“Once the delivery is done, they have finished”: A qualitative study of perspectives on postnatal care referrals by traditional birth attendants in Ebonyi state, Nigeria

Background: While 79% of Nigerian mothers who deliver in facilities receive postnatal care within 48 h of delivery, this is only true for 16% of mothers who deliver outside facilities. Most maternal deaths can be prevented with access to timely and competent health care. Thus, the World Health Organization, International Confederation of Midwives, and International Federation of Gynecology and Obstetrics recommend that unskilled birth attendants be involved in advocacy for skilled care use among mothers. This study explores postnatal care referral behavior by TBAs in Nigeria, including the perceived factors that may deter or promote referrals to skilled health workers. Methods: This study collected qualitative data using focus group discussions involving 28 female health workers, TBAs, and TBA delivery clients. The study conceptual framework drew on constructs in Fishbein and Ajzen’s theory of reasoned action onto which we mapped hypothesized determinants of postnatal care referrals described in the empirical literature. We analyzed the transcribed data thematically, and linked themes to the study conceptual framework in the discussion to explain variation in TBA referral behavior across the maternal continuum, from the antenatal to postnatal period. Results: Differences in TBA referral before, during, and after delivery appear to reflect the TBAs understanding of the added value of skilled care for the client and the TBA, as well as the TBA’s perception of the implications of referral for her credibility as a maternal care provider among her clients. We also found that there are opportunities to engage TBAs in routine postnatal care referrals to facilities in Nigeria by using incentives and promoting a cordial relationship between TBAs and skilled health workers. Conclusions: Thus, despite the potential negative consequences TBAs may face with postnatal care referrals, there are opportunities to promote these referrals using incentives and promoting a cordial relationship between TBAs and skilled health workers. Further research is needed on the interactions between postnatal maternal complications, TBA referral behavior, and maternal perception of TBA competence. The study conceptual framework drew on constructs in Fishbein and Ajzen’s theory of reasoned action [7, 8]. This theory predicts behaviors that are under volitional control, that is where the person exercises control over the behavior. Thus, we applied it to advocacy for skilled care among TBAs. This theory assumes that behavior is predicted by intention, which is determined by attitudes toward and perceptions of social norms regarding the behavior [8]. Attitude is determined by the individual’s beliefs about the results of performing the action, referred to as behavioral beliefs, and the value placed on these results. Thus, if a person believes that the results will be positive, the theory predicts a positive attitude towards that behavior. However, if a person believes that the results will be negative, the theory predicts a negative attitude towards that behavior. Perceptions of social norms are determined by normative beliefs, that is a person’s beliefs about whether significant individuals approve or disapprove of their behavior. If a person believes that these significant individuals think he or she should perform an action, he or she will hold a positive subjective norm. However, if a person believes that these significant individuals think he or she should not perform an action, he or she will hold a negative subjective norm [8]. This study collected data primarily using focus group discussions (FGDs), which are useful for studying phenomena at an aggregate level and for assessing collective opinions [9]. FGDs can also aid in developing interventions to address public health problems that have local meaning and utility. In this case the study team intended to decipher if there was justification for and to inform the design of a field experiment on performance-based monetary incentives for TBA referrals in postnatal care. The study was conducted in July 2016 in Ebonyi State, South-Eastern Nigeria, where about 1 in 2 mothers does not receive postnatal care within the first two days of childbirth [3]. As part of a larger mixed methods study, the study team purposively selected 128 wards in Ebonyi State that had at least one primary health care facility with a health care provider offering maternal postnatal care. We identified these wards using the national facility census list and via consultations with officials in the State Ministry of Health. Each selected ward also had to have at least one TBA who lived and worked there. We recruited FGD participants from these wards. The recommended sample size for focus group discussions is 6–12 participants, as group interviews are difficult to manage above 12 [10]. The study team decided on FGD participant categories based on the key stakeholder groups involved in referrals for postnatal care. Thus, we purposively selected 10 TBAs, 10 TBA clients, and 8 health care providers from communities that had primary health care facilities offering postnatal care services by a skilled provider and at least one resident TBA (we intended to interview 10 health care providers, 2 of whom arrived at the interview venue after the discussion had held). At recruitment, potential participants were informed that the study was aimed at understanding postnatal care practice in their communities. Health care providers were recruited face-to-face through a monthly meeting in the primary health care board in the state capital. TBAs were identified in partnership with health care providers, and recruited face-to-face in the 2 cases where they were based in the state capital. Recruitment over the phone was done for 8 TBAs who lived outside the state capital. We recruited TBA clients by asking recruited TBAs for the name of at least one past client, who was then contacted by the study team. Recruitment over the phone was done for 8 TBA clients who lived outside the state capital. TBA clients were required to be beyond 42 days after the culmination of their first pregnancy so that they would have had the opportunity to choose to attend at least one postnatal visit. TBA clients with multiple past pregnancies were therefore qualified to join the study as well. None of the recruited participants declined the invitation to the study. We held one FGD with 8 health workers, another with 10 TBAs, and a final FGD with 10 TBA delivery clients (9 of whom came for the discussion with their newborns). Discussions were held in a quiet, secluded location in the State Teaching Hospital. Focus groups were conducted in either English or Igbo language depending on participant consensus. The discussions were audio-recorded with the consent of participants and facilitated using topic guides by an experienced qualitative researcher. One member of the research team took notes describing the group interaction. The topics in the discussion guide were informed by the review of the empirical literature and study conceptual framework and differed by participant type (see Additional file 1). The research team translated the audio-recordings in Igbo language into English, and transcribed all the recordings verbatim in Microsoft Word. The transcripts were anonymized with pseudonyms and entered into NVivo V.11. Data analysis began with reading of the transcripts at least twice to achieve immersion. Then, codes were derived by identifying phrases that captured key concepts on which majority of the group agreed on or did not object to, via open coding. That is, we developed codes based on the meaning that emerged from the data. We also identified perspectives that deviated from the group consensus and highlighted them in the analysis. Two team members independently coded the study transcripts. Codes that were similar were grouped into themes. For example, individual factors that influenced client preference for TBAs in the antenatal period were coded separately, and grouped subsequently into one theme [11]. In the discussion, we linked study themes to the study conceptual framework, to explain variation in TBA referral behavior across the maternal continuum, from the antenatal to postnatal period. The two members of the research team involved in facilitating the group discussions and analyzing the data are both medical doctors who have experience providing care at the primary, secondary, and tertiary level, and who have graduate training in public health. To prevent the researchers from imposing opinions informed by their background on the discussants, the study team and experts from other disciplines deliberated on the choice of questions and probes used in the topic guide, as well as the themes from the analysis. These deliberations ensured that the leading questions and probes were avoided in the topic guides and that the discussion was facilitated, analyzed, and interpreted to reflect the opinions of participants rather than the biases of the researchers. The study findings have been reported in line with the consolidated criteria for reporting qualitative research (COREQ) [12].