Optimizing clinical trial design using prospective cohort study data: a case study in neuro-urology

Abstract

Study Design Simulations using data from a prospective cohort study. Objectives To illustrate how prospective cohort data can be employed in randomized controlled trial (RCT) planning to assess feasibility and operational challenges, using TASCI (Transcutaneous tibial nerve stimulation in patients with Acute Spinal Cord Injury to prevent neurogenic detrusor overactivity: a nationwide randomized, sham-controlled, double-blind clinical trial) as a case study. Setting Spinal cord injury rehabilitation centers in Switzerland. Methods TASCI is nested in the multicenter Swiss Spinal Cord Injury Cohort Study (SwiSCI), which prospectively includes patients with acute spinal cord injury. In simulations, data from 640 patients, collected by SwiSCI, were used to investigate different scenarios of patient eligibility and study consent, as well as the performance of the randomization list. Descriptive analysis was used to describe the population of interest and the simulation results; multivariable logistic regression analysis was performed to identify predictors of discharge within the TASCI intervention time period. Results The recruitment target of 114 patients is obtainable within the originally envisioned three-year time period under the most favorable recruitment scenario examined. The distribution of the primary prognostic factor produced imbalance in the randomization lists and informed further discussion of the cut-off values used in stratification. Influxes of patients resulted in overlapping intervention periods for multiple participants, which guided resource allocation. Early discharge was related to the primary prognostic factor and study center, but is only anticipated in about 8% of participants. Conclusions Prospective cohort data are a very valuable resource for planning RCTs.