Methods and Applications of Statistics in Clinical Trials, Volume 2: Planning, Analysis, and Inferential Methods

Chapter 38

Sample Size Considerations for Morbidity/Mortality Trials

Joanna H. Shih

38.1 Introduction

Sample size and power determination plays a crucial role in the design of clinical trials. To ensure that a trial is adequately sized so that the true treatment effect can be detected with a desired power, it is important to evaluate the assumptions used in the sample size calculation procedure. For morbidity/mortality trials, time-to-event outcome is usually used in the primary analysis. Because the rate of outcome event for this type of study is usually not very high, survival times for a large number of study participants are censored at the end of the study. Therefore, survival analysis is applied to time-to-event outcomes. Various closed-form sample size formulas have been proposed for the analysis of survival end point. Most of them are formed on the basis of restrictive assumptions and do not take into consideration factors that may affect the estimate of the treatment effect. Factors typically encountered in clinical trials include drop-in, drop-out, staggered entry, loss to follow-up, lags in treatment effects, and informative noncompliance. It is important to consider these issues in planning, because they recognize some practical problems encountered in the conduct of the research and ensure that intent-to-treat analyses are properly powered. This article reviews the sample size calculation methods for testing the treatment effects in two-arm survival trials. For each ...