"background": "Municipal infrastructure asset systems in sub-Saharan Africa face severe financial constraints, yet robust methodologies for evaluating the cost-effectiveness of maintenance and rehabilitation interventions are lacking. Current diagnostic approaches are often retrospective and fail to account for heterogeneous asset conditions and contextual operational factors. ", "purpose and objectives": "This article presents a novel methodological framework for conducting randomised field trials (RFTs) to diagnose the cost-effectiveness of municipal infrastructure asset management. The primary objective is to provide a replicable protocol for generating comparative evidence on intervention strategies under real-world conditions. ", "methodology": "The proposed RFT methodology clusters infrastructure assets into statistically balanced blocks based on covariates like age and soil type, followed by random assignment of different maintenance protocols. Cost-effectiveness is measured via a longitudinal performance-cost ratio. The primary analysis employs a generalised linear mixed model: \ (Y{it) = \0 + \1 Tit + ^\ \\ + ui +, where Yit is the cost-effectiveness ratio for asset i at time t, Tit is the treatment indicator, are covariates, and uᵢ is a random intercept. Inference uses cluster-robust standard errors. ", "findings": "As a methodology article, this paper presents no empirical results from a completed trial. However, the designed framework indicates that a minimum detectable effect size of 0. 35 standard deviations is achievable with 80% power for a trial involving 15 municipalities, each with 20 asset clusters, assuming an intra-cluster correlation coefficient of 0. 10. ", "conclusion": "The structured RFT methodology provides a rigorous, evidence-based alternative to observational studies for infrastructure diagnostics. It enables causal inference on cost-effectiveness, directly informing capital allocation and maintenance policy. ", "recommendations": "Municipal engineers and asset managers should adopt this RFT framework for piloting new interventions
Kirabo et al. (Tue,) studied this question.