A comparative study of community verification processes in the context of performance-based financing in Mali and Burkina Faso

PurposeThis article compares the processes of community verification (CV) and user satisfaction surveys during the implementation of performance-based financing (PBF) in Mali and Burkina Faso.Design/methodology/approachThe authors adopted a qualitative approach based on a multiple-case study design. Data were collected from August 10 to 25, 2017, in Mali, and from January to May 2016 in Burkina Faso. In Mali, 191 semi-structured interviews were conducted with investigators (people who collect information from health centre users in the communities, using survey tools), users, users' relatives, and health workers in three of the 10 health districts in the Koulikoro region. In Burkina Faso, 241 non-participatory observation sessions were recorded in a research diary, and 92 semi-structured interviews and informal discussions were conducted with investigators, community verifiers, users, PBF support staff at the national level, and administrative staff in one of the 15 health districts involved in PBF. The data were analysed inductively.FindingsIn both Mali and Burkina Faso, the delayed availability of survey forms led to a delay in starting the surveys. In Mali, to get off to a quick start, some investigators went to health centres to conduct the sampling with their supervisors. In both countries, investigators reported difficulties in finding certain users in the community due to incorrect spelling of names, lack of telephone details, incomplete information on the forms, common or similar sounding names within the community, and user mobility. There was little interference from health workers during user selection and surveys in both countries. In both countries, many surveys were conducted in the presence of the user's family (husband, father-in-law, brother, uncle, etc.) and the person accompanying the investigator. Also in both countries, some investigators filled in forms without investigating. They justified this data fabrication by the inadequate time available for the survey and the difficulty or impossibility of finding certain users. In both countries, the results were not communicated to health centre staff or users in either country.Research limitations/implicationsCV and user satisfaction surveys are important components of PBF implementation. However, their implementation and evaluation remain complex. The instruments for CV and user satisfaction surveys for PBF need to be adapted and simplified to the local context. Emphasis should be placed on data analysis and the use of CV results.Originality/valueThere are similarities and differences in the CV process and user satisfaction surveys in Mali and Burkina Faso. In Mali, the data from the user satisfaction survey was not analyzed, while in Burkina Faso, the analysis did not allow for feedback. The local non-governmental organisations (NGOs) that carried out the CV were pre-financed for 50% of the amount in Mali. In Burkina Faso, community-based organisations (CBOs) were not pre-financed. The lack of financing negatively impacted the conduction of the surveys. In Mali, fraudulent completion of survey forms by interviewers was more common in urban than in rural areas. In Burkina Faso, the frauds concerned consultations for children under five years of age. In Burkina Faso, the survey form was not adapted to collect data on the level of satisfaction of the indigent.Key messagesThere were similarities and differences in the community verification (CV) processes in Mali and Burkina Faso.In both Mali and Burkina Faso, tracing users within their community was difficult for several reasons, including incorrect or incomplete information on forms, common or similar names, and user mobility.In both countries, there was no feedback on the results of the CV process to health centre staff or users.Survey forms were falsified by investigators in both countries. In Mali, falsification was more common in urban than in rural areas. In Burkina Faso, falsification was more often observed for consultations for children under five years of age.