosy India

1999â&#x20AC;&#x201C;2000

Leprosy detection and treatment integrated into PHC (Comprehensive Rural Health Project and community workers )

Leprosy detection and treatment through vertical delivery

General; 5 selected model villages + 5 randomly selected villages in the integrated PHC area

General; 5 vertical villages matched to the intervention villages

Self-reported stigmatization measures: selfesteem, family membersâ&#x20AC;&#x2122; attitude, economic and social interaction

In the integrated PHC approach social stigma was minimal for leprosy patients compared to the vertical approach: ability to freely discuss the disease was 78.6% vs. 10%; the proportion of patients who continued to work after diagnosis was 60% vs. 20%; the number of patients invited to social functions was 100% vs. 50%

Kasturiaratchi 2002

Sri Lanka

1999â&#x20AC;&#x201C;2001

Curative health facilities, dermatological clinics and Central Leprosy Clinics (CLCs) diagnosed, treated and educated communities

n/a

Preschool children and postpartum women

n/a

Prevalence; new case load

Increase in prevalence from 0.6/10,000 to 0.8/10,000; increase in new case load from 0.89/10,000 to 1.2/10,000

37

Arole 2002

(continued )

Name

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

Parkash 2003

India

1997–1999

Leprosy case detection and treatment through PHC

Leprosy case detection and treatment through NLEP

3 PHC centers serving a population of 239,142, located in 2 districts

2 Leprosy Control Units serving the remaining population of 1,065,020 in the intervention areas

Prevalence; new case detection rate (NCDR); proportion of multibacillary (MB), child and deformity among newly detected cases; cure/ treatment completion; caseholding

Results Negligible differences between study (S) and control areas (C) before and after integration in prevalence (S: 2.1/10,000– 2.8/10,000; C: 1.7/10,000– 1.4/10,000) and NCDR (S: 7.7/ 10,000–5.6/10,000; C: 6.2/ 10,000– 5.3/10,000).

38

Appreciable differences in the study– but not control– area before and after in child (S: 36.6%–24.6%; C: 30.1%–32.5%), MB (S: 15.8%–33.7%; C: 17.2%–17%) and deformity (S: 1.9%–4.3%; C: 3%–3.3%) among new cases.

Appreciable differences in MB caseholding between the study and control areas after integration (S: 57.4%; C: 86.7%)

India

1993–1996 and 1998– 2001

Leprosy case detection and treatment through PHC

n/a

Postintegration: patients in two blocks in Gudiyatham Taluk, Tamil Nadu

Preintegration: patients in two blocks in Gudiyatham Taluk, Tamil Nadu

Case detection through voluntary reporting

Voluntary reporting was 349 before integration (1993–1996) and 302 three years after integration (1998–2001)

Deitchler 2004

Bangladesh, Cambodia, China, Indonesia, Laos, Philippines, South Africa, Sri Lanka, Thailand, Vietnam, India, and Myanmar

n/a

Distribution of vitamin A, iodine (in some cases on immunization days) and iron supplements (mainly through antenatal care)

n/a

Preschool children and postpartum women

n/a

n/a

n/a

Gross 2006a

Peru

2001

Health professionals and communities involved in education and distribution of multimicronutrient supplements

n/a

20082 women and adolescent girls (12–44 yr and 8081 children (<5 yr)

n/a

Coverage of multimicronutrient supplements

Coverage rate was 47% for women and adolescent girls and 88% for children

Rao 2002

Nutrition

39

(continued )

40

Name

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

Gross 2006b

Peru

2001

Accompanying communication program to the multimicronutrient supplementation program

n/a

20082 women and adolescent girls (12–44 yr) and 8081 children (<5 yr)

n/a

Effectiveness of the communication program: knowledge, beneficiaryreported compliance, acceptability of supplements

Increased knowledge (e.g. 60% of interviewees named specific foods with high micronutrient content); compliance over 3 months was 90%; acceptance of the supplements was 89%–100%

Hossain 2005

Bangladesh

1996–2002

BINP: education and food supplementation to children with severe malnutrition or growth failure, by communitybased nutrition centers

No major nutrition projects activated in the control areas

Children <5 yr and carers in 4554 households in areas with BINP in place for 6 years

Children <5 yr and carers in 2266 households in the adjacent areas

Rates of malnutrition: underweight (WAZ), stunting (HAZ), wasting (WHZ); caring practices

No observed significant differences in rates of malnutrition between the project and control areas (equal to 0.7% for severe low WAZ, 0.8% for HAZ and 0.1% for WHZ)–the target of reducing the prevalence of severe malnutrition by 40% was not achieved; mothers in project areas reported significantly better caring practices than mothers in nonproject areas (e.g.: 58.7% vs. 23.8% took iron tablets during pregnancy)

Results

Peru

2001

Multimicronutrient supplementation program

n/a

20082 women and adolescent girls (12–44 yr) and 8081 children (<5 yr)

n/a

Routine cost per year; effectiveness (prevalence of anemia); costeffectiveness

Annual cost per community member was US$1.51; effectiveness was equal to 10.7% for women and 14.5% for children; costeffectiveness ratio was US$0.12 per 1% of prevented anemia per community member– higher than that of the current fooddistribution programs

Lechtig 2006b

Peru

2001

Multimicronutrient supplementation program

n/a

20082 women and adolescent girls (12–44 yr) and 8081 children (<5 yr)

n/a

n/a

n/a

Lopez DeRomana 2006

Peru

2002

Multimicronutrient supplementation program– only the fourth campaign

n/a

448 households

418 households

Children's weight; children's height; children's heightfor-age and weight-for-age z-scores; hemoglobin concentrations; BMI

No detected effect on children's weight or height after 8 weeks; protective effect on the hemoglobin levels of the three target groups compared to the respective control (children: 108.5 vs.

