Content area
Abstract
Cluster randomized trials, which often enroll a small number of clusters, can benefit from the use of constrained randomization to balance potentially prognostic covariates during the design phase. In covariate-constrained randomization, a final randomization scheme is selected from a list of randomizations that are known to balance chosen cluster-level covariates between the trial arms. An ongoing cluster randomized trial aims to investigate the efficacy of Targeted Indoor Residual Spraying (TIRS) in the prevention of symptomatic dengue, chikungunya, and Zika in children in Mérida, Mexico. This study utilizes covariate-constrained randomization to balance two treatment arms across four specified covariates and geographic sector. As some selected clusters may have been subsequently found unsuitable in the field, we desired a strategy to substitute new clusters while maintaining balance. We developed an algorithm that successfully identified a set of clusters that maximized the average minimum pairwise distance between clusters in order to reduce contamination and balanced the specified covariates both before and after substitutions were made. Simulations were performed to explore the limitations of this algorithm. Additionally, previous literature has addressed the suitability of adjusting an analysis for the covariates that were balanced in the design phase when the outcome is continuous or binary. Here we extended this work to time-to-event outcomes. We conducted a simulation study to assess type I error rates and power between simple randomization and constrained randomization using both prognostic and non-prognostic covariates. We analyzed the data using a semi-parametric Cox proportional hazards model with robust variance, a mixed effects Cox model, and a permutation test utilizing deviance residuals. We also present an analysis of statistical validity for the constrained randomizations. Finally, we present an exploration of analysis methods to account for the hierarchical clustering structure of the TIRS Trial data.