Sample Size Calculations for Clinical Trials with a Survival Endpoint

Steven Julious – Medical Statistics Group, School of Health and Related Research

The three most important things in any clinical investigation are: design, design and design. A sample size calculation is just one aspect of a study design. This presentation will describe the calculations to estimate the sample size where the primary endpoint is a survival endpoint. Initial basic calculations will be introduced to calculate the number of events required in the trial. The talk will then describe the steps for calculating total sample size to have the estimated number of events. This total sample size will depend on the: duration of recruitment; duration of follow up and rate of censoring. It will be highlighted how the sample size for the same effect size can be reduced depending on duration of follow up both in total and to the period of recruitment.

Reference
SA Julious. Estimating sample sizes in clinical trials. Chapman and Hall, London, 2010