Reflections on the unintended consequences of the promotion of institutional pregnancy and birth care in Burkina Faso

The policy of institutional delivery has been the cornerstone of actions aimed at monitoring and achieving MDG 5. Efforts to increase institutional births have been implemented world-wide within different cultural and health systems settings. This paper explores how communities in rural Burkina Faso perceive the promotion and delivery of facility pregnancy and birth care, and how this promotion influences health-seeking behaviour. A qualitative study was conducted in South-Western Burkina Faso between September 2011 and January 2012. A total of 21 in-depth interviews and 8 focus group discussions with women who had given birth recently and community members were conducted. The data were analyzed using qualitative content analysis and interpreted through Merton’s concept of unintended consequences of purposive social action. The study found that community members experienced a strong pressure to give birth in a health facility and perceived health workers to define institutional birth as the only acceptable option. Women and their families experienced verbal, economic and administrative sanctions if they did not attend services and adhered to health worker recommendations, and reported that they felt incapable of questioning health workers’ knowledge and practices. Women who for social and economic reasons had limited access to health facilities found that the sanctions came with increased cost for health services, led to social stigma and acted as additional barriers to seek skilled care at birth. The study demonstrates how the global and national policy of skilled pregnancy and birth care can occur in unintentional ways in local settings. The promotion of institutional care during pregnancy and at birth in the study area compromised health system trust and equal access to care. The pressure to use facility care and the sanctions experienced by women not complying may further marginalize women with poor access to facility care and contribute to worsened health outcomes. Situated in West Africa, Burkina Faso is among the world’s poorest countries and has a high burden of maternal deaths, with an estimated maternal mortality ratio of 400 per 100 000 live births in 2013 [13]. In Burkina Faso births with skilled attendants take place in health facilities with few exceptions. Hence, the promotion of facility care has been the core effort aiming to reduce maternal mortality. A primary objective in the Ministry of Health’s (MoH) strategic plan to reduce maternal mortality is to increase the proportion of women giving birth with skilled assistance from 50 to 80% between 2006 and 2015 [14]. Among the factors that limit the utilization of facility care during pregnancy and at birth in Burkina Faso are distance to the health facility, financial constraints, and women’s limited decision-making power [15–17]. In this context, a subsidiary policy for pregnancy and birth care has been implemented since 2006 to reduce financial barriers to facility care [14,18]. Poor quality of care in primary health facilities has also been proposed as an explanation of frequent home births, nevertheless users’ assessment of care remains largely favourable [19–22]. The study was conducted in two health districts in the South-Western part of Burkina Faso, Banfora and Mangodara. The annual number of expected deliveries for these health districts in 2011 was 24 500 for a population of approximately 500 000 [23]. The proportion of deliveries taking place with a skilled attendant was 67% in Banfora and 59% in Mangodara [23]. At the time of the study, the area had 39 primary health centres (Centres de santé et de promotion sociale, CSPS) and one regional referral hospital in Banfora town. In the study area, subsistence farming is prevalent and maternal literacy remains very low. A cohort study among pregnant women in the area indicated that 83% had never attended school [24]. The main spoken language is Dioula. The data collection lasted from September 2011 to January 2012, as part of a study on the quality of facility birth care in four health centres in the Banfora region. Assuming that facility care would differ between urban and rural areas and also taking into consideration the monthly number of births, one urban, one semi-urban and two rural facilities were purposively selected to achieve maximum diversity. According to health district data, the health centres had an assisted delivery rate varying from 48 to 77% [25,26]. The health centres varied in size, and had from 2–12 health workers with different levels of training. Their infrastructure also varied substantially; some had electricity and running water, while others relied on torches as the only source of light; and water was provided from wells situated up to one kilometre from the health centre. A total of 21 in-depth interviews (IDIs) and 8 focus group discussions (FGDs) with women who recently experienced childbirth, their partners and community members were conducted, Table 1. A research assistant trained in sociology and fluent in Dioula and French recruited the participants in the IDIs and FGDs. She was assisted by community health workers in semi-urban and rural communities in the areas covered by the four health centres. Participants were purposively selected for the interviews, on the basis that they or their partner had given birth within the last three months. The age of the interviewees ranged from 18 to 42 years, they had none to 13 living children, and lived from one to 20 km from their local health centre. A good majority relied on subsistence farming, and only a handful had attended school. The recruitment of informants ended at the point of data saturation. Both IDIs and FGDs were conducted in Dioula; AM conducted the IDIs with the research assistant as an interpreter, while the research assistant facilitated the FGDs in Dioula with AM as an observer. The IDIs took place in the interviewees’ home, while the FGDs took place outdoor in a public place in the community where the participants lived. Both IDIs and FGDs lasted between 45 and 90 minutes. The interview guides included open-ended questions about practices during pregnancy and childbirth, the place of birthing and the personal, as well as community perceptions on the care provided in the health centres, S1–S3 Interview Guides. The co-authors contributed to the development of the interview guides, which were translated from French to Dioula by a certified Dioula translator. Both IDIs and FGDs were recorded and transcribed verbatim in Dioula before translation into French. During fieldwork, AM carefully read the transcripts and discussed the meaning of the verbatim transcripts and the culturally embedded expressions with the research assistant. After data collection, the transcripts were examined by drawing upon qualitative content analysis [27]. After familiarization with the dataset, initial codes were identified in the interviews. These codes were grouped into categories and subsequently into themes. For example, the quote ‘If you don’t do the weighing [attend antenatal care (ANC)] she [health worker] will say “Why haven’t you come to be weighed [attended ANC]. It’s when your child is sick you’re coming” She growls like that. She will care for you, but she will disrespect you while caring for you.’ will be grouped into the category imposing a sanction by use of verbal reprimands and consequently into the theme sanctions for not using the pregnancy and childbirth services as prescribed. Ethical clearance was provided by the national health research ethics committee of the Ministry of Health, Burkina Faso (Ref 2011-9-57, Comité d’éthique pour la Recherche en Santé, Ministère de la Santé, Ouagadougou, Burkina Faso). The Banfora regional health directorate Chief and the Heads of Banfora and Magodara health districts provided administrative authorisations. As a great majority of the study participants were illiterate, the research assistant would read a written consent form in Dioula before signed or thumb-printed informed consent was obtained from all interviewees. When names have been used, these were changed to preserve anonymity.