Performance-based financing (PBF) has attracted considerable interest from governments and aid agencies in low-income countries as a means to increase productivity and quality of health-care providers. In Africa alone, more than 35 countries are in the process of introducing payment methods that reward performance. The Burundi Government has gradually rolled out PBF across all provinces between 2006 and 2010. Our aim was to evaluate whether this has led to changes in the quantity and quality of health-care services provided.
We exploited the staggered rollout of PBF to apply a difference-in-differences approach using data from 2006 (baseline), 2008, and 2010. Cross-sectional data were collected among a random sample of 3200 households and panel data among 75 randomly selected health-care facilities across intervention and control provinces.
We found that PBF increased the probability of women delivering in an institution by 21 percentage points (p=0·0001), the probability of using antenatal care by 7 percentage points (p=0·067), and the use of modern family planning services by 5 percentage points (p=0·002). The overall facility quality score as constructed by external audits increased by 46%, but no effect of PBF was found on the quality of care as reported by patients. We also found no effects of PBF on vaccination rates, and the satisfaction with waiting times decreased. There is no strong evidence for unequal effects of PBF across socioeconomic groups.
The introduction of PBF in Burundi has led to significant increases in mother and child care utilisation and in the quality scores of health-care facilities. Given the rising popularity of this financing strategy, more robust evidence on its effects is warranted.
We thank the BMG Innovation fund, the Rotterdam Global Health Initiative, and Cordaid Netherlands for funding this research. EvdP is a Postdoctoral Fellow of the Netherlands Organisation for Scientific Research—Innovational Research Incentives Scheme—Veni project 451-11-03.
IB and EvdP conducted data analysis and writing of the abstract. RS, OB, and GL conducted data collection, data analysis, and implementation of the intervention. FvdL contributed to the implementation of the intervention, and EvD coordinated the study.