Knowledge Brief Health, Nutrition and Population Global Practice COMMUNITY RESULTS-BASED FINANCING IN HEALTH PRACTICE: REFLECTIONS ON IMPLEMENTATION FROM EXPERIENCES IN SIX COUNTRIES Jean-Benoît Falisse, Petra Vergeer, Joy Gebre Medhin, Maud Juquois, Alphonse Akpamoli, Paul Jacob Robyn, Walters Shu, Michel Zabiti, Rifat Hasan, Bakary Jallow, Musa Loum, Cédric Ndizeye, Michel Muvudi, Baudouin Makuma Booto, Jean Claude Taptue Fotso, Séverin Sokegbe, Ibrahim Magazi and Gil Shapira April 2017 KEY MESSAGES: The term ‘community Results-Based Financing’ (cRBF) has been used to qualify a range of schemes whereby community actors such as community organizations, community health workers (CHW), and health facility committees (HFC) are contracted to facilitate access to ‒and sometimes directly provide‒ preventative, promotional, and curative health-care services. o It is too early to assess the effects of such experiences in the countries this brief focusses on (i.e. Benin, Cameroon, the Gambia, the Republic of Congo, and DR Congo), but comparing and discussing those schemes reveal aspects that are key in implementation, among others:  Pre-cRBF community engagement in health-care varies a lot; successful implementations of cRBF have built on those features and peculiarities.  Timely payment is crucial in a context where community actors often live in poverty; forms of pre-payment may improve retention and motivation.  Central to quality is the training and monitoring of community actors, which is easily undermined by low commitment of district officers and chief nurses. Certification and focus on the lower levels of ‘cascading’ training may improve quality, as well as testing the knowledge of community actors.  Information and Communication Technology is not a panacea for improving data collection and analysis: it requires a strong system in place, simple tools, and trained, supervised, and monitored actors‒three conditions rarely met in the field. cRBF is not always well integrated into health information systems.  The choice of indicators and bonuses is often a top-down decision, more community engagement may be desirable but has to be accepted by the Ministry of Health. cRBF schemes are part of wider community health policy reforms and represent an entry door to re-vitalize the often neglected sector of community health. In recent years, the emergence of ideas such as ‘close to level RBF; it is seen as helping with some of its bottlenecks client systems’ and ‘community accountability’ have such as the lack of information from and for users, issues renewed interest for community actors, and for new in reaching remote populations, and difficulties in strategies to organize and fund them. One such strategy, enhancing health-related behavioral change at community whereby community actors are paid based on the activities and household levels. they undertake, is ‘community-based Results-Based Financing’ (cRBF). The approach seeks to bolster demand While a growing literature is analyzing the successes and for health-care services through the introduction of shortcomings of RBF at the health-facility level, almost incentives for community-based actors. cRBF is regularly nothing exists regarding community RBF. This brief looks presented as a necessary complement to health facility- at cRBF implementation, relying on interviews, Page 1 HNPGP Knowledge Brief • discussions, and workshops with practitioners involved in example, distance from health facility) inform the payment six developing cRBF experiences in Cameroon, The of an extra bonus. In other countries, neglecting the Gambia, Benin, DR Congo, Rwanda, and the Republic of variance in efforts to perform the contracted activities has, Congo. It seeks to shed light on success and difficulties reportedly, led to unease with contracted community encountered in the implementation of cRBF programs or actors. policies. cRBF systems are recent but the question of their Who is the community in ‘community RBF’? The term resilience to emergency situations ‒such as epidemics‒ is cRBF is used to categorize a wide variety of schemes already being raised. Part of the answer may lie in contracting of a community actor on the basis of mechanisms discussed below: fluid payment systems, performance indicators ‒e.g. the delivery of a service. This integration of cRBF into national health policies, and strong community actor can be a group of or individual Community quality assurance mechanisms. Health Workers (in the Republic of the Congo, Benin, Cameroon, and Rwanda), a Health Facility Committee (in 2. TIMELY INCENTIVES AND PAYMENTS DR Congo), or any other local committees (Voluntary With community actors often living in situations close to Support Groups in The Gambia). Do cRBF schemes poverty, the timeliness of payment is fundamental. In their work? It is too early to say: most schemes are still in their first months, most schemes have experienced important infancy. An impact evaluation of the Rwandan scheme delays that led not only to discontent but also attrition of the found mixed effects. Ongoing schemes in Cameroon, (often trained) community actors willing to participate in Congo, and The Gambia integrate rigorous impact cRBF. Slight changes in the cRBF architecture