Diarrheal illness creates a large and persistent disease burden among children in low-income countries; mortality estimates attribute 10% of global deaths among children < 5 to diarrhea. Fecal-oral pathogens (believed to be the most important causes of diarrhea) are transmitted from feces to new hosts through a variety of complex, environmentally mediated pathways that interact with each other and are also influenced by human behavior. Sanitation interventions that prevent the spread of pathogens from feces into the environment (e.g. provision of latrines) could block these transmission pathways and reduce the burden of diarrheal disease. Despite the potential importance of sanitation in reducing diarrhea, all studies to date of sanitation interventions on child gastrointestinal infections in rural settings have used non-randomized designs. Additionally, few studies have included detailed monitoring of sanitation-related behaviors and microbiological assessment of the complex pathogen transmission pathways that lead from feces to child illness.
Aims : The principal aims of this study are to assess the impact of improvements in sanitation infrastructure and household sanitation practices on fecal contamination along pathogen transmission pathways in rural Bangladesh, and to assess how these pathways mediate the impact of sanitation improvements on diarrhea and parasitic infections in children < 24 months old. We hypothesize that improved sanitation will reduce contamination of drinking water, household soil, and children's hands; we also posit that these pathways will mediate any reductions in diarrhea prevalence and parasitic infections caused by the intervention. Methods: We will test our hypotheses by nesting microbiological and behavioral measurements within WASH Benefits, a randomized controlled trial currently underway in rural Bangladesh funded by the Gates Foundation and led by our team. We will visit a total of 720 households from the trial's sanitation and control arms quarterly for 2 years, starting 12 months after intervention. At each visit, we will collect samples from tubewells, ponds, stored water, soil from the child play area, soil near tubewells and from children's hands for analysis of fecal indicator bacteria. During the last two sampling rounds, we will analyze samples for prominent fecal pathogens and perform source tracking to distinguish fecal contamination from humans versus animals. We will monitor sanitation practices through spot checks and passive infrared sensor measurements. These data will leverage the health outcomes measured in the existing trial protocol (diarrhea and helminth and protozoan infections in stool). We will compare fecal contamination between the study arms using intention to treat and complier average causal effect estimators. We will estimate the contribution of contamination to health outcomes through individual pathways by analysis of direct and indirect effects. Our team (UC Berkeley, Stanford, ICDDR,B) has extensive experience collaborating on studies of water, sanitation and hygiene in Bangladesh and other low-income countries on projects funded by the NIH, CDC, DFID, USEPA and the Gates Foundation.
Poor sanitary conditions are common in less developed countries and can lead to contamination of the environment with fecal pathogens, contributing to a large and persistent burden of diarrheal disease in these settings. Sanitation interventions that divert feces away from the environment through the provision of toilets and the adoption of adequate sanitary behaviors are thought to reduce the risk of child diarrhea and parasite infections. This randomized trial will test: (i) whether providing households with a package of improved toilets, tools to remove feces from their compounds, and a behavior change program will reduce fecal contamination of their environment, and (ii) which specific transmission pathways account for the reduced risk (if any) of diarrhea or parasite infections in children < 24 months in rural Bangladesh.
| Arnold, Benjamin F; Ercumen, Ayse (2016) Negative Control Outcomes: A Tool to Detect Bias in Randomized Trials. JAMA 316:2597-2598 |
| 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 |
