Bulletin of the World Health Organization

Economic evaluation of neonatal care packages in a cluster-randomized controlled trial in Sylhet, Bangladesh

Amnesty E LeFevre, Samuel D Shillcutt, Hugh R Waters, Sabbir Haider, Shams El Arifeen, Ishtiaq Mannan, Habibur R Seraji, Rasheduzzaman Shah, Gary L Darmstadt, Steve N Wall, Emma K Williams, Robert E Black, Mathuram Santosham, Abdullah H Baqui & for the Projahnmo Study Group


To evaluate and compare the cost-effectiveness of two strategies for neonatal care in Sylhet division, Bangladesh.


In a cluster-randomized controlled trial, two strategies for neonatal care – known as home care and community care – were compared with existing services. For each study arm, economic costs were estimated from a societal perspective, inclusive of programme costs, provider costs and household out-of-pocket payments on care-seeking. Neonatal mortality in each study arm was determined through household surveys. The incremental cost-effectiveness of each strategy – compared with that of the pre-existing levels of maternal and neonatal care – was then estimated. The levels of uncertainty in our estimates were quantified through probabilistic sensitivity analysis.


The incremental programme costs of implementing the home-care package were 2939 (95% confidence interval, CI: 1833–7616) United States dollars (US$) per neonatal death averted and US$ 103.49 (95% CI: 64.72–265.93) per disability-adjusted life year (DALY) averted. The corresponding total societal costs were US$ 2971 (95% CI: 1844–7628) and US$ 104.62 (95% CI: 65.15–266.60), respectively. The home-care package was cost-effective – with 95% certainty – if healthy life years were valued above US$ 214 per DALY averted. In contrast, implementation of the community-care strategy led to no reduction in neonatal mortality and did not appear to be cost-effective.


The home-care package represents a highly cost-effective intervention strategy that should be considered for replication and scale-up in Bangladesh and similar settings elsewhere.