Sample Size and Data Monitoring for Clinical Trials with Extremely Low Incidence Rate

• Activity Number: 350
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Shih-Ting Chiu*+ and Shein-Chung Chow
• Companies: Duke University and Duke University School of Medicine
• Keywords: Low incidence rates ; Precision analysis ; Maximum error margin ; Data safety monitoring ; Bayesian approach

Abstract

In clinical trials, a pre-study power analysis for sample size calculation is often performed for selecting an appropriate sample size that will achieve a desired power (i.e., the probability of correctly detecting the clinically meaningful difference if such a difference truly exists) at a pre-specified level of significance. In practice, it is expected that a greater sample size is needed for detecting a smaller difference. A much larger sample size is required for detecting a relatively small difference, especially for those clinical studies with extremely low incidence rates. Thus, sample size calculation based on pre-study power analysis may not be feasible in practice. In this case, as an alternative, we propose to justify a selected sample size based on a precision analysis and a sensitivity analysis. A recommended step-by-step procedure for sample size determination in clinical trials with extremely low incidence rate is given. A statistical procedure for data safety monitoring based on probability statement during the conduct of the clinical trial is also proposed.