Spillover effects of a combined water, sanitation, and handwashing intervention in rural Bangladesh: a randomized controlled trial

Abstract

BackgroundWater, sanitation, and handwashing (WSH) interventions may confer indirect benefits (\"spillovers\") on neighbors of recipients by interrupting pathogen transmission. We measured geographically local spillovers in WASH Benefits, a cluster-randomized trial in rural Bangladesh, by comparing outcomes among neighbors of intervention vs. control participants.\n\nMethodsWASH Benefits had randomly allocated geographically-defined clusters to a compound-level intervention (chlorinated drinking water, upgraded sanitation, and handwashing promotion) or control and followed children for two years. We enrolled neighboring children age-matched to trial participants that would have been eligible for WASH Benefits had they been conceived slightly earlier or later. After 28 months of intervention, we quantified fecal indicator bacteria in toy rinse and drinking water samples, measured soil-transmitted helminth infections, and recorded caregiver-reported diarrhea and respiratory illness. Neither fieldworkers nor participants were masked. Analysis was intention-to-treat.\n\nResultsWe enrolled neighbors of WASH Benefits participants in 90 control (N=900) and 90 intervention clusters (N=899). Neighbors characteristics were balanced across arms. The prevalence of any detectable E. coli in tubewell samples was lower for neighbors of intervention vs. control (prevalence ratio=0.83; 0.73, 0.95). There was no difference in E. coli and coliform prevalence between arms for other environmental samples. Disease prevalence was similar in neighbors of intervention vs. control participants: Ascaris (prevalence difference [PD]=0.00; -0.07, 0.08), hookworm (PD=0.01; -0.01, 0.04), Trichuris (PD=0.02; -0.02, 0.05), diarrhea (PD=0.00; -0.02,0.03), respiratory illness (PD=-0.01; -0.04, 0.03).\n\nConclusionsWe found spillover effects of a compound-level combined WSH intervention for tubewell water contamination but not for child health outcomes.\n\nKey MessagesO_LIWater, sanitation, and handwashing (WSH) interventions may confer indirect benefits (\"spillovers\") on neighbors of recipients by interrupting pathogen transmission, reducing environmental contamination, or spurring the adoption of health behaviors.\nC_LIO_LIWe conducted a randomized trial in rural Bangladesh to measure whether neighbors of a compound-level WSH intervention improved hygiene behaviors and had lower prevalence environmental contamination, soil-transmitted helminth infection, diarrhea, and respiratory illness among children under 5 years after two years of intervention.\nC_LIO_LIWe did not find evidence of intervention adoption or improved hygiene behavior among neighbors of a WSH intervention delivered for 2 years.\nC_LIO_LIThe WSH intervention reduced fecal contamination of neighbors tubewell water but did not lead to spillovers for other proximal measures of contamination in the domestic environment or for child health outcomes. For proximal spillover effects to translate to distal spillover effects, improvements in neighbors health behaviors may have been necessary.\nC_LI