March 009
Evidence, Policy and Action
Cases from the Western Pacific Region
WHO Library Cataloguing in Publication Data
Promoting Health and Equity : Evidence, Policy and Action : Cases from the Western Pacific Region 1. Health services accessibility. 2. Health policy. 3. Evidence-based practice. 4. Western Pacific
ISBN 987 92 9061 427 2 (NLM Classification: W 76)
© World Health Organzaton 009 All rghts reserved.
The desgnatons employed and the presentaton of the materal n ths publcaton do not mply the expresson of any opnon whatsoever on the part of the World Health Organzaton concernng the legal status of any country, terrtory, cty or area or of ts authortes, or concernng the delmtaton of ts fronters or boundares. Dotted lnes on maps represent approxmate border lnes for whch there may not yet be full agreement.
The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by the World Health Organzaton n preference to others of a smlar nature that are not mentoned.
Errors and omssons excepted, the names of propretary products are dstngushed by ntal captal letters.
The World Health Organzaton does not warrant that the nformaton contaned n ths publcaton s complete and correct and shall not be lable for any damages ncurred as a result of ts use.
Publcatons of the World Health Organzaton can be obtaned from WHO Press, World Health Organzaton, 0 Avenue Appia, 1211 Geneva 27, Switzerland (tel: +41 22 791 2476; fax: +41 22 791 4857; e-mail: [email protected]).
Requests for permsson to reproduce WHO publcatons, n part or n whole, or to translate them – whether for sale or for noncommercial distribution – should be addressed to WHO Press, at the above address (fax: +41 22 791 4806;
e-mail: [email protected]). For WHO Western Pacific Regional Publications, request for permission to reproduce should be addressed to Publications Office, World Health Organization, Regional Office for the Western Pacific, P.O.
Box 2932, 1000, Manila, Philippines, fax: +632 521 1036, e-mail: [email protected]
v
9
47
63
8
99
115
9
43 157 Acronyms
Improving health equity through the use of evidence: cases from the Western Pacific Region
Health financing strategies to improve access to health services for the poor in Cambodia: from pilot to policy and action—
A case study of Health Equty Funds
Research, nterventon desgn and polcy mplementaton of the New Rural Cooperatve Medcal Scheme n Shandong, Chna Health Care Fund for the Poor in Viet Nam: how evidence and poltcs came together
Scaling up primary health care in the Lao People’s Democratic Republic usng evdence from a long-term prmary health care development project Promoting health equity: evidence, policy and acton—
The New Zealand experence
The development and targetng of malara control nterventons for populations in high transmission areas of Cambodia: the influence of research on polcy and practce
Public-private mix DOTS: a strategy to engage all health care providers in tuberculosis control and significantly increase access to DOTS services n the Phlppnes
Geographc equty n dstrbuton of scarce dalyss resources n Malaysa Promotng health equty through capacty buldng of prmary health care workers n Mongola
ADB Asan Development Bank
AIDS Acquired immunodeficiency syndrome
AusAID Australan Agency for Internatonal Development CBHI Communty-based health nsurance
CDC Centers for Dsease Control CHS Commune health staton
CMBS Camboda Malara Baselne Survey CNM Natonal Malara Centre
CUP Comprehensive and Unified Policy for Tuberculosis Control DfID Department for Internatonal Development
DOT Drectly-observed treatment
DOTS Drectly-observed treatment, short-course therapy DPMU Dstrct project management unt
DRF Drug revolvng fund EC European Commsson
EDAT Early dagnoss and approprate treatment EPI Expanded programme on mmunzaton EVIPNet Evdence-Informed Polcy Network GDP Gross domestc product
GIS Geographc nformaton system
GRET Groupe de Recherche et d’Echanges Technologiques HEF Health equty fund
HEPR Hunger Eradcaton and Poverty Reducton programme HIS Health nformaton system
HNZ Housng New Zealand
HSPI Health Strategy and Polcy Insttute HSRP Health Sector Reform Project
HSSP Natonal Health Strategc Plan 003-007 ITN Insectcde-treated bednet
JICA Japan Internatonal Cooperaton Agency LLIN Long-lastng nsectcdal net
MAF Medcal assstance fund
MDG Mllennum Development Goal MSF Médecns Sans Frontères
NCMS New Cooperatve Medcal Servce NGO Non-governmental organzaton
NGPES Natonal Growth and Poverty Eradcaton Strategy NHC Natonal Health Commttee
pACT Pre-packaged artemsnn-based combnaton therapy PHC Prmary health care
PhlCAT Phlppne Coalton aganst Tuberculoss PhlHealth Phlppne Health Insurance Corporaton
PhlTIPS Phlppne Tuberculoss Intatve for the Prvate Sector PHO Prmary Health Organsaton
PMU Project management unt PPMD Publc-Prvate Mx DOTS
PPMU Provncal project management unt PPP Purchasng power party
PRSP Poverty Reducton Strategy Paper RDT Rapd dagnostc test
RII Relatve ndex of nequalty RM Malaysan rnggt
SARS Severe acute respratory syndrome SCA Save the Chldren Australa
SEDP6 6th Soco-economc Development Plan SII Slope ndex of nequalty
SRD Standardzed rate dfference SRR Standardzed rate rato SWAp Sector-wde approach TB Tuberculoss
TBA Tradtonal brth attendant TFR Total fertlty rate
TWG-H Techncal Workng Group-Health UNICEF United Nations Children’s Fund
VHLSS Vetnam Household Lvng Standard Survey VHV Vllage health volunteer
VND Vetnamese Dong VSS Vetnam Socal Securty WHO World Health Organzaton
Improving health equity through the use of
evidence: cases from the Western Pacific Region
Steve Fabrcant
Introduction
Health equty and the barrers to achevng t n developng countres have been a major subject of research for many years, resultng n a growng evdence base on polces and actons to promote health equty. Despte efforts over the past two decades, the evdence shows that nequaltes are ncreasng rather than decreasng n many countres. Ths may partly be due to faulty polcy decsons that have ncreased some of the barrers to access to health servces faced by the poor. It has become clear that a better evdence-based approach to health polcy s needed, wth equty as ts focus.
At the same tme, understandng s also nadequate on how the growng evdence base on promotng equty n health can be best used by polcy-makers. Health polcy development n countres s ncreasngly supported by research, nvolvng a range of stakeholders ncludng academc nsttutons, government thnk tanks, NGOs and consumer groups. Partners from other nternatonal organzatons, academa, and research networks are also engaged n strengthenng the evdence base for health polcy and acton, ncludng that focusng on equty-, gender-, and poverty-related ssues n health. However, the lnks between evdence and polcy-makng are nconsstent and varable. There s, thus, stll a need for stronger lnks between evdence and health polcy-makng and mplementaton.
