Assessing the Effectiveness of a Water Filter Technology as an Arsenic Mitigation Strategy in West Bengal

The Unilever Pureit As+ filter will be assessed for chemical and microbiological performance and health impact using a parallel, household-randomised controlled trial (RCT) design over a four-month follow-up period.

This study will follow an open (non-blinded) design because filtration technologies are particularly challenging to blinding (Boisson et al., 2010). Equal randomisation will be applied (i.e. 1:1). Randomisation will be undertaken through a lottery organised in each participating village.

The study will be undertaken in two phases; an initial phase will involve a baseline survey in the identified affected communities. Those households identified as eligible during the baseline survey will then be randomised into the intervention or control group and the intervention will then be distributed. After a 2-4 week adaptation period, control and intervention arms will be followed at monthly visits for a period of four months.

The baseline study will collect information on demographics, sanitation facilities, hygiene practices, drinking water sources and treatment practices, diet of household members, knowledge of arsenic contamination and arsenicosis, and other factors that could potentially impact the outcomes of interest. Additionally, data on behavioural factors that have been observed to be associated with Household Water Treatment (HWT) use will be collected to predict intervention use in the intervention arm. During the baseline survey, water samples of source water and stored drinking water will be collected and tested for total thermotolerant coliforms (TTC) and total arsenic concentration. The female head of the household will be invited to provide a sample of urine for total urinary arsenic testing. The female head of the household will be followed because it is expected that she will be the household member that i. is mainly responsible for drinking water management, and ii. is at home most of the day and thus making it logistically feasible to obtain spot urine samples throughout the study.

At monthly intervals, participating households will be visited to assess filter use using self report as well as objective indicators. Additionally, in each round of follow-up, drinking water samples will be taken in all participating households for both microbiological and arsenic testing. Additionally, the female head of household will be invited to provide spot urine samples for total urinary arsenic at each round of follow up.

For arsenic testing samples of stored drinking water will be collected and analysed for total arsenic. Additionally, in intervention households, samples of source water will also be collected to assess the percentage removal of Arsenic by the Pureit AS+ filter. Water samples will be collected in polyethylene bottles pre-washed with nitric acid water and nitric acid (0.1% v/v) will be added after collection at the laboratory as a preservative (Roychowdhury, 2008). Samples will be stored in ice box until arrival to the lab and stored at 4°C until shipment to the laboratory for analysis. Total water arsenic will be measured by Flow Injection Hydride Generation Atomic Absorption Spectrometry (FI-HG-AAS) and or Inductively coupled plasma mass spectrometry (ICP-MS).

For microbiological testing, in each round of follow-up, in intervention households water samples will be collected of i. stored drinking water (whether this is from the intervention filter or another container), ii. in a 25% random sample, samples will also be taken of influent water (source water) and effluent water out of the filter (if they keep their drinking water elsewhere). For control households, water samples will be collected of i. stored drinking water and, ii. in a 25% random sample, from the water source. Untreated drinking water will be collected to determine technology performance by comparing effluent and influent water quality. For the stored drinking water, the householder will be asked if there is any drinking water in the house at that moment and a sample will be collected from the vessel that the householder identifies as being the main container used at that moment to fill a cup or glass for drinking. All samples for microbial analysis will be collected in sterile Whirl-Pak™ bags containing a tablet of sodium thiosulfate to neutralize any disinfectant. Samples will be placed on ice and processed within 4-6 hours of collection to assess TTC levels. Microbiological assessment will be performed using a DelAgua field incubator (University of Surrey) in accordance with Standard Methods (Eaton et al., 2005).

Acceptability and major limitations with the filtration technology will be assessed during the final follow-up visit. The acceptability of the technology will be evaluated by conducting household surveys, in-depth interviews (IDI) and focus group discussions (FGDs). A subsample of male and female heads of households will be asked to participate in FGDs at the end of the study to provide feedback on the technology including perceived benefits and limitations.