BAYESIAN INFERENCE FOR HUMAN IMMUNODEFICIENCY VIRUS (HIV) CLINICAL PROGRESSION: MODELLING CURRENT CLUSTER OF DIFFERENTIATION 4 (CD4) WITH PATIENT-LEVEL COVARIATES

Abstract

This study aimed to model current CD4 lymphocyte counts among people living with HIV by examining the effects of age, sex, baseline CD4 levels, and changes in body weight. A Bayesian regression framework was applied to clinical data to incorporate prior information and quantify uncertainty in parameter estimates. The findings indicate that baseline CD4 count was the strongest predictor of current CD4 levels. At the same time, age showed a moderate positive association, and the effects of sex and short-term weight change were smaller and less certain. Overall, the results demonstrate that Bayesian regression provides a robust and informative approach for understanding CD4 dynamics and supporting clinical decision-making in resource-limited settings.