and evaluation mechanisms and initial results should be increased experience of the implementers have usually led available in two to five years. At this stage, the main to resorbing those delays. Pre-payment mechanisms also discussion we can have is around the implementation of help. In Benin, health facilities reluctantly accepted to pre- cRBF. What can be learnt from those new schemes? pay CHWs monthly but, eventually, the pre-payment Together with country teams, seven areas have been improved the motivation and retention of CHWs. identified as crucial. The idea is not to provide any ‘recipe’, but to identify significant issues and questions raised in Whether there should be a fixed, non-results based, part in contexts that vary in terms of (c)RBF architecture and the remuneration of the cRBF community actor is debated. structure, health system effectiveness, and socio- Most countries have suppressed this fixed part because it economic and political environment. was not providing strong enough incentives. A fixed part may, however, help with retention. The experience of 1. CONTRACTING COMMUNITY STRUCTURES Cameroon has showed higher attrition of CHWs after the Different rationales underpin the architecture of the fixed part was suppressed. At the same time, it may make different cRBF schemes; they are intimately linked to local community actors a de facto part of the MoH payroll ‒a contexts. Understanding the existing community structures problem in some countries. The issue of the fixed part ties and collaborating with them has been crucial to the design into two other important issues. and implementation of most cRBF schemes. For example, the Republic of Congo barely had households visits and First is the question of the ‘amount’ of indicators contracted CHWs prior to cRBF, hence, in consultation with the actors routinely achieve: in Benin it was estimated that, at different stakeholders, the rainbow programme (the local the end of the day, payment does not vary hugely between cRBF scheme) set up a system of household visits by CHWs (but it varies between health zones), while in newly trained CHWs. Conversely, in Benin and Cameroon, Cameroon it was reported that what varies most is the CHWs have existed for a long time but their activities have bonus based on technical quality. Second is whether being often been inconsistent, the approach has therefore been a contracted cRBF actor is, in practice, a full-time activity. to train them on a package of activities and sub-contract The amount of money cRBF actors can expect to earn them through their local health centers. In The Gambia, varies hugely between countries. In Cameroon, many community meetings helped decide that the best option consider the price of indicators too low, but being a would be to contract the existing, and already well- contracted CHW can nonetheless provide a sizeable functioning, village Voluntary Development Committees. complementary income to people living in situations close to poverty. This is less the case in Benin or with the HFC Clear and simple contracts have helped implementation in members of DR Congo and the Committee members in the various cases: they entail being explicit with the The Gambia. It is important to note that the living conditions community about the phasing out or discontinuation of of community actors does not seem to play too important a programs. What is more, a good communication should role when deciding the price of indicators. In most cases, also be sensitive to the characteristics of the local the price of an indicator is set according to (1) the total communities. Good examples of this include the case of available budget and (2) the potential positive and The Gambia where the premium is adjusted to the size of cumulative externalities of the activity (in The Gambia, a the community and DR Congo where equity criteria (for pricing consultant helped the exercise). Page 2 HNPGP Knowledge Brief session in Cameroon). District teams are a prime candidate 3. PARTICIPATORY SELECTION OF INDICATORS to take on this type of role and to provide additional training The choice of indicators is a clear policy tool used to set and refreshers, and they have been supported in that health-care priorities, a responsibility of the Ministry of sense in DR Congo. However, in other contexts, local Health (MoH). Since a smooth and successful governments have high turnover, limited capacity for implementation of cRBF requires the collaboration of many supervision and have displayed very little enthusiasm to actors, most schemes include discussion mechanisms that monitor CHWs, leading cRBF implementers to consider bring together different divisions within the MoH as well as alternative channels such as municipalities. financial and technical partners. In most cases, this process is consultative rather than deliberative but Quality is hard to check given (1) the often high number of ultimately, because a series of cRBF schemes are, at least community actors and (2) the fact that most activities are partly, externally funded, donors have a decisive influence scattered across the health center’s responsibility areas. on the choice and pricing of indicators. The community and This does not mean that it