Background: Uganda’s maternal mortality remains unacceptably high, with thousands of women and newborns still dying of preventable deaths from pregnancy and childbirth-related complica-tions. Globally, Antenatal care (ANC) attendance has been associated with improved rates of skilled births. However, despite the fact that over 95% of women in Uganda attend at least one ANC, over 30% of women still deliver at home alone, or in the presence of an unskilled birth attendant, with many choosing to come to hospital after experiencing a complication. We explored barriers to women’s decisions to deliver in a health care facility among postpartum women in rural southwestern Uganda, to ultimately inform interventions aimed at improving skilled facility births. Methods: Between December 2018 and March 2019, we conducted in-depth qualitative face-to-face interviews with 30 post-partum women in rural southwestern Uganda. The purposeful sample was intended to represent women with differing experiences of pregnancy, delivery, and antenatal care. We included 15 adult women who had delivered from their homes and 15 who had delivered from a health facility in the previous 3 months. Women were recruited from 10 villages within 20 km of a regional referral hospital. Interviews were conducted and digitally recorded in a private setting by a trained native speaker to elicit experiences of pregnancy and birth. Translated transcripts were generated and coded. Coded data were iteratively reviewed and sorted to derive descriptive categories using an inductive content analytic approach. Results: Regardless of where they decided to give birth, women wished to deliver in a supportive, respectful, responsive and loving environment. The data revealed six key barriers to women’s decisions to deliver from a health care facility: 1) Fear of unresponsive care, fueling a fear of being neglected or abandoned while at the facility; 2) fear of embarrassment and mistreatment by health care providers; 3) low perception of risk associated with pregnancy and childbirth; 4) preferences for particular birthing positions and their outcome expectations; 5) perceived lack of privacy in public facilities; and 6) perceived poor clinical and interpersonal skills of health providers to adequately explain birthing procedures or support expectant or laboring women and their newborn. Conclusion: Anticipation of unsupportive, unresponsive, disrespectful treatment, and a perceived lack of tolerance for simple, non-harmful traditions prevent women from delivering at health facilities. Building better interpersonal relationships between patients and providers within health systems could reinforce trust, improve patient–provider interaction, and facilitate useful information transfer during ANC and delivery visits. These expectations are important considerations in developing supportive health care systems that provide acceptable patient-friendly care. These findings are indicative of the vital need for midwives and other health care providers to have additional training in the role of communication and dignity in delivery of quality health care.
In this study, we used qualitative interview study design to explore and understand barriers to women’s decisions and choices to deliver in a health care facility among postpartum women in rural southwestern Uganda. The study was conducted between December 2018 and March 2019 in the rural southwestern Uganda’s Mbarara district, one of the highest maternal mortality ratio of 489 per 100,000 women Uganda16 with a projected dense population of 524,400, 17 sub-counties, 83 parishes and 757 villages.17 Uganda’s public health system is organized into seven tiers with national and regional referral hospitals, general district hospitals and four levels of community health centers. At the county level are health center IVs (HCIV), sub-county level are health center IIIs (HCIII), the health center IIs (HCII) at the parish level and village level (HC1) that is operated by Village Health Teams (VHTs). The VHTs are community volunteers identified by their community members and are given basic training on major health programs so they can in turn mobilize and sensitize communities to actively participate in utilizing the available health services.18 According to the Uganda Ministry of Health, VHTs also act as an important link between the communities and health facilities, and can provide treatment of uncomplicated diseases like malaria, pneumonia, worm infestations, diarrhea and mass drug administration for Neglected Tropical Diseases. VHTs mobilize communities during specific health campaigns and community disease surveillance activities through active data collection and reporting. Staffing and available services vary across the four levels: HCIII and HCIV should offer Emergency Obstetrics Care (EMOC), whereas HCI and HCII serve as low resource referral units which are not able to provide EMOC and have no ambulances and blood transfusion services.19 In total, there are about 10 public facilities within a 20 km radius from Mbarara Regional Referral Hospital, the main teaching hospital for Mbarara University of Science and Technology. Private providers operate in parallel to the public system to provide maternal health care. A purposeful qualitative sampling strategy was used to construct a sample of postpartum women with varied knowledge and experiences of pregnancy, antenatal care, and childbirth. A trained research assistant initiated contact to one of