Most maternal and perinatal deaths could be prevented through timely access to skilled birth attendants. Women should access appropriate obstetric care during pregnancy, labor, and puerperium. Maternity waiting homes (MWHs) permit access to emergency obstetric care when labor starts. This study compared maternal and perinatal outcomes among MWH users and nonusers through a retrospective cohort study. Data were collected through obstetric chart reviews and analyzed using STATA version 15. Of the 8144 deliveries reported between 2015 and 2019, 1305 women had high-risk pregnancies and were included in the study. MWH users had more spontaneous vaginal deliveries compared to non-users (38.6% versus 16.8%) and less cesarean sections (57.7% versus 76.7%). Maternal morbidities such as postpartum hemorrhage occurred less frequently among users than non-users (2.13% versus 5.64%). Four women died among non-users while there was no death among users. Non-users had more stillbirths than users (7.68% versus 0.91%). The MWH may have contributed to the observed differences in outcomes. However, many women with high risk pregnancies did not use the MWH, indicating a probable gap in awareness, usefulness, or their inability to stay due to other responsibilities at home. Use of MWHs at scale could improve maternal and perinatal outcomes in Rwanda.
The study was conducted at Ruli Hospital (RH) in Gakenke District, Rwanda. Geographically, Gakenke District has a mountainous landscape with an average altitude of 1788 m above sea level. According to the Rwanda Health Management Information System, the district’s maternal mortality ratio and skilled birth attendance were 325 per 100,000 live births and 96.7%, respectively by December 2019. About 50% of households walk for more than 1 h to reach a nearby health facility. Ruli Hospital serves a population of about 110,548 inhabitants (projections from the 2012 census) and also receives patients from neighboring districts. The hospital has a bed capacity of 179 and provides a wide range of services including comprehensive emergency obstetric and newborn care for eight health centers and eight health posts within their catchment area. During the study period, the medical staff in the maternity ward consisted of one medical officer, five nurses, and nine midwives. There was no obstetrician. Furthermore, the hospital has an operating theatre, a laboratory, a medical imaging unit, and blood transfusion services. In 2011, Matres Mundi, an international non-governmental organization, supported RH to set up a MWH within the hospital premises, which by the end of 2019 had received over 700 pregnant women. It is the only MWH in Rwanda. While staying in the MWH, pregnant women received obstetric care; psycho-social care; peer support; and education on proper nutrition, breastfeeding, and birth preparedness. Additionally, they were taught hands-on skills such as gardening, cooking, making handcrafts, and knitting baby items. All medical services offered are paid for by the community-based health insurance scheme (commonly known as “mutuelle de sante”) and pregnant women contributed 10% of the cost. A facility-based retrospective cohort study was conducted. Over a period of 5 years (January 2015 to December 2019), 8144 women gave birth in RH. Among those, 1305 women were eligible for staying in the MWH while 6839 women were not eligible. Eligibility for admission to the MWH was determined by a medical officer on duty based on predefined criteria that consisted of the following: a problem related to the pregnancy e.g., history of an abortion; a caesarian section (CS); prolonged labor, as well as problems during the current pregnancy, including preterm premature rupture of membranes, antepartum hemorrhage, reduced fetal movement, pre-eclampsia, etc. and one of the following conditions: at least 36 weeks of amenorrhea, place of residence that was far from the hospital (more than 3 h of walking), having no one to take care of the woman at home, victim of gender-based violence, being in social economic category 1 or 2, as well as clinician decision. Among the 1305 eligible women, 329 MWH-users (women who delivered after having stayed at the MWH) and 976 non-users (women who delivered but had not stayed at the MWH) were compared on maternal and perinatal outcomes. Non-users were identified based on the assumption that they met the same admission criteria as users (see Table 1). We identified MWH users and non-users by following the elimination process as shown in Figure 1. Inclusion and exclusion of users and non-users. Indications for admission to the MWH. * preterm premature rupture of membranes, ** cephalopelvic disproportion, *** intrauterine fetal death. Data were collected from hard copy medical records by two trained research assistants using a pre-designed data collection form (see Supplementary File, Figure S1) that was designed using Kobo Toolbox (version 2018) and installed on Android tablets under the supervision of the principal investigator (ET). Information from files included sociodemographic characteristics, ANC, indication of MWH and non-MWH admissions, obstetric history, mode of delivery, complications during delivery, outcome of delivery (maternal and perinatal), etc., which were abstracted from obstetric charts. Data were exported from Kobo Toolbox to MS Excel 2016 and then to STATA 15 for cleaning and analysis. Demographic characteristics were compared between MWH users and non-users using chi-square statistic. Univariable logistic analysis was conducted to determine associations between dependent and independent variables. Crude Odds Ratios (cOR) and their 95% confidence intervals (CI) were calculated, and significant outcomes with p value < 0.05 were considered for multivariable logistic regression. Adjusted Odds Ratios (aOR) and their 95% CI were calculated to take care of potential confounding and identify variables that showed statistically significant differences in the two groups. The ORs were adjusted for age, parity, occupation, and ownership of health insurance. Ethical clearance was obtained from the Rwanda National Ethics Committee (protocol code 335/RNEC/2020).