Supporting data and simulation code for predicting event accrual in randomized clinical trials using Gompertz modelling
Description
This dataset accompanies the manuscript “Predicting event accrual in ongoing randomized controlled trials – lessons learned from an early analysis of the Scandinavian trial of uncomplicated aortic dissection therapy.” The files contain Gompertz mortality model parameters derived from Danish population mortality data and baseline demographic characteristics of the first 50 SUNDAY trial participants, together with Stata code for simulating mortality and predicting event accrual. The simulation code evaluates different recruitment sample sizes, mortality hazard ratios and follow-up durations using 10,000 repetitions per scenario. It calculates the distribution of expected deaths and the probability of reaching the prespecified target of 80 events. An accompanying Excel workbook provides calculations of upper mortality estimates based on exact Poisson confidence limits and the corresponding expected number of deaths. The files support the methodological analyses presented in the manuscript. They do not contain individual-level participant records and do not independently reproduce every analysis reported in the manuscript.
Files
Institutions
- Aarhus UniversityCentral Jutland, Aarhus