The World Health Organzaton (WHO) s ncreasngly asked to help strengthen mechansms that link research and policy-making in developing countries. For example, in May 2005, resolution WHA 58.34 called for better use of health research and health information, as well as better knowledge management, to support evdence-nformed health polcy and practce, specifically requesting WHO to assist in the development of more effective mechanisms to brdge the dvde between knowledge generaton and ts use, ncludng the translaton of health research findings into policy and practice. At the fifty-seventh session of the WHO Regional Commttee held n September 006, a mnsteral round table dscusson was held on the topc Translaton of Research nto Polcy and Health Care Practce, and WHO was requested to lead efforts to strengthen the capacty for better utlzaton of research results n natonal health polcy-makng. Ths s one of the strategc drectons for WHO n the perod 008–03.
To respond in part to this need, the WHO Western Pacific Regional Office convened the High-Level Meeting on Promoting Health Equity: Evidence, Policy and Action, from 16 to 8 October 007 n Phnom Penh, hosted by the Royal Government of Camboda, wth the objectve of gvng partcpants the opportunty to exchange experences n evdence-based polcy-makng, and to dentfy ways to promote the more systematc use of equty research n health polcy and acton.
Health mnstres and other stakeholders engaged n the evdence-to-polcy process n the Western Pacific Region were invited to submit case studies that illustrate how the process has worked n ther country. Nne cases were presented, representng experences n eght countres.
In the discussions that followed, participants identified key factors that can strengthen the capacty for evdence-based polcy-makng and develop a culture of nvestng n and actng upon nformaton and evdence. The meetng also provded useful deas for promotng evdence- based debate, analyss and polcy development for health equty over the longer term.
Ths book comples all nne cases. Ths ntroductory chapter comprses a synthess of the cases and the lessons learned from them. Comparng and contrastng the dfferent cases, it identifies key factors that stimulate policy-relevant research and better use of evidence for polcy-makng. These lessons can be especally valuable for developng countres, where resources are lmted and better understandng of the trade-offs mpled by alternatve polcy choces s needed.
Equity and barriers to access to health services
Overall gans n health can occur amd persstent, and even wdenng, nequaltes between soco-economc groups and areas. To narrow health gaps under a condton of resource constrants requres mprovng the health of the poorest and dong so at a rate that outstrps that of the wder populaton. Effectve pro-equty health polcy can acheve absolute and relatve mprovement n the health of the poorest groups, and n the factors that determne ther access to health care and ther exposure to rsk factors.
It s now well understood that poor, vulnerable and socally excluded groups have a hgher burden of dsease, whle at the same tme havng worse access to and lower utlzaton of health servces, a phenomenon descrbed by the “nverse care law”.[] Evdence from the Western Pacific region presented at the High-Level Meeting confirmed that households in the lowest ncome quntle, and those n rural areas, use fewer health servces than those n higher income quintiles or in urban areas. These groups face financial, geographical, and socio- cultural barrers to equtable access to health servces. Health systems can attempt to reduce these obstacles, but often fal to respond adequately.
The tools polcy-makers have at ther dsposal for reducng these access barrers nclude directing public subsidies to poor and excluded groups, either spatially or by targeting specific health condtons, makng greater use of exstng prvate sector provders, and montorng health system performance. However, public subsidies have been shown to actually benefit the better-off more than the poor, unless there s an explct pro-poor focus, accompaned by effectve regulaton of the prvate sector. Polcy decsons should be based on approprate
“local” evdence n order to establsh a pro-poor balance, among sub-sectors, such as between hosptal and prmary care, between urban and rural health servces, and between basc care that helps many and costly specialized care that benefits a few. All of the case studies presented here entaled approaches desgned to target and reduce these barrers, and most showed through montorng and evaluaton steps that they were pro-poor.
Financial barriers are especially significant where health systems rely on user fees at the tme of servce and where rsk-poolng and pre-payment schemes have not been establshed.
Not only do the drect costs of health care deter many sck people from seekng care, but other costs such as transport and food are mportant barrers to access, as are opportunty costs (e.g., income lost while seeking care or assisting a family member.) The Asia Pacific region has the hghest percentage of out-of-pocket payment for health servces of any of the WHO regons, and the reducton of ths barrer s a regonal prorty for WHO. Three of the cases descrbe different approaches to targeting financial assistance to poorer households.
Geographcal barrers result from the concentraton of health facltes and health workers n urban areas and n areas wth adequate transport. The poor and dsadvantaged tend to lve n the least-served areas, whch also suffer the worst envronmental condtons. Health workers generally prefer urban, hgher ncome settngs where there are more opportuntes for them and ther famles. The remoteness of some areas remans a barrer even n some better-off countres, and the cost of reachng such populatons wth adequate care can be qute hgh.
Soco-cultural barrers may be based on socal status or a consequence of the poverty and powerlessness assocated wth lower status, ncludng ethncty, gender and other socal factors.
Mnorty groups often have language and cultural dfferences from the majorty populaton, ncludng health workers, whch consttutes a barrer to seekng or recevng adequate care.
Gender s frequently a determnant of access to care n many settngs, wth women and grls receiving insufficient or delayed services more often than men or boys.
Poor responsveness of the health system can be a problem, even where health facltes are avalable. Compared to complants by those from better-off localtes, complants by the poor and dsadvantaged are less lkely to be heard and acted on. Common ssues nclude nconvenent locaton of health facltes or workng hours, rude or abusve health workers, more frequent problems n mantanng adequate stocks of medcnes, and mssng or malfunctonng equpment. Several of the cases presented here show how more resources were focused on specific diseases or environmental factors that especially affect the poor.
Several others ncreased or montored the resources allocated to geographcal areas havng a hgh percentage of poor or ethnc mnorty households.
Country examples of how evidence was used to stimulate pro-poor health policy-making
A major goal of polcy research s to acheve a systematc understandng of what, when, and how research should feed nto the development of polcy. Such an understandng of how research can contrbute to pro-poor polces, and systems to put t nto practce, could mprove health as well as development outcomes.
Most governments in the Western Pacific Region already have explicit pro-poor health polces n place, often as part of poverty reducton strateges, strateges for meetng the Mllennum Development Goals, or n statements of natonal equty prncples. Consequently, none of the case studes n ths collecton attempt to provde more evdence for the need to have pro-poor and pro-equty polces. Rather, they all focus on crucal detals of why and how research was done and how the evdence was translated nto polces and acton.
Case studes are tradtonally assocated wth busness school, law school and socal scence, but can be used n any dscplne to explore how ssues and prncples nteract n real-world stuatons, and ncreasngly are beng used to study development ssues. Case studes can be a useful learnng tool from whch to draw lessons for adaptaton and use n other contexts.
