Effect of a participatory intervention with women's groups on birth outcomes in Nepal: cluster-randomised controlled trial
Dharma S Manandhar, David Osrin, Bhim Prasad Shrestha, Natasha Mesko, Joanna Morrison, Kirti Man Tumbahangphe, Suresh Tamang, Sushma Thapa, Dej Shrestha, Bidur Thapa, Jyoti Raj Shrestha, Angie Wade, Josephine Borghi, Hilary Standing, Madan Manandhar, Anthony M de L Costello, MIRA Makwanpur trial team
The Lancet · 2004 · DOI: 10.1016/S0140-6736(04)17021-9
Countries: Nepal
What Was Studied
This cluster-randomised controlled trial tested whether a community-based participatory intervention with women's groups could reduce neonatal mortality in a poor, rural, mountainous district of Nepal. Researchers pair-matched 42 village development committees (VDCs) in Makwanpur district into 21 pairs based on topography, ethnic-group distribution, and population density, then randomly selected 12 pairs (24 VDCs, average population ~7,000 each, covering ~170,000 people across 1,600 km²) and randomly allocated one VDC per pair to intervention or control. In each intervention cluster, a locally recruited female facilitator convened monthly women's-group meetings across nine wards, guided by an action-learning cycle adapted from Bolivia's Warmi project, in which groups identified local perinatal problems and designed their own strategies (community funds, stretcher schemes, clean delivery kits, home visits). A closed cohort of 28,931 married women of reproductive age (15-49) was placed under monthly surveillance for pregnancy and birth outcomes; the 2-year analysis period ran from November 2001 to October 2003. The primary outcome was neonatal mortality rate; secondary outcomes included stillbirths, maternal deaths, and uptake of antenatal, delivery, and newborn-care practices. Analysis was by intention to treat using hierarchical logistic models accounting for clustering.
What They Found
Neonatal mortality was 26.2 per 1,000 livebirths in intervention clusters (76 deaths among 2,899 livebirths) versus 36.9 per 1,000 in control clusters (119 deaths among 3,226 livebirths) — a 30% relative reduction (adjusted odds ratio 0.70, 95% CI 0.53–0.94). The maternal mortality ratio was 69 per 100,000 livebirths in intervention clusters versus 341 per 100,000 in controls, an 80% relative reduction (adjusted odds ratio 0.22, 95% CI 0.05–0.90), though this was a secondary rather than pre-specified primary outcome, based on only 13 total maternal deaths. Stillbirth rates did not differ significantly (24.6 vs 23.3 per 1,000 births; odds ratio 1.06, 95% CI 0.76–1.47). Only 8% of eligible women ever attended a group meeting, but the groups attracted 37% of newly pregnant women, and intervention clusters showed large gains in antenatal care uptake (55% vs 30%; odds ratio 2.82), institutional delivery (7% vs 2%; odds ratio 3.55), and clean delivery-kit use (19% vs 5%; odds ratio 4.59) relative to controls. A companion cost-effectiveness analysis found the intervention cost US$3,442 per newborn life saved (US$4,397 including health-system-strengthening costs) and US$111 per life-year saved — well below the World Bank's US$127-per-DALY cost-effectiveness benchmark cited in the paper.
What This Means for Nepal
This is arguably the single most policy-relevant Nepal-specific RCT in the maternal and newborn health evidence base — it is the empirical foundation for the participatory women's-group model later replicated and meta-analysed across South Asia and Africa. Two decades on, Nepal's neonatal mortality rate has plateaued at 21 per 1,000 since 2016 (NDHS 2022) even as institutional delivery has climbed to 79.4%, suggesting diminishing returns from supply-side expansion alone. Nepal already funds a ~50,396-strong FCHV cadre (DoHS AHR 2080/81) whose mothers'-group activities are frequently described as inconsistent and under-resourced — the same 'sporadic' pattern this 2004 trial found among pre-existing FCHV-linked women's groups before dedicated facilitation revived them. MoHP and provincial health directorates should pilot a trained, incentivised facilitator role — modelled on this trial's design — layered onto the existing FCHV network in the highest-burden districts (Karnali and Sudurpashchim, where physician and consultant posts remain critically understaffed, and Madhesh, where facility density is lowest), rather than assuming FCHV presence alone will replicate the trial's mortality reduction. A cost-effectiveness re-assessment using current Nepali prices would also help decide whether to fund a facilitator cadre nationally, since the original US$111-per-life-year-saved estimate is now more than two decades out of date.
Contextualisation
The MIRA Makwanpur trial is the foundational evidence behind Nepal's community-mobilisation model for maternal and newborn health: a locally recruited female facilitator convening monthly women's groups to identify and solve perinatal problems. The approach is a direct forerunner to the mothers'-group activities Nepal now expects from its ~50,396 Female Community Health Volunteers (FCHVs) (DoHS AHR 2080/81) — indeed, the trial itself found that FCHV-initiated women's groups already existed in some communities but met only 'sporadically' until a dedicated facilitator revived them. At baseline (2001), national institutional delivery was near the trial's own recorded rate of ~13% trained attendance; it has since climbed to 79.4% (NDHS 2022), largely credited to the Aama safe-delivery incentive programme launched in 2005. Yet neonatal mortality has stagnated at 21 per 1,000 live births since 2016 (NDHS 2022), suggesting that further reductions require the kind of demand-side, community-level intervention this trial tested rather than continued supply-side expansion alone.