To strengthen the health care delivery, the Burundian Government in collaboration with international NGOs piloted performance-based financing (PBF) in 2006. The health facilities were assigned - by using a simple matching method - to begin PBF scheme or to continue with the traditional input-based funding. Our objective was to analyse the effect of that PBF scheme on the quality of health services between 2006 and 2008. We conducted the analysis in 16 health facilities with PBF scheme and 13 health facilities without PBF scheme. We analysed the PBF effect by using 58 composite quality indicators of eight health services: Care management, outpatient care, maternity care, prenatal care, family planning, laboratory services, medicines management and materials management. The differences in quality improvement in the two groups of health facilities were performed applying descriptive statistics, a paired non-parametric Wilcoxon Signed Ranks test and a simple difference-in-difference approach at a significance level of 5%. We found an improvement of the quality of care in the PBF group and a significant deterioration in the non-PBF group in the same four health services: care management, outpatient care, maternity care, and prenatal care.The findings suggest a PBF effect of between 38 and 66 percentage points (p<0.001) in the quality scores of care management, outpatient care, prenatal care, and maternal care. We found no PBF effect on clinical support services: laboratory services, medicines management, and material management.The PBF scheme in Burundi contributed to the improvement of the health services that were strongly under the control of medical personnel (physicians and nurses) in a short time of two years. The clinical support services that did not significantly improved were strongly under the control of laboratory technicians, pharmacists and non-medical personnel.

Background: Performance-based financing (PBF) was first implemented in Burundi in 2006 as a pilot programme in three provinces and was rolled out nationwide in 2010. PBF is a reform approach to improve the quality, quantity, and equity of health services and aims at achieving universal health coverage. It focuses on how to best motivate health practitioners. Objective: To elicit physicians’ and nurses’ experiences and views on how PBF influenced and helped them in healthcare delivery. Methods: A qualitative cross-sectional study was carried out among frontline health workers such as physicians and nurses. The data was gathered through individual face-to-face, in-depth, semi-structured interviews with 6 physicians and 30 nurses from February to March 2011 in three hospitals in Gitega Province. A simple framework approach and thematic analysis using a combination of manual technique and MAXQDA software guided the analysis of the interview data. Results: Overall, the interviewees felt that the PBF scheme had provided positive motivation to improve the quality of care, mainly in the structures and process of care. The utilization of health services and the relationship between health practitioners and patients also improved. The salary top-ups were recognized as the most significant impetus to increase effort in improving the quality of care. The small and sometimes delayed financial incentives paid to physicians and nurses were criticized. The findings of this study also indicate that the positive interaction between performance-based incentive schemes and other health policies is crucial in achieving comprehensive improvement in healthcare delivery. Conclusions: PBF has the potential to motivate medical staff to improve healthcare provision. The views of medical staff and the context of the area of implementation have to be taken into consideration when designing and implementing PBF schemes.

Background: Several developing countries, especially in Africa, have implemented performance-based financing (PBF) schemes with the aim of improving healthcare provision. PBF was first implemented in Burundi in 2006 as a pilot programme in three provinces and was rolled out nationwide in 2010.Objective: To enrich existing studies on Burundi in three ways. Firstly, by evaluating the effect of PBF on maternal care at primary and hospital levels; secondly, on the possession of maternity logbooks for maternal care records; and thirdly, how the amount of subsidies influences healthcare outputs.Design: We used data from repeated cross-sectional surveys in 500 households (intervention group: 225; control group: 275) conducted in 2006 and 2008. A total of 274 women, aged 15–49, who had recently given birth, were interviewed about the use of maternal healthcare and the possession of maternity logbooks. We performed a difference-in-differences analysis using pooled cross-sectional survey data from 2006 and 2008.Results: We found that PBF is associated with an increased institutional deliveries probability of 39.5 percentage points (p < 0.01) – a relative improvement of 81.8%. Institutional deliveries probability increased significantly only at health centre level by 33.6 percentage points (p < 0.01), a  relative rise of 80.6%. There is an indication of a positive spillover effect of PBF on the possession of maternity logbooks. We found no PBF effect on the number of antenatal care visits and anti-tetanus immunization.Conclusions: Our findings suggest that institutional delivery highly improved because it came from a low baseline and its unit payment was relatively high, leading health workers to promote its use. The fact that deliveries mainly increased in health centres and not in hospitals may be explained by the context of how health delivery is organized in Burundi. Health policymakers have to determine the appropriate financial incentives that best influence the improvement of each health service.