Initiatives supporting evidence informed health system policymaking in Cameroon and Uganda: a comparative historical case study

Background There is a scarcity of empirical data on institutions devoted to knowledge brokerage and their influence in Africa. Our objective was to describe two pioneering Knowledge Translation Platforms (KTPs) supporting evidence informed health system policymaking (EIHSP) in Cameroon and Uganda since 2006. Methods This comparative historical case study of Evidence Informed Policy Network (EVIPNet) Cameroon and Regional East African Community Health Policy Initiative (REACH-PI) Uganda using multiple methods comprised (i) a descriptive documentary analysis for a narrative historical account, (ii) an interpretive documentary analysis of the context, profiles, activities and outputs inventories and (iii) an evaluative survey of stakeholders exposed to evidence briefs produced and policy dialogues organized by the KTPs. Results Both initiatives benefited from the technical and scientific support from the global EVIPNet resource group. EVIPNet Cameroon secretariat operates with a multidisciplinary group of part-time researchers in a teaching hospital closely linked to the ministry of health. REACH-PI Uganda secretariat operates with a smaller team of full time staff in a public university. Financial resources were mobilized from external donors to scale up capacity building, knowledge management, and linkage and exchange activities. Between 2008 and 2012, twelve evidence briefs were produced in Cameroon and three in Uganda. In 2012, six rapid evidence syntheses in response to stakeholders’ urgent needs were produced in Cameroon against 73 in Uganda between 2010 and 2012. Ten policy dialogues (seven in Cameroon and three in Uganda) informed by pre-circulated evidence briefs were well received. Both KTPs contributed to developing and testing new resources and tools for EIHSP. A network of local and global experts has created new spaces for evidence informed deliberations on priority health policy issues related to MDGs. Conclusion This descriptive historical account of two KTPs housed in government institutions in Africa illustrates how the convergence of local and global factors and agents has enabled in-country efforts to support evidence-informed deliberations on priority health policy issues and lays the ground for further work to assess their influence on the climate for EIHSP and specific health policy processes.


Background
Poor access to health interventions and poor performance of health systems are consistently an issue of concern for national and global stakeholders as we approach the year 2015, and many predict the failure to achieve the targets set for health Millennium Development Goals -MDGs -particularly in sub Saharan countries [1][2][3]. Fostering evidence informed health system policymaking (EIHSP) in low-and middle-income countries (LMICs) has become a priority for the United Nations and development agencies striving to bridge the "know-do" gap that undermines progress towards the health MDGs. As a consequence, several agencies are providing financial support to tens of initiatives worldwide in that regard [4,5].
Knowledge Translation Platforms (KTPs) are such an initiative that brings together policymakers, researchers and other stakeholders including civil society for evidence informed deliberations on health priorities. KTPs are conceived as knowledge brokering enterprises building from the integrated model for linking research to policy [6][7][8][9]. Examples in Africa are Evidence Informed Policy Network (EVIPNet) and the Regional East African Community -Health Policy Initiative (REACH-PI) together involving twelve countries [10,11] with EVIPNet Cameroon and REACH-PI Uganda being amongst the most active [12]. The premise of such efforts is that the use of research evidence for health system policymaking will yield positive public health and social impacts [13].
While several case-studies have gathered evidence on the impact of health technology assessment units and government support units in high-income countries [14,15] and few case-studies of embedded knowledge translation strategies within research to policy projects in LMICs [16], there is a scarcity of empirical data on institutions devoted to knowledge brokerage and their influence [17][18][19]. Boaz and colleagues [20],concluding a systematic review, called for the development of new conceptual frameworks and methods to orient future evaluations of interventions designed to promote research use, including knowledge brokers, networks, and linkage and exchange programmes.
The lack of systematic documentation of the KTPs in LMICs prevents learning from these social innovations in countries synonymous with scarcity. The objective of this paper is to describe and interpret the history, the infrastructure, the activities, and the outputs of two pioneering KTPs in Cameroon and Uganda established since 2006.