111.7 g/L; women: 123.9 vs.125.6 g/L); no detected effect on women and adolescents’ mean BMI

41

Lechtig 2006a

(continued )

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

Amin 1996

Sierra Leone

1988–1992

EPI (neonatal tetanus, measles, pertussis, diphtheria, polio, whooping cough and tuberculosis)

n/a

Mothers and children in two rural districts

n/a

Immunization coverage; infant and child morbidity and mortality Immunization

coverage was >60%; decline in infant and child mortality rate from 162/1,000 to 77/1,000; the most frequently mentioned disease symptoms at the time of death of an infant were–in order–fever, cough, troubled breathing and diarrhea

Bonu 2003

India

1993–1999

Pulse Polio Immunization (PPI) campaign

n/a

Sample of approximately 23,000 women and their surviving children aged 12–35 months (around 6,400)

n/a

Completion rates for polio immunization; coverage of non-polio EPI vaccines; equity in coverage of polio and nonpolio EPI vaccination

Increase in coverage of polio vaccine (I dose: from 48% to 73%, III doses: from 34% to 45%, indicating a dropout rate of 28%); no significant change in immunization coverage of all 5 doses of nonpolio EPI (from 17.7% to 18.5%)

Koop 2001

Macedonia

April–May 1999

Two mass EPI campaign in refugee camps and weekly mobile immunization clinics for new arrivals and children who missed the first campaign

n/a

All <4 years children living in 7 refugee camps who didn’t have an up-to-date vaccination card

n/a

Immunization coverage

Immunization coverage rates in 4/5 camps were >89% during the first campaign; coverage rates of weekly clinics was >90%

Name

Results

Immunization

42

Child Health & Development–IMCI Adam 2005

Tanzania

1999

IMCI

Amaral 2004

Brazil

2002

IMCI standard

Usual care

43

2 rural IMCI districts

2 neighboring non-IMCI districts

Total economic costs of start up and implementation of IMCI; incremental costs of introducing and running IMCI

Lower costs associated with IMCI (US$ 11.19) compared with routine care (US$ 16.09)–44% difference, of which 6% could be attributed to IMCI; no additional staff required; higher cost-effectiveness of IMCI compared with standard care

20 selected low-income families in municipalities with a population 5,000–50,000

Neighboring municipalities in which there were no IMCI-trained workers

Quality of care provided to children <5

Quality of care under IMCI significantly improved compared to standard care (e.g. weight checked against growth chart 77.3% in IMCI vs. 36.3% in control facilities; checking for at least 3 danger signs 38.3% in IMCI vs. 0.6% in nonIMCI; evaluation of feeding practices 35.4% in IMCI vs.

10.1% in non-IMCI) (continued )

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

Armstrong Schellenberg 2004

Tanzania

1999–2002

IMCI adapted to reflect national child-health policies and local terminology

Usual care

2 districts in which IMCI had been implemented for 2 years

2 districts where IMCI had not yet been implemented

Children’s survival and health; economic cost

Improved case management (95% of children were checked for cough, diarrhea and fever in IMCI vs.

36% in control facilities); 13% lower mortality rate in children <5yr in IMCI districts; similar costs of care under IMCI and conventional casemanagement (US$ 11.19 vs. US$ 16.09–US$ 8.30 vs.

US$ 8.76, excluding hospital-level costs); high cost-effectiveness of IMCI

Bryce 2005a

Tanzania

1999

IMCI

Usual care

2 districts in Tanzania where IMCI was implemented in 1997 (i.e. 2 years before the study was conducted)

2 districts where IMCI had not yet been implemented

Quality of care; economic costs

IMCI was associated with significant improvements in quality of care (e.g. proportion of children who were managed correctly was 65% under IMCI and 16% under usual care); lower costs (total cost per child in 1999 was US$7.86 in IMCI vs. US$8.34 in non-IMCI); improved efficiency (cost per child visit managed correctly was US$4.02 in IMCI vs. US$25.70 in non-IMCI)

44

Name

Results

45

Bryce 2005b

General

n/a

IMCI

n/a

n/a

n/a

Implementation of guidelines, quality of care at first-level facilities, health system supports, utilization of health services, child mortality, coordination

Overextension of the guidelines to settings other than the original ones (developing countries with infant mortality >40/1000 live births); improving the quality of care in first-level facilities alone was not sufficient to increase low utilization levels; wide variability in IMCI implementation among the MCE-IMCI sites

Chopra 2005

South Africa

2001â&#x20AC;&#x201C;2002

IMCI adapted to local epidemiological profile (â&#x20AC;&#x201C;measles and malaria; + HIV/AIDS and asthma)

n/a

General; randomly selected clinics within IMCI district

n/a

Index of integrated assessment; primary and secondary indicators as identified by WHO analysis plan

Large improvement in the overall assessment of children (e.g. assessment of danger signs in children: 7% before vs. 72% after IMCI); increased efficiency of care (e.g. rational prescribing: 62% before vs. 84% after IMCI); no change in level of counseling and absence of improvement in caregiver understanding of key practices (e.g. advice to return immediately if the child was not able to drink/breast feed: 32.5% before vs.

35.5% after) (continued )

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

El Arifeen 2004

Bangladesh

2001–2004

IMCI with adapted case management guidelines

Usual care

20 randomly selected facilities in IMCI district

20 paired facilities in area without IMCI

Quality of care (adherence to IMCI guidelines); care seeking behavior; utilization of governmental health facilities

Improvements in quality of health care (mean index of correct treatment for sick children was 54 in IMCI and 9 in control facilities); increase in use of first-level facilities (0.6 visits per child at baseline and 1.9 after 21 months of IMCI); 19% of sick children in IMCI and 9% in control facilities were taken to a health worker

Huicho 2005a

Peru

1996–2001

IMCI training for doctors and nurses, adapted to epidemiological/cultural context (from 11 to 7 days and reduced supervision)

n/a

All 34 districts

n/a

IMCI health worker training; supervision activities; health facility support; utilization of child health services

Increase in the overall number of health workers trained in IMCI until 1999– decrease in 2000 and 2001; overall training coverage of 10.3% for doctors and nurses; average of 4.7 annual attendances per child in 2000; improved performance (e.g. children checked for 3 general danger signs: 35–40% vs.

