Predictors of household drinking water E. coli contamination: Population-based results from rural areas of Ghana, Malawi, Mozambique, Niger, Rwanda, Uganda, and Zambia.

Abstract

Background: In sub-Saharan Africa, rural areas have lower rates of access to safe drinking water compared to urban areas. We investigated predictors of Escherichia coli contamination in drinking water of rural households in Ghana, Malawi, Mozambique, Niger, Rwanda, Uganda, and Zambia.

Methods: We used a population-based, cluster randomized sampling design to select rural households in each country. Household interviews on water access, sanitation, and hygiene (WaSH) practices and demographic characteristics were conducted and water samples from every fifth household were collected and enumerated for E. coli. Negative binomial regression models with survey sampling weights were run to evaluate predictors of E. coli contamination.

Results: A total of 18,747 rural household surveys (2,378-2,804 per country) were conducted and a total of 3,848 water samples (460-660 per country) were collected. Of surveyed rural households, 61-78% of households had high (11-100 E. coli cfu/100 mL) or very high (>100 cfu/100 mL) risk water quality in Ghana, Niger, and Uganda. Statistically significant WaSH predictors associated with lower E. coli incidence rates included using an improved-type primary water source (Mozambique), storing water in a narrow-mouthed container or container with a spigot (Niger), having continuous water supply during the dry season (Ghana), paying for water service (Rwanda), having soap or ash at handwashing points (Mozambique), having an improved-type household sanitation facility (Malawi), and attaining an education level greater than primary school (Niger and Zambia).

Conclusion: This study highlights the variability in WaSH access between rural areas of the study countries in association with microbial drinking water quality.