BAYESIAN POISSON VECTOR AUTOREGRESSIVE MODEL TO EXPLORE THE DYNAMIC INTERRELATIONSHIP AMONG HIV/AIDS, TUBERCULOSIS, AND HEPATITIS IN NIGERIA

This study employs a Bayesian Poisson Vector Autoregressive (BAPVAR) model to investigate the dynamic interrelationships and mutual influences among HIV/AIDS, Tuberculosis (TB), and Hepatitis (HPT) diseases in Nigeria from 2004 to 2023. By capturing the temporal dependencies and forecast error variances, the analysis reveals complex interactions that underscore the significance of integrated disease control programs.

The trend analysis revealed a statistically significant decrease in the prevalence of all three diseases over the study period. For each year that passed, the expected log count of HIV cases decreased by approximately 3.9%, TB cases by 8.2%, and Hepatitis cases by 2.2%. The BaP-VAR/Log-VAR models showed that all three diseases exhibit strong and statistically significant meaning the prevalence in the previous year is a powerful predictor of the current year's prevalence. A significant cross-disease dynamic was identified: a 1% increase in TB prevalence in the prior year was associated with a 0.3762% decrease in HIV prevalence in the current year. Other cross-effects between the diseases were not found to be statistically significant. The forecast error variance decomposition showed that while each disease's own shocks were the main driver of its variance, shocks to TB prevalence explained a substantial and growing portion of the variance in HIV prevalence over time.