# Determinants of CD4 count recovery among severely immunosuppressed HIV patients initiated on antiretroviral therapy: a prospective cohort study in KwaZulu-Natal, South Africa

## TL;DR

This study examines how long it takes for HIV patients with low CD4 counts to recover after starting treatment and identifies factors that influence recovery time.

## Contribution

The study identifies specific demographic and clinical factors affecting CD4 recovery in HIV patients in South Africa using survival analysis techniques.

## Key findings

- Males had a 41% lower chance of CD4 recovery compared to females.
- Rural care was associated with a 49% higher chance of CD4 recovery than urban settings.
- TB co-infection reduced the chance of CD4 recovery by 39.5%.

## Abstract

For people living with HIV (PLWH), CD4 count serves as an effective indicator of response to antiretroviral therapy (ART) and a predictor of morbidity and mortality before and after ART initiation. In PLWH on ART, changes in CD4 count reflect immunological response to treatment. CD4 count recovery after ART initiation is defined as achieving a CD4 cell count of ≥500 cells/mm3. This study utilised survival analysis techniques to determine time to CD4 count recovery and to identify determinants among PLWH with severe immunosuppression who were initiated on ART in KwaZulu-Natal, South Africa.

Secondary data were collected by CAPRISA from June 2004 to August 2013, with monthly hospital visits. Time to CD4 recovery was calculated from ART initiation to the first CD4 count ≥500 cells/mm3. Participants were followed until recovery; those who did not recover, were lost to follow-up, or died were censored. Kaplan–Meier curves estimated median survival time, while Cox and Weibull regression models identified risk factors associated with CD4 recovery.

Among 2,528 participants (median age 32 years), 1,803 had viral load <400 copies/mL, 1,589 were females, and 524 had TB. By the end of the study, 727 achieved CD4 recovery. Kaplan–Meier estimates showed median time to recovery of nearly 40 months for females and 59 months for males. The Weibull model outperformed the Cox model. Male participants had a 41% lower hazard of recovery than females [HR: 0.589, 95% CI: 0.497–0.698], while rural care was associated with a 49% higher hazard of recovery than urban settings [HR: 1.494, 95% CI: 1.228–1.818]. TB-coinfected participants had a 39.5% lower hazard [HR: 0.605, 95% CI: 0.489–0.748], whereas those with viral suppression had a 49% higher hazard of recovery [HR: 1.489, 95% CI: 1.213–1.826].

Delayed CD4 recovery among males, urban residents, TB-coinfected participants, and those with unsuppressed viral load (>400 copies/mL) underscores the need for targeted, differentiated HIV care strategies to accelerate immunological recovery and reduce HIV-related morbidity. Policymakers should prioritise male-focused interventions, strengthened TB–HIV integrated services, intensified adherence and viral load monitoring, and context-specific interventions in urban settings to accelerate immunological recovery and reduce HIV-related morbidity.