The countres represented n these case studes cover the spectrum of economc development, from the two wth lowest per capta ncomes, to the thrd wealthest country in the Western Pacific Region, and the other five countries filling in the low-middle to high- mddle ncome ranges. Ther levels of health sector development and polcy-makng cover a similar wide range. All the cases identify and show evidence about specific equity issues.
Some studes go nto somewhat less detal about how the evdence was presented to polcy- makers and how the policy process was influenced by the evidence, but enough is given to enable comparsons.
Gudelnes for case study preparaton (see Annex ) were sent n advance to help ensure that the cases would be relevant to the health equty focus of the Hgh-Level Meetng and to the agenda of examnng the evdence-to-polcy process. Most of the cases were prepared and presented through collaboraton between a range of stakeholders, ncludng local and nternatonal researchers, polcy-makers, donors, and cvl socety organzatons. The nne country cases presented at the Hgh-Level Meetng (by meetng sesson) and compled n this collection are:
Session 1: Health care financing
• Health financing strategies to improve access to health services for the poor in Cambodia: from pilot to policy and action—a case study of Health Equity Funds. Pro-poor health financing strategies developed n Camboda n the 990s led from user fees to contractng. The case examnes steps taken to improve equity in health financing that culminated in the wide use of health equty funds (HEF).
• Research, intervention design and policy implementation of the New Rural Cooperative Medical Scheme in Shandong, China. Ths paper descrbes research and polcy-makng about adjustng the premium, subsidy, and benefit package at county level for the national insurance system for rural households, the New Cooperatve Medcal Servces (NCMS) scheme.
• Health Care Fund for the Poor in Viet Nam: how evidence and politics came together. Ths paper s an overvew of the evoluton of the natonal health nsurance scheme and examnes how an effective and efficient way of subsidizing membership for the poor was devised, evaluated, and became official policy.
Session 2: Primary health care
• Scaling up primary health care in the Lao People’s Democratic Republic using evidence from a long- term primary health care development project. Ths s a case study of a communty-based, non-governmental organzaton (NGO)-sponsored project and how the results were dssemnated and partally replcated. It ncludes an analyss of the roles of large donors and NGOs n polcy decsons.
• Promoting health equity: evidence, policy and action—the New Zealand experience. Ths study descrbed research into health problems of the Māori population and how they were effectively targeted by nterventons to mprove communty-based prmary health care servces and housng condtons.
Session 3: Communicable diseases
• The development and targeting of malaria control interventions for populations in high transmission areas of Cambodia: the influence of research on policy and practice. This paper describes how field surveys resulted n operatonal changes n how malara treatment s provded, and how nsectcde-treated bednets are subsdzed and targeted.
• Public-Private Mix DOTS: a strategy to engage all health care providers in tuberculosis control and significantly increase access to DOTS services in the Philippines. Ths paper dscusses why and how prvate provders were ncorporated nto the natonal tuberculoss control programme, based on research that suggested the publc sector alone could not acheve targets.
Session 4: Health systems
• Geographic equity in distribution of scarce dialysis resources in Malaysia. Ths research compares the provncal concentraton of publc, chartable, and prvate resources and determnes that publc funds subsdze the poor more than the well-off.
• Promoting health equity through capacity building of primary health care workers in Mongolia. Ths case documents how external resources were not used effectvely to support rural health servces, whch led to a change n strategy for n-servce tranng of md-level staff by developng a local fellowshp programme.
How evidence was used for policy-making in the cases
Previous research has identified factors that encourage good policy-making in terms of the relevance, effectiveness, efficiency, and timeliness of policies. The country cases in ths collecton llustrate several dfferent types of polcy mpact. The small sample and many varables do not support a rgorous analyss, but factors n each case can be hghlghted wth the degree of uptake of evdence used as a measure of mpact.
Overall, research evdence was used to make a concrete change n polcy at some level n seven of the nne cases. However, varatons are observed across countres. The use of evdence ranges from rapd adopton as natonal polcy, to changes n how exstng polcy was mplemented, to very slow or no uptake. In Malaysa, for example, polcy change dd not result, because t was not an ntended outcome of the research.
In Viet Nam, evidence was used to support a new central government health financing polcy and gude development of a programme that establshed a system of subsdzed health nsurance cards for the poor. The polcy decson does not rgdly spell out detals, leavng some to be resolved by future research and plannng.
In the Phlppnes, the evdence was used to show a need for a publc-prvate partnershp strategy, and later that the plots of that approach were effectve and should be replcated.
In Chna, evdence convnced polcy-makers to allow the use of Medcal Assstance Funds to pay NCMS nsurance premums for the poor. Most recommendatons were accepted and implemented at local level. In Cambodia, evidence that health equity funds were efficient and equtable resulted n acceptance and replcaton, but although HEFs are recognzed as a tool for achieving equity and poverty reduction, they have not become an official financing policy.
Several cases show how evdence led to polcy changes amng to mprove the mplementaton of exstng programmes. In Chna, the evdence nformed the recommendatons to revse NCMS subsidies and to add benefits at the county level. In Cambodia evidence resulted in a decson to extend the geographcal elgblty for free bednets and to rely on vllage workers to dstrbute antmalaral drugs n the least accessble areas, leavng the socal marketng programme to focus on other endemc areas. In Mongola, the evdence supported a shft n tranng polcy from overseas fellowshps to local fellowshps.
In several cases, research was used to verfy that an nterventon was pro-poor. In Malaysa, for example, the geographcal dstrbuton of dalyss facltes was montored over tme, confirming that there was a relative increase in the use of government facilities. Research on equity funds in Cambodia verified that subsidies were targeted accurately and efficiently.
The uptake of health care cards was montored n Vet Nam n order to dentfy operatonal problems.
In some cases, the evdence was not mmedately adopted n polcy or acton. In the Lao People’s Democratic Republic, the rural project’s community-based strategy was not replicated for several years, partly because evidence of the project’s success was not considered to be strong, and because polcy-makng favoured large-scale, top-down health development projects.
In Chna the recommendaton to ncrease ndvdual premums was not mmedately acted on n two of the three countes because local managers were reluctant to ncrease premums, whch they feared would result n lower NCMS enrolment. In the Phlppnes, several elements of the Comprehensive and Unified Policy for TB Control were not implemented, for lack of financing and support.
Contemporary models of the research-to-policy process
Polcy-makng s not lnear and s rarely a unque or explct set of decsons, but evolves through multple nteractons and dfferent sources of knowledge, usually engagng a range of actors and stakeholders. Especally n the arena of health polcy and equty, t s a multdmensonal process that can ental teratons n the form of populaton surveys, plot projects, evaluatons, welfare analyss, and other complementary studes. There may be no clear dstncton between researchers and polcy-makers, often because of the small pool of
concerned stakeholders and lmted resources. Fnally, producton of knowledge s not always limited to a set of specific findings, because of time lags in research and new policy issues beng rased through research n other countres.
