The Efficacy of Maternity Waiting Homes in Decreasing Maternal and Perinatal Mortality in Low-Income Countries – A Systematic Review

Abstract

Maternal and perinatal mortality remains significantly high in low‐income countries with over 800 deaths per day of women around childbirth. Greater than 90% of such deaths occur in low‐income countries. The concept of maternity waiting homes (MWH) was reintroduced to aid in decreasing maternal and perinatal mortality. Since the previous Cochrane Review in 2012 on maternity waiting homes, there have not been any published randomized controlled studies. Do observational studies on MWHs demonstrate decreased maternal and perinatal mortality in low‐income countries when compared with the standard of care? We searched for primary articles that reported maternal and perinatal deaths as major outcomes in studies who compared MWHs to other methods such as direct hospital admits, we also investigated cesarean delivery rates. Search engines used were: Cochrane Review, Medline and CINAHL. Meta‐analyses and forests plots were formulated using MedCalc Software. Systematic review was drafted using MOOSE guidelines for meta‐analysis and systematic reviews of observation. Seven articles met criteria for this study. The maternal mortality rate for MWH was 105/100,000 and 1,066/100,000 for non‐MWH, Relative Risk (RR) 0.145 (95% Confidence Interval (CI) 0.062 to 0.204). Perinatal mortality rate was 60/1,000 in MWH compared to 65/1,000, RR 0.782 (CI 0.602 to 1.120) in non‐MWH. Stillbirth rate was 18/1,000 in MWH and 184/1,000 in non‐MWH, RR 0.204 (CI 63.88 to 94.08). Neonatal mortality rates were 16/1,000 in MWH and 15/1,000 in non‐MWH, RR 0.862 (CI 0.392 to 1.628). Cesarean deliveries rate was 24/100 for MWH and 18/100 in non‐MWH, RR 1.229 (CI 1.226‐1.555). MHWs statistically decreased maternal death, stillbirths and increased cesarean delivery rates. Overall, the observation nature of the study designs introduces selection biases that may have altered the results of the studies. No randomized trials have been done to date. We suggest cluster‐randomized studies to further evaluate the effect of MWHs.