Modelling CD4 counts before and after HAART for HIV infected patients in KwaZulu-Natal South Africa

Background: This study aims to make use of a longitudinal data modelling approach to analyze data on the number of CD4+cell counts measured repeatedly in HIV-1 Subtype C infected women enrolled in the Acute Infection Study of the Centre for the AIDS Programme of Research in South Africa.
Methodology: This study uses data from the CAPRISA 002 Acute Infection Study, which was conducted in South Africa. This cohort study observed N=235 incident HIV-1 positive women whose disease biomarkers were measured repeatedly at least four times on each participant.
Results: From the findings of this study, post-HAART initiation, baseline viral load, and the prevalence of obese nutrition status were found to be major significant factors on the prognosis CD4+ count of HIV-infected patients.
Conclusion: Effective HAART initiation immediately after HIV exposure is necessary to suppress the increase of viral loads to induce potential ART benefits that accrue over time. The data showed evidence of strong individual-specific effects on the evolution of CD4+ counts. Effective monitoring and modelling of disease biomarkers are essential to help inform methods that can be put in place to suppress viral loads for maximum ART benefits that can be accrued over time at an individual level.

Keywords
Random-effects model; spatial covariance structure; CD4+ count; HAART; CAPRISA.