By Josephine Borghi

For many low- and middle-income country governments it is no longer a question of whether to introduce performance-based funding but how to. This is a result of the large scale promotion and funding of performance-based schemes by a number of donors who see this as a mechanism to ensure that aid money achieves results. There is also some support for performance-based funding as a lever for systems change. However, the evidence base from low- and middle-income countries remains weak.  Personally, I need more convincing.

Performance-based funding in Cambodia

A study recently published by Matsuoka et al. provides new insights into the effect of a GAVI-funded performance-based financing system to improve maternal and child health services in one operational district of Cambodia. Based on facility service statistics, Matsuoka et al. found that the introduction of GAVI funding led to an increase in antenatal care visits (up to 4 visits) and vaccination coverage. However, as the performance payment relies on facility reports, and in the absence of independent verification of these reports, the change in service use could be driven by the simple incentive to over-report, or ‘gaming’.

What about the effects on service quality? 

Although quality of care was not explicitly measured, Matsuoka et al. indicated that system constraints had not been overcome by the scheme. Notably, midwifery capacities and knowledge remained limited, and equipment, medical supplies and drugs shortages prevailed.

Why did utilisation increase, if indeed it did?

The authors suggest that longer opening hours may be partly responsible. Greater financial autonomy and revenue resulted in health workers investing in drugs and supplies in some instances to improve facility resource availability.  The scheme was felt to strengthen relations between district and facility and improve motivation of providers, reducing the likelihood of private practice – both valuable achievements. This all sounds plausible; however, I still have a nagging doubt over whether the increase was driven by changes in reporting practices, or whether it mirrors increases that would have happened anyway.

Lessons learnt 

The paper highlights a number of important lessons for performance-based financing in low-income settings. The authors suggest that to improve quality of care, quality indicators need to be built into the performance contract. A critical question for me here is how do we measure and verify quality? The authors underline the importance of independent verification to ensure facility reporting is accurate, yet achieving this may not be so straight forward in low-income settings or without cost.

I agree with the authors that performance-based funding cannot be expected to transform the health system, and that it should, if at all, be integrated with broader health system strengthening strategies. Matsouka et al.’s findings of improved relations among facility and district staff and some increased responsiveness to population needs are encouraging. It will be interesting to see to what extent these effects are sustained. I am also curious as to whether greater financial autonomy and increased supervision would not achieve the same results, in the absence of performance contracts, at considerably lower cost.

On a final note, I would urge future evaluations to triangulate data; we really need household survey data to validate changes in service use for performance-based funding schemes that rely on facility service statistics. I also have to ask my favourite question: are these schemes cost-effective and if so, how can they be sustained by governments when donor funds run out?

Dr Josephine Borghi is Health Economics Editor of Health Policy and Planning and Senior Lecturer in Health Economics and Policy at the London School of Hygiene and Tropical Medicine.

 

Further reading:
http://www.thecochranelibrary.com/details/file/1479345/CD007899.html
http://www.rbfhealth.org/news/item/results-based-financing-health-rbf-whats-all-the-fuss
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3040374/pdf/BLT.10.077339.pdf