Methods
We conducted a comparative historical case study of two KTPs within their contexts using multiple methods [21]. EVIPNet Cameroon housed at the Central Hospital, Yaoundé and REACH-PI Uganda housed at Makerere University College of Health Sciences, Kampala were identified for their exemplarity as pioneers in Central and Eastern Africa during the period 2001-2012. This post Millennium Summit timeframe was retained in order to investigate two six-year periods before and after the launching of both initiatives in 2006. The investigation comprised (i) a descriptive documentary analysis to provide a narrative historical account, (ii) an interpretive documentary analysis of the context and the profile/activities and outputs inventories and (iii) an evaluative survey of stakeholders exposed to evidence briefs for policy and invited to policy dialogues. The authors stand as insiders intervening either as policymaker (POZ), knowledge broker leading a KTP secretariat since its inception (POZ, NKS) or investigator in the Supporting the Use of Research Evidence for policy in African health systems -SURE research project (POZ, NKS, JNL and GT) [www. global.evipnet.org/sure] and Knowledge Translation Platforms Evaluation-KTPE research project (POZ, NKS, JNL) [22]. As a group of authors with different levels of engagement in the KTP activities, we strived to maintain as much neutrality and objectivity by combining different sources of data, online discussions on and several iterations of the draft manuscript.

Document review
We conducted a qualitative descriptive and interpretive archival review of both initiatives. All the available documents were requested from the KTPs' secretariats, the research coordinators of KTPE project at McMaster University in Hamilton, Canada and SURE project at the Norwegian Knowledge Centre for Health Sciences in Oslo, Norway. We searched the websites of EVIPNet (www.global.evipnet.org), EVIPNet Cameroon (www. cdbph.org), and the Uganda clearinghouse for health policy and systems research (www.uchpsr.org) for any relevant documents or activities. We equally conducted a structured documentary review of poverty reduction/ eradication strategic papers and health sector strategic plans produced in Cameroon and Uganda during the period 2001-2012 to capture the political, social and economic contexts and salient features of both health systems. These documents were obtained from the respective ministries of health (Table 1). We extracted relevant data featuring the contexts, the institutional arrangements, the activities and outputs of KTPs.

Evaluative survey
In both countries, policymakers, researchers and other stakeholders likely to be involved in or affected by policy decisions on the issue addressed by an evidence brief (policy brief) were invited to a deliberative dialogue (policy or stakeholder dialogue). All dialogue participants were surveyed about the evidence brief that was pre-circulated for the dialogue and about the dialogue itself. Seven evidence briefs and five dialogues were concerned in Cameroon. Three evidence briefs and two policy dialogues were concerned in the case of Uganda. The questionnaires available both in French and English were developed as part of the KTPE study. Each questionnaire comprised three or four sections to depict how helpful each of the features of the brief/dialogue were, how well the brief/dialogue achieved its intended purpose, items based on theory of planned behaviour constructs, and questions about respondents' professional experiences. Further details on the instruments can be accessed online at http://www. researchtopolicy.org/KTPEs/KTPE-overview.
The coding of the features of the brief/dialogue based on the electronic copies of the dialogue summary and/or report was checked with core members of each KTP secretariat. Descriptive statistics were used to examine respondents' overall assessments of brief/dialogue and their features and to profile the assessments of each feature of the brief/dialogue, each of the brief and the dialogue as a whole, and respondents' intentions to act on what they had learned.

Analytical framework
To systematically describe the KTPs, we elaborated an analytical framework (Table 2) from a purposive review of writing including frameworks, concepts and theories pertaining to knowledge brokerage and the integrated model for knowledge translation [6][7][8][9]. The latter is underpinned by social learning theory and planned behaviour change model geared at addressing barriers and facilitators to research use by policymakers within the "two-community" thesis [23]. Several scholars have explained the poor use of research evidence into policymaking by the differences of cultures across the research community and the policy community thus establishing the foundations of the knowledge brokerage, and linkage and exchange models [6][7][8][9][24][25][26][27][28]. From the political sciences, we draw from the health policy analysis triangle [29,30], policy networks [31][32][33] and the critical drivers of policymakinginstitutions, interests, ideas and external factors [34,35].
The framework combines the three functions of a knowledge brokering enterprise [28] with the domains and elements to assess country efforts to link evidence to action [9] as well as activities and outputs deemed to influence the policy context, process and content [29,30] and the critical drivers of policymaking and to eventually intersect with contextual factors such as political and health systems and policy networks. We described and analyzed the health systems according to their governance, financial and delivery arrangements, as well as health technology provisions [35]. We used the interpretive constant comparison of KTPs within their contexts to highlight similarities and differences.

