Novel Approach of Multistate Markov Chains to Evaluate Progression in the Expanded Model of Non-alcoholic Fatty Liver Disease

Abstract

A global increase in the prevalence of obesity and type 2 diabetes is strongly connected to an increased prevalence of non-alcoholic fatty liver disease (NAFLD) worldwide. In this article, the progression of the NAFLD process is modeled by continuous time Markov chains (CTMCs) with nine states. Maximum likelihood is used to estimate the transition intensities among the states. Once the transition intensities are obtained, the mean sojourn time and its variance are estimated, and the state probability distribution and its asymptotic covariance matrix are also estimated. A hypothetical example based on a longitudinal study assessing patients with NAFLD in various stages is discussed. The mean time to absorption is estimated, and the other abovementioned statistical indices are examined. In this article, the maximum likelihood estimation (MLE) function is utilized in a new approach to compensate for the missing values in the follow-up period of patients evaluated in longitudinal studies. A MATLAB code link is provided, at the end of the article, for the estimation of the transition rate matrix and transition probability matrix.