Experience of RBF to date
The evidence from RBF for health is mixed. Given the diversity of RBF schemes, there is significant variability between projects. The design of RBF programs is highly context-specific. Health indicators and performance incentives are largely dependent upon the specific health needs and risk factors of the population, the health system and infrastructure in the country, and the priorities of stakeholders. Limited and/or inadequate monitoring and evaluation further complicate efforts to review RBF programs. A Cochrane review of the effects of paying for performance in health care in LMICs concluded that more robust and comprehensive studies are needed, a finding supported by other studies of RBF as well [
21‐
24]. More extensive monitoring and evaluation is particularly necessary for showing results in the long-term [
25]. Despite the need for additional evidence, it has been suggested that RBF can be an important tool for addressing efficiency and accountability, as well as health sector reform more broadly [
26].
Rwanda has perhaps the best-known and most rigorously evaluated RBF program. Evidence from Rwanda is promising. The impact evaluation of Rwanda’s PBF program determined the effect on fourteen maternal and child health indicators [
27]. While some indicators showed more positive results than others, overall results demonstrated that RBF can help to increase service delivery and quality of care, improving health sector performance [
28]. Various other PBF programs have demonstrated positive or mixed results, such as PBF schemes in the Democratic Republic of Congo, Egypt, and Burundi [
29‐
31]. Less successful programs, such as the Tanzania PBF scheme, have provided important lessons [
32].
PBC programs, like PBF, have generally produced encouraging results. A PBC scheme in Cambodia found that contracting NGOs was effective in increasing service delivery [
33]. Promising results have also come from other countries, such as PBC programs in Liberia and Afghanistan [
34,
35]. A PBC scheme in Southern Sudan, however, experienced significant implementation challenges [
36], and in Uganda various survey rounds found no impact from PBF schemes. Lessons from Uganda include the need for substantial enough incentives directly tied to results, autonomy for health facilities to make decisions and to use resources, and rigorous verification of results [
37].
Progresa/Oportunidades in Mexico is the largest and most cited CCT program. This program provided incentives to households in the form of cash transfer payments to women conditional on engaging in certain behaviors related to child health, nutrition and education. Evaluation has shown success in improving child health indicators, among other areas [
38]. A similar program in Brazil,
BolsaFamilia, however, produced no positive results in terms of health status, but succeeded in reducing poverty, suggesting that demand-side incentives alone may not be enough to improve health outcomes [
39,
40]. Other CCT programs, such as the Janani SurakshaYojana program in India, have shown positive yet mixed results [
41]. Psychological research has demonstrated in clinical trials that incentives can be effective in changing behaviors related to smoking and drug use as well, however, these findings must also be proven in population-based interventions [
42].
While both PBF and PBC programs have produced mixed results, on the whole there is much encouraging evidence. Demonstrated successes from some programs and lessons taken from others, suggest that both of these approaches could be applied to NCDs. CCT programs, have been successful at reducing poverty, which has important health implications. CCT programs could have potentially significant effects on risk factors for NCDs, including diet, exercise, and tobacco use, as well as preventive measures for certain types of cancers. The benefit of using supply-side interventions in tandem with demand-side incentives, suggests that the use of CCTs for NCDs may be most effective when combined with other supply-side programs.
Applicability of RBF to NCDs
While most RBF programs have focused on Millennium Development Goal (MDG)-related health issues, usually on maternal and child health (MCH), a few have included NCDs. In Belize, a PBF program designed to scale up access and improve quality of health services targeted chronic illnesses in addition to prenatal and postnatal care. Evaluation of the six month pilot showed an increase in use of primary care and the diagnosis and treatment of diabetes and hypertension, the country’s top two causes of death [
43,
44]. In Abu Dhabi, the Weqaya (or prevention) program has used performance incentives for NCDs to improve quality and measurable health indicators through the use of Disease Management Programmes (DMPs) to address compliance with clinical care and behavior change related to specific risk factors at the patient level. As the program was initiated in 2006, the long-term effects are yet to be seen [
45].