Study context
Cameroon and Uganda political systems are marked by their presidential regimes strongly anchored in traditional ruling systems bolstering the ethnic diversity with 220 and 56 ethnic groups respectively. The Head of State in each country has been in office since the 1980's. The Parliaments are dominated by a large majority from the Head of State's political party and technocrats play a pivotal role during health policymaking. The thrust of development policies has been the achievement of MDGs following the Millennium Summit and the African Union resolutions to speed up health investments and align them with health MDGs targets with both countries eligible for grants from a diversity of global health initiatives. Efforts were engaged to strengthen national health research systems leading to establishing a division in charge of health research in the ministry of health in Cameroon since 2002 and increasing financial support to Uganda National Health Research Organization (UNHRO) since 2008.
Since the mid 1990's, health decentralization was initiated in both countries to align with the health district framework established by the African Regional Office of WHO. The tiered health systems are mixed; state owned health services coexist with private health facilities operating in poorly regulated environments. The ministry of health is the overarching health authority in addition to the inter-sectoral governing bodies of priority health programs established in response to global health initiatives (e.g.; expanded programme of immunization, control programs for AIDS, tuberculosis and malaria, reproductive health, neglected tropical diseases, etc.). The major changes observed include: i)the abolition of user fees in Uganda in 2001 and the promotion of community based health insurance in Cameroon since 2004; ii) the tangible efforts towards actual decentralization of health authority to provincial/regional and district authorities in both countries starting in 2001; iii) the promotion of universal access to HIV/AIDS care including antiretroviral therapy since 2003; iv) the universal access to malaria control interventions since 2002 and; v) the scaling up of reproductive health programs in line with the African Union's campaign to accelerate the reduction of maternal mortality in 2009. Table 3 summarizes the political and health systems and main indicators of health MDGs. While Cameroon ranked as lower middle income and Uganda as low income, the maternal mortality ratio (MDG 5) has worsened in the former while improving in the latter. Neither country will reach the health MDGs targets by 2015.  Three Ministers, three Director General and changes of high ranking civil servants in health policy and planning units

Tiered health system features
Health system governance arrangements National ministry of health + inter-sectoral governing bodies for public health programmes. 10 provincial delegations and 143 districts with dialogue structures poorly functional.
National ministry of health + inter-sectoral governing bodies for public health programmes.
National ministry of health + intersectoral governing bodies for public health programmes.12 regional directions and 87 districts. Dialogue structures linked to different levels of local governments.
10 regional delegations and 178 districts with municipal leaders holding leadership positions in health district management boards.
Health financial arrangements User fees under a fee-for-service scheme in government owned facilities. The Government raise some funds from the general tax system and overseas development aid. Civil servants are paid by the central government but also receive bonus based on user fees. Private clinics operate under a poorly regulated fee-for-service scheme.
User fees under a fee-for-service scheme. 98% out of pocket payments. Despite a national strategy to promote community-based health insurance, coverage is below 2%. Rising petty corruption in state owned facilities.
Abolition of the user-for-service scheme in 2001 in government owned facilities. Civil servants are paid by the central government.
Service delivery arrangements Community health volunteers provide some benevolent primary health care services. Free preventive services in government health facilities. Private clinics operate under a fee-for-service scheme and pharmacies. Faith based and not for profit NGO health facilities operate under a subsidized fee-for-service scheme. Traditional healers and informal health facilities.
Technologies, medicines and vaccines A national procurement system for essential and generic medicines coexists with dedicated procurement systems for vertical priority health programs (vaccines, ART). Private medicines wholesalers operate under a poorly regulated environment in which drugs prices are free. Private medical equipment firms. Historical account  Table 5 outlines the KTPs' infrastructure. EVIPNet-Cameroon secretariat has been operating with a multidisciplinary group of part-time researchers and research assistants (e.g.; public health, economy, anthropology, sociology, epidemiology, clinical sciences) trained as brokers. While two scientists have remained engaged the whole time, a turn-over was noted amongst researchers and assistants. REACH-PI Uganda has been operating with a smaller group of full time staff of public health experts trained as brokers. A social scientist trained as broker left after 12 months and the number of brokers went from one to six between 2009 and 2012. The initial stakeholder analyses during grant preparation laid the groundwork for participatory priority setting exercises and validation of the respective programs of work thus creating the enabling environment for mutually beneficial exchange amongst knowledge brokers, policymakers, researchers, and other stakeholders. Both KTPs were established as demonstration projects informed by existing theoretical frameworks and were guided by a monitoring and evaluation framework that has enabled this description. They were conceived as problem solving enterprises, operating under the "learning through doing" principle.  Table 6 provides an account of the activities and outputs in terms of capacity building, knowledge management and linkage and exchange. The human capital for EIHSP was increased by more than thirty training workshops in Cameroon, Uganda and other countries (Kenya, Tanzania, Burkina Faso, Mali, Zambia, and Mozambique) to jointly build capacity for policymakers, researchers, civil society groups and media on EIHSP. Almost five hundred stakeholders were sensitized or trained by both KTPs including five Africans and four Canadians enrolled in the joint doctoral program in health policy and knowledge translation at Makerere University and McMaster University respectively.

