Discussions on pay-for-performance programs (P4P) and results-based financing (RBF) are often linked to discussions on intrinsic vs extrinsic motivation, job satisfaction, and a gap between knowledge and performance. However, among many of the studies published on P4P programs, most notably the study Paying Primary Health Care Center for Performance in Rwanda which demonstrated the impact of these programs on quality of care and coverage, very few focus on mechanisms through which such results are achieved or not achieved. Existing evaluation studies of RBF programs do not provide rigorous evidence on the impact of these programs on health worker motivation. While it is important to focus on service delivery results, such as increased number of antenatal care visits or institutional deliveries, understanding if and how P4P influences health workers will shed light on mechanisms through which contingent rewards change provider attitudes and behaviors, helping us to get a glimpse into what has sometimes been referred to as the “black box.”
In this post I would like to share some important lessons on health worker motivation and P4P based on my research in Afghanistan.
In the summer of 2010, Afghanistan launched an RBF project in 11 of its 34 provinces with the aim of improving coverage of maternal and child health services within the existing health system. 442 facilities across the 11 provinces were randomized into treatment and control arms. Facilities in the treatment arm were supposed to receive quarterly payments for a set of indicators such as number of ANC visits, number of children receiving DPT3 etc., adjusted by the overall quality of care score. Facilities were free to decide on the way these payments would be distributed among the health workers. Unlike in Rwanda, the control arm did not receive additional funding or support.
It appears that financial incentives in the form of P4P do not necessarily improve health worker motivation, and in some cases can actually undermine it. According to the results of the study of the P4P program in Afghanistan, no statistically significant changes were observed in health worker motivation in the intervention group as compared to the control group (p-value>0.05). Contrary to our expectations, external motivation did not increase among health workers in facilities that were assigned to receive P4P. Some explanation can be found in the design and implementation details of the Project Below I will highlight a few key areas that were found to influence motivation. Examining these can provide important lessons for future programs.
CHW Visiting a woman Health Services Support Project. Photo © Nasrat Ansari
It’s all about expectations!
Heightened expectations were not met due to the relatively small size of incentive payments in proportion to total salaries. While excitement about RBF is good, we need to manage expectations and not create false ones. When preparing RBF program and discussing it with stakeholders, there should be a clear message that the overall RBF resource envelope will be constrained with what is feasible for the country to adopt after the project support is no longer there. RBF does not mean large increases in salaries, it simply means that some portion of salaries will reflect performance, however it is measured.
Delays in payments and the link between performance and rewards
Long payment delays, mostly due to verification requirements, undermined the perceived link between performance and payments. In countries with weak information systems we may want to first invest in upgrading these in order to minimize additional reporting and verification requirements. Whatever system is set up, verification should not become a long burdensome process that delays making of payments. We should balance between the costs and time requirements of verification system and its benefits.
Investments in infrastructure and medical supplies
The main obstacles to good performance my lie not in lack of motivation but in lack of means to perform. In Afghanistan, I met several male health workers who stayed through the Taliban times and continued treating female patients even though it was strictly forbidden and they could have been executed on the spot. Giving these doctors and nurses means to perform, such as facilities with reliable electricity and water supply, has to be the first step in improving their performance. In the RBF program in Afghanistan, there were no accompanying efforts to improve working conditions at facility level. In Rwanda, an average of 23% of P4P funds were allocated for overall facility-level investments such as infrastructure and medical supplies. This is a common feature in many RBF programs that was not implemented in Afghanistan.
“Motivation contracts” and other tools to ensure transparency and perception of a fair process
Facility management was free to decide on the intra-facility distribution rules for the P4P funds, and while increased provider autonomy is usually seen as being positive, it seems that certain mechanisms (such as clearly spelled out contracts, staff meetings, and bulletin boards) should be put in place first in order to ensure transparency and perception that the process is fair. There were three types of within-facility distribution schemes in Afghanistan: (1) in proportion to existing salary, (2) based on the contribution to performance indicators, and (3) equal amount to all facility staff. Based on MOPH administrative data, it was estimated that only a third of health workers in this study came from facilities where within-facility distribution was related to performance. At the same time, there were no clear formulas for estimating performance and contribution of individual staff members, and health workers did not have “motivation contracts” as in Burundi. Autonomy should be accompanied by capacity-building of facility managers and accountability mechanisms.
Almost half of health workers from treatment facilities stated they did not receive P4P. Thus, they either did not receive any payments or were not aware of having received them. One reason for this confusion was that bonus payments were paid out on the same account and at the same time as regular salaries. This may have made sense from the administrative point of view, but not from the point of view of health workers: Imagine suddenly receiving a larger salary than usual one quarter but not the next one. What would be your reaction, especially if you did not understand that the extra payment you received earlier was related to your facility’s performance?
These seem like small details but they highlight the importance of careful planning and the need to look at the entire health system before implementing P4P. With so much excitement around the concept, it is easy to miss practical details.