Reflecting this understanding of policy-making, the relationship between research and polcy s also no longer thought of as a lnear or purely ratonal process n whch research findings are shifted from the ‘research sphere’ over to the ‘policy sphere’, where they then have some impact on policy-makers’ decisions.
The “Polcy Wheel” (Fgure ) as developed by Dr. Don Matheson and colleagues n New Zealand depcts the teratve and contnuous nature of ths process. It serves as a gude for each of the polcy development stages, from problem assessment to polcy change and mplementaton.
The crucial question engaging policy researchers is: why are some ideas that circulate in the research–polcy arena pcked up and acted on, whle others are gnored and dsappear? Several useful frameworks or models have been used to evaluate the process through whch research leads to polcy and acton. None of these has as yet attempted to be predctve n the sense that weghts can be gven to factors and the probablty of polcy uptake then calculated. It can be reasonably sad that all plausble models are currently of equal precson and utlty, varyng chiefly in the definition of the factors rather than on their relative importance, which are not yet known emprcally.
These models have helped dentfy crtcal factors assocated wth effectve use of hgh- qualty evdence. Most are based on experence from ndustralzed countres, however, and the dversty of cultural, economc, and poltcal contexts n the less developed countres makes it difficult to draw valid generalizations and lessons from them. In addition, international
Figure 1: The Policy Wheel
Develop optons
Gather nformaton Consult Assess the problem Introducton
8
7
6
5
4
3
A good place to start
Montor, revew and evaluate
Implement
Formulate advce
Assess the optons
actors continue to exert much influence on health research and policy processes in developing countres. Nevertheless, some observatons on the research-to-polcy process can be made.
Most models and frameworks propose common factors which determine the influence of evidence on policy: 1) the context in which the policy and evidence are situated; 2) the content, timeliness, and credibility of the evidence; and 3) actors and the interaction between them, to which can be added the effect of external influences. These frameworks can illuminate significant aspects of the case studies presented in this book. Some models discussed at the High-Level Meeting are summarized here:
. The RAPID Framework[] (Figure 2) is based on a comparative analysis of 50 cases on the research-to-policy process in the international development field. Research uptake is seen as a function of the interaction of context, evidence, and links. External influences also bear on all of these.
. A varaton on the above model, the “4K framework”[3] was used by researchers n one of the case studies presented here (“Health financing strategies to improve access to health services for the poor in Cambodia: from pilot to policy and action—A case study of Health Equity Funds”). The 4K framework describes four stages: Stage 1 of exploitation of existing knowledge; Stage 2 of creating of new knowledge or innovation; Stage 3 of disseminating the new knowledge or evidence brokering; and Stage 4 of adopting and usng the new knowledge. Each stage was analysed by determnants of context, content and actors, smlar to those of the RAPID model.
3. A framework by Lavs et al.[4] overlaps with the RAPID framework’s context and linkages or the 4K framework’s context and actors. Four alternative models of the demand for and use of research evdence are descrbed, as stemmng from ) a “push” by research producers or purveyors, ) a “pull” by research users, 3) an “exchange” between sngle groups of research producers and users, or 4) “ntegrated” efforts that nvolve such knowledge translaton platforms as the Evdence-Informed Polcy Network (EVIPNet).
The Evidence, credblty, methods, relevance, use, how the message
s packaged and communcated etc.
Links between polcymakers and other stakeholders, relatonshps, voce
trust, networks, the meda and other
ntermedares etc.
Political The Context - poltcal structures / processes, nsttutonal pressures, prevalng concepts, polcy streams and wndows etc.
External Influences Internatonal factors, economc and cultural influences, etc.
Figure 2: The RAPID Framework: Context, Evidence and Links
Source: Court and Young 2002.
Lessons from the processes described in country case studies
Only a few of the cases here llustrate the polcy process n ts entrety. Instead, most focus on one or more phases or stages of the overall process. The Chna case study descrbes a farly thorough but narrowly focused assessment of problems wth the NCMS, but dd not dscuss wider issues of health equity and financing. The Cambodia and Viet Nam financing reforms were done in the context of fairly comprehensive evaluations of equity and financing, but these were the background to the HEF and Health Care Fund for the Poor (HCFP) strateges and not drect parts of the polcy process. A more rgorous approach to problem assessment can be made, as shown n a recent descrpton[5] of equty polcy n South Afrca. Most of the cases also did not include in significant detail steps 4 and 5 of the policy wheel, in which options are developed and evaluated. The Cambodia financing study comes the closest but, in actuality, the health equity funds were introduced as a complement to other financing modes and did not replace the exstng modes. Most of the cases do descrbe the evaluaton and consultaton steps n the process farly thoroughly, n part because the nstructons for case preparaton stressed ths.
The nne cases n ths collecton are dverse but also have elements n common. The uptake of evdence by polcy-makers can be seen as the outcome of the research-to-polcy process, or as a dependent varable. Independent varables can be descrbed n terms of factors that map on to the context, evidence, links, and external influences, used n the RAPID model and others.
The case studes can be compared usng the followng expanded categores, whch are discussed below, and also summarized in Table 1 on page 20:
1. Context (including external influences) a. The reasons the research was carred out
b. The primary focus of the polic(ies) the case study influenced
c. Specific access barrier or equity issue and the type of equity analysis used in generating the evdence
. Evdence (or content)
a. The type of research done and the qualty of evdence produced
b. Whether the research evdence provded an operatonal soluton to a problem 3. Links and actors (including external influences)
a. The types of collaboraton between researchers and polcy-makers b. Stakeholder nteractons that affect the uptake of research for polcy.
1. Context
Context covers a number of mportant enablng factors[6] that affect the degree to whch research has an mpact on polcy. It ncludes health and equty status, prevalng opnons (local and from other countres), exstng polcy commtments, related dscourse among polcy- makers, and the extent of demand for new deas by polcy-makers and socety generally (“pull”
factors). The exstence and nature of pror research on a gven polcy ssue s another mportant context factor. These contextual influences are summarized in Table 1 on page 20.
Context (a): Why was the research done?
Because they were delberately selected as such, there was a commtment to mprovng equty n all the countres and case studes. Indvdual research goals may have been determned by socal and poltcal envronment, the exstng health system stuaton and capacty, and the demand for polcy change. The latter factor s relevant to the push /pull/ exchange/ integrated model of Lavs et al., and also to the lnkages between stakeholders.
In all of the cases, prior evidence of the lack of equity from pre-existing research was an mportant stimulus for finding effective interventions. The processes that were described in Cambodia and Vet Nam made extensve use of studes that had found evdence of unequal provson and uptake of publc subsdes and dfferentals n health status. Such evdence also exsted n Chna.
