Cluster Randomized Trial of Low-cost Microbiological Water Test Kits

The proposed trial utilizes a cluster randomized controlled methodology, with clustering occurring at the village level. The intervention that will be tested in this cluster randomized controlled trial is informed by the development of a simple, low-cost water quality test kit developed and piloted by J. Brown and colleagues in Nagpur, Maharashtra.

There will be two intervention arms and one control/comparison arm. In all selected villages, the study team will work with local government to conduct a village-level information session on water quality and key water safety behaviours. Informational material on water safety behaviours will be distrubted to all houseohlds following this information session. In the control group, households will be visited to provide water safety informational cards and collect a drinking water sample to be analyzed in the lab. They will not receive the results of the analysis until the conclusion of the study. Households in Intervention Arm 1 (Standard Testing) will have their water sampled and analyzed in the lab, and results of the analysis will be delivered to the households. They will receive informational cards describing water purification strategies. Households in Intervention Arm 2 (Test Kits) will be provided with the materials and training needed to perform water quality testing in their own household using the newly developed testing kits. They will also receive informational cards describing water purification strategies.

Following the initial village-level information session, the study team will randomly sample households for participation in the trial. Participating households will complete a short survey on household demographics and water handling, storage, and use behaviours and a water sample will be taken and tested in the lab for fecal indicator bacteria. At an unannounced follow-up visit, roughly 1 - 2 months after enrollment, households will complete a short survey about changes in water handling, storage, and use behaviours and a second sample of household stored drinking water will be collected for laboratory analysis. The effect of the cluster randomized controlled trial on water quality will be determined by comparing arithmetic mean E. coli count from samples of household water and changes in water safety behaviours between baseline and endline.

Microbiological drinking water quality will be assessed at two time points: baseline and endline. At each visit, a 330 ml sample of household water will be collected for analysis; we will ask participants to fill the sample container as if it were a drinking cup for a child living in the household. Samples will be kept cold (on ice in a cooler) until delivery to the laboratory and thereafter stored at 4oC until processing, to begin within 8 hours of the sampling event.