Sample Size Calculation for Comparability Studies Under Nested Mixed Models

• Activity Number: 86
• Type: Contributed
• Date/Time: Sunday, August 9, 2015 : 4:00 PM to 5:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Qianqiu Li* and Bill Pikounis and Todd Yeager
• Companies: Johnson & Johnson and Janssen R&D and Janssen R&D
• Keywords: Sample Size Calculation ; Comparability ; Equality and Equivalence ; Nested Mixed Models ; Variance Components ; Estimation Accuracy

Abstract

Experimental designs are critical in qualification and comparability evaluation of a new manufacturing site or testing lab. This paper presents sample size estimation approaches via between-site percent mean differences and site-to-site variability in multilevel designs involving operator and testing day as random effects nested within site. Given ratios of variance components with respect to between-site variance, numbers of operators and testing days are determined given acceptable powers from testing percent mean differences in equality and equivalence comparisons and by prespecified estimation accuracy of the site-to-site variability.