5–4%); expected synergies between health facility and community interventions were not realized

46

Name

Results

47

Huicho 2005b

Peru

n/a

IMCI

n/a

All 34 districts

n/a

Health services utilization; coverage; impact (mortality and nutritional status)

Average of 4.8 visits per child/year; 2/3 of all children had all basic vaccines recorded on their cards; IMCI training coverage was 10.2% in 2000; in 2000, the average cumulative reduction of under-5 mortality rate was 40.3% of the 1996 level; non-significant correlations between IMCI training coverage and outpatient utilization, vaccine coverage, mortality or malnutrition indicators

Masanja 2005

Tanzania

1999–2002

IMCI

Usual care

2 district with IMCI

2 districts without IMCI

Equity: morbidity, nutritional status, immunization coverage, caretaker knowledge and behavior, home management of illness (household wealth index)

Improvements in child health, not at the expense of equity: equity differentials for 6/10 child health indicators (underweight, stunting, measles immunization, access to treated and untreated nets, treatment of fever with antimalarials) significantly improved in IMCI (e.g. concentration index for stunting improved from −0.102 to −0.032 in IMCI vs. –0.122 to –0.133 in non-IMCI) (continued )

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

Naimoli 2006

Morocco

2000

Outpatient facilities in IMCI where training of HC workers lasted 12 days (instead of 11) and guidelines were adapted

Usual care

Ill children attending a facility where workers received IMCI training

Ill children attending a facility where workers didn’t receive IMCI training

Index of overall guideline adherence; appropriateness of antibiotic prescription

Improved quality of care: adherence to guidelines was 79.7% in IMCI vs. 19.5% in non-IMCI facilities, and correct prescription of antibiotics was 60.8% in IMCI vs.

31.3% in non-IMCI facilities

Pariyo 2005

Uganda

July— December 2000, 2001 and 2002

IMCI case management training to qualified and (later) auxiliary staff

No training in IMCI provided

Children <5 yr; Health workers managing children who received training in IMCI

Health workers managing children who did not receive the training

Factors mediating the effect of IMCI on the quality of care to children: performance of health workers; training coverage; health facility support indicators

Increase in IMCI training coverage from 7.7% to 74.1%; health workers checked the vaccination status of nearly 50% of sick children; the index of availability of the 4 essential vaccines was >70%; <20% of children needing immunizations were vaccinated

Zhang 2007

China

2004–2006

IMCI training in township hospitals, village health centers (no adaptations specified)

n/a

146 sick children, 136 mothers and 74 health facilities from townships of the preintervention set

Preintervention: 550 sick children, 542 mothers and 3345 health facilities in various townships

Presence of basic equipment and supplies; inappropriate prescribing of drugs; presence of basic drug supplies; satisfaction of care givers

Significant improvements in: availability of basic equipment and drug supplies (e.g. children’s scales: from 27.8% to 90.5%), rational use of drugs (inappropriate use of antibiotics: from 59.3% to 6.2%), satisfaction of care givers (from 83.4% to 95.6%)

48

Name

Results

Child Health & Development–ICDS

49

Agarwal 2000

India

1987–1993

ICDS services plus nutritional supplementation provided to malnourished pregnant women

ICDS without food supplementation provided to malnourished pregnant women

Pregnant malnourished women in 28 villages in the ICDS block Harahua

Pregnant malnourished women in 21 villages in the adjacent block

Anthropometric measurements: maternal height, weight and midarm circumference, birth weight; hemoglobin level

Improvements in the supplemented ICDS group compared with the unsupplemented ICDS group: percentage of low birth weight deliveries was 14.4% vs. 20.4%; rate of newborns weighted >3 kg was 16.2% vs. 11%; 100 g extra gain in maternal weight; 0.3 weeks increase in gestational age

Dubowitz 2007

India

n/a

Dular’ Program: ICDS plus training of local resource groups and Anganwadi workers

Usual ICDS

Women and children (0–36 months); 200 from Dularintensive villages and 229 from Dularregular ones

Women and children (0–36 months); 315 from nonDular villages (receiving regular ICDS only)

Use of iodized salt; children’s immunization status; antenatal care in pregnancy; feeding of colostrums; feeding of breast milk; use of delivery kit for childbirth

Statistically significant differences in all measured outcomes between Dular and non-Dular villages (e.g. iodized salt use: 95.8% vs.

53.2%; prenatal care: 91.1% vs.

70.7%; colostrum fed: 82% vs. 20.7%; use of delivery kit for childbirth 48.2% vs.

6.5%) and, although less pronounced, in a number of measures between Dular-intensive and Dular-regular villages (e.g. colostrum fed: 95% vs. 82%) (continued )

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

Gupta 1984

India

1981â&#x20AC;&#x201C;1982

Standard package of ICDS services provided by Anganwadi centers

Non ICDS care

Children <6 years in ICDS block of Rae Bareli, Uttar Pradesh

Children <6 years in nonICDS areas

Impact on feeding practices, growth and development; prevalence of malnutrition; utilization of health care services

Breast feeding practices were almost equal (99.8% ICDS vs.

99.4% non-ICDS), but ICDS performed better than non-ICDS on the other indicators: higher weight-forage (e.g. Kg of 2 yr male children: 9.8 vs. 8.1); timely milestones of development (e.g. teething: 7.2 months vs. 8.5); lower prevalence of severe grade of protein energy malnutrition (3.2% in ICDS vs.

14.8% in non-ICDS); higher (but not satisfactory) immunization rates (e.g. oral polio vaccine coverage: 54.3% vs. 0.3%); use of antenatal care (61.3% vs. 19.9%)

Kapil 1999

India

1996â&#x20AC;&#x201C;1997

Anganwadi community centers provided therapeutic supplementary feeds and advise on breastfeeding and nutrition

n/a

61 severely malnourished children visiting the centre, followed for 1 year of ICDS

n/a

Grade of malnutrition

Significant reduction in grades of undernutrition from 7.7% to 4.5% after 12 months of intervention

50

Name

Results

51

Lal 1980

India

1976â&#x20AC;&#x201C;1979

Anganwadi community centers provided nutritional and immunization services

n/a

Children <6 years and pregnant women in 10 anganwadi centers in ICDS block of Kathura

n/a

Nutritional and health status; utilization of therapeutic and prophylactic nutrients; utilization of immunoprophylaxis

Reduction in prevalence of severe degrees of malnutrition in <6 yr children from 17.6% to 8.4%; increase in utilization of therapeutic and prophylactic nutrients (iron and folic acid: children <6yrâ&#x20AC;&#x201C;from 2.3% to 58%; pregnant womenâ&#x20AC;&#x201C;from 9.2% to 35%); increase in immunization coverage (e.g. all 3 doses of DPT: from 6.7% to 69.9%)

Saiyed 2000

India

n/a

ICDS services provided at Anganwadi centers: immunization, vitamin A, iron and folic acid supplementation, growth monitoring, supplementary food provision and referral for serious illness

n/a

Children <3 years in 9 anganwadi centers located in slum areas of urban Baroda, Gujarat

n/a

Nutritional status measured as height-for-age, weight-for-age and weight-forheight; episodes and duration of illness

Children who utilized all four services (immunization, vitamin A, monthly weighing, and food supplementation) fully rather than partially showed improvements in their nutritional status and in the number of episodes and duration of illness (no illnesses: 72% vs.

