Bayesian approach in estimating risk determinants of infectious diseases

Abstract

HIV/AIDS has had greatest adverse impact on TB. People coinfected with TB and HIV are at high risk of dying from TB. Similarly the relationship between HIV and STIs is complex as they are both sexually transmitted. In South Africa, no coherent analysis of determinants of HIV and TB has been done at a national level and thus this study seeks to mend this gap. Results will therefore provide a better understanding of the spread of these two epidemics in South Africa. The dataset used is from a population based second-generation surveillance survey conducted by the HSRC in 2005. The survey was conducted in 15 000 households from 1000 enumerator areas. The results show that sex, race, age, education and health are important determinants of HIV and TB. Furthermore, the spread of HIV is clustered within an enumerator area. Thus intervention programmes should target enumerator area level for their maximum effects.