Thesis - Statistical Methods for Clinical Data: Survival Analysis, Longitudinal Regressions, and Bayesian Monitoring
Abstract
Implementation of clinical trials is a necessary step in increasing medical knowledge, such as providing information about the efficacy of an innovative medical device, procedure, or a medication. To establish the efficacy, human participants are carefully selected based on their characteristics suitable for the study. They are randomly assigned to either a treatment or control group and are monitored and measured over time to detect any physical changes. Clinical data obtained this way is vital in determining the efficacy of the tested product. In this thesis, we give an overview of a broad range of statistical methodology used in analysis of clinical data. We present techniques from survival analysis, longitudinal regression modeling, and Bayesian monitoring of clinical trials. For each method, we discuss theoretical framework and illustrate with an application to a suitable data set.