Modeling Time to Death of HIV Infected Patients on Antiretroviral Therapy the case of Hossana Queen Elleni Mohammad Memorial Hospital, South Ethiopia

Abstract

Human Immunodeficiency Virus (HIV) is the virus that causes Acquired Immune Deficiency Syndrome (AIDS). HIV attacks and destroys certain types of white blood cells that are essential to body's immune system, the biological ability of the human body to fight infections. . The main aim of this study is modeling the factors that affect survival time of HIV infected patients by using Cox ph and parametric survival regression models. This study is a retrospective cohort study based on data from the ART clinical in Hossana Queen Elleni Mohamad Memorial Hospital , south Ethiopia. All HIV positive patients who are 15 years old and above placed under ART in between February 2011 to January 2016 were population in this study. The analytical methodologies were used the Kaplan-Meier and Log Rank Test to estimate Descriptive analysis , Cox’s regression model was employed to identify the covariates that have a statistical significant effect on the survival time of HIV infected patients and exponential, weibull, log logistic and log-normal survival regression models were applied to compare efficiency of the models. The overall mean estimated survival time of patients was 51.5 months. The Cox Proportional Hazards regression Model result revealed that baseline weight, ART adherence, baseline CD4 count, WHO clinical stage, level of education, substance use and TB co-infection of patients are the major factors that affect significantly survival time of HIV infected patients. For future researchers on this area should apply Weibull survival regression model because Weibull distribution is unique that means only one that simultaneously both proportional and accelerated and also it predicts well the covariate that are associated with high risk of mortality.