Men’s roles in care seeking for maternal and newborn health: A qualitative study applying the three delays model to male involvement in Morogoro Region, Tanzania

Background: Increasing the utilization of facility-based care for women and newborns in low-resource settings can reduce maternal and newborn morbidity and mortality. Men influence whether women and newborns receive care because they often control financial resources and household decisions. This influence can have negative effects if men misjudge or ignore danger signs or are unwilling or unable to pay for care. Men can also positively affect their families’ health by helping plan for delivery, supplementing women’s knowledge about danger signs, and supporting the use of facility-based care. Because of these positive implications, researchers have called for increased male involvement in maternal and newborn health. However, data gathered directly from men to inform programs are lacking. Methods: This study draws on in-depth interviews with 27 men in Morogoro Region, Tanzania whose partners delivered in the previous 14 months. Debriefings took place throughout data collection. Interview transcripts were analyzed inductively to identify relevant themes and devise an analysis questionnaire, subsequently applied deductively to all transcripts. Results: Study findings add a partner-focused dimension to the three delays model of maternal care seeking. Men in the study often, though not universally, described facilitating access to care for women and newborns at each point along this care-seeking continuum (deciding to seek care, reaching a facility, and receiving care). Specifically, men reported taking ownership of their role as decision makers and described themselves as supportive of facility-based care. Men described arranging transport and accompanying their partners to facilities, especially for non-routine care. Men also discussed purchasing supplies and medications, acting as patient advocates, and registering complaints about health services. In addition, men described barriers to their involvement including a lack of knowledge, the need to focus on income-generating activities, the cost of care, and policies limiting male involvement at facilities. Conclusion: Men can leverage their influence over household resources and decision making to facilitate care seeking and navigate challenges accessing care for women and newborns. Examining these findings from men and understanding the barriers they face can help inform interventions that encourage men to be positively and proactively involved in maternal and newborn health.