The full cost of kdney dalyss treatment was known to be unaffordable for many Malaysans, and the disadvantaged status of Māori in New Zealand had been thoroughly documented. Rural Mongolans were known to be dsadvantaged n terms of dsease ncdence and avalablty of health services, as were people in a remote province in the Lao People’s Democratic Republic.
It was also found that tuberculoss sufferers n the Phlppnes are much poorer than average, as are malara vctms n hyperendemc areas of Camboda.
Commitment to improving equity and the current policies that affect it was another man reason for the research descrbed n several case studes. Evdence of worsenng health condtons and international scrutiny of China’s health system led to more attention by the central government focused on equty ssues and especally rural health care. The low rembursement rates were based on poltcal consderatons because leaders needed to be seen as dong somethng about health, and a small benefit package is easily delivered to a large population while keeping the premum low. The goal of the research was to mprove the NCMS n terms of equty, qualty, and efficiency using evidence-based interventions. The China case is an example of research beng pulled by demand (by natonal prortes and from a mult-country polcy research project), and t also nvolved consderable exchange between researchers and polcy-makers.
The Camboda study of health equty funds descrbed how research and plot project evaluations established HEFs as an accepted health financing strategy. Evidence showing that adequate publc health servces could not be provded free of charge for all populaton groups prompted the Ministry of Health to identify and test alternative financing strategies, including user fees wth exemptons, contractng, health equty funds and communty-based health nsurance. The polcy process reled heavly on research on nterventons that took place over nearly a decade.
The Viet Nam study documents the evolution of pro-poor national health financing policies n response to deteroraton n equty followng the transton to a free market system. User fees were ntroduced n 989 n response to a shortfall n fundng for the health system. An overall pro-poor polcy and goal of unversal nsurance coverage had been stated clearly n polcy documents, and a preventon-orented health system was functonal. There were mechansms n place for nteractng wth the poor locally, and the Vet Nam Health Insurance Agency was n favor of expandng coverage by means of subsdes. In both Vet Nam and Camboda, the need to attack the causes of poverty and the strong support of donors facltated the research.
In Mongola, rural-urban mgraton deprved the rural areas of amentes as well as traned human resources. Evaluatons of the exstng tranng strategy found that t was not meetng the needs of rural areas, and identified potential benefits of a local fellowship programme.
Ths strategy was tested and accepted as a component of health human resource polcy.
To monitor implementation of existing policy: A secondary purpose of the research n Vet Nam was to monitor poverty reduction effects of the evolving health financing policy and whether the health cards were targeted accurately. The research descrbed n the Malaysa case study was ntended to verfy that the exstng polcy of pro-poor and pro-rural publc health provson was beng followed n practce. The Phlppnes PPMD project, Camboda malara nterventons, and Camboda HEF development reled heavly on evaluatons of plot projects to gude polcy development or change. The Mongola case study descrbes a prelmnary assessment of the new local fellowshp tranng program.
To provide support and guidance for disease control programmes was the prncpal am of the Cambodia malaria case study, which was specifically aimed at improving the efficiency and effectveness of preventve and curatve nterventons mplemented under the Natonal Malara Programme. The Phlppnes PPMD study focused on combatng tuberculoss by broadenng the provder base for the exstng TB control programme.
To inform or refine donor assistance was another secondary reason for several studes. The Cambodia health financing study showed that Health Equity Funds are an effective and efficient way to target donor funding at the poor. The Mongolia study demonstrated that the shft n WHO fundng to local fellowshps was effectve n reachng a larger number of rural workers, should be expanded, and also has attracted other potental donors. The Lao People’s Democratic Republic study described how evidence was intended to encourage the Government and donors to fund the expanson of the communty-orented prmary health care strategy. The Camboda malara case study and the Phlppnes PPMD studes were used to justfy contnued fundng from the Global Fund and other donors.
Context (b): What specific equity objectives were addressed in the studies?
Poverty reducton as a natonal polcy goal was the focus of several country cases. The Chna NCMS case nvolved a key element of the natonal poverty reducton strategy. As n Mongolia and Viet Nam, China’s economic transition has resulted in rural areas becoming relatvely dsadvantaged. Implementaton of the NCMS s decentralzed and the research and policy work was specific to six counties in two provinces, but it is likely that national policy will be revsed on the bass of these studes.
The Viet Nam case describes the steps that led to an official decision to subsidize social health nsurance as part of the natonal poverty reducton strategy. Ths work was carred out on the natonal level, wth attenton gven to dfferences between provnces n ther ablty to finance the scheme.
Equty and poverty were ntensvely researched n Camboda, and the health sector plays an mportant role n the poverty reducton strategy. The research descrbed n the case study confirmed that health equity funds can reduce out-of-pocket expenditures in most situations.
HEFs are mentoned as a possble poverty reducton strategy n natonal plans, and government funds have been allocated to HEFs, but they do not yet have formal status n natonal polcy.
Reducng urban-rural nequalty as a natonal polcy goal was mplct n the Chna case, since the NCMS is a national programme to benefit rural households. The rural population was found to spend a hgher proporton of ncomes on health and receve less publc subsdy than urban resdents. The NCMS s ntended to both assure access to all rural resdents va near- unversal nsurance coverage, and to revve the publc health care n rural areas by channelng provncal and local subsdes to publc health facltes.
Inequality is addressed in the Malaysia case through research confirming that the national pro-rural and pro-poor polcy of subsdzng health servces was beng mantaned. The methodology used measured geographcal equty of dalyss as a proxy for ncome or vertcal equty.
The Mongola case was a response to a shortage of rural health workers, where evdence generated guded the desgn of a cost-effectve programme of local long-term tranng. The new local tranng strategy s mplemented at the natonal level.
The primary health care development programme described in the Lao People’s Democratic Republc case mproved access to health servces n a remote rural provnce. The project was talored to needs of local people, but many of the same condtons are found n other rural areas. The natonal health strategy prortzes the development of prmary health care n rural areas, but the case showed that communty-based approaches are not always consdered to be replcable.
Reducing morbidity from specific poverty-related diseases and achieving MDG targets was the goal of the Phlppnes PPMD project, launched n order to mprove TB case detecton rates. Smlarly, the Camboda malara studes came about n response to hyperendemcty and hgh mortalty from malara n several areas of the country. Malara ncdence n the three lowest ncome quntles was several tmes more than n the two hghest ones.
Engaging and monitoring the non-state (private and NGO) sector was another polcy goal of the Phlppnes PPMD Drectly-Observed Treatment Short-Course (DOTS) strategy. The prvate sector was found to be an mportant TB servce provder, but the qualty of treatment was often low. Partcpaton of prvate provders opened ways for local partnershps, NGOs and corporatons to become drectly nvolved n TB control, and also local governments as the payers n the devolved health system.