Few RBF-NCD programs may have been initiated because the epidemiological nature of NCDs does not easily lend itself to RBF. Programs aimed at NCDs often involve major health behavior change and long-term interventions. Incorporating RBF into these programs is more difficult, compared to one-time or short-term interventions, such as immunizations [
46]. A RBF expert from the World Bank made the distinction that RBF focuses on both health outputs and outcomes. It is more challenging, however, to have specific measurable indicators for health outcomes that may occur over long time periods, which is the case for many NCD indicators. This is due to the need to provide incentives on the supply-side on a regular basis (i.e. monthly or quarterly) when using PBF. Therefore, incentives to health facilities are often linked to more short-term outputs, such as immunizations, antenatal care, institutional deliveries, growth monitoring, and family planning. For example, a NCD health intervention focused on tobacco use might use smoking prevalence as measured by population-based survey data as an indicator. Tying incentives to percentage decreases in smoking prevalence, however, would be difficult. Instead, it might be more effective to use the number of people enrolled in smoking cessation programs as an indicator, allowing incentive payments to providers to be made quarterly.
Rena Eichler and Ruth Levine note the potential of RBF for NCDs based on evidence from developed-country settings [
46]. For example, a supply-side intervention for diabetes in the United States demonstrates the potential of performance incentives for chronic conditions [
46]. Demand-side incentives have been used to effect behaviors to reduce consumption of tobacco and alcohol but have not been successful in the long-term [
46]. The minimal available evidence suggests that supply-side RBF programs for NCDs are likely to be more effective than demand-side programs.
Although little evidence is available on RBF for NCDs, analysis of RBF programs for MCH, communicable diseases, and health systems strengthening (HSS) provides a reasonable basis for applying RBF concepts to NCDs. NCD programs would benefit from improvements in service delivery, quality, and health sector performance that may result from RBF programs directed at MCH or HSS [
3]. For example, improved access to essential medicines and primary care could mean increased adherence to chronic disease medications and increased rates of screening for certain types of cancers. Improved health care quality could include indicators for health IT, which could in turn improve chronic disease management. Broader health system effects from RBF could have meaningful implications for NCDs.
Despite many challenges, application of RBF principles to NCDs has great potential. A summary of common themes from key informant interviews can be seen in Table
3. Expert testimony highlighted the potential of RBF and cautioned about the challenges, such as the long-term effects of NCDs and health behaviors external to the health system. Even these behaviors such as diet and exercise, may lend themselves to incentive programs. The potential for RBF, key informants noted, depends on country-specific priorities and risk factors. An expert on RBF from the World Bank saw potential in using RBF for preventive and health promotion services, two areas into which NCDs fall. It is important to integrate NCD indicators into existing RBF schemes where possible. There is also potential for linking efforts to reduce NCDs to a RBF mechanism as part of a larger package of essential health services. Kathy Kantengwa, an expert on PBF from MSH, stated that RBF can apply to any service delivery, and therefore could be used for early detection of NCDs and management. She stated:
“The RBF principles for communicable disease can be adapted to the NCDs”[
47].
Table 3
Common themes from semi-structured interviews
1 | Key elements of successful RBF programs | Political commitment, government ownership, buy-in of stakeholders |
| | Clearly defined rules, understanding of indicators; accountability, verification of indicators |
| | Measuring and evaluation |
| | Design of program, piloting and testing; participatory approach |
| | Flexibility in implementation; communication, transparency, sustainability |
2 | Areas of health for which RBF traditionally used | Maternal and child health; MDGs 4 and 5 |
| | Health service delivery, primary care, quantity and quality of services |
3 | Potential use of RBF for NCDs | Application of RBF to any service delivery |
| | Incentivizing preventive and health promotion activities; national, institutional, and individual levels |
| | Part of package of essential health services; combining efforts for communicable and non-communicable diseases |
4 | Challenges in taking a RBF approach | Variation in capacity of donor agency representatives |
| | Use of RBF as panacea, depletion of resources; unintended consequences |
| | Insufficient ownership and accountability; corruption |
| | Technical assistance-intensive to establish new/sustainable systems |
| | Complexity of RBF; significant time for design and implementation |
| | Skepticism about RBF mechanisms |
5 | Potential for US involvement with RBF for NCDs | Collect best practices from RBF; assess epidemiological situation |
| | Engage stakeholders; take participatory approach |
| | Pilot programs to test applicability of RBF for NCDs; increase funding for NCDs |
| | Include NCDs as part of package of essential health services; avoid dichotomy between communicable and non-communicable diseases |
The challenge for NCDs arises in determining which services need to be purchased, how and by whom, as well as how to measure the results. Patricio Marquez, Lead Health Specialist at the World Bank suggested using existing RBF programs to identify the entry points to cover some of the interventions that may have an impact on the onset of NCDs [
48].
