Improving the feasibility of stepped-wedge cluster randomised trials: a mixed methods enquiry

Stepped-wedge cluster randomised trials are methodologically complex, and can be logistically challenging, due to the staggered implementation of the intervention. These trials may therefore be more likely than other trials, to encounter issues that make them unable to successfully meet their objectives. Consequently, feasibility studies might be particularly useful for stepped-wedge cluster randomised trials. However, it has not been known what specific aspects of this trial design affect its feasibility, nor to what extent the feasibility of these trials has been investigated. The overarching aim of this thesis, was to improve the feasibility of stepped-wedge cluster randomised trials, by: understanding how the feasibility of these trials has been investigated; identifying the aspects of this trial design that affect its feasibility; investigating a solution to an identified issue; and making recommendations to improve the feasibility of trials of this design. This was achieved through a mixed methods enquiry. Feasibility studies, designed to inform stepped-wedge cluster randomised trials, were identified. The objectives of these studies were examined and the quality of reporting of those that had been published was assessed. The objectives of these studies were rarely specific to the chosen study design. Aspects of the stepped-wedge cluster randomised trial design that could affect its feasibility, were identified from individuals with an involvement, or interest, in trials of this design, through an online questionnaire and interview study. The process for stepped-wedge cluster randomised trials, of balancing prognostic covariates during the randomisation, was identified as a cause for concern for those involved in trials of this design. The use of covariate constrained randomisation for stepped-wedge cluster randomised trials, was therefore investigated using a simulation study. Based on the findings of this thesis, recommendations were made to improve the feasibility of future stepped-wedge cluster randomised trials.