New Covariate-Adjusted Response-Adaptive Designs and Associated Methods for Statistical Inference

George Washington University, Washington DC

Investigators

Abstract

Precision medicine is an approach that allows physicians to tailor a treatment regimen based on an individual patient's characteristics (which could be biomarkers or other covariates). With today's modern technology, it is much easier to identify important biomarkers (usually called predictive biomarkers) that may associate with certain diseases and their treatments. To design an efficient clinical study of precision medicine, one should incorporate these useful predictive biomarkers. In this project, the investigator undertakes a systematic and comprehensive study of feasible designs and statistical inference related to clinical trials of precision medicine. The project will introduce and study innovative designs and statistical methods and will investigate how to implement the proposed designs and statistical methods in some real clinical trials. The objective of this project is to introduce new designs and to provide novel statistical methods for adaptive randomized clinical trials of precision medicine. Recently, advances in genetics have permitted identification of genes (biomarkers) that are linked to certain diseases. These biomarkers (called predictive biomarkers) could be used to predict the response of a specific treatment and the confirmation of their predictive power relies on carefully designed clinical studies. To develop precision medicine, groundbreaking designs are needed for clinical trials so that predictive biomarkers can be incorporated to facilitate treatment selection. Advance and innovative statistical methods are required to match special features of clinical trials of precision medicine. This project focuses on the designs of clinical trials and the corresponding statistical inference for precision medicine. The investigator introduces innovative classes of covariate-adjusted response-adaptive designs, studies the statistical properties of these adaptive designs, develops new methods for statistical inference and obtains their properties, and applies these methods to practical problems.