39%; mean days of illness: 1.8 vs. 3.9) (continued )

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

Tandon 1988

India

1976–1984

Anganwadi community centers coordinated EPI and ICDS services to improve immunization coverage

EPI without ICDS

Children <6 yr and pregnant/ lactating women in areas with ICDS of 3–5 or 8 years duration

Children <6 yr and pregnant/ lactating women in areas without ICDS support

Immunization coverage

Significant improvements in immunization coverage for children by about 100% in 3–5 yr and 125% in 8 yr ICDS programs (e.g.: BCG: 22.6% in EPI, 39.6% in 3–5 yr ICDS, 50.5% in 8 yr ICDS) and pregnant women (TT: 24.5% in EPI, 44.4% in 3–5 yr ICDS, 32.1% in 8 yr ICDS)

Tandon 1992

India

n/a

Anganwadi community centers coordinated EPI and ICDS services to improve immunization coverage

EPI without established ICDS

5367 children (1–2 yr) and 5111 mothers in 36 ICDS projects in place for at least 5 years

2018 children (1–2 yr) and 1890 mothers in 18 newly sanctioned ICDS projects

Immunization coverage

Increased immunization coverage in ICDS projects compared with non-ICDS ones for children (BCG: 65% vs. 22%, DPT: 63% vs. 28%, polio 64% vs. 27%) and pregnant women (TT: 68% vs.

40%)

52

Name

Results

Trivedi 1995

India

1990–1993

Standard package of ICDS services provided by Anganwadi centers

Care-as-usual; no ICDS services

709 children between 1–6 years in 6 anganwadi centers within ICDS blocks in rural area of Indore, Madhya Pradesh

500 children between 1–6 years in 5 matched villages in nonICDS area

Nutritional status (grade of malnutrition); immunization status (BCG, DPT, polio, measles)

No significant difference in nutritional status between ICDS and non-ICDS areas (e.g. normal and grade I malnutrition: 74.3% and 72.4%); lower immunization coverage in ICDS compared with non-ICDS areas (BCG: 80.2% vs.

88.8%; DPT: 79.5% vs. 94.4%; polio: 88% vs. 95.3%; measles: 45.7% vs. 62%)

Extended analysis of Phillips1984

n/a

General; 2 blocks of 20 villages each

n/a

Contraceptive prevalence rate as function of MCH services

The addition of 18 months to the time series reported by Phillips (1984) did not change substantive results: intensification of MCH services in 1982 did not have a positive incremental impact on the contraceptive prevalence rate (measles/tetanus vaccine and antenatal care/TBA training variables did not have significant coefficients for explaining contraceptive prevalence rate)

Family Planning and Maternal and Child Health DeGraff 1986

Bangladesh

1982–1984

53

(continued )

Setting

Time Frame

Description of Intervention

Description of Control

Intervention Group

Control Group

Outcome Measure

Douthwaite 2005

Pakistan

1993–2002

Lady Health Worker Program (LHWP): promotion and provision of door-to-door FP and primary health care services

Usual care

Women from randomly selected households and LHWs in LHWP area

Women from sampled communities not served by the LHWP

Current use of reversible modern family planning methods

Significantly higher current use of reversible modern contraceptive methods under LHWP (women using reversible modern methods: 13% in LHWP, 7% in control areas–national average of 10%), even after controlling for socio-economic factors

Phillips 1984

Bangladesh

1982–1983

Stepwise introduction of limited or intensive MCH services into existing FP by family welfare centers and community health workers

Usual care

General; 2 blocks of 20 villages each

General; 2 blocks of 20 villages each

Contraceptive prevalence rate as function of MCH services

No significant difference between an incremental addition of MCH services and a basic package of comprehensive FP with minimal MCH services on contraceptive prevalence rate

Simmons 1991

Bangladesh

1978–1985

Matlab FP and Health Services Project (FPHSP): communitybased services (focus only on the limited package of services)

Government of Bangladesh FP Program: compartmentalized service orientation

A large, fairly homogeneous region divided into two parts (intervention and control)

A large, fairly homogeneous region divided into two parts (intervention and control)

Costs of the service delivery system

The Matlab Project is more expensive than the government's FP program, but it is also more effective (e.g. cost-effectiveness ratio under Model III: average cost per birth prevented ranged from $171 to $220 in treatment areas and from $240 to $298 in control areas)

54

Name

Results

Tuladhar 1982

55

Nepal

1975â&#x20AC;&#x201C;1978

Village health workers providing FP/MCH and numerous other services more frequently (e.g. nutrition education, immunization)

FP/MCH and fewer comprehensive and preventive services delivered less frequently

Pregnant/ lactating women in 2 districts where services had been integrated in 1976

Pregnant/ lactating women in 2 districts matched to the intervention districts

Knowledge of FP; use of and intention to use FP services; child survivorship; infant mortality

Compared to the vertical program, the integrated model reported: smaller impact on knowledge of FP (increase in the number of women who had knowledge of FP: 90% in V vs. 22% in I); higher impact on intentions to use FP services (decrease of 13% vs. 35%); relatively greater impact on child survivorship (1% vs. 4%); significantly smaller impact on infant mortality (decrease of 47% vs. 25%)

