Menu
Menu
Menu
Menu
Background: South Sudan has one of the highest maternal mortality ratios in the world, at 789 deaths per 100,000 live births. The majority of these deaths are due to complications during labor and delivery. Institutional delivery under the care of skilled attendants is a proven, effective intervention to avert some deaths. The aim was to determine the prevalence and explore the factors that affect utilization of health facilities for routine delivery and postnatal care in Torit County, South Sudan. Methods: A convergent parallel mixed method design combined a community survey among women who had delivered in the previous 12 months selected through a multistage sampling technique (n = 418) with an exploratory descriptive qualitative study. Interviews (n = 19) were conducted with policymakers, staff from non-governmental organizations and health workers. Focus group discussions (n = 12) were conducted among men and women within the communities. Bivariate and multivariate logistic regression were conducted to determine independent factors associated with institutional delivery. Thematic analysis was undertaken for the qualitative data. Results: Of 418 participants who had delivered in the previous 12 months, 27.7% had institutional deliveries and 22.5% attended postnatal care at least once within 42 days following delivery. Four or more antenatal care visits increased institutional delivery 5 times (p < 0.001). The participants who had an institutional delivery were younger (mean age 23.3 years old) than those who had home deliveries (mean age 25.6 years). Any previous payments made for delivery in the health facility doubled the risk of home delivery (p = 0.021). Women were more likely to plan and prepare for home delivery than for institutional delivery and sought institutional delivery when complications arose. Perceived poor quality of care due to absence of health staff and lack of supplies was reported as a major barrier to institutional delivery. Women emphasized fear of discrimination based on social and economic status. Unofficial payments such as soap and sweets were reported as routine expectations and another major barrier to institutional delivery. Conclusion: Interventions to stop unofficial payments and discrimination based on socio-economic status and to increase access to ANC, delivery services and PNC are needed.
A convergent parallel mixed method design [17] comprising a community based- cross sectional survey and an exploratory qualitative study was used. Equal importance was given to qualitative and quantitative data which were combined in the interpretation phase. The study was conducted between March and December, 2016. To explore the barriers for accessing maternal health services, the Gabrysch and Campbell framework [10] rooted in the three delays model [18] was used. The framework addresses four dimensions: socio cultural factors (traditional beliefs, norms, gender dynamics), perceived needs and benefits (knowledge, information, perceived quality of care), economic factors (ability to pay) and physical accessibility (distance, transport, roads). The study reported here is part of a larger research project supporting participatory interventions with women and communities on the one hand and health workers and facilities on the other in two neighboring conflict-affected settings: Torit, South Sudan and Gulu, Uganda. The project was initiated by the staff of two hospitals with some existing links; St Mary’s Lacor Hospital in Gulu and Torit State Hospital in South Sudan. Both hospitals had played central roles in maintaining some health care over decades of war, and both sought to engage more effectively and responsively with community priorities and to learn from and with the other setting’s experiences. The study was conducted in two payams of Torit County in Imotong state, Republic of South Sudan: Nyong and Himodonge. Payams, in South Sudan, are administrative areas of at least 25,000 residents. Several payams constitute a county which in turn constitute a state. Payams are themselves composed of smaller administrative units called bomas. Each boma includes several small villages or hamlets. Nyong and Himodonge are among the eight payams of Torit County. Each has five bomas. Their total projected population was 61,297 in 2016 with 2212 pregnancies expected annually [19]. Nyong Payam in which the state capital, Torit, is located was the most populated with 49,419 inhabitants while Himodonge Payam had 11,878 residents. Torit town is located in south eastern South Sudan, about 150 km from Juba, the capital city of South Sudan. While Nyong Payam is considered urban, there are distinct neighborhoods and villages within the Payam. There are five public health care levels in South Sudan: Primary Health Care Units (PHCUs), Primary Health Care Centers (PHCCs), County Hospitals, State Hospitals and Teaching Hospitals. PHCUs are the lowest level facilities and provide preventive, promotional and curative services but not delivery services which are offered at all the other levels [16]. At the time of the study there were only three public health facilities above the PHCU level and therefore able to conduct deliveries in the study area. A sample size of 383 was calculated for a 95% confidence level, using OpenEpi version 3 for sample size calculation for the proportion of women who deliver in the facilities assuming a hypothesized facility delivery rate for South Sudan of 21% [7] and design effect (for cluster surveys) of 1.5, and allowing 10% for missing data. The survey participants were selected among the women who had given birth in the previous 12 months preceding the survey. A four-stage stratified sampling technique was used to select the participants. It was designed to capture urban and rural populations and people living at varying distances from a health facility staffed by skilled birth attendants. First, three out of the eight payams of Torit County (Kudo, Nyong and Himodonge) were purposively selected for the study. Nyong is an urban payam, Himodonge is a rural payam, and Kudo is a remote payam, about 2 h outside of Torit town during the dry season and often inaccessible during heavy rains. Kudo was excluded at the time of the survey because of insecurity resulting from renewed open conflict in the area in July 2016. In the second stage, three bomas out of the five in each payam were selected: the one closest, at middle distance and farthest from the facility offering skilled birth attendance. These distances were about 0.5, 