The am of the Malaysa case was to determne f the rapd growth of prvate and chartable dalyss servces was affectng the overall pro-poor polcy n health servce provson. The prvate sector was found to concentrate dalyss centres n economcally developed states where patents could afford to pay. Whle the publcly provded share of dalyss has decreased, the
publc sector stll focuses on the poorer states but the NGO sector surprsngly parallels the prvate sector n locatng dalyss centres.
The Lao People’s Democratic Republic case illustrates collaboration between an NGO and the provncal health department n the development of a communty-based prmary health care programme, but the Mnstry of Health gves more prorty to large-scale, top-down, donor-funded health development projects. In contrast, the New Zealand case descrbes polcy changes that utlzed the effectve role of Prmary Health Organzatons, non-governmental units that specifically focus on the health needs of local communities.
Context (c): Which access barriers were addressed in the country cases?
Financial barriers to access to health services were the focus of five country cases. Despite the contnuous development of pro-poor health polces n Vet Nam, ncludng mandated exemptons and subsdzed health nsurance, prvate expendture and out-of-pocket payments remaned hgh. Lmted fundng for subsdes and lack of strong commtment from local governments compounded the problem of unaffordablty of publc servces. In response, the government ntroduced the Health Care Fund for the Poor, whch provdes health nsurance subsidies financed from the central budget.
The Camboda Health Equty Funds case addressed the problem of catastrophc payments directly by introducing local financing schemes that cover user fees at hospitals for the poor. A demographc survey n 999 found that the poor had much lower use of health servces than the rch, suffered more llness, and spent a much hgher proporton of household ncome on health care.
The Chna study found that hgh hosptal co-payments were a barrer to utlzaton, but also that most non-poor people could afford to pay more than the current NCMS premum.
Evdence from a baselne household survey guded recommendatons for adjustng NCMS premium and benefits and for using a separate fund to subsidize insurance premiums for the poor.
In the Phlppnes, TB s largely a dsease of the poor and TB drugs are expensve. The case descrbes a strategy for usng prvate provders to ncrease coverage of free DOTS dagnoss and treatment. In Malaysa, free publc provson of kdney dalyss was ntroduced, to overcome the potentially catastrophic costs of this treatment. The case verifies that public sector provson s stll pro-poor.
Overcoming geographical barriers to access was a major focus of the Lao People’s Democratic Republc case, where a hghly decentralzed approach succeeded n provdng servces to a remote mountanous area. In Camboda, malara s endemc n remote forested areas, and baselne research was used to develop a programme of vllage malara workers to dagnose and treat suspected cases n these areas. Other evdence led to expanded free bednet dstrbuton n those areas. An assessment n the Phlppnes showed that DOTS coverage was nadequate n some of the largest urban areas. Establshment of prvate sector DOTS unts n underserved areas helped to ncrease the natonal case detecton rate by 8%.
Barriers caused by social exclusion are hghlghted by the New Zealand case, whch descrbes how the Māori population was shown to have worse health outcomes than other groups, largely due to dscrmnaton and excluson. A new prmary health care programme was desgned to overcome these barrers, complemented by a targeted, multsectoral, housng mprovement programme. The needs of specific social groups were also addressed by the Cambodia malara nterventons snce the hyperendemc forested areas are largely populated by ethnc mnortes.
Socal excluson s closely ted to the problem of health system unresponsiveness. The prmary health care strategy described in the Lao People’s Democratic Republic case included addressng the needs of ethnc mnortes usng health workers who spoke ther languages and were senstve to ther belefs and customs, and provded outreach servces to communtes with difficult access to fixed services. Inability to stem the outflow of rural health workers in Mongola was one of the reasons for persstent hgh nfant and chld mortalty and morbdty rates. The failure of the official exemption systems that led to reforms in Cambodia and Viet Nam s another example of how health systems do not respond to the needs of the poor.
2. Evidence and communication
Analyss of standards n health polcy research, especally that related to equty, s at an early stage of development. Varous “herarchy of evdence” systems have been descrbed.
Fgure 3 shows one such typology of research evdence. In general, hgher-level analyss and systematc revews are consdered superor forms of evdence than case studes.[7]
Figure 3: Hierarchy of evidence: ranking of research evidence evaluating health care interventions
Effectiveness Appropriateness Feasibility
Excellent • Systematc Revew
• Mult-centre studes • Systematc Revew
• Mult-centre studes • Systematc Revew
• Mult-centre studes Good • RCT
• Observatonal studes • RCT
• Observatonal studes
• Interpretatve studes
• RCT
• Observatonal studes
• Interpretatve studes Fair • Uncontrolled trals wth
dramatc results
• Before and after studes
• Non-randomzed controlled trals
• Descrptve studes
• Focus groups • Descrptve studes
• Acton research
• Before and after studes
• Focus groups Poor • Descrptve studes
• Case studes
• Expert opnon
• Studes of poor methodologcal qualty
• Expert opnon
• Case studes
• Studes of poor methodologcal qualty
• Expert opnon
• Case studes
• Studes of poor methodologcal qualty Source: Evans D. 2003.
However, classifications such as the above and others[8] are not very relevant to equty polcy development and publc health nterventons, where tools such as double-blnd trals are not generally feasble. Case studes are usually more common than most of the other, stronger forms of evdence. Much can be learned from well done case studes, and especally from revews based on several case studes n the context of a sngle country.
The influence of research on policy is much increased if it has topical relevance and operatonal usefulness. The most mportant factors enablng uptake of research evdence were found by Court and Young to be whether the evidence was presented effectively as a practcal soluton to a problem and was credble n terms of research approach. A “knowledge pyramd”[9] wth actonable messages at the apex s applcable to equty polcy.
Actonable messages should deally be based on entre bodes of research knowledge, not just ndvdual studes, and stakeholders should try to acheve consensus on the messages that are conveyed. An optmum strategy may be to focus knowledge-transfer efforts at the apex of the knowledge pyramd usng actonable messages, whle contnung to buld a sold knowledge base. The varous types of evdence generated n the nne cases, and ther operatonal usefulness, are summarzed n Table on page 0.
Evidence and communication (a): What kinds of research methods and evidence were used in the cases?
Varous types of evdence were generated by the research dscussed n the country cases. A range of methodologcal approaches was taken n response to the challenge of assessng equty ssues. There was also varaton n the ntensty of the research and the resources requred.
These n turn affected the overall tme frame of the research-to-polcy process.
Baseline studies were the foundaton of much of the evdence used n the Chna polcy development process. Townshps and countes were selected randomly for three research actvtes, a health survey of 3,339 households, an organzatonal analyss, and focus group discussions. A 2003 national survey also provided evidence of the high financial burden from hospital co-payments, finding that hospital costs were high and reimbursements low.