Haiti

1985â&#x20AC;&#x201C;2000

GHESKIO clinic providing voluntary counseling, testing and care to HIV/AIDS children and adults, and services for other communicable diseases

n/a

HIV/AIDSaffected families, health care workers after accidental HIV/AIDS exposure, female rape victims

n/a

Demographics and n. individuals for counseling/ testing; n. clients receiving each integrated service; effect of HIV/AIDS prevention; HIV/AIDS transmission

Increase in the number of new people seeking voluntary counseling and testing (from 142 to 8175); care was provided to 17% of the new patients who asked for AIDS treatment, 6% of TB, 18% of STI, 19% of FP; HIV/AIDS transmission between discordant couples was 0 infections/100 follow-up years; vertical transmission was 11 infections/ 100 live births

HIV/AIDS/AIDS Peck 2003

APPENDIX C

Contextual Factors

57

Priority Intervention

Sustainability Fiscal Space, Frailty

Opportunities Critical Events, Synergy, Technology/Innovation

Desirability Political Economy, Socio-Cultural Factors

■ Small dengue outbreaks in nonendemic area

■ Dengue control low priority for communities ■ Resistance to change amongst vector control personnel ■ Provision of adequate services at municipal level required

Dengue Cuba (Toledo Romani and others 2007)

58

Malaria ■ Program decentralized

Colombia (Rojas, Botero, and Garcia 2001) Schistosomiasis Brazil (Coura Filho and others 1992)

■ Municipal capacity

Burundi (Engels, Ndoricimpa, and Gryseels 1993; Engels, Sindayigaya, and Gryseels 1995)

■ Reduced disease burden

Cameroon (Bausch and Cline 1995; Cline and Hewlett 1996)

China (Sleigh and others 1998a, 1998b, 1998c)

■ Civil unrest following a military coup attempt, disturbed Primary Health Care services ■ Development projects that might expand snail habitats and transmission ■ Urinary schistosomiasis as an entry point into the PHC system ■ Easy to recognize, diagnosis and treat

■ Strong government commitment to national PHC system

■ High priority for all levels of government

Saudi Arabia (Ageel and Amin 1997; Jarallah and others 1993)

■ Reduced disease burden ■ Advanced system of PHC with >90% population coverage

■ General rise in living standards and socio-economic conditions, improved sanitation and water supplies, and medical care ■ In remote communities up to 30% of school-aged children not enrolled in school: not reached by school based approaches

Uganda (Kabatereine and others 2006)

Leprosy Sri Lanka (Kasturiaratchi, Settinayake, and Grewal 2002)

■ Reduced disease burden

■ High degree of political support from all levels of health administration

India (Rao and others 2002; Thakar and Kumar 2003)

■ Reduced disease burden

■ Resistance to change amongst NLEP personnel; prior attempt at integration failed

59

Nutrition Nutrition–Peru (Gross, Mamani Diaz, and Valle 2006a; Gross, Valle, and Mamani Diaz 2006b; Lechtig and others 2006a, 2006b; Lopez de Romana and others 2006)

■ Evidence of high prevalence of irondeficiency anemia

■ Rapid urbanization

■ High transmission of wild poliovirus

■ Rural areas with persistent inequities in immunization coverage by caste, gender, wealth and religion ■ Priority for the international agenda

Immunization India (Bonu, Rani, and Baker 2003)

(continued )

Opportunities Critical Events, Synergy, Technology/Innovation

Desirability Political Economy, Socio-Cultural Factors

Macedonia (Koop, Jackson, and Nestel 2001)

■ Emergency situation with high risk of vaccine-preventable diseases’

■ Military conflict

Sierra Leone (Amin 1996)

■ Endemicity of common childhood diseases and inadequate rural health services

■ Military conflict ■ Widespread poverty ■ Priority for the international agenda

Priority Intervention

Sustainability Fiscal Space, Frailty

60 Child Health & Development–IMCI

■ Possible conflict between the integrated case management approach and other clinical rule systems already in use.

Morocco (Naimoli and others 2006)

Peru (Huicho and others 2005a, 2005b)

■ Good network of health facilities with high utilization rates

South Africa (Chopra and others 2005)

■ Good facility infrastructure and management support

Tanzania (Armstrong Schellenberg and others 2004; Masanja and others 2005)

■ High utilization rates of public health facilities

■ Overlapping child health programs and competition for resources.

■ IMCI introduced concurrently with measures designed to improve district health systems management

■ Period of rapid political change ■ Poor governance and lack of political commitment

Child Health & Development–ICDS ■ ICDS introduced in pursuance of National Policy for Children

India–ICDS (Agarwal and others 2000; Gupta and others 1984; Kapil and others 1999; Lal 1980; Saiyed and Seshadri 2000) ■ ICDS and EPI target same group of children

India–ICDS + EPI (Tandon and others 1992; Tandon and Sahai 1988) India–ICDS + “Dular” (Dubowitz and others 2007)

■ Builds on existing ICDS infrastructure

■ Strong leadership

Family Planning ■ Many years of experience ■ Dedicated human and physical resources and organizational ability

Pakistan–LHWP (Douthwaite and Ward 2005)

■ Overburdened health system ■ Significant funding shortfalls

61

Bangladesh–FPHSP (DeGraff and others 1986; Phillips and others 1984; Simmons, Balk, and Faiz 1991)

■ High population growth rates hindering economic growth and development ■ Presence of MCH contributes to legitimacy of FP, activating latent demand. Reduction in child morbidity and mortality leads to change in reproductive motives

HIV/AIDS/AIDS Haiti–“GHESKIO” (Peck and others 2003)

■ HIV/AIDS VCT attracts a population with high rates of co-morbidity. Reciprocally, on-site services attract more people to VCT

References Adam, T., F. Manzi, J.R. Armstrong Schellenberg, L. Mgalula, D. de Savigny, and D.B. Evans. 2005. “Does the Integrated Management of Childhood Illness Cost More than Routine Care? Results from the United Republic of Tanzania.” Bulletin of the World Health Organization 83(5):369–77.

Agarwal, K.N., D.K. Agarwal, A. Agarwal, S. Rai, R. Prasad, S. Agarwal and T.B. Singh.

2000. “Impact of the integrated child development services (ICDS) on maternal nutrition and birth weight in rural Varanasi.” Indian Pediatrics 37(12):1321–27.