is impossible: district local-level actors who lack representation in regional and assessment exists in all schemes and community national fora are the ones who face most difficulties having monitoring is being experimented in Cameroon and Benin their voice heard. with local Community-Based Organizations (CBOs), who check entries in CHW registers used by the CHWs and All schemes except Cameroon’s cRBF have deliberately assess user satisfaction (the same surveys have been, chosen to operate with less than ten indicators to reduce reportedly, used to identify areas of improvement for health costs, increase the monetary incentive attached to facility staff). This system requires well-trained and literate individual indicators, improve data quality and verification CBOs that are not always in high supply in rural areas. but also to set clear policy priorities and maintain the focus, Testing the knowledge CHWs have of the different and thereby quality, of the work of the community actors. activities they are supposed to undertake is another possible –and interesting– way to monitor quality and is an Most schemes focus on indicators related to health option that is being tested in Cameroon. High caliber promotion and patient referral, with apparent success in CHWs would, then, be those who are, for instance, able to some areas such as the community detection and referral correctly identify an illness in the community and refer the of tuberculosis in Benin. In countries like Cameroon and patient to the health center. Benin, there is an interest for potentially integrating more service delivery (including curative activities in Cameroon) 5. SIMPLE DATA SYSTEMS in the cRBF package of activities. This is seen as In a perfectly functioning cRBF scheme, quality issues potentially beneficial for the population that is not in would be identified in real time, by an efficient data-based frequent contact with health facilities but would also monitoring system. This, however, appears to be quite overstretch the role of the CHWs and begs questions about challenging. Too often, (data collection) tools are too their level of training and qualification. It also risks to complex to be meaningfully used at the community-level exacerbate a real challenge in the field: to ensure that the and there is little and sometimes no integration between CHWs do not overstep their assigned role and only provide cRBF data collection and other health information systems the services they are meant to provide and are trained for. ‒with some notable exceptions such as Benin and The That indicators should be SMART (Specific, Measurable, Gambia where cRBF data collection is fully integrated into Accepted, Relevant and Time-bound) appears obvious to DHIS II. most practitioners involved in cRBF who called for indicators to be as specific as possible. However, at the The Gambia cRBF scheme has innovated in data collection same time, arises the question of what is left apart, not and verification with the use of survey-based Lot Quality contracted because it is hard to measure. Assurance Sampling (LQAS). The model is quite appealing at first glance but the solution may not fit other cases: it 4. ENSURING FEASIBILITY AND QUALITY does not provide detailed community-level data; it is costly, Quality comes, first and foremost, from the appropriate and it is technically complicated to implement (population training of the contracted community actors. This has been size is needed, for instance). For most cases, this will not raised as a possible issue in many of the schemes. Indeed, be a panacea. Conversely, the belief that Information and the ‘cascading’ or ‘snowball’ training model that is used, Communication Technologies (ICT) automatically improve and understandably so given the high number of actors data collection and quality seems often misplaced. For this involved, has too often meant that training at the to happen, the data collection mechanism in place must be community level is of poor quality ‒when it is precisely that strong and simple enough to be accessible for people with level that is in contact with the users. Solutions to this potentially low technological literacy. Training is also problem include (1) proper certification by third-party required as shown by the examples of The Gambia and (Benin) and badge schemes that signal quality training and Benin. These conditions are simply not met in many cases, (2) supervision mechanisms and re-cap session (yearly making the use of ICTs at best a waste of energy ‒and at Page 3 HNPGP Knowledge Brief • worst creating more problems by adding another layer of community health sector that is often neglected. By doing complexity in a system many already find complicated. so, cRBF also asks questions about the health (and RBF) system, among others in terms of roles of the community Ultimately, practitioners are still exploring ways to best use actors or quality insurance. community-level data to inform and influence activities and policies. At the micro-level of a health facility, a huge, There is no one-size-fits all and successful largely untapped, potential is the use of data analysis by implementations have built on local realities and institutions local nurses and CHWs to identify