Activities and outputs
Following the priority setting exercises, both KTPs have produced 15 evidence briefs for policy. Preparing evidence briefs was very labour intensive as few evidence briefs have required two full time equivalent knowledge brokers during one year. Between 2008 and 2012, EVIPNet Cameroon prepared 12 evidence briefs and REACH-PI Uganda prepared three evidence briefs. In line with the SURE grant plans, a mechanism to prepare rapid  Implementation of SURE projectmore details in Table 6 Implementation of SURE projectmore details in Table 6 IRCI Knowledge translation workshop Launch of the Rapid Response Service  Preparation of an evidence brief on task shifting to optimize roles for mother and child health. Piloting of the rapid mechanism to respond to urgent needs of evidence of officials in the ministry of health Linkage and exchange Presentations with officials in the ministry of health to elicit support. Two policy dialogues on governance for health district development and for scaling up malaria control interventions Policy dialogue on task shifting to optimize the roles of healthcare providers for mother and child health. Rapid evidence syntheses to respond to health stakeholders' urgent needs 2011 Capacity building Workshops for policy makers, researchers, civil society representatives and media Collaborative program for doctoral studies in health policy and knowledge translation conduct policy relevant trials and Cochrane reviews in collaboration with the SACC. EVIPNet Cameroon has contributed to Cochrane Collaboration's efforts to translate its products into French. In terms of linkage and exchange, EVIPNet Cameroon organized seven policy dialogues informed by pre-circulated evidence briefs. The policy dialogues were jointly convened by the KTP secretariat and the Ministry of Health. The selection of participants was informed by the stakeholder analysis. Participants deliberated on scaling up access to artemisinin-based combination therapy (ACT), scaling up malaria control interventions, improving governance for health district development, retention of human resources for health in rural areas, scaling up community-based health insurance, improving antenatal care services coverage, improving access to and quality of care in the accident and emergency departments. REACH-PI Uganda organized three dialogues on scaling up access to ACT, task shifting for maternal and child health and, improving skilled birth attendance. EVIPNet Cameroon and REACH-PI Uganda played a central role organizing the first international forum on EIHP in LMICs (27)(28)(29) August 2012, Addis Ababa, Ethiopia) whose 121 participants were from 27 countries including 17 African countries. Participants were policymakers, international bureaucrats, knowledge brokers, researchers, civil society groups, and media.
Stakeholders' perspectives on the evidence briefs and policy dialogues Table 7 summarizes the results of the survey of readers of ten evidence briefs. The respondents largely agreed that the briefs achieved their purpose of presenting the available research evidence on a high-priority policy issue in order to inform a policy dialogue where research evidence would be just one input to the discussion. The different design features of the brief were highly appreciated but respondents expressed lower satisfaction with the brief not concluding with any recommendations. Table 8 features the results of the survey of participants attending ten dialogues. All respondents felt the dialogues achieved their purpose of a full discussion of relevant considerations about a high-priority policy issue in order to inform action and the different features of how the dialogues were designed were considered very helpful including that the dialogue was informed by a pre-circulated evidence brief.

