Similarity and Distinction Between Preclinical and Clinical Longitudinal Studies in Statistical Modeling

• Activity Number: 336
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Shubing Wang*+
• Companies: Merck
• Keywords: Longitudinal data analysis ; Preclinical ; Linear mixed effects model ; Baseline adjustment ; multiplicity adjustment

Abstract

Longitudinal data present two types of variances: the between-subject and within-subject variances. Linear mixed effects models provide a flexible framework to identify and estimate the two types of variances. They have become the standard method for longitudinal data analysis in both preclinical and clinical settings. Being similar in practical experimental design, the statistical modeling in the two settings can be dramatically different due to the even subtle distinct nature of the studies. In this talk, we highlight the similarity in biological hypothesis, power analysis and randomization. The extensive discussions on how statistical model selection, missing data strategy, baseline and multiplicity adjustments in the two settings need delicate special care, based-on an exclusive set of examples.