Modeling the survival in patients with HIV by the presence of competing risks for death: sub-distribution and cause-specific hazard approach

In the presence of competing risks, patients with human immunodeficiency viruses (HIV) experience death by various causes, including co-infection with acquired immune deficiency syndrome (AIDS) with tuberculosis (TB), AIDS, and other causes in the follow-up. This study aimed to model the survival in patients with HIV in the presence of these competing causes of death utilizing sub-distribution hazard (SDH) and cause-specific hazard (CSH) models to overcome biased estimates of the classical analyses. In this longitudinal study, patients with HIV+ diagnosis (n = 2328) were recruited from Imam Khomeini and Zamzam consulting centers from 2003 to 2012. In the presence of the competing causes of death, the SDH and CSH models evaluated the effect of underlying predictors on the cumulative incidence and instantaneous hazards, respectively, with the cmprisk package in R4.1 software. The median survival time of patients with AIDS+TB, AIDS and other causes were 7.79 (SE .84), 11.57 (SE .98), and 14.1 (SE .91), respectively. In the SDH model, CD4(350+) [AIDS: sub-distribution hazard ratio (SHR) = .13, 95% confidence interval(CI) = (.08–.19)); AIDS+TB .10(.04–.25)], antiretroviral therapy (ART) [AIDS .44(.32–.61); AIDS+TB .57(.31–.99); other .07(.02–.23)], isoniazid prophylaxis therapy (IPT) [AIDS .47(.28–.78); AIDS+TB .08(.01–.58)], and cotrimoxazole prophylaxis therapy (CPT) [AIDS .38(.22–.68)], were inversely related to hazard of death, while being a male [AIDS 2.62(1.574.39); AIDS+TB 10.43(2.32–46.83); Other 9.48(1.95–45.99)] was directly related to hazard of death. The CSH model resulted in similar estimates except for CD4(350+) which was inversely related to hazard of death by other causes. Taking into account the strong association of CD4(350+), ART, IPT, CPT, and being a male with the hazard of mortality caused by the competing causes of death in patients with HIV, in both SDH and CSH models, designing sex-specific policymaking and interventional programs are recommended to prolong the survival of patients with HIV. The future treatment program can utilize the results.