Abstract
Community health workers (CHWs) are a vital part of the health infrastructure in Uganda and in many other low- and middle-income countries. While the need for CHWs is clear, it is less clear how they should dispense health products to maximize the health benefits to their community. In this study, we assess the cost-effectiveness of several competing CHW distribution strategies in the context of treatment for child diarrhoea. We used data from a four-armed cluster-randomized controlled trial to assess the cost-effectiveness of (1) free distribution of oral rehydration salts (ORS) via home deliveries prior to diarrhoea onset (free delivery arm), (2) free distribution via vouchers where households retrieved the treatment from a central location (voucher arm), (3) a door-to-door sales model (home sales arm) and (4) a control arm where CHWs carried out their activities as normal. We assessed the cost-effectiveness from the implementor's perspective and a societal perspective in terms of cost per case treated with ORS and cost per disability-adjusted life year (DALY) averted. Free delivery was the most effective strategy and the cheapest from a societal perspective. Although implementor costs were highest in this arm, cost savings comes from households using fewer resources to seek treatment outside the home (transport, doctor fees and treatment costs). From the implementors' perspective, free delivery costs $2.19 per additional case treated and $56 per DALY averted relative to the control. Free delivery was also extremely cost-effective relative to home sales and vouchers, but there was a large degree of uncertainty around the comparison with vouchers. Free distribution of ORS by CHWs prior to diarrhoea onset is extremely cost-effective compared to other CHW distribution models. Implementers of CHW programmes should consider free home delivery of ORS.