Estimates of Information Growth in Longitudinal Clinical Trials

Abigail B. Shoben, University of Washington - Seattle Campus
Kyle Rudser, University of Minnesota
Scott S. Emerson, University of Washington


In group sequential clinical trials, it is necessary to estimate the amount of information present at interim analysis times relative to the amount of information that would be present at the final analysis. If only one measurement is made per individual, this is often the ratio of sample sizes available at the interim and final analyses. However, as discussed by Wu and Lan (1992), when the statistic of interest is a change over time, as with longitudinal data, such an approach overstates the information. In this paper, we discuss other problems that can result in overestimating the information, such as heteroscedasticity and correlated observations. We demonstrate that when using an inefficient estimator on unbalanced data, the true information growth can be nonmonotonic across interim analyses.