Shumba, Khumbo2021-10-192021-10-192020https://hdl.handle.net/10539/31729A research report submitted in partial fulfilment of the requirements for the degree of MSc in Epidemiology in the field of Epidemiology and Biostatistics to the Faculty of Health Sciences, School of Public Health, University of the Witwatersrand, Johannesburg, 2020BACKGROUND: While reports suggest that some HIV-infected patients on a protease inhibitor (PI)-based second-line antiretroviral therapy (ART) achieve viral re-suppression after experiencing virological failure, their risk of experiencing virological rebound (subsequent virological failure) is unclear. This study estimated the incidence rates and investigated predictors of virological rebound among these patients in Johannesburg, South Africa. METHODS: A retrospective cohort study design was used to analyse routinely collected adult (≥18years) patient data from Themba Lethu Clinic for patients switched to PI-based second-line ART between 1st January 2012 and 31st December 2016. The analysis was restricted to patients who experienced virological failure and achieved viral re-suppression within six months while on PI-based second-line ART. The follow-up period was 24 months, and censoring occurred at the time of death, transfer to another HIV treatment site, loss to follow-up, or at the dataset closure date (24 months from the viral re-suppression date). Baseline socio-demographic and clinical characteristics were summarised with stratification by sex. Kaplan Meir survival analyses were used to estimate incidence rates and the probability of surviving virological rebound stratified by selected covariates. Predictors of virological rebound were identified using Cox proportional hazard models. RESULTS: Out of 2371 patients initiated on PI-based second-line ART, a total of 246 patients (57.4% males and 42.6% females) experienced initial virological failure and achieved viral re-suppression in a median (IQR) of 3.7 (2.8-4.6) months. The probability of surviving virological rebound was significantly higher for patients who received at least five enhanced adherence counselling (EAC) sessions compared to those who received less than five (log-rank test: p-value <0.0001). The overall incidence rate of virological rebound was 23.20 (95% CI: 18.38-29.27) per 100 person-years. However, the incidence rate was higher for males (Males: 24.39; 95% CI: 17.15-34.68 per 100 person-years vs Females: 22.35; 95% CI: 16.39-30.47). The risk of virological rebound was higher for underweight patients (BMI: <20kg/m2) compared to those with a normal BMI (BMI: 25-29.9 kg/m2), (aHR: 3.08; 95% CI:1.16-8.13). Patients with a low CD4 count (<350 cells/μl) had a higher risk compared to those with a higher CD4 count (>500 cells/μl) (aHR: 3.68; 95% CI:1.10-12.33). Receiving at least five adherence counselling sessions seemed to reduce the risk by 89% compared to receiving less than five (aHR: 0.11; 95% CI:0.06-0.20). CONCLUSION: Virological rebound is highly incident among HIV-infected patients on PI-based second-line ART who re-suppress after experiencing initial virological failure. Enhancing patient-centred interventions to improve ART adherence, CD4 count, and BMI in these patients, are essential for the reduction of switches to more expensive and limited third-line ART regimens.enThesis