Abstract

Background. Every year hundreds of thousands of women and more than two million newborn babies die from causes that skilled care could prevent. Most of these deaths occur in poor, rural and hard-to-reach communities in low- and middle-income countries (LMICs). Community-based midwifery models and similar programmes bring trained care closer to women, but it is not clear how much they reduce deaths and illness.

Objectives. To assess whether community-based midwifery models and related community-level maternal and newborn programmes reduce (1) maternal mortality, (2) neonatal mortality and (3) maternal and neonatal illness (morbidity) in underserved settings.

Methods. We searched seven databases and grey-literature sources without language or date limits (PROSPERO CRD420261386288; reported according to PRISMA 2020). Two reviewers independently screened records, extracted data and assessed risk of bias (Cochrane RoB 2 for trials; Newcastle-Ottawa Scale for quasi-experimental studies). Because the programmes, populations and outcome measures differed too much to combine statistically, we synthesised results narratively, following the SWiM guidance.

Results. Of 3,214 unique records, 13 studies met the criteria: six cluster-randomised trials (Nepal, India, Bangladesh [two], Ghana, Malawi) and seven quasi-experimental studies (Pakistan, Kenya and five from Nigeria). Eight studies measured neonatal deaths; three measured maternal deaths. All six trials estimated fewer newborn deaths in the intervention areas, but the reduction was statistically convincing in only three: women's groups in Nepal (odds ratio 0.70, 95% CI 0.53 to 0.94) and India (0.68, 0.59 to 0.78), and home visits by community health workers in Sylhet, Bangladesh (risk ratio 0.66, 0.47 to 0.93). Trials in Ghana, Malawi and a second Bangladeshi trial found small reductions that could be due to chance. Maternal mortality results were inconsistent and rested on very few deaths. Evidence on illness was sparse. None of the five Nigerian studies measured mortality.

Conclusions. Community programmes that reach many families and combine women's groups or home visits with links to skilled care can save newborn lives, but the benefit is not automatic. The evidence that community-based midwifery specifically reduces deaths in Nigeria and sub-Saharan Africa is thin, and well-designed evaluations that measure deaths are urgently needed.