A substantial body of evidence suggests that water, sanitation, and hygiene (WASH) interventions can effectively reduce diarrheal illness and other infectious diseases. While numerous studies have measured effects on children receiving WASH interventions, no studies have measured their effects (spillovers) on non-recipients who live adjacent to intervention recipients. Spillovers are important to understand because, when present, the estimated effects and cost effectiveness of interventions will be biased downwards. Although a rich, model-based literature describes spillovers of infectious disease interventions, few studies have measured spillovers empirically. We propose a randomized study to quantitate spillovers from WASH interventions. Our results will inform prioritization of interventions and guide allocation of public funds.
Specific aim : We propose to measure spillovers of water, sanitation, and hygiene interventions in an existing, large, rigorously designed trial: the WASH Benefits trial. Funded by the Bill & Melinda Gates Foundation, this trial aims to measure the individual and combined effects of water, sanitation, and hygiene interventions on child health and development. It is a cluster-randomized, controlled trial with si treatment arms and a double- sized control arm carried out in rural Bangladesh. WASH Benefits is the ideal platform upon which to design a study of spillovers because of its rigorous design, it large size (N=5,760 newborns), and the range of individual interventions it evaluates. We hypothesize that children who live in close proximity to compounds that receive a combined sanitation, hand washing, and water treatment intervention--compared to children who live in close proximity to control compounds (no intervention)--will have: 1) lower prevalence of diarrhea, 2) lower prevalence and intensity of infection of soil transmitted helminths, and 3) lower prevalence of respiratory illness. Methods: We will collect additional data from the existing combined intervention (sanitation+handwashing +water) and control arms of the WASH Benefits trial. For each WASH Benefits household, we will locate the nearest household with children 0-59 months of age that are not enrolled in WASH Benefits and collect data in that household. Our primary outcomes are caregiver-reported 7-day diarrhea and respiratory illness among children 0-59 months as well as soil transmitted helminth infection among children 0-59 months (the same age as the WASH Benefits cohort). Our findings will document either the presence or absence of spillovers of WASH interventions. Innovation: This will be among the first studies to empirically measure spillover effects with an experimental design, and it will be the first study t do so for WASH interventions. Our innovative and interdisciplinary design along with our findings will contribute to methodology in multiple disciplines (e.g. economics, political science)
Water, sanitation, and hygiene interventions could prevent illness among and disrupt disease transmission to not only children directly receiving the interventions but also children in geographic proximity to intervention recipients. Knowledge of these interventions' effects on geographically proximate children will improve our estimates of interventions' population-level impact and cost effectiveness; such information is needed to optimally allocate resources for health interventions. Our proposed cluster-randomized study will measure the effects of water, sanitation, and hygiene interventions in rural Bangladesh on the health of children neighboring intervention recipients.
| Benjamin-Chung, Jade; Amin, Nuhu; Ercumen, Ayse et al. (2018) A Randomized Controlled Trial to Measure Spillover Effects of a Combined Water, Sanitation, and Handwashing Intervention in Rural Bangladesh. Am J Epidemiol : |
| Arnold, Benjamin F; Ercumen, Ayse; Benjamin-Chung, Jade et al. (2016) Brief Report: Negative Controls to Detect Selection Bias and Measurement Bias in Epidemiologic Studies. Epidemiology 27:637-41 |