Interpretive synthesis
Both initiatives are equipped with research units operating as national knowledge brokering institutions with regional influence. Within the two health systems, a network of local and global experts has created new spaces for inclusive evidence informed deliberations amongst policymakers, researchers and stakeholders on highpriority health policy topics related to MDGs. The interaction between the KTP secretariats and ministries of health and other stakeholders enabled the identification of priorities for evidence briefs as well as evidence gaps. Both initiatives have progressively expanded to cover the array of operations of a knowledge brokerage enterprise namely capacity building, knowledge management, and linkage and exchange. Applications to funders and advocacy meetings have enhanced their visibility and provided enabling resources towards institutionalization and sustainability.
The evidence briefs and rapid evidence syntheses prepared generally aligned with health policy and systems priorities to achieve the health MDGs. The technical and consensual natures of the topics addressed and the problem-driven approach have contributed to a high level of satisfaction amongst all categories of stakeholders. The mechanisms to address stakeholders' urgent needs of evidence within days and weeks were well received. The briefs and syntheses have provided evidence-based problem frames, policy options and implementation strategies yielding potential changes in two of driving forces in policymaking namely interests and ideas.
This historical account illustrates how the convergence of local and global factors and agents has enabled the implementation of in-country efforts to support EIHSP related to health MDGs. It also illustrates how the differences in historical background, institutional anchorage, contexts and funding sources have led to differences in activities and outputs of these KTPs. The diversity of grant arrangements and the differences in institutional arrangements and planning cycles as well as the stability of health technocrats explain the differences in evidence outputs and the contrasted uptake of the rapid response mechanism. EVIPNet Cameroon was more prolific 13. Was subjected to a review by at least one policymaker, at least one stakeholder, and at least one researcher (called a "merit" review process to distinguish it from "peer" review, which would typically only involve researchers in the review) 6.4 0.8 6.1 1.3 The ratings are on a Likert scale from 1 to 7 (least useful = 1 and most useful = 7) for question 1 to 13. The lowest rating (5.4) was for the briefs not concluding with particular recommendations. These are mean values for seven evidence briefs in Cameroon and three evidence briefs in Uganda.
in preparing evidence briefs and organizing policy dialogues because of the closer ties with the ministry of health thus allowing working concurrently on several evidence briefs. REACH-PI Uganda prepared more rapid evidence syntheses and aligned the production of evidence briefs to the SURE project arrangements. This historical account equally illustrates the unpredictability of the course of events during the initial decade of these initiatives conceived of as demonstration projects. Initial priority settings have been readjusted to align with changes in leadership within ministries of health and global funding opportunities (e.g. in Cameroon shifting from nutrition and chronic non communicable diseases to health district governance and health financing based on the grant from the AHPSR). Contributions from governments have remained in kind. In this documentary review, we failed to identify any empirical evidence on the influence or impact of the KTPs on the country climate for research use or on specific policy processes beyond the policy deliberations.

Principal findings
The infrastructure, activities, outputs and outcomes of both initiatives encompass the full array of activities of knowledge brokerage enterprises and they have experimented at various levels the three key functions of such enterprise: capacity building, knowledge management and, linkage and exchange [8,28]. Indeed, both KTPs have trained almost 500 policymakers, researchers and stakeholders to facilitate researcher push and user pull [9]. More than 100 tailored evidence syntheses were produced, disseminated and made openly available online. Inclusive consultations were organized to identify high-priority policy issues related to health MDGs and structured stakeholder mappings laid the ground work to convene 10 deliberative dialogues informed by research evidence.
The historical account and the critical analysis of actors of these social experiments feature the influence of policy learning/diffusion in the establishment of policy networks and epistemic communities [31][32][33]. Leading researchers from northern universities linked with African researchers to create a new momentum for EIHSP, developing and testing new resources and tools to popularize knowledge translation activities across Anglophone and Francophone Africa. They exemplify the relevance of the recommendations formulated based on analysis of similar institutions in other settings [35].