Ageel, A.R.M., and M.A. Amin. 1997. “Integration of schistosomiasis-control activities into the primary-health-care system in the Gizan region, Saudi Arabia.” Annals of Tropical Medicine and Parasitology 91(8):907–15.

Amaral, J., E. Gouws, J. Bryce, A.J. Madeiro Leite, A.L.A. da Cunha and C.G. Victora. 2004.

“Effect of Integrated Management of Childhood Illness (IMCI) on health worker performance in Northeast-Brazil.” Cadernos de Saude Publica 20 Suppl 2:S209–19.

Amin, R. 1996. “Immunization coverage and child mortality in two rural districts of Sierra Leone.” Social Science and Medicine 42(11):1599–1604.

Armstrong Schellenberg, J.R., T. Adam, H. Mshinda, H. Masanja, G. Kabadi, O. Mukasa, T. John, S. Charles, R. Nathan, K. Wilczynska, L. Mgalula, C. Mbuya, R. Mswia, F. Manzi, D. de Savigny, D. Schellenberg, and C. Victora. 2004. “Effectiveness and cost of facility-based Integrated Management of Childhood Illness (IMCI) in Tanzania.”

Lancet 364(9445):1583–94.

Arole, S., R. Premkumar, R. Arole, M. Maury and P. Saunderson. 2002. “Social stigma: a comparative qualitative study of integrated and vertical care approaches to leprosy.”

Leprosy Review 73(2):186–96.

Atun, R., S. Bennett and A. Duran. 2008. When do vertical (stand-alone) programmes have a place in health systems? Copenhagen: World Health Organization.

Atun, R., T.E. de Jongh, F.V. Secci, K. Ohiri and O. Adeyi. 2009 (forthcoming). “ Integration of Targeted Health Interventions into Health Systems: A Conceptual Framework for Analysis.” Health Policy and Planning.

Bausch, D., and B.L. Cline. 1995. “The impact of control measures on urinary schistosomiasis in primary school children in northern Cameroon: A unique opportunity for controlled observations.” American Journal of Tropical Medicine and Hygiene 53(6): 577–80.

Bonu, S., M. Rani, and T.D. Baker. 2003. “The impact of the national polio immunization campaign on levels and equity in immunization coverage: evidence from rural North India.” Social Science and Medicine 57(10):1807–19.

Briggs, C.J., and P. Garner. 2006. “Strategies for integrating primary health services in middle- and low-income countries at the point of delivery.” Cochrane Database of Systematic Reviews (2).

63

64

World Bank Working Paper

Bryce, J., E. Gouws, T. Adam, R.E. Black, J.A. Schellenberg, F. Manzi, C.G. Victora and J.P. Habicht. 2005a. “Improving quality and efficiency of facility-based child health care through Integrated Management of Childhood Illness in Tanzania.” Health Policy and Planning 20:169–76.

Bryce, J., C.G. Victora, J.P. Habicht, R.E. Black, R.W. Scherpbier, and MCE IMCI Tech Advisors. 2005b. “Programmatic pathways to child survival: results of a multi-country evaluation of Integrated Management of Childhood Illness.” Health Policy and Planning 20:15–117.

Chopra, M., S. Patel, K. Cloete, D. Sanders, and S. Peterson. 2005. “Effect of an IMCI intervention on quality of care across four districts in Cape Town, South Africa.” Archives of Disease in Childhood 90(4):397–401.

Cline, B.L., and B.S. Hewlett. 1996. “Community-based approach to schistosomiasis control.” Acta Tropica 61(2):107–19.

Coura Filho, P., R.S. Rocha, MFF de Lima e Costa, and N. Katz. 1992. “A municipal level approach to the management of schistosomiasis control in PeriPeri, MG, Brazil.”

Revista Do Instituto De Medicina Tropical De Sao Paulo 34(6):543–48.

Cueto, M. 2004. “The origins of primary health care and selective primary health care.”

American Journal of Public Health 94(11):1864–74.

DeGraff, D.S., J.F. Phillips, R. Simmons, and J. Chakraborty. 1986. “Integrating healthservices into an MCH-FP program in Matlab, Bangladesh–an analytical update.” Studies in Family Planning 17(5):228–34.

Deitchler, M., E. Mathys, J. Mason, P. Winichagoon, and MA. Tuazon. 2004. “Lessons from successful micronutrient programs. Part II: program implementation.” Food and Nutrition Bulletin 25(1):30–52.

Douthwaite, M., and P. Ward. 2005. “Increasing contraceptive use in rural Pakistan: An evaluation of the Lady Health Worker Programme.” Health Policy and Planning 20(2): 117–23.

Dubowitz, T., D. Levinson, J.N. Peterman, G. Verma, S. Jacob, and W. Schultink. 2007.

“Intensifying efforts to reduce child malnutrition in India: An evaluation of the Dular program in Jharkhand, India.” Food and Nutrition Bulletin 28:266–73.

El Arifeen, S., L.S. Blum, D.M. Hoque, E.K. Chowdhury, R. Khan, R.E. Black, C.G. Victora, and J. Bryce. 2004. “Integrated Management of Childhood Illness (IMCI) in Bangladesh: early findings from a cluster-randomised study.” Lancet 364(9445): 1595–1602.

Engels, D., J. Ndoricimpa, and B. Gryseels. 1993. “Schistosomiasis-mansoni in Burundi– progress in its control since 1985.” Bulletin of the World Health Organization 71(2): 207–14.

Engels, D., B. Sindayigaya, and B. Gryseels. 1995. “Sustainability of schistosomiasis case detection based on primary health care.” Transactions of the Royal Society of Tropical Medicine and Hygiene 89(6):599.

Gross, U., M. Mamani Diaz, and C. Valle. 2006a. “Effectiveness of the communication program on compliance in a weekly multimicronutrient supplementation program in Chiclayo, Peru.” Food and Nutrition Bulletin 27(4, Suppl. 1):130–42.

Gross, U., C. Valle, and M. Mamani Diaz. 2006b. “Effectiveness of distribution of multimicronutrient supplements in children and in women and adolescent girls of childbearing age in Chiclayo, Peru.” Food and Nutrition Bulletin 27(4, Suppl.1):122–29.

