Automatic Chlorination and Child Health in Urban Bangladesh

Communicable Diseases, Diarrhea

Water treatment, Child growth, Enteric infection

Municipal water networks within industrialized countries typically rely on centralized treatment to manage piped water quality. Optimal water quality at the tap, however, requires well-maintained piped distribution networks, and performs best when piped systems are fully pressurized. In low-income cities such as Dhaka, water distribution networks are inadequately maintained and typically supply intermittent service; as such, they are vulnerable to recontamination during negative pressure events. Among populations accessing these types of improved water sources in urban settings (e.g. shared taps), it is unknown if consistent treatment to provide chlorinated water at the point of collection would have a significant health benefit. Furthermore, almost all previous studies of water treatment interventions in low-income countries have been unblinded with self-reported diarrhea as the main outcome, casting doubt that reported impacts of water disinfection on diarrhea are not due entirely to social desirability bias. Stanford University in collaboration with icddr,b will conduct a randomized evaluation to assess the impact on access to automatically chlorinated water on water quality and child health.

Investigators will conduct a blinded cluster randomized controlled trial to evaluate the health and economic impacts of having access to automatically chlorinated water. The unit of randomization will be shared water points that typically serve 20-200 households. Shared water points connected to holding tanks compatible with the water treatment technology, and serving more than 4 households with at least one child under five, will be identified. Households accessing eligible water points as their primary drinking water source will be enrolled before installation of chlorine devices, and a baseline survey will be conducted of water quality, diarrhea prevalence, and health care expenditures. Following this baseline, households will be randomly assigned to control or treatment groups. The chlorination devices will be installed at the treatment group water points, while a doser containing vitamin C (and no chlorine), will be installed in the control group. The free chlorine dosing target will be below \<1ppm to preserve blinding. All households will be surveyed every 2-3 months for a total follow up period of 14-16 months (5-7 survey rounds, budget permitting). Objectives: 1. To evaluate the impact of an automated chlorination system on microbial stored drinking water quality, residual chlorine, user satisfaction, user perceptions of water taste and smell, under-five child diarrhea (longitudinal prevalence) compared to a control group. 2. Compare the marginal additional cost (per person served) of installing and maintaining an automated chlorination system integrated with the current water supply infrastructure in low-income areas of Dhaka. 3. Assess the impact of an automated chlorination system on hospital visits and health care expenditures. 4. To measure secondary outcomes of the impact of an automated chlorination system including under-five child weight-for-age (WAZ), under-five height-for-age (HAZ), as well as levels of C-reactive protein and immunoglobin G in serum samples collected from children under five (these are objective indicators of infection, such as repeat diarrheal episodes). Analysis: The primary analyses will be intent-to-treat (investigators will analyze differences in outcomes between the treatment and control groups, with groups defined by their random allocation). Investigators will also conduct a secondary analysis comparing outcomes between intervention and control, where the intervention group is defined as those households that had free chlorine residual detected in their stored drinking water (treated on the treated analysis).

Bangladesh