Use, Microbiological Effectiveness and Health Impact of a Household Water Filter Intervention in Rural Rwanda-A Matched Cohort Study

Published In

International Journal of Hygiene and Environmental Health

Document Type


Publication Date



Unsafe drinking water is a substantial health risk contributing to child diarrhoea. We investigated impacts of a program that provided a water filter to households in rural Rwandan villages. We assessed drinking water quality and reported diarrhoea 12–24 months after intervention delivery among 269 households in the poorest tertile with a child under 5 from 9 intervention villages and 9 matched control villages. We also documented filter coverage and use. In Round 1 (12–18 months after delivery), 97.4% of intervention households reported receiving the filter, 84.5% were working, and 86.0% of working filters contained water. Sensors confirmed half of households with working filters filled them at least once every other day on average. Coverage and usage was similar in Round 2 (19–24 months after delivery). The odds of detecting faecal indicator bacteria in drinking water were 78% lower in the intervention arm than the control arm (odds ratio (OR) 0.22, 95% credible interval (CrI) 0.10–0.39, p < 0.001). The intervention arm also had 50% lower odds of reported diarrhoea among children <5 than the control arm (OR = 0.50, 95% CrI 0.23–0.90, p = 0.03). The protective effect of the filter is also suggested by reduced odds of reported diarrhoea-related visits to community health workers or clinics, although these did not reach statistical significance.



Persistent Identifier