The effects of skill-based health education-A randomised-controlled intervention in primary schools in rural Bangladesh

Abstract

This paper evaluates the impact of skill-based health education (SBHE) on children's hygiene practices and health in rural Bangladesh. Over one year, SBHE was delivered weekly to primary schools through a randomised-controlled trial (RCT) by locally recruited trained para-teachers. The SBHE and soap provision interventions, provided in a cross-cutting manner, were randomly assigned to 180 schools stratified by two school types. Data were collected at both the school and child levels, involving 40 students in grades 1-4 at baseline and expanding to 50 students in grades 1-5 at endline, ten students per grade. The study tracked the same pupils, supplemented by additional and replacement students, resulting in effective sample sizes of 7,192 at baseline and 8,992 at endline. The results indicate that SBHE significantly improved hygiene practices; the average treatment effects on overall hygiene practices was 0.22 SD [0.14-0.31] (p < 0.001), and those on health/hygiene knowledge was 0.44 SD [0.33-0.55] (p < 0.001). While improvements in cold-related symptoms among the SBHE school children were marginally significant (-0.05 SD [-0.10, -0.01], p = 0.09), the overall trends indicated general improvement in health as well as healthy behaviours across all schools, irrespective of treatment status. Additional analysis incorporating inter-school spillover effects provided robust evidence of beneficial healthy practice externalities, extending beyond treatment school children. A cross-cutting soap provision treatment, although implemented imperfectly, did not show any standalone positive health-related effects. Nonetheless, our cost-effectiveness analysis indicated the economic viability of SBHE, particularly when accounting for spillover effects. This study is registered in the AEA RCT registry (No.0004265) and the ISRCTN registry (No.18002856).

Fig 1. CONSORT flow diagram: study profile with participant flow and the randomisation scheme.