An observational retrospective study was conducted over a 5-year period to assess survival and predictors of death in people with HIV-positive serology undergoing antiretroviral treatment with first-line regimens at the Military Hospital of Nampula, Mozambique. We collected data from 332 patient records. Kaplan-Meier boundary product estimator, log-rank, Gehan-Breslow, Tarone-Ware, time-dependent Cox models and estimates of hazard ratios (HR), with 95% confidence interval (CI) were calculated. Meantime survival for females and males was 54.8 months [95% CI 50.32-55.40] and 49.7 months [95% CI 45.89-53.53], respectively. Cox regressions indicated higher death rates significantly or potentially associated with: male sex (HR = 1.3; [95% CI 0.7-2.39]); suspected diagnosis reported only by the physician (HR = 3.6; [95% CI 1.8-7.4]); disease stages III (HR=1.2 [95% CI 0.3-3.6]) or IV (HR 1.4 [95% CI 0.4-5.8]); first TCD4+ lymphocyte count lower than 350 cells per ml (HR = 3.2; [95% CI 0.9-11.2]) or between 350-500 cells per ml (HR = 1.3; [95% CI 0.3-5.8]); or do not present cells count (HR = 3.6; [95% CI 1.2-10.2]). The above variables were significant for HIV prognosis and as predictors of death and should be considered in the clinical care of these patients.
Keywords: Cox regression; HIV/AIDS; Kaplan–Meier; Survival; epidemiology.