In this study, which we conceived of as a companion experiment to research that Aquaya and UC Berkeley researchers conducted in Kenya (with support from the aid group CARE), we examined the point-of-use (POU) water treatment preferences of 800 households in a densely populated mixed income area of Dhaka, Bangladesh.  We cycled 600 households through two-month trials of each of four POU treatment methods: a liquid chlorine disinfectant, a tablet chlorine disinfectant, a chlorine-based disinfectant/flocculant powder mixture, and a silver-impregnated porous ceramic filter.  The remaining 200 households served as controls.

Household uptake of all the products was consistently below 30% (considerably lower than the approximately 70% uptake that we observed among subsistence farmers in rural western Kenya).  When households used the products, they did generate significant increases in drinking water safety as compared to the control households.  In addition, we observed less microbial water contamination in households that used the chlorine-based products than in households that used the non-chemical ceramic filter.  At the same time, the filter – which did not affect taste and odor – was the most highly used of the tested products (a pattern that is consistent with our observations from rural Kenya).

We offer several possible explanations for the lower product uptake in Bangladesh than in Kenya.  First, much of the Kenyan population that we surveyed intermittently relies on turbid surface water sources that are visibly contaminated, whereas the urban population we studied in Bangladesh has access to improved, water sources.  Second, our Kenyan study region was the target of multiple water, sanitation, and hygiene education campaigns over the years, including the use of point-of-use water treatment products, whereas our study site in Bangladesh has received fewer previous interventions.  Third, we suspect that the Kenyan study population of rural subsistence farmers was more likely to view the water treatment products as “aspirational” goods than were the urban Bangladeshis we surveyed in Dhaka.

In any case, the generally low adoption of these POU products in Dhaka, where fecal contamination of drinking water is a significant public health concern, was sobering. Most theories of health decision-making identify consumers’ lack of information regarding the dangers of untreated drinking water, coupled with product cost, as the central barriers to the uptake of POU products.  Our intervention addressed these obstacles by 1) setting price to zero during the free product trials; and 2) making multiple educational household visits.  Although our informational messages had a minor effect, the overall low usage of these products indicates that POU water treatment products are not likely to reduce the incidence of waterborne illness among poor and working-class inhabitants of Bangladesh in the near future.

You can access our full journal article from the Bangladesh study here.