the VHTs based in each of the study villages. These VHTs in each of the 10 villages located within 20 km from Mbarara Regional Referral Hospital were identified from the list of VHTs provided by the district health office. The VHT contact person helped to identify women who had had their last delivery within the last 3 months. Trained research assistants then initiated a telephone contact to the identified women to seek permission to visit them for an interview. All women contacted were willing to participate in the study. Research assistants explained what the study is all about and obtained voluntary written informed consent from all eligible participants in the local language in a private area of their homes, communities or study office. All consenting participants gave written informed consent, or for those who could not write, a thumbprint was made on the consent form. We identified three women from each of the 10 villages with different facility or home birthing experiences. We purposively selected and recruited a total sample of 15 women who had delivered from their homes and 15 who had delivered from a health facility for their last delivery, and within the last 3 months. The places for facility or home deliveries were regardless of where their previous births happened, bringing a wide mix and range of different experiences, circumstances and decision-making processes from the participating women. Eligible women were: 1) adults of childbearing age (18–49 years); 2) had delivered a child in the previous 3 months; 3) had access to a mobile phone; and 4) were able and willing to give informed consent. Data collection consisted of individual, open-ended interviews with each of the study participants (N=30 interviews). A preliminary interview guide was developed and pilot tested by the primary author amongst five women in one of the 10 participating villages. The guide was revised based on the results of the pilot test. Topics included in the final version of the guide were: 1) perceptions of pregnancy and childbirth; 2) experiences of previous pregnancy or pregnancies; 3) experiences of ANC; 4) engagement with health care providers within a facility; 5) social support; and 6) childbirth experiences. Individualized probes were used to elicit details corresponding to each topic. Field notes and observations were collected by research assistants and included in interview debriefs. As the interviews were conducted and different field notes and observations filed, emerging content was continuously reviewed by the primary and senior authors to sharpen the interview questions and identify new probes. Demographic information (eg, age, occupation, educational background) was collected usually at the outset of each interview. All interviews took place in a private location mutually agreed upon by the participant and the interviewer. Interviews were conducted in the local language (Runyankole), and digitally recorded. Interviews lasted 60–90 minutes. Qualitative interviews were digitally recorded with the participant’s permission and transcribed. A Ugandan research assistant transcribed the interviews from the local language directly to English. The research team comprised of seven senior investigators, inclusive of epidemiologists (ECA, CO), obstetrician (GRM), a medical anthropologist (NCW), maternal/reproductive health expert (LTM), nurse (JN) and a health informatics specialist (AM). Based on our previous research and work experience in maternal health in Uganda, the team sought to explore barriers for low utilization of maternity services in Uganda. Two independently hired (male and female) research assistants were trained to conduct research in human subjects. They are both social scientists. These two research assistants generated transcripts but were not involved in concept development or coding of data. This multi-disciplinary team leveraged on their expertise and experience with maternal health issues in Uganda to design, conduct, analyze and present findings from this study. The aim of this qualitative data analysis was to inductively construct categories describing barriers to facility-based delivery. Analysis began with repeated review of transcripts to identify relevant content. The identified content served as the basis for developing a coding scheme. Coding was done in three stages namely: 1) open coding to identify and describe women’s ideas, meaningful expressions, phenomena or incidents highlighting their experiences during pregnancy and childbirth; 2) axial coding to relate and label codes or data that shared concepts, dimension and properties (relationship identification); and 3) selective coding to delimit coding to the identified core variables/concepts from the data (Strauss and Corbin, 1998). Data were coded with the aid of the qualitative data management software, NVivo10 (Melbourne, Australia). Coded data were iteratively reviewed and sorted to identify themes (repeated patterns in the data). Categories were then developed to describe each identified theme. Categories consisted of descriptive labels, elaborating text to define and specify each category’s meaning, and illustrative quotes taken from the qualitative data. Data analysis was done jointly by ECA, EA, CO, JN and GRM. Both JN and ECA coded five sampled transcripts and compared the results. Together with GRM and CO, we resolved disagreements until we were satisfied with the consistency in our coding. We aimed at ensuring consistency in coding.
N/A