Baselne surveys also underpnned the nterventon trals n the case related to the Camboda malara programme, the Mongola rural human resource case, and the New Zealand prmary
Figure 4: Knowledge pyramid
Source: Lavis et. al 2006.
Actonable messages
Systematc revews of research Indvdual studes, artcles and reports Basc, theoretcal and methodologcal nnovatons
health care case. In the last, it was shown that Māori people have poorer access than other groups to health servces, poorer qualty of care wthn the health system and worse health outcomes for most dsease groups. In Vet Nam, equty was researched n depth through a partcpatory poverty assessment that found that user fees at hosptals mposed a hgh financial burden on poor households, and through two living standards measurement surveys that showed that the poor utlze servces less than the rch and capture a smaller share of public subsidies. The HCFP policy process was guided by several studies of the new financing polces, ncludng some based on large valdated survey datasets and done by reputed local and foregn researchers. The evdence and polcy recommendatons took place over a perod of 8 to 12 years, with evaluations at several stages of the development process. The finding of significantly higher utilization of public health services by insurance beneficiaries supported universal health insurance and identified shortcomings in existing financing arrangements, culminating in Decision 139 to increase insurance benefits, coverage, and central government fundng.
The Cambodia HEF case similarly used evidence from prior health financing interventions as well as several studes of at least twenty HEF plot projects. Evdence was consdered relable, and showed that HEFs helped reduce financial barriers to access to timely and needed care for poor patents, whle mantanng the ncome of publc health facltes. The data requred significant analysis and interpretation since it consisted of a large number of studies.
Concrete proposals for polcy changes or nterventons were derved from the research evdence n all of these cases. The evdence was consdered to be credble because of the professonal status of the researchers and the apparent valdty of the data.
Pilot interventions and their evaluations were used to generate credble evdence n some of the case studes. Ths evdence tends to be hghly focused and easy to understand, whch may explan why these types of analyses have resulted n polcy changes relatvely quckly. Ths was true of the plot vllage malara worker nterventons n Camboda, where an evaluaton demonstrated that t was cost-effectve enough to be scaled up. In contrast, the other element of the Camboda malara case, a malarometrc survey of sample communtes, dd not nvolve a plot, but by showng that malara rsk outsde the forest areas was hgher than had been prevously beleved, t led drectly to a change n the bednet dstrbuton coverage area and reallocaton of programme resources. In the Phlppnes case, analyss of plots at several stes showed that PPMD was effectve and recommended that t should be scaled up. Assessment of the Mongola Local Fellowshp Tranng Programme showed that the strategy was wdely accepted by health workers and other stakeholders, and was successful n that all local tranng fellows returned to ther place of work and some ntated new communty health programmes.
The Lao People’s Democratic Republic case was based on evaluations of each of the four phases of the project. Evaluations were built into the project and were carried out by qualified researchers, but n the vew of some polcy-makers the lack of an ntal baselne survey was a weakness. The replcablty of the project was also questoned by polcy-makers.
Evidence and communication (b): Was a practical solution to an operational problem recommended?
Operatonal recommendatons were not the only goal of the research n some cases.
Several types of polcy recommendaton came from the research evdence.
In China, specific recommendations on the need to increase premiums and benefits were communcated to county-level managers, as well as a proposal to use the Medcal Assstance Fund to pay NCMS premums for the poor. The Camboda malara studes focused on operatonal problems and offered detaled recommendatons for resolvng them. The Camboda health financing reforms and Health Equity Fund evaluations produced a concrete recommendation that the Mnstry of Economy and Fnance should allocate a budget for these decentralzed funds, whch ncluded basc polcy gudelnes for the HEF mechansm, but left some polcy ssues unresolved, awatng further research. Ths was true n New Zealand as well, where t was recommended that communty-run prmary health care provders be gven lattude to deal with local health issues. Another set of studies identified specific housing issues, and did make recommendatons to correct the observed problems.
In the Phlppnes case, snce the strategy of nvolvng the prvate sector was developed and piloted first and then evaluated, the analysis resulted only in a recommendation that the approach should be officially adopted by the Department of Health. Similarly in Mongolia, the strategy of local fellowshp tranng was developed ndependently, wth subsequent research verfyng that t was effectve. Ths bascally happened n the case of health cards for the poor n Vet Nam as well. Early research found a need to assst the poor, and several approaches were successively implemented and evaluated, leading to refinements and the eventual development of the HCFP polcy.
The Malaysa case study was a one-off evaluaton made after a long process of development of dalyss servces, wth the actual research completed n a short tme. It resulted n no new recommendatons, other than contnung the exstng polcy because t was found that the government provson was stll pro-poor even as the prvate and NGO sectors expanded ther servces.
The Lao People’s Democratic Republic case study focuses on the recommendation that the prmary health care approach and strategy tred n Sayaboury provnce should be scaled up. Detals were made avalable about the successful project elements but the project was not replcated as the NGO had recommended, n part because some of the evdence was felt to be unconvncng, but also because t was thought that the presence of a hghly sklled expatrate advser was essental to ts success and could not be replcated.
3. Links between stakeholders and the role of external influences
The nature of collaboraton between researchers and polcy-makers s consdered to be a key determnant of the success of the overall evdence-to-polcy process. Ths ncludes the qualty of the lnks and feedback processes nherent to the process, some of whch are descrbed n the cases. Issues of trust, legtmacy, openness and formalzaton of networks are understood
to be mportant, as s the role of “translaton” of techncal content nto smple language and its effective communication. Often, intermediary organizations and networks influence formal policy guidance documents, which, in turn, influence officials. The RAPID framework emphasizes the mportance of lnks through communtes, networks and ntermedares (e.g. the meda and lobbyng groups) n effectng polcy change. The lnkage model of Lavs et al. hghlghts advantages of the “exchange” and “networked” processes.
Polcy uptake has been found to be greatest f the research programme has a clear communcaton strategy from the start, and f the results are packaged n famlar concepts. It is often difficult to convince policy-makers of the value of more theoretical research if it is not clearly lnked to polcy applcatons. The sources and conveyors of nformaton, the way new messages are packaged (e.g., couched n famlar or unfamlar terms) and targeted can affect how polcy documents are perceved and utlzed. Contnuous nteracton leads to greater chances of successful communcaton than a smple or lnear approach.
The cases llustrate varous types of collaboraton arrangements between stakeholders, as well as several varatons n the way stakeholders collaborated. The detals of who dd the research, who funded the research, and whch stakeholders were nvolved n dssemnatng the evdence and generatng polcy recommendatons, are shown n Table . In addton to the researchers, the stakeholders included central and local health and finance authorities, NGOs that were concerned with specific health issues and/or communities, and donors or donor consortia. The researchers were from academc nsttutons, Government (executve and research branches), and consultants employed by donors.
