CTS: a clinical trials simulator.

We describe an interactive clinical trial simulation program (CTS), which provides continuous graphical and numerical output as a simulation runs. CTS is capable of simulating trials with two treatment groups and a binary outcome with options for: (1) up to five prognostic subgroups; (2) one of three methods of randomization; (3) equal or unequal allocation of patients to treatment groups; (4) a null or alternative hypothesis; (5) common methodological problems including biased referrals, false positive and false negative outcome reports, non-compliance, dropouts, crossovers, and co-interventions; (6) six methods of dealing with each of these problems; (7) the type I error rate, and (8) stratified or unstratified analysis. In addition, the current results of a trial in progress may be used as the starting point of the simulation to estimate the power of the trial if carried to completion. A few thousand simulations of a user specified scenario is typically entered and executed in about 5 minutes, permitting a student or clinical investigator to explore a variety of scenarios.