Background The huge proportion of child marriage contributes to high rates of pregnancies among adolescent girls in Bangladesh. Despite substantial progress in reducing maternal mortality in the last two decades, the rate of adolescent pregnancy remains high. The use of skilled maternal health services is still low in Bangladesh. Several quantitative studies described the use of skilled maternal health services among adolescent girls. So far, very little qualitative evidence exists about attitudes and practices related to maternal health. To fill this gap, we aimed at exploring maternal health care-seeking behavior of adolescent girls and their experiences related to pregnancy and delivery in Bangladesh. Methods and Findings A prospective qualitative study was conducted among thirty married adolescent girls from three Upazilas (sub-districts) of Rangpur district. They were interviewed in two subsequent phases (2014 and 2015). To triangulate and validate the data collected from these married adolescent girls, key informant interviews (KIIs) and focus group discussions (FGDs) were conducted with different stakeholders. Data analysis was guided by the Social-Ecological Model (SEM) including four levels of factors (individual, interpersonal and family, community and social, and organizational and health systems level) which influenced the maternal health care-seeking behavior of adolescent girls. While adolescent girls showed little decision making-autonomy, interpersonal and family level factors played an important role in their use of skilled maternal health services. In addition, community and social factors and as well as organizational and health systems factors shaped adolescent girls’ maternal health care-seeking behavior. Conclusions In order to improve the maternal health of adolescent girls, all four levels of factors of SEM should be taken into account while developing health interventions targeting adolescent girls.
This was a prospective qualitative study in which data were collected from married adolescent girls in two phases. Multiple data sources were used to triangulate and validate the findings including in-depth interviews (IDIs) with married adolescent girls, key informant interviews (KIIs) and focus group discussions (FGDs) with different stakeholders. This study was conducted in Rangpur district in Rangpur division, Bangladesh. Most recent data show that Rangpur division has the highest rate (37%) of teenage pregnancy in Bangladesh [19]. We purposively selected married adolescent girls residing in three sub-districts of Rangpur district: Mithapukur, Kaunia and Badarganj. Socio-economic conditions, cultural practices and beliefs and access to maternal health services are quite similar for the people living in these three sub-districts. Community health workers from BRAC (an international development organization based in Bangladesh) have been delivering door-to-door family planning and maternal care services in almost every village in the three sub-districts of Rangpur. In addition, LAMB [a non-governmental organization (NGO)] has been providing free ANC, postnatal care (PNC) and delivery services via its Safe Delivery Unit (SDU) at Badarganj sub-district. Qualitative data were collected from a wide range of respondents. In addition to married adolescent girls, the main study population, we collected data from community health workers, community people, family members of adolescent girls (mothers-in-law and husbands), representatives from the government, NGOs and health providers. Table 1 shows a list of study participants and data collection methods. We collected data purposively from different types of respondents to obtain rich data. In-depth interviews (IDIs) were conducted with married pregnant and non-pregnant adolescent girls in two phases. During the first phase (December 2014), pregnant adolescents were asked about their knowledge, perception and practices related to maternal healthcare services and their intended delivery places and methods. Non-pregnant adolescents were interviewed about their knowledge, perception and practices related to family planning methods and intention of childbearing. During the second phase of the study (December 2015), the same participants were asked about their experiences during pregnancy and delivery care, whether they had become pregnant or not, and if any, what their experiences were with maternal healthcare services, such as ANC. For both groups, the information collected in the first phase was combined with that of the second phase. Four female research assistants (anthropologists, experienced in conducting IDIs and FGDs) collected data from the adolescent girls during these two phases. Research assistants were trained to conduct interviews in a way that biases were reduced (i.e. dominant respondent bias, shyness bias). KIIs were conducted with representatives of the government, NGOs, and hospital personnel who had been working in a public hospital in Rangpur district. Finally, three FGDs were conducted with community health workers, members of a village maternal health committee, and adolescents’ mothers-in-law in order to validate the data gathered via IDIs and KIIs as well as to explore common practices and barriers to the use of maternal health services. BRAC field staff working on a maternal health project in Rangpur district supported the research team in identifying married pregnant and non-pregnant adolescent girls in the community. The interview guides were pre-tested in Rangpur Sadar Upazila, Rangpur district and adapted. All topic guides were developed in English and translated into Bangla, before pre-testing. Due to logistical issues (e.g. time constraint, difficulties to find respondents to gather in a place) the FGD topic guide could not be piloted. We analyzed data with the help of MAXQDA 11 software using the Social-Ecological Model (SEM) as an initial coding guide [30]. The SEM is a theory-based framework which considers the complex interplay of multiple levels of a social system and interactions between individuals and environment within this system. The SEM thus adequately facilitated the exploration of adolescent girls’ experiences, integrating their intrapersonal, partner-related, family, community and socio-cultural contexts to produce one behavioral outcome regarding maternal health care-seeking behavior. Guided by the objectives of the study and the SEM, an initial coding framework was generated after reading a subset of the transcripts. Newly emerging text segments or codes in subsequent transcripts were inductively added to the framework to build our model of factors influencing maternal health care-seeking behavior (Fig 1). When new codes or themes were added to the framework, all data were re-scrutinized to assess their relevance. The data from IDIs, KIIs and FGDs were scrutinized several times to obtain a sense of the whole. Researchers with different backgrounds provided input to the analysis to increase its validity. The research protocol was approved by the Institutional Review Board (IRB) of the Institute of Tropical Medicine (ITM), Antwerp and the Ethical Review Committee (ERC) of the James P. Grant School of Public Health at BRAC University, Bangladesh. Written informed consent was obtained from all the participants. However, because of cultural issues (i.e. respondents feeling uncomfortable to sign) and participants’ illiteracy levels, from a few respondents verbal consents were obtained. Written informed consent was documented through a signature on a ‘participant information sheet and informed consent’ form and verbal informed consent was documented via audio recording. Respondents aged below 18 provided assent, while written consent was sought for them from their legal guardians/husbands. Confidentiality was strictly maintained: only the researchers had access to the data and no personally identifying information was kept that could personally identify respondents after the research had been completed.
N/A