Background: In Malawi, as in many low-and middle-income countries, health facility committees (HFCs) are involved in the governance of health services. Little is known about the approaches they use and the challenges they face. This study explores how HFCs monitor the quality of health services and how they demand accountability of health workers for their performance. Methods: Documentary analysis and key informant interviews (7) were complemented by interviews with purposefully selected HFC members (22) and health workers (40) regarding their experiences with HFCs. Data analysis was guided by a coding scheme informed by social accountability concepts complemented by inductive analysis to identify participants’ perceptions and meanings of processes of social accountability facilitated by HFCs. Results: The results suggest that HFCs address poor health worker performance (such as absenteeism, poor treatments and informal payments), and report severe misconduct to health authorities. The informal and constructive approach that most HFCs use is shaped by formal definitions and common expectations of the role of HFCs in service delivery as well as resource constraints. The primary function of social accountability through HFCs appears to be co-production: the management of social relations around the health facility and the promotion of a minimum level of access and quality of services. Conclusions: Policymakers and HFC support programs should take into account the broad task description of HFCs and integrate social accountability approaches in existing quality of care programs. The study also underscores the need to clarify accountability arrangements and linkages with upward accountability approaches in the system.
This study is part of a larger research project on social accountability in maternal health service delivery in Mzimba North and South districts, situated in the Northern Region of Malawi. Data collection took place between April 2015 and June 2016. The district was purposefully selected in the context of a partnership between the researchers, an NGO and district authorities. Ethical approval was obtained from the National Health Science Research Committee of the Ministry of Health in Malawi (NHSRC#15/03/1398). Our study goal was to gain a comprehensive understanding of experiences with, and perceptions on, the role of HCACs as social accountability interfaces and the approaches HCACs use to address poor service quality and performance in rural health centres. Although HCACs oversee all services offered by health centres, the focus of this study was on maternal health services. A total of 41 HCACs were eligible in Mzimba district, based on their association to a rural health centre (not urban health centres or rural health posts) and a minimum level of functionality (according to district health authorities). Out of these 41 HCACs, we selected all HCACs in the Northern part of the district (n = 12) and a sample of HCACs in the Southern part of the district. This district is divided into six clusters of health centres that are constructed by the DHMT for their supervision visits on the basis of the geographical concentration of health centres. Out of the six clusters, we randomly selected three that hosted a total of 10 HCACs associated with rural health centres, resulting in 22 included sites. The proportion of births assisted by skilled health personnel in the Northern Region is reported to be 90.6% in 2015 [29]. Maternal healthcare, like most other care, is provided free of user fees. The health centres in the study sites had an average of two skilled birth attendants, below the planned four. The number of deliveries per year per health centre (2014) ranged from 36 to 723 with a mean of 326 deliveries per year. The goal of data collection was to gather evidence from multiple sources on social accountability approaches, perceptions, outcomes and contexts associated with these approaches. In each HCAC site, we collected available minutes from HCAC meetings, letters pertaining communication between HCAC and health workers and we conducted semi-structured interviews with HCAC members, health workers and District Health Management Team (DHMT) members. A facility checklist was used to document the conditions in the health centre and evidence on the presence of social accountability tools, such as scorecards, monitoring sheets, complaint and suggestion boxes, or a phone number to call in case of a complaint. Two key informants, an NGO representative and a researcher, provided guidance and feedback on the data collection tools, choice of participants and interpretation of data during analysis. They were both community development professionals and had experience with the set-up and training of HCACs. The interview guides for both HCAC and health workers covered a range of topics including a general section on perceived roles and functioning of HCACs and a specific section to obtain perceptions on, and examples of, approaches to the monitoring of quality of services, complaint and feedback processes. The guide consisted mostly of open questions but also included some statements with yes/no answers for probing purposes (e.g. ‘HCACs should monitor health workers’) and some closed questions to cover HCAC characteristics. Concepts from Bovens’ conceptual framework were combined with concepts as defined in the HCAC training manual, in particular on complaint and feedback management (see Table 1). Scans from relevant documents were obtained from HCAC chairpersons and secretaries. We collected a total of 12 HCAC meeting minutes. Additional interviews were held with DHMT members to get their views on, and experiences with, HCACs in relation to the performance of health workers. Questions included their role in supporting HCACs, and supervision and accountability relations in the district health system. The DHMT shared copies of minutes of three district-level meetings on HCAC, and two copies of letters from HCACs addressed to them. Interview guides The data collection at health centre level was divided between two researchers (JM and CI). During 2 months, they jointly visited all sites. The travel schedule to visit health centres was set based on the participants’ availability, taking into account the accessibility and relative distance between health centres. In each site, two or three interviews were held, one with the HCAC chairperson and one or two with health workers. Interviews were conducted with informed consent and lasted a maximum of 60 min. Interviews with most HCAC members were in the local language Tumbuka, assisted by a translator while interviews with health workers were in English. Interviews with DHMT members were divided between the three researchers (EL, JM, CI). In each site, participants were purposefully sampled to represent the HCAC, health centre management (the officer in-charge who is usually a clinician and the contact person for the HCAC) and health workers (registered or enrolled nurse-midwives). Nurses were targeted because they are most involved in maternal healthcare, the focus of the larger research project. The DHMT provided contact data of the HCAC chairperson and the health centre in-charge and participants were invited by phone or letter. Table 2 provides an overview of the interview participants. A total of 62 interviews were conducted in 22 health centre sites. Participants During data collection, observations on interview data and emerging findings per site were noted down. Regular reviews were conducted between the three researchers (JM, CI and EL) to check consistency in data collection. All interviews were digitally recorded after consent and transcribed by local transcribers and the researchers. Interviews in Tumbuka were translated into English. The field notes, transcripts and documents were all uploaded in the qualitative data analysis program Maxqda (version 11). We used deductive thematic analysis methods during the first phase of data analysis, guided by Bovens’ concepts of an accountability process, applied to all data sources. This analysis supported the identification and analysis of patterns in the ways in which HCACs facilitate social accountability and the outcomes associated with this process [30]. Three researchers did this separately for each available data source per HCAC and according to a codebook with predefined codes in the MaxQDA database. Data from all sites were then combined as the focus was on identifying common patterns and descriptions rather than differences. The three researchers discussed the common themes, and EL integrated the analysis to a final set of themes and sub-themes. EL also performed a second analysis on context, including the identification of perceptions, explanations and motivations of participants and contextual data in the interview and documentary data.