2 and 6 km respectively for the bomas in Nyong Payam and 0.5, 6 and 26 Km for those in Himodonge Payam. In the third stage, three villages or hamlets were selected by simple random sampling from a list of the villages in each selected boma. On the interview day, vehicles dropped interviewers at an estimated geographical central location of the village, from which interviewers used chain or sequential referral technique [20] and home health promoters to identify the next household that qualified for the interview and seeking to capture all or most women normally residing in that village or hamlet who had delivered in the previous 12 months. Home health promoters in South Sudan are community members identified by the communities and supported by the government to promote community health and facilitate linkage to the health facilities. For the qualitative study, five categories of participants bringing diverse perspectives and experiences to the study topic were identified for purposive sampling [21]: (1) policy-makers from State Ministry of Health (SMoH) and the national MOH, (2) staff of NGOs and humanitarian partner organizations, (3) health managers (county health department officials and director of hospital), (4) health workers (physicians, midwives, nurses, clinical officers), and (5) community members including women, men and community leaders. In consultation with the local research team, a list of participants for IDIs was prepared including policymakers, health managers and NGO staff; and interviews were scheduled. FGDs were also planned at the community level in consultation with the village chiefs and HHPs who helped to mobilise the communities. FGDs with the healthcare providers were planned in consultation with the facility in-charges. The sample size was not predetermined but rather sought to achieve data saturation. In total, 19 in depth interviews and 12 Focus Group Discussions (FGDs) were conducted (see Table Table11). Number of interviews and FGDs per group of participants A pretested, standardized adapted survey tool for maternal and child health (MCH) care (from University College London (UCL)) (Additional file 1) was used with permission of UCL and administered by trained data collectors to obtain the quantitative data. Community entry was enhanced by working with home health promoters as guides. Prior to the administration of the questionnaire, informed consent was obtained. Individual interviews were conducted in the home of the respondent in a convenient location offering privacy. Demographic information was collected on women who had given birth in the previous one year; then questions on socio-cultural, economic and physical barriers that affected the decision making to access care for routine delivery and postnatal care were asked, together with questions on perceptions of benefit or need to seek care. To gather qualitative data, in- depth interviews and FGDs were conducted to explore the determinants influencing utilization of health services for institutional deliveries and postnatal care. While individual interviews and FGDs with policy makers, health managers, staff of humanitarian partner organizations and health workers were conducted in English, FGDs with the community members were conducted in the local language (Lotuko) by local research assistants familiar with the social and cultural context of the study area. The local research assistants included two women and one male. All of them were trained in basic qualitative research methods by an anthropologist and public health specialist (LB) who has worked with the team and in Torit County since 2015. The individual interviews were conducted by the research team members who are co-authors of this paper (CZ, EO, LB). Interview guides for FGD and in- depth interviews were designed according to the research objectives and the conceptual framework. FGDs were recorded and notes were taken systematically during the in -depth interviews as initial attempts to digitally record interviews were resisted by some respondents. The duration of the interviews and FGDs was between 45 min and one hour. Quantitative data were entered using Epi data software and analysis done using SPSS version 21. Categorical data were summarized into proportions and comparisons made using chi square tests. For continuous data, the mean or median and interquartile ranges were calculated. Bivariate analysis was conducted to assess the association between the place of delivery and the socio-cultural, economic and physical variables as well as perception of need or benefit for institutional delivery. Correlations between different age groups and different payams were assessed using Student’s t test. Significant factors on bivariate analysis were entered into a multivariate logistic regression model to determine the independent variables associated with institutional delivery. The confidence level was set at 95% and all statistical tests were considered significant at a p-value ≤0.05. In depth interviews and FGDs were transcribed and translated from the local language to English by local research assistants. Both were coded with QSR International’s Nvivo software version 11.3.2. The categories and sub categories of themes were organized according to the framework, the research objectives and emergent themes from empirical data. A mixed thematic (inductive/ deductive) approach was used to analyse the data [22]. Data sources (health providers, managers, policy makers, communities) and data collection methods (interviews and FGDs) were triangulated to enhance the internal validity of the study [23, 24]. Ethical approvals for this study were obtained from the ethical committee of the Ministry of Health, Republic of South Sudan and the University of Montreal Hospital Research Centre (CRCHUM, Canada). Informed consent was obtained from all participants. Permission to record the FGDs was sought. Confidentiality and anonymity of the participants were maintained throughout the study process. Data were kept under lock and key only accessed by the research team. We explicitly adopted the guidelines proposed for qualitative research in conflict-affected settings by the ReBUILD Consortium [25]. In particular, these guidelines caution that audio or video-recording interviews and FGDs can sometimes be problematic both scientifically (through limiting the open expression of perspectives) and ethically (through creating a fear of possible reprisals should the recordings be shared) where trust has been eroded in conflict-affected settings.
