Design Considerations and Analytical Framework for Reliably Identifying a Beneficial Individualized Treatment Rule

• Activity Number: 208 - Personalized and Precision Medicine
• Type: Contributed
• Date/Time: Tuesday, August 10, 2021 : 1:30 PM to 3:20 PM
• Sponsor: Biometrics Section
• Author(s): Charles H Cain* and Thomas Murray and Kyle D Rudser and Alexander J Rothman and Anne C Melzer and Anne M Joseph
• Companies: University of Minnesota and Division of Biostatistics, University of Minnesota and University of Minnesota and University of Minnesota and Minneapolis VA Health Care System and University of Minnesota
• Keywords: Individualized Treatment Rule; Personalized Medicine; Clinical Trials; LASSO; Sample Size

Abstract

An individualized treatment rule (ITR) formalizes personalized medicine by assigning treatment as a function of patients' clinical information, which contrasts with a static treatment rule that assigns everyone the same treatment. Much of the focus on ITRs revolves around identifying rules that are close to a theoretical optimal rule, which could lead to identifying rules that perform worse than the optimal static rule particularly in the absence of substantial effect heterogeneity. This limitation motivates new methods to select the penalty parameter in a LASSO model such that the static rule is identified with high probability in the absence of treatment effect heterogeneity and considerations for sample size regarding reliable identification of a beneficial ITR using a Monte Carlo integration based calculation of the probability to identify a beneficial ITR—an ITR that performs better than the optimal static rule. Results in both simulation and application to the PLUTO study show that the probability is low in trials where the main effect is larger relative to the interaction effect and tends to be higher when a higher proportion of people benefit from an ITR.