Application of stochastic processes to participant recruitment in clinical trials

Abstract

The allocation of sufficient time for participant recruitment is one of the fundamental aspects in planning a clinical trial. This paper illustrates how a Poisson process can be used to determine an optimal period of time for participant recruitment by simulating Poisson arrival into a clinical trial. The simulation study provides the means to generate of an empirical probability density function for the recruitment time based on time-dependent changes in the accrual rate. From this empirical distribution, a clinical trial recruitment period can be planned to provide a high level of confidence (e.g., 90% probability) of enrolling the sample size within the planned amount of time given the simulation assumptions.