Online Program

Friday, February 20
CS10 Special Designs Fri, Feb 20, 2:00 PM - 3:30 PM
Borgne

Analysis Plans for Doubly Repeated Measures Designs (302919)

Robbie A. Beyl, Pennington Biomedical Research Center 
*Jeff Burton, Pennington Biomedical Research Center 
William Johnson, Pennington Biomedical Research Center 

Keywords: AUC, Multivariate model, Mixed model, OGTT, doubly repeated measures

Doubly repeated measures designs involve v visits, with each visit consisting of t time points. An example of this setting is the oral glucose tolerance test (OGTT), in which glucose is measured at several time points following ingestion of a glucose solution.The OGTT is carried out before and after the administration of a treatment. Comparing the change in the shape of the glucose curve from baseline to follow-up between two or more treatment groups is the primarily goal. A common approach used by nonstatistical researchers is to ignore the baseline visit and use area under the curve (AUC) analysis to compare group curves at the follow-up visit only. Alternatively, one may analyze this type of data using a multivariate model or a linear mixed model for repeated measures. For each method, we will go over assumptions and advantages/disadvantages when using complete data and when only using follow-up data. Additionally, using both simulated and real data, the performance of hypothesis tests under each method will be shown and compared. Last, we will go over approaches to explaining each method to nonstatistical researchers.