Modeling of CD4 counts change and survival times of HIV naive patients on ARVs

Abstract

This study was a longitudinal study on 400 ARV naive HIV positive patients started on ARVS.. The study was aimed at modeling of cd4 counts change and survival times of these patients. 66.7% of the study participants were female, 62.8% had education above primary level with 43% earning below KSH 2,000. The mean baseline CD4 T cell count and weight is 134.1 and 62.5 kgs respectively. Linear mixed model was used in the modeling of CD4 T cells change before and after start of ARVS. CD4 T cell at the time of enrollment was found to be the only covariate influencing change in CD4 T cell before start of ARVS, whereas time since start of ARVS, patient age and gender were the covariates influencing the CD4 count change after start of taking ARVS. Classification (regression) trees were used to model TB incidence after start of ARVS. Patients' baseline weight was found to influence the incidence of TB. Hence the patients were classified to either having low weights (below 60 kgs) or normal weights (60kgs and above). Incidence was found to be greater for those with low baseline weight