Penerapan Analisis Regresi Logistik dan Analisis Survival Pada Data Masa Terapi Antiretroviral Penderita HIV

Abstract

AIDS is still a deadly disease. The new treatment has been found to suppress the replication of HIV that is antiretroviral therapy (ART). ART can increase the value of CD4 cell count then improve the immunity of patient. Based on a number of research the rate’s progression of the virus infected in each person is varies. There are many factors influence. This study aims to apply a statistical analysis on the data of survival time the HIV patient who use an antiretroviral therapy, compare the analysis result and determine the factors that significantly affect the survival time of ART. There are logistic regression analysis and survival analysis. Data of Survival time of the patient with ART are measured from the