Measuring Agreement in Method Comparison Studies with Skew-Normal Mixed Models

• Activity Number: 71
• Type: Contributed
• Date/Time: Sunday, July 31, 2011 : 4:00 PM to 5:50 PM
• Sponsor: Biometrics Section
• Author(s): Dishari Sengupta and Pankaj Choudhary*+
• Companies: The University of Texas at Dallas and The University of Texas at Dallas
• Keywords: Agreement evaluation ; Concordance correlation ; EM algorithm ; Multiple comparisons ; Repeated measurements ; Total deviation index

Abstract

The primary aim of a method comparison study is to evaluate agreement between two or more methods of measurement of a continuous variable. It is common to model the data from method comparison studies using a normal mixed model. But the assumption of normality for random effects and errors does not always hold. To deal with this issue, we consider a generalization of the normal mixed models where the random effects and errors follow skew-normal distributions, which includes normality as a special case. This model is fitted using a version of EM algorithm. To estimate the extent of agreement between the measurement methods, the resulting model parameter estimates are plugged into a measure of agreement, and large-sample considerations are used to obtain a confidence interval for the agreement measure. This methodology can accommodate repeated measurements from each measurement method, more than two measurement methods, and unbalanced designs. The methodology is illustrated by a applying it to a real dataset. Results of a simulation study are also reported.