ASSESSMENT OF CONFIRMABILITY IN CLINICAL TRIALS AND INDIVIDUALISED MEDICINE DEVELOPMENT USING STATISTICAL TOLERANCE REGIONS

Abstract

This thesis presents assessment of confirmability and individualised medicine development in two-arm randomised clinical trials framework. A novel clinically meaningful model, ETZ model, is proposed for assessing confirmability of study results. We aim to increase the phase transition success rate, that a promising Phase 2 result will have a high probability of being confirmed in a Phase 3 study. Therefore, the drug developer decides whether to invest in additional resources to reduce the variability of a particularly impactful component. A key innovation of our research is to show ETZ variabilities can be estimated from three variances rou- tinely in randomised clinical trials. An app is developed for assessing of each ETZ variability component on confirmability. The idea of our individualised medicine research is that targeting a subgroup is only possible when variability between patients within each treatment arm is small and the measurement error is small. A key innovation is that we consider the patient- to-patient variability in individualised medicine. We also come up with an innovative pivot quantity. In addition, we discuss about the derivation of covariance-variance matrix in mixed model repeated measurements to obtain more appropriate estimates.

Date of Award	25 Nov 2022
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Yang Han (Main Supervisor) & Jianxin Pan (Co Supervisor)