Resistance to policy change and how to overcome it
Despte the exstence of clear evdence, efforts to mprove equty often challenge structures of socal and economc power, leadng to poltcal resstance that can mpede change. Bureaucratc factors, nsttutonal pressures, and vested nterests can also dstort polces durng mplementaton.
The exstence of a degree of openness and cvl and poltcal freedoms can sometmes effectvely counter these other contextual factors. In several of the cases, the advantages of close collaboraton were demonstrated clearly.
Stakeholders from the polcy arena n the cases dscussed here manly nclude publc health and finance officials, and to a lesser extent, civil society through NGOs. There is little mention of mass publc support or opposton to polces. In these cases, resstance to polcy change comes from specific vested interests or bureaucratic factors. Overcoming such opposition requires effectve communcaton between researchers, polcy-makers and other stakeholders. Two cases that describe positive policy outcomes—China and the Philippines— emphasize the importance of personal and group communcaton strateges. Other successful processes, such as n the Vet Nam and Camboda HEF cases, rely more on formal polcy dalogues between natonal counterparts and donors, and, n the New Zealand case, wth cvl socety stakeholders.
In Chna, researchers accorded top prorty to mantanng a formal relatonshp wth polcy- makers through the project, whch facltated good communcaton. All the stakeholders were ncluded n desgnng the study. Polcy-makers from several government departments were nvolved
n desgnng the nterventon packages and shared the responsblty for ther mplementaton.
Local-level polcy-makers and managers dscussed and evaluated the results n publc meetngs.
Researchers provded contnung support to the polcy-makers durng the post-research phases.
There was relatively little direct influence by donors; the project was monitored by WHO and other donors, and EVIPNet also partcpated. There was some resstance from polcy-makers and managers, whose performance ratngs could be affected f enrollments decreased due to hgher ndvdual premums. On the other hand, there was lttle opposton to rasng publc (county) subsdes to NCMS, or to usng a Medcal Assstance Fund that s under dfferent management and budget from the NCMS to subsdze premums for the poor. In addton, the central government announced an ncrease n NCMS provncal subsdes.
In the Phlppnes case, there was ntal opposton from the publc health establshment to nvolvng the prvate sector n TB control, but the successful plot projects and a collegal research and polcy atmosphere helped overcome ths. The research was carred out by the Department of Health and WHO n collaboraton wth the major project mplementer (PhlCAT). Personal contacts enhanced the exchange of nformaton and polcy dscussons, and the lnes between researchers, mplementers, and polcy-makers were not consstently fixed. The financial participation of the stakeholder PhilHealth to provide private physician rembursement for TB outpatent treatment was also crtcal to the success of the PPMD.
Internatonal partners and donors and donor fundng have a specal role both n stmulatng research and enablng research to have an mpact on polcy. For example, WHO techncal nputs, and broad ncentves such the Global Fund, the Mllennum Development Goals (MDGs), and the poverty reducton strategy paper (PRSP) process, have had substantal mpacts.
In Vet Nam, government at all levels was concerned about the affordablty of nsurance premums by the poor, but the Mnstry of Fnance was reluctant to ncrease health spendng by subsdzng the poor through the nsurance scheme. Mnstry of Health polcy researchers used evdence from World Bank, Asan Development Bank, and WHO studes to formulate recommendatons, wth close contact between all stakeholders mantaned through the process.
In addton, research was funded and guded by the major donors n several cases. The Camboda malara case was based on donor-funded research, as was the development of the local fellowshp scheme n Mongola. In the Mongola case, there was lttle opposton, perhaps because continuing the overseas fellowships would have benefited relatively few people. The Sector-Wde Approach used n Camboda provded an exchange envronment, to use the termnology of the model developed by Lavs et al., n whch donors to the health sector were able to coordinate their inputs into equity research and pilot financing reforms.
Evaluatons by credble researchers and organzatons were wdely dssemnated. Formal and nformal meetngs, workshops, conferences and study tours helped to nform polcy-makers.
There was lttle opposton to the expanson because the HEFs are seen as complementary to contracting, currently the most important financing scheme. Being locally generated, HEFs enjoyed ownershp among health polcy-makers and were easy to understand and mplement, wthout threatenng the nterests of any stakeholder. Communty-based nsurance are seen as havng some advantages, but have had relatvely lttle support.
Table 1: An overview of the nine country cases
Case Context Evidence and its
dissemination Linkages and external
influences
Policy outcome Health financing
strateges to mprove access to health servces for the poor in Cambodia: from plot to polcy and action—A case study of Health Equty Funds
Strong
government equty commitment;
previous financing reforms, extensve exstng research, many HEF plots.
Many plots were evaluated;
HEFs are shown to be efficient.
Operatonally useful findings.
Sector-wde approach, avalablty of donor and NGO fundng for research and plots.
Mnstry of Health and academc researchers nvolved.
HEFs have been wdely adopted but are not yet official policy.
Research, nterventon desgn and polcy mplementaton of the New Rural Cooperatve Medcal Scheme n Shandong, Chna
Need to mprove NCMS; focus on rural areas;
multnatonal research project.
Evdence
consdered sound.
Many meetngs and workshops to dscuss and develop feasble nterventons.
Partcpaton of EVIPNet and support from key donors.
Plot nterventons evaluated.
Premum, benefit, Medical Assstance Fund polces were changed.
Health Care Fund for the Poor n Vet Nam: how evidence and poltcs came together
Strong equty commitment;
previous financing reforms and evaluatons, Mnstry of Fnance supports ncreasng subsdy, functonng PHC and identification of poor.
Qualty research over several years in health financing.
Contnuous polcy dalogue on health nsurance.
Avalablty of donor fundng for research.
Mnstry of Health and academc researchers nvolved. Mnstry of Fnance got good evdence to increase financing.
Decson 39 was enacted at Prme Mnsteral level. Beneficiary identification is beng steadly mproved.
Scalng up prmary health care n the Lao People’s Democratc Republc usng evdence from a long-term prmary health care development project
Weak emphass on PHC n Master Plan; large “top- down” projects receve Mnstry of Health attention;
NGOs have lttle influence.
Some evdence was consdered weak. Relance on documents to dssemnate evdence.
Major project donor support stopped.
Replcaton consdered difficult.
There was acceptance n prncple but dssemnaton has been slow
In New Zealand and Malaysa, research unts wthn the Mnstry of Health played a major role n the entre process In the New Zealand case, a research unt n the Mnstry of Health developed the evidence used in the primary health care intervention for Māori health, and an intersectoral collaboration that benefited both ministries led to the housing intervention. The Malaysa country case was also developed by a research unt wthn the Health Mnstry. In the Lao People’s Democratic Republic case, the research was initiated and funded by the NGO and bilateral donor that implemented the project. The provincial health office participated in the research.
(Table contnued on next page)