and act upon local and adjusted and learned from the field. There remain problems. This, however, requires additional training and many more questions and unknowns than answers about may require skills that are simply not available at that level. cRBF and it will be crucial that experiences are well documented and researched in order to push the global 6. OUTREACH AND ACCESS reflection on the topic forward. In addition to the need to CHWs systems that have been ‘revitalized’ through cRBF soundly evaluate the impact of cRBF schemes, further may help reach populations that were not reached in the research is required to better understand a series of key past. In Cameroon for instance, newly incentivized CHWs issues, among others: (1) the link between cRBF and other started visiting nomadic people when they pass in the demand-side approaches (including vouchers); (2) the health centres’ area of responsibility, reportedly improving links between cRBF and primary health-care approaches, vaccination coverage among that population. Further, also in terms of coordination of the different stakeholders cRBF schemes are said to be open to the entire community (and in terms of community participation in cRBF –and few of them have chosen to devote energies to verification); (3) the role of non-monetary incentives for targeting specific vulnerable groups, this despite cRBF CHWs; (4) community feedback mechanisms, including being a potentially powerful tool to doing exactly this. One those enabling vulnerable and marginalized groups to be exception is Benin, where CHWs receive a higher premium heard in the public sphere; as well as the mechanisms that when they visit indigents in their home. Home visits are empower communities and make them recognized; (5) the known to be an opportunity for increasing health extent to which CHWs can be used as frontline providers, awareness among the most vulnerable. especially for family planning (for example, distributing pills); (6) the mechanisms capable of compelling 7. HEALTH SYSTEM INTEGRATION households into action following a CHW’s household visits, The cases of Benin, the Republic of Congo, and Cameroon and more broadly, the best ways to incentivize behavioral illustrate how cRBF can be an opportunity to 'rationalize' change and community action; (7) potential systems for the CHWs and HFC members in terms of clarifying their regular community-level data collection, monitoring, and roles and ensuring that they have adequate skills, but also quality control –and the role of technology in them; (8) the bringing them closer to health centers. cRBF approaches effects and sustainability of projects undertaken with or are generally horizontal, in the sense of an integration in support by community subsidies; and (9) the medium and the local health system. It remains unclear how new vertical long-run viability of cRBF schemes, both within programs would fit in such a model but cRBF could be a communities and as national schemes, and in terms of both platform for interventions relying on CHWs. community financing and support from local actors. The Gambia and Rwanda cRBF cases also present Several ongoing studies on CRBF schemes are being interesting features in terms of integrating traditional actors implemented in the context of the portfolio of impact into the health system through the contracting of evaluations funded through the Health Results Innovation ‘cooperatives’ of birth-attendants (Rwanda) or groups that Trust Fund. In the Republic of Congo and the Democratic include them (The Gambia). Republic of Congo, the research will focus on incentivizing health facilities to conduct home visits. A study in Further Learning Cameroon will evaluate a mechanism through which health centers subcontract community health workers. In The Community engagement is crucial for expanding the Gambia, a model of performance payments to community coverage of essential health services in a cost-efficient organizations will be studied. The studies employ both manner, especially in situations with limited fiscal space for quantitative and qualitative methods and are expected to health. cRBF schemes are an eclectic mix of approaches be completed in the next three years. that attempt to bring in more funding to community actors through various incentive systems. It is too early to tell This HNP Knowledge Brief is based on the 2016 report ‘Community-based whether the different approaches covered in this brief are Results-based Financing in Practice: A Discussion Piece’; the cRBF effective. What seems clear at this stage is that cRBF workshop held in Harare 18-20 September 2016; an online workshop held with country teams on 19 January 2017, and a series of individual approaches lead to re-thinking, and possibly re-vitalizing, a consultations with and between country team. The Health, Nutrition and Population Knowledge Briefs of the World Bank are a quick reference on the essentials of specific HNP-related topics summarizing new findings and information. These may highlight an issue and key interventions proven to be effective in improving health, or disseminate new findings and lessons learned from the regions. For more information on this topic, go to: www.worldbank.org/health. Page 4