Repeated Outcomes Regression Methods

Abstract

In clinical research repeated measures in a single subject are common. In its simplest version it is a statistical model only assessing the differences between within subject differences as outcomes without any predictors. More complex data are increasingly present in clinical research. The current chapter provides analysis methods for repeated outcomes including predictors either continuous or categorical. Also analysis models where within subject comparisons receive fewer degrees of freedom for the benefit of better sensitivity to test between subject differences are reviewed. Finally, multivariate models with batteries of more than a single set of repeated measures are reviewed. Some of the models reviewed are pretty complex, but increasingly used and sometimes indispensable in modern clinical data analysis.