"background": "Rural water treatment systems in sub-Saharan Africa are critical for public health, yet their operational performance and associated public health risks are often poorly quantified. Existing assessments are typically cross-sectional, limiting causal inference about the efficacy of diagnostic and intervention strategies. ", "purpose and objectives": "This case study aimed to develop and apply a randomised field trial methodology to rigorously diagnose performance failures and measure the risk reduction achieved through a standardised maintenance intervention for rural water treatment facilities. ", "methodology": "We conducted a clustered, randomised controlled trial involving 120 community water treatment systems. Facilities were randomly allocated to control or intervention groups. The intervention comprised a structured performance diagnostic followed by targeted corrective maintenance. Primary outcomes were the reduction in technical performance failures and the associated microbiological risk, measured via log-reduction values (LRV) for E. coli. Performance was modelled using a generalised linear mixed model: \ (P (Y{ij=1) ) = \0 + \1 Tij + uj + eij, where Yij is a failure indicator for system *i* in cluster *j*, Tij is the treatment assignment, and uⱼ are random cluster effects. ", "findings": "The intervention significantly reduced the probability of a major performance failure by 47 percentage points (95% CI: 38 to 56). The mean LRV improvement in the intervention arm was 1. 8 log units, indicating a substantial reduction in microbial risk. Uncertainty in the treatment effect was quantified using robust standard errors clustered at the facility level. ", "conclusion": "The randomised trial format proved to be a methodologically robust approach for generating causal evidence on engineering interventions in rural water infrastructure. Structured diagnostics coupled with maintenance can yield significant and measurable improvements in system performance and public health protection. ", "recommendations": "Implement periodic, structured performance diagnostics as part of routine maintenance protocols. Funding models should support evidence-based interventions validated through field trials. Further research should
Jean de Dieu Uwimana (Wed,) studied this question.