Clearing the Global Health Fog

65

Gupta, S.B., B.C. Srivastava, V. Bhushan, and P. Sharma. 1984. “Impact of the integrated child-development services in Uttar-Pradesh.” Indian Journal of Medical Research 79(March):363–72.

Hossain, S.M.M., A. Duffield, and A. Taylor. 2005. “An evaluation of the impact of a US$60 million nutrition programme in Bangladesh.” Health Policy and Planning 20(1):35–40.

Huicho, L., M. Davila, M. Campos, C. Drasbek, J. Bryce, and C.G. Victora. 2005a. “Scaling up Integrated Management of Childhood Illness to the national level: achievements and challenges in Peru.” Health Policy and Planning 20(1):14–24.

Huicho, L., M. Davila, F. Gonzales, C. Drasbek, J. Bryce, and C.G. Victora. 2005b. “Implementation of the Integrated Management of Childhood Illness strategy in Peru and its association with health indicators: an ecological analysis.” Health Policy and Planning 20:132–41.

Jarallah, J.S., S.A. al-Shammari, T.A. Khoja, and M. al-Sheikh. 1993. “Role of primary health care in the control of schistosomiasis. The experience in Riyadh, Saudi Arabia.”

Tropical and Geographical Medicine 45(6):297–300.

Kabatereine, N.B., E. Tukahebwa, F. Kazibwe, H. Namwangye, S. Zaramba, S. Brooker, J.R. Stothard, C. Kamenka, S. Whawell, J.P. Webster, and A. Fenwick. 2006. “Progress towards countrywide control of schistosomiasis and soil-transmitted helminthiasis in Uganda.”

Transactions of the Royal Society of Tropical Medicine and Hygiene 100(3):208–15.

Kapil, U., M. Tandon, D. Nayar, S. Ramachandron, R. Yadav and S.N. Dwivedi. 1999.

“Nutritional rehabilitation of severely malnourished children at domiciliary level through the integrated child development services (ICDS) scheme: A perspective study.” Journal of Tropical Pediatrics 45(2):124–25.

Kasturiaratchi, N. D., S. Settinayake, and P. Grewal. 2002. “Processes and challenges: How the Sri Lankan health system managed the integration of Leprosy Services.” Leprosy Review 73(2):177–85.

Katabarwa, M.N., P. Habomugisha, F.O. Richards, and D. Hopkins. 2005. “Communitydirected interventions strategy enhances efficient and effective integration of health care delivery and development activities in rural disadvantaged communities of Uganda.” Tropical Medicine & International Health 10(4):312–21.

Koop, D.G., B.M. Jackson, and G. Nestel. 2001. “Results of the expanded program on immunization in the Macedonian refugee camps.” American Journal of Public Health 91(10):1656–59.

Lal, S. 1980. “Better primary health care services utilization through integrated child development service scheme in Haryana.” Indian Journal of Pediatrics 47(387):293–96.

Lechtig, A., R. Gross, O. Aquino Vivanco, U. Gross, and D. Lopez de Romana. 2006a.

“Lessons learned from the scaling-up of a weekly multimicronutrient supplementation program in the Integrated Food Security Program (PISA).” Food and Nutrition Bulletin 27(4, Suppl.1):160–65.

Lechtig, A., R. Gross, J. Paulini, and D. Lopez de Romana. 2006b. “Costs of the multimicronutrient supplementation program in Chiclayo, Peru.” Food and Nutrition Bulletin 27(4, Suppl.1):151–59.

Lopez de Romana, D., S. Verona, O. Aquino Vivanco, and R. Gross. 2006. “Protective effect of multimicronutrient supplementation against anemia among children, women, and adolescent girls in lower-income areas of Chiclayo, Peru.” Food and Nutrition Bulletin 27(4, Suppl.1):143–50.

66

World Bank Working Paper

Magnussen, L., J. Ehiri and P. Jolly. 2004. “Comprehensive versus selective primary health care: Lessons for global health policy.” Health Affairs 23(3):167–76.

Masanja, H., J.A. Schellenberg, D. De Savigny, H. Mshinda, and C.G. Victora. 2005.

“Impact of Integrated Management of Childhood Illness on inequalities in child health in rural Tanzania.” Health Policy and Planning 20:177–84.

Naimoli, J. F., A. K. Rowe, A. Lyaghfouri, R. Larbi and L. A. Lamrani. 2006. “Effect of the Integrated Management of Childhood Illness strategy on health care quality in Morocco.” International Journal for Quality in Health Care 18(2):134–44.

Newell, K. W. 1988. “Selective primary health-care–the counter revolution.” Social Science & Medicine 26(9):903–06.

Pariyo, G.W., E. Gouws, J. Bryce, G. Burnham, and IMCI Impact Study Team Uganda.

2005. “Improving facility-based care for sick children in Uganda: training is not enough.”

Health Policy and Planning 20:158–68.

Parkash, I., and P. S. Rao. 2003. “Study on integration of the National Leprosy Eradication Programme into primary health care services: a pilot project.” Indian Journal of Leprosy 75(1):25–35.

Peck, R., D.W. Fitzgerald, B. Liautaud, M.M. Deschamps, R.I. Verdier, M.E. Beaulieu, R. GrandPierre, P. Joseph, P. Severe, F. Noel, P. Wright, W.D. Johnson, and J.W. Pape.

2003. “The feasibility, demand, and effect of integrating primary care services with HIV voluntary counseling and testing–Evaluation of a 15-year experience in Haiti, 1985–2000.” Journal of Acquired Immune Deficiency Syndromes 33(4):470–75.

Phillips, J.F., R. Simmons, J. Chakraborty, and A.I. Chowdhury. 1984. “Integrating healthservices into an MCH-FP program–lessons from Matlab, Bangladesh.” Studies in Family Planning 15(4):153–61.

Rao, P.S.S.S., N. Gift, G.S. Rao, P. Samuel, and R.S. Bushanam. 2002. “Elimination of leprosy: the integration of leprosy related activities into the general health services of Tamil Nadu.” Leprosy Review 73(2):123–29.

Rojas, W., S. Botero, and H.I. Garcia. 2001. “An integrated malaria control program with community participation on the Pacific Coast of Colombia.” Cadernos de Saude Publica 17(Suppl.):103–13.