Experts on NCDs identified various approaches for utilizing RBF mechanisms. Rachel Nugent, an expert on NCDs from the University of Washington, noted that the need for patient involvement in managing NCDs lends itself to incentive programs for both prevention and treatment related to NCD risk factors. Incentives could be provided to eat and exercise correctly, for tobacco cessation, or for adherence to medication [
49]. Patricio Marquez suggested using HIV programs as an opportunity to target certain forms of cancer, using maternal health schemes to also affect cervical cancer or using TB interventions as the entry point for tobacco control programs [
48]. Other examples of NCDs with infectious components include Hepatitis B and liver cancer, diabetes and TB, and lymphoma and malaria [
50]. An important aspect of MCH programs is childhood malnutrition, a leading risk factor for NCDs such as diabetes. Given the interconnectedness of communicable diseases, MCH and NCDs, health intervention efforts could be complementary. In the words of Patricio Marquez:
“It’s time to start thinking of the patient as a whole. Let’s stop seeing the patient or populations by disease. Because the interconnection between communicable and non-communicable diseases is there. So the question is how we adopt a global health strategy that avoids the dichotomy. By doing that we will be able to prevent the proliferation of government programs that in some cases create distortions”[
48].
Other key informants stressed the importance of taking a policy-level approach, often in addition to individual-level interventions. Dr. Sameh El-Saharty, a Senior Health Policy Specialist at the World Bank, listed several interventions that could benefit from a RBF strategy including banning tobacco advertisements and smoking in public places, encouraging exercise and seatbelt use, and incentivizing industries to use healthier foods. At the institutional or sub-national level the same mechanisms could be used to incentivize institutions to adopt programs or revise procedures, such as clinical protocols or health insurance programs, and to introduce screening programs for conditions such as cancer and cardiovascular disease. RBF could be used at the sub-national level to incentivize states, regions, or districts, or on the supply-side to provide incentives at the service delivery level [
51]. From supply-side to demand-side, national-level to individual-level, and NCD-focused to basic health service-focused programs, experts agreed that RBF could be used to affect NCDs in resource-limited countries.
Key elements of successful RBF programs
Experience from RBF programs has provided lessons useful when designing NCD programs using RBF mechanisms. RBF experts specified numerous key factors in designing and implementing successful RBF programs, such as political commitment, a participatory approach, clearly defined rules and targets, strong verification systems, and robust measuring and evaluation. A summary of these themes can be seen in Table
3. One expert on RBF from the World Bank said:
“RBF is so different from traditional input-based financing, in order for it to work, you need political commitment, and you need commitment from technical levels of Ministry of Health and Ministry of Finance, to be able to make the whole thing work. People really have to be convinced… And that’s not unique to RBF, but it’s more important in RBF because it’s such a different system.”
Various other elements were stated as being key to developing a successful program. Testing mechanisms on a small scale through pilot projects is crucial in identifying potential challenges early on. Joseph Naimoli, an expert from USAID/CDC, stated:
“They spend a considerable amount of time designing these programs and getting them right. And also, they spend some time in piloting and testing these things in a smaller geographic area – like at the district level or multiple districts – and spend some time understanding what the potential obstacles are and trying to work those obstacles out. I think those kinds of programs are the most successful”[
52].
Flexibility in implementation and the ability to adapt to context-specific conditions, having basic health infrastructure in place, and developing open communication are also central.