Inconsistent use of household water treatment and safe storage (HWTS) systems reduces their potential health benefits. Ultraviolet (UV) disinfection is more convenient than some existing HWTS systems, but it does not provide post-treatment residual disinfectant, which could leave drinking water vulnerable to recontamination. In this paper, using as-treated analyses, we report on the field efficacy of a UV disinfection system at improving household drinking water quality in rural Mexico. We further assess the risk of post-treatment contamination from the UV system, and develop a process-based model to better understand household risk factors for recontamination. This study was part of a larger cluster-randomized stepped wedge trial, and the results complement previously published population-level results of the intervention on diarrheal prevalence and water quality. Based on the presence of Escherichia coli (proportion of households with ≥ 1 E. coli/100 mL), we estimated a risk difference of -28.0% (95% confidence interval (CI): -33.9%, -22.1%) when comparing intervention to control households; -38.6% (CI: -48.9%, -28.2%) when comparing post- and pre-intervention results; and -37.1% (CI: -45.2%, -28.9%) when comparing UV disinfected water to alternatives within the household. We found substantial increases in post-treatment E. coli contamination when comparing samples from the UV system effluent (5.0%) to samples taken from the storage container (21.1%) and drinking glasses (26.0%). We found that improved household infrastructure, additional extractions from the storage container, additional time from when the storage container was filled, and increased experience of the UV system operator were associated with reductions in post-treatment contamination. Our results suggest that the UV system is efficacious at improving household water quality when used as intended. Promoting safe storage habits is essential for an effective UV system dissemination. The drinking glass appears to represent a small but significant source of recontamination that is likely to impact all HWTS systems.

Complier average causal effects (CACE) estimate the impact of an intervention among treatment compliers in randomized trials. Methods used to estimate CACE have been outlined for parallel-arm trials (e.g., using an instrumental variables (IV) estimator) but not for other randomized study designs. Here, we propose a method for estimating CACE in randomized stepped wedge trials, where experimental units cross over from control conditions to intervention conditions in a randomized sequence. We illustrate the approach with a cluster-randomized drinking water trial conducted in rural Mexico from 2009 to 2011. Additionally, we evaluated the plausibility of assumptions required to estimate CACE using the IV approach, which are testable in stepped wedge trials but not in parallel-arm trials. We observed small increases in the magnitude of CACE risk differences compared with intention-to-treat estimates for drinking water contamination (risk difference (RD) = -22% (95% confidence interval (CI): -33, -11) vs. RD = -19% (95% CI: -26, -12)) and diarrhea (RD = -0.8% (95% CI: -2.1, 0.4) vs. RD = -0.1% (95% CI: -1.1, 0.9)). Assumptions required for IV analysis were probably violated. Stepped wedge trials allow investigators to estimate CACE with an approach that avoids the stronger assumptions required for CACE estimation in parallel-arm trials. Inclusion of CACE estimates in stepped wedge trials with imperfect compliance could enhance reporting and interpretation of the results of such trials.

Low adoption and compliance levels for household water treatment and safe storage (HWTS) technologies have made it challenging for these systems to achieve measurable health benefits in the developing world. User compliance remains an inconsistently defined and poorly understood feature of HWTS programs. In this article, we develop a comprehensive approach to understanding HWTS compliance. First, our Safe Drinking Water Compliance Framework disaggregates and measures the components of compliance from initial adoption of the HWTS to exclusive consumption of treated water. We apply this framework to an ultraviolet (UV)-based safe water system in a cluster-randomized controlled trial in rural Mexico. Second, we evaluate a no-frills (or "Basic") variant of the program as well as an improved (or "Enhanced") variant, to test if subtle changes in the user interface of HWTS programs could improve compliance. Finally, we perform a full-cost analysis of both variants to assess their cost effectiveness (CE) in achieving compliance. We define "compliance" strictly as the habit of consuming safe water. We find that compliance was significantly higher in the groups where the UV program variants were rolled out than in the control groups. The Enhanced variant performed better immediately postintervention than the Basic, but compliance (and thus CE) degraded with time such that no effective difference remained between the two versions of the program.