Repeated Measures Analyses in Clinical Trials With Titration Visits

摘要

In longitudinal clinical studies, after random izuiion at baseline,subjects are followed for a preiod of time for development of symptoms A mixed model for repealed measures (MMRM) can be used to analyze such sata To accommo-dale sufety and tolerability in studies,the treatment hs to be ktitrated to the optimal dose un-Then subjects will stay on therir optinal dose un-til the end of the study. In an MMRM analysis onr may attempt to ignore the titration visits because including them could add extra vari-ability unnecessarily. However, when patients drop out during the titration period, ignoring the tirtation visits would exclude these patients from the analyses In this article we evaluate the impact of excluding and including tiotration vis its in an MMRM analysis by a simmulation stusy We evaluadte the approavhes based on the vias and the coverage accuracy of the confisence in terval. The results suggest that excluding titra-tion visits may result in undercovered confi-dence inervals and biased estimates