Background: In many low to middle income countries, traditional birth attendants (TBAs) play various roles (e.g., provision of health education, referral to hospitals, and delivery support) that can potentially improve women’s access to healthcare. In Tanzania, however, the formal healthcare systems have not acknowleded the role of the TBAs. TBAs’ contributions are limited and are not well described in policy documents. This study aimed to examine the perspectives of both TBAs and skilled birth attendants (SBAs) to clarify the role of TBAs and issues impacting their inclusion in rural Tanzania. Methods: We used a qualitative descriptive design with triangulation of investigators, methods, and data sources. We conducted semi-structured interviews with 15 TBAs and focus group discussions with 21 SBAs in Kiswahili language to ask about TBAs’ activities and needs. The data obtained were recorded, transcribed, and translated into English. Two researchers conducted the content analysis. Results: Content analysis of data from both groups revealed TBAs’ three primary roles: emergency delivery assistance, health education for the community, and referrals. Both TBAs and SBAs mentioned that one strength that the TBAs had was that they supported women based on the development of a close relationship with them. TBAs mentioned that, while they do not receive substantial remuneration, they experience joy/happiness in their role. SBAs indicated that TBAs sometimes did not refer women to the hospital for their own benefit. TBAs explained that the work issues they faced were mainly due to insufficient resources and unfavorable relationships with hospitals. SBAs were concerned that TBAs’ lacked formal medical training and their actions could interfere with SBAs’ professional work. Although there were no between-group interactions at the time of this study, both groups expressed willingness to collaborate/communicate to ensure the health and lives of mothers and babies. Conclusions: TBAs and SBAs have different perceptions of TBAs’ knowledge and skills, but agreed that TBAs need further training/inclusion. Such collaboration could help build trust, improve positive birth experiences of mothers in rural Tanzania, and promote nationwide universal access to maternal healthcare.
We utilized a descriptive cross-sectional design with a qualitative approach. Data were collected through individual interviews with TBAs and focus group discussions with SBAs working in a district hospital in rural Tanzania. To ensure a comprehensive understanding of the studied phenomena, triangulation is recommended for qualitative studies [16]; therefore, we used triangulation of investigation, methods, and data sources. Investigators included both Japanese and Tanzanian researchers to involve insider and outsider perspectives. We sourced data by exploring the perspectives of both TBAs and SBAs. The different methods were chosen in consideration of the environment where participants would feel comfortable, making it easy for them to talk. The study setting was a community in the Korogwe District, which is located in Northeast Tanzania, a rural area of the country. Tanzania comprises 940,000 km2. In the past four decades, its population has grown by approximately four times (more than 50,000,000), owing to high birth rates and a decrease in mortality rates [3, 17]. The district of Korogwe is spread over 3756 km2 and comprises 132 villages [17]. The main economic activities in the region include agriculture, horticulture (both of which are performed with the natural resources found in the region), and game parks. In 2012, the Korogwe district had a rural population of 242,038 people and an urban population of 68,308 people [18]. It has one public and two private hospitals, 59 dispensaries, and three health centers. With the support of a local collaborator, the first and second author conducted interviews with TBAs in one of the villages located in a mountain area; the nearest town is 10 km away from the village. They conducted focus group discussion with SBAs in a private room of a district hospital. We used purposeful sampling to identify and collect data from individuals who were information-rich in terms of research purposes. We chose to include both TBAs and SBAs to incorporate the perspectives of both sides. For TBAs, the inclusion criteria were that they must (1) be an active TBA, (2) be able to read and speak Kiswahili, and (3) agree to participate in the study. In our study, SBAs comprised of nurses, midwives, and doctors working in the region. They were invited to participate if they met the inclusion criteria: (1) working in a maternity ward (or working close to this ward), (2) able to speak Kiswahili, and (3) provided consent to participate in the study. For both groups, the exclusion criteria were (1) having never met TBAs, and (2) having no specific perceptions toward this group. To perform the interviews, the second author asked the village leaders to invite TBAs who worked in their communities to participate in this study. Since we had no previous knowledge of the number of TBAs in the area, we had initially planned to interview all TBAs who eventually appeared in the interview site. Owing to TBAs not being publicly recognized as professionals, we considered that performing group interviews would be a difficult and not very suitable task; hence, we planned individual interviews among this group. The semi-structured interview guide was created by the first author and later analyzed by the third author to determine if any other questions should be included according to the relevant available literature. The English versions of the interview guides for SBAs and TBAs are attached as supplemental files 1 and 2. After completing the development of the interview questions, the second author translated them from English to Kiswahili. When participants arrived at the interview site, we explained the aims and procedures of the study (including how we would record our conversations) and then asked for their permission to record the interviews. The first author led the interviews in English; the second author acted as an interpreter who translated between English and Kiswahili. The finalized semi-structured interview guide included questions related to TBAs’ activities, recently conducted deliveries, the support they provided to pregnant women, referral cases, their perceptions of TBAs’ roles, their lives aside from the TBA work, and their relationships with healthcare personnel and institutions. TBA participants came to the village by foot or motorbike. Data collection took place in a classroom of a school in the village. All interviews were conducted with assistance from the second author, acted as an interpreter, who translated between English and Kiswahili. Nonetheless, during the interviews, participants would eventually speak in Kisambaa (the local language); however, the second author (interpreter) understands both languages, so translation was not hindered. The interviews were digitally recorded, transcribed into Kiswahili, and translated into English by the same interpreter. Two authors (the first and fifth authors) reviewed all of the translated data. For SBAs, we planned focus group discussions (FGDs) so that participants could share their perceptions toward maternal & child health and the roles of TBAs working in the community. The first and second authors acted as the facilitator and the interpreter, respectively. The discussion topics included (1) hospital daily maternal care situations and the corresponding issues and solutions, and (2) the possibility of a collaboration with the TBAs working among pregnant women in rural areas. The FGDs were conducted in Kiswahili, recorded with participants’ consent, and transcribed (in the Kiswahili language). The transcription was later translated from Kiswahili into English by the second author; the fifth author and a Tanzanian assistant checked its accuracy and corrected any issues. Data collection took place in December 2015 for the TBA and August 2016 for the SBAs. The qualitative content analysis was guided by the checklist from Elos and Kyngas’ [19] to increase trustworthiness. As they suggest, inductive content analysis was used due to the limited number of previous studies dealing with the phenomenon. The authors put the data in the matrix, which was constructed based on interview aims. For TBA interviews, two authors (the first and third authors) discussed the possible categories and ways to summarize them until they achieved consensus. For the focus group discussions, two authors (the first and fourth authors) discussed the categories until consensus was achieved. Then, similar codes were grouped into sub-categories, and similar sub-categories were grouped into categories. After these two analyses were completed, the first author merged and sorted the final categories according to their similarities and differences. The merged results were shared with the research group (all authors) and received agreement. Ethical reviews and approvals were obtained from the 1) Institutional Review Board at St. Luke’s International University, Tokyo, Japan (14–040); 2) Director of the Korogwe District Council; 3) National Institute for Medical Research, Tanzania (NIMR/HQ/R.8/Vol.IX/1604); and 4) Tanzania Commission for Science and Technology (COSTECH) (No.2013–273-NA-2013-101).