Strengths
This is the first historical account of what is constitutive of two KTPs housed in government institutions in sub Saharan Africa. It contributes empirical knowledge on the feasibility and practicality of enhancing the technical The ratings were on Likert scales from 1 to 7 (least useful = 1 and most useful = 7) for question 1 to 12. The highest rating (6.6) was for the dialogue addressing a high-priority policy issue in Cameroon and the lowest rating (5.7) was for the dialogue providing an opportunity to discuss who might do what differently.
capacity of policymakers, researchers and stakeholders; preparing evidence briefs, syntheses and summaries; providing evidence related services; convening dialogues and creating space for evidence informed deliberations on high-priority health policy topics. This investigation complements the lessons learned from the Zambian Forum for Health Research (ZAMFOHR) case study [36], a national nongovernmental organization spearheading knowledge translation efforts in Zambia. It also enriches the recent gathering of lessons learned on KTPs by providing a longitudinal perspective on what constitutes a KTP and how it operates in two LMICs [18]. The study offers an insiders' perspective as two authors have a deep knowledge and understanding of the context in Cameroon and Uganda and the authors have been involved with both KTPs from the inception phase through the current state of operations. This study also provides a grounded feedback to the chorus of voices calling for support to EIHSP and the recommended strategies for facilitating the uptake of research into policy in LMICs [37][38][39][40][41][42]. Finally, by providing an historical insight on institutional arrangements of KTPs, this study contributes empirical evidence to the call for new conceptual frameworks and methods to orient evaluation of efforts to support EIHSP in LMICs [20]. In that regard, the analytical framework used for this study lays the ground work for further political sciences informed perspectives on KTPs to comprehend their influence and impact.

Limitations
This study presents three main limitations. First, the study is restricted to describing what is constitutive of the two KTPs since their inception; an assessment of their influence on specific policy processes and the climate for EIHP is still awaited to empirically inform the efforts to explicate KTPs using sound political sciences perspectives. Second relates to the nature of retrospective qualitative archival review, the exclusive use of official documents might have overlooked challenges experienced by the KTP implementers particularly the informal networks to navigate the health bureaucracies, to engage with officials and gain their support over time. The restriction to the KTPs' archives might have prevented the capture of the external players' views and perspectives and particularly the funding agencies. Last relates to the insiders' narrative as recall bias and social desirability yield potential negative effects on neutrality and objectivity.

Implications for local and global policymakers, stakeholders and researchers
This empirical documentation can inform the development of new initiatives with three implications: (i) those planning to establish initiative to support EIHSP in LMICs should carefully consider opportunities for national and international collaborations to mobilize political support from government officials and funding agencies; (ii) the critical role of participatory processes during priority setting exercises, stakeholders dialogue and needs assessment so as to secure commitment from both national policymakers and global players investing in health sector development; (iii) establishing an initiative to support EIHSP requires committed and skilled human resources to cope ably with intense and somehow stressful endeavour and to navigate the complex interfaces of knowledge to policy and action with a long term perspective. This study equally provides a strong basis on which researchers can attune their efforts in developing and validating robust methods and tools to evaluate the effects and influence of KTPs [20].Indeed, the framework developed by Lavis and colleagues [9] to assess country efforts to link research to policy and used elsewhere [18] provide descriptive categories for efforts (e.g.; climate, research production, push efforts, facilitate user-pull, user-pull, exchange, and evaluation) engaged by a given country but doesn't provide tools to assess the influence of such efforts. Further, the framework developed by Ward and colleagues [28] on what constitutes a knowledge brokering enterprise while accounting for the three main functions (e.g.; capacity building, knowledge management, and linkage and exchange) fails to account either for the effects and influence on drivers of policymaking (e.g.; institutions, interests and ideas) or the intersection with contextual factors during policymaking in environments permeate by cross jurisdictional learning. The need to have further reflection on the appropriate evaluative framework of KTPs remains valid [20].
The rising numbers of skilled individuals in EIHSP and the availability of contextualized evidence resources imply that national and global players investing in health sector development in Africa should create the enabling environment (e.g.; new rules and regulations, incentives) for and foster effective management and use of the human capital for policy analysis and research during health system planning and programming.

Conclusion
This descriptive historical account of two KTPs housed in government institutions in Africa illustrates how the convergence of local and global factors and agents has enabled in-country efforts to support evidence-informed deliberations on high-priority health policy issues and lays the ground for further work to assess their influence on the climate for EIHSP and specific health policy processes.
Abbreviations EIHSP: Evidence informed health system policymaking; EVIPNet: Evidence informed policy network; KTP: Knowledge translation platform; LMIC: Low-and middle-income country; MDG: Millennium development goal; REACH-PI: Regional east african community health policy initiative.