The study took place in five districts of Morogoro Region: Gairo (at time of study, this district was part of Kilosa district), Kilosa, Morogoro Rural, Mvomero, and Ulanga districts. This setting is marked by male-dominated household decision making and elevated maternal mortality (see Table 1). A comparison of Tanzania’s 2010 DHS data and 2015/2016 DHS data suggests that women’s involvement in decision making has improved, although 27.5% of men remain the primary decision makers related to women’s health care, and the percentage of women reporting challenges accessing health care increased during this time [37, 38]. At the same time, maternal mortality decreased slightly between 2010 and 2015/2016, but remains a notable problem in Tanzania [37, 38]. Summary of Decision Making and Access to Health Care Indicators from Demographic and Health Surveys 2015/ 2016 Burundi [39] Kenya [39] 2014/ 2015 Tanzania reports similar decision-making patterns as other countries in the East African Community including Burundi, Kenya, Rwanda, and Uganda, although in Kenya and Uganda about half of women report deciding how their own earnings are used compared to a little over one third of women in Tanzania (see Table ​Table1)1) [37–39]. Barriers to accessing health services are also ubiquitous in these countries, with the problem of getting money for treatment cited by the highest percentage of women in each country [37, 38]. Although government policy stipulates the provision of free maternal and child health services in Tanzania, the policy has been difficult to implement [40]. A study conducted in western Tanzania found that among women who delivered at a public facility, 62.5% paid for services, with an average cost for delivery of 3840 Tanzania Shillings (TZS) (USD3.08 at the time of the study) [41]. This cost increased to 6268 TZS (USD5.03 at the time of the study) when components of health expenses including transportation, drugs, and supplies were considered [41]. This study was part of a larger evaluation of an integrated maternal and newborn healthcare program implemented in partnership between the Ministry of Health and Social Welfare (MoHSW) and Jhpiego. The program sought to increase facility-based care during pregnancy and delivery in Morogoro Region, Tanzania. Further program details are available from Bishanga et al. (2018) [42] and Lefevre et al. (2015) [43]. Study participants were male respondents who resided in one of the five focus districts of Morogoro Region and were partners of women who met the study criteria of giving birth within the previous 14 months and not experiencing complications during delivery. The researchers purposefully selected the 14-month period to reduce recall bias among study respondents and to allow time for outcomes measured as part of the larger study, such as re-initiation of contraceptive use among women [44] and use of postnatal care services [45]. The research excluded men whose partners experienced complications during delivery in order to avoid skewing the data toward more intensive health service requirements beyond routine childbirth. The sample included men whose partners delivered both at health facilities and in the community. Researchers selected respondents from villages near (< 3 km) and far (≥3 km) from health centers, as identified by facility-based staff. Twenty-seven men meeting these criteria provided informed consent before being interviewed. The ethical review boards of Johns Hopkins University School of Public Health in Baltimore, USA and Muhimbili University of Health and Allied Sciences in Dar-es-Salaam, Tanzania provided ethical approval for this study. Research assistants (RAs) received 5 days of training on maternal and newborn health, qualitative methods, and research ethics, followed by piloting and tool revision. The RAs, all bilingual Swahili and English speakers, included male and female social science graduate students and teachers. Strategies to collect trustworthy data followed the approach of naturalistic inquiry put forth by Lincoln and Guba (1985), including prolonged engagement and analyst triangulation (see Table 2) [46]. The RAs conducted semi-structured in-depth interviews with the 27 men over a period of 2 months. Interview guides probed on experiences of care seeking during pregnancy, delivery, and the postpartum period, as well as how men view their role in maternal and child health. RAs and respondents were gender-matched to diminish gender-related influences that could affect data quality. The data collection phase included daily debriefing sessions amongst the research team [47]. Regular discussion during these debriefings provided opportunities to facilitate analyst triangulation, identify when to seek follow-up interviews, and iteratively refine data collection tools. Debriefings resulted in the decision to conduct follow-up interviews with three men in order to gain clarity and to further probe on responses insufficiently addressed in earlier interviews. All interviews ranged in length from 35 to 99 min, with an average length of 69 min. Each interview was digitally recorded, checked for quality, and transcribed verbatim in Swahili. Study rigor as informed by Lincoln and Guba (1985) and McMahon and Winch (2018) – Debriefingsb were conducted each night throughout data collection and involved all members of the data collection team sharing, comparing, amplifying, or refuting one another’s findings. Findings from debriefings were presented to Tanzania-based researchers, program implementers, and policymakers engaged in maternal health programs for feedback. Debriefing memos formed the basis for an audit trail of the study. – Members of the research team who undertook data collection also participated in data analysis. During analysis, at least two analysts analyzed each theme report and compared interpretations. – Throughout analysis, in the event of discrepancies, two senior researchers who were present throughout data collection and led most debriefings, weighed in and determined a way forward (highlighting opportunities for re-translation of interviews as necessary) – All final results were reviewed by the full research team (including most data collectors) aAs informed by Lincoln and Guba 1985 [41]; bAs informed by McMahon & Winch 2018 [42] Data analysis followed a modified framework approach [48]. Analysis started in the field during debriefings with the research team, when themes began to emerge [42]. Co-authors, during a data analysis workshop, further drew out themes from a selection of qualitatively rich transcripts – an inductive approach [49] – and later applied a questionnaire based on those themes a priori to all transcripts through a deductive approach [50]. Throughout the data analysis process, the authors returned to analyst triangulation to facilitate trustworthy conclusions from the data (see Table ​Table2)2) [46]. To begin the inductive phase of analysis, the lead author and last author reviewed translated portions of transcripts identified during initial analysis conducted in ATLAS.ti, a qualitative data management software [51]. They chose three transcripts that were illustrative of the experiences of male respondents. Swahili-speaking co-authors (three of whom also participated in data collection) analyzed these transcripts in full in their original Swahili to retain their original context and meaning. The analysts and lead author then discussed themes that emerged from the coded transcripts to triangulate findings. Analysts used Gulliford et al’s (2002) framework on access to health care [52], identified based on congruent themes, to inform the creation of a questionnaire deductively applied to all interview transcripts [50]. The lead author compiled the results of the questionnaire into theme reports. At least two analysts further analyzed each theme report to continue to triangulate findings and outline similarities and differences emerging across transcripts. Analysts then prepared summary reports including the translation of selected quotations into English. When translations differed between reports, an author re-translated the text. Routine dialogue took place among workshop participants to corroborate the findings across researchers [50].

N/A

• Dima Health