Modeling the Progression of HIV/AIDS by a Hidden Markov Model(HMM)

• Activity Number: 534 - Contributed Poster Presentations: Section on Statistics in Epidemiology
• Type: Contributed
• Date/Time: Wednesday, August 1, 2018 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Sanam Sanei* and Shanglun Li and Le Le Bao
• Companies: Penn State and Penn State and Penn State University
• Keywords: HIV Progression; Hidden Markov Model; CD4 cells measurement

Abstract

CD4 cells progression rate and patients' death rate are key indicators of the progression of HIV disease. There are two facts about CD4 cell counts; first, there is a substantial measurement error due to imprecise measurement techniques. Second, the number of CD4 cells varies over time because of the short-term variations in the immune system. These two factors lead to uncertainty about the progression state of HIV disease and make a hidden Markov model (HMM) an appropriate approach to uncover the disease progression. We use an HMM framework and devise a new MCMC estimation method to determine transition probabilities across different states of disease progression. We apply the model to a dataset of a panel of more than 30000 Chinese patients over 7 years. We gain insights about HIV progression, treatment effect, and death rate across various subgroups of patients based on demographic information.