Saiyed, F., and S. Seshadri. 2000. “Impact of the integrated package of nutrition and health services.” Indian Journal of Pediatrics 67(5):322–28.

Simmons, G.B., D. Balk, and K.K. Faiz. 1991. “Cost-effectiveness analysis of family-planning programs in rural Bangladesh–evidence from Matlab.” Studies in Family Planning 22(2):83–101.

Sleigh, A., S. Jackson, X. Li, and K. Huang. 1998b. “Eradication of schistosomiasis in Guangxi, China. Part 1: Setting, strategies, operations, and outcomes, 1953–92.” Bulletin of the World Health Organization 76(4):361–72.

———. 1998a. “Eradication of schistosomiasis in Guangxi, China. Part 2: Political economy, management strategy and costs, 1953–92.” Bulletin of the World Health Organization 76(5):497–508.

———. 1998c. “Eradication of schistosomiasis in Guangxi, China. Part 3. Community diagnosis of the worst-affected areas and maintenance strategies for the future.” Bulletin of the World Health Organization 76(6):581–90.

Tandon, B.N., N. Gandhi, B. Kumar, R.K.P. Sinha, V.K. Karan, R.G. Bajaj, D.K. Yadava, A. Rouf, S.P. Sharma, M. Ahmed, C.K. Sasidharan, K. Rajan, K. Thomas, M. Ketkar,

Clearing the Global Health Fog

67

K.K. Kaul, N.R. Bhandari, S. Upadhyay, T. Guha, M. Natu, D.M. Meshram, R.B. Patel, S.U. Warerkar, M.K. Paul, A.C. Patowary, G. Camotim, R. Thakur, D. Sharma, S.P. Mehta, E.Y. Singh, L.I. Devi, D. Roy, Rothangliana, D.K. Srivastava, K.K. Vadhera, V. Mohan, A. Chandrabhushanam, J. Subramaniyam, and B. Roy. 1992. “Immunization coverage in India for areas served by the integrated child-development services program.” Bulletin of the World Health Organization 70(4):461–65.

Tandon, B.N., and A. Sahai. 1988. “Immunization in India–contribution of integrated child-development services scheme to Expanded Program of Immunization.” Journal of Tropical Pediatrics 34(6):309–12.

Toledo Romani, M.E., V. Vanlerberghe, A. Baly Gil, E. Ceballos, L. Valdes, M. Searret, M. Boelaert, and P. van der Stuyft. 2007. “Towards active community participation in dengue vector control: results from action research in Santiago de Cuba, Cuba.”

Transactions of the Royal Society of Tropical Medicine and Hygiene 101(1):56–63.

Trivedi, S., B.C. Chaparwal, and S. Thora. 1995. “Utilization of ICDS scheme in children one to six years of age in a rural block of central India.” Indian Pediatrics 32(1):47–50.

Tuladhar, J.M., and J. Stoeckel. 1982. “The relative impacts of vertical and integrated FP/MCH programs in rural Nepal.” Studies in Family Planning 13(10):275–86.

Walsh, J.A., and K.S. Warren. 1979. “Selective primary health-care–interim strategy for disease control in developing-countries.” New England Journal of Medicine 301(18): 967–74.

Warren, K.S. 1988. “The evolution of selective primary health-care.” Social Science & Medicine 26(9):891–98.

Wisner, B. 1988. “GOBI versus PHC–some dangers of selective primary health-care.”

Social Science & Medicine 26(9):963–69.

World Bank and World Health Organization. 2006. “High Level Forum on the Health Millennium Development Goals: Selected Papers 2003–2005.”

Zhang, Y.F., Y.H. Dai, and S. M. Zhang. 2007. “Impact of implementation of Integrated Management of Childhood Illness on improvement of health system in China.” Journal of Paediatrics and Child Health 43:681–85.

Eco-Audit Environmental Benefits Statement The World Bank is committed to preserving Endangered Forests and natural resources. We print World Bank Working Papers and Country Studies on 100 percent postconsumer recycled paper, processed chlorine free. The World Bank has formally agreed to follow the recommended standards for paper usage set by Green Press Initiativeâ&#x20AC;&#x201D;a nonprofit program supporting publishers in using fiber that is not sourced from Endangered Forests. For more information, visit www.greenpressinitiative.org.

In 2008, the printing of these books on recycled paper saved the following: Trees*

Solid Waste

Water

355 16,663 129,550 *40' in height and

6â&#x20AC;&#x201C;8" in diameter

Pounds

Gallons

Net Greenhouse Gases Total Energy

31,256

247 mil.

Pounds CO2 Equivalent

BTUs

Clearing the Global Health Fog is part of the World Bank Working Paper series. These papers are published to communicate the results of the Bankâ&#x20AC;&#x2122;s ongoing research and to stimulate public discussion.

A longstanding debate on health system organization relates to the benefits of integrating programs that emphasize specific interventions into mainstream health systems to increase access and improve health outcomes. This debate has long been characterized by polarization of views and ideologies, with protagonists for and against integration arguing relative merits of each approach. Recently, the debate has been rekindled due to substantial rises in externally-funded programs for priority health, nutrition, and population (HNP) interventions and an increase in international efforts aimed at health system strengthening.

However, all too frequently these arguments have not been based on hard evidence. In this book we present findings of a systematic review that explores a broad range of evidence on: (i) the extent and nature of integration of targeted health programs that emphasize specific interventions into critical health systems functions; (ii) how the integration or non-integration of health programs into critical health systems functions in different contexts have influenced program success; and (iii) how contextual factors have affected the extent to which these programs were integrated into critical health systems functions.

The findings provide a new synthesis of evidence to inform the debate on health systems and targeted interventions. In practice a rich mix of solutions exists. While the discussion on the relative merits of integrating health interventions will no doubt continue, discussions should move away from the highly reductionist approach that has polarized this debate.

World Bank Working Papers are available individually or on standing order. Also available online through the World Bank e-Library (www.worldbank.org/elibrary).

ISBN 978-0-8213-7818-2 THE WORLD BANK 1818 H Street, NW Washington, DC 20433 USA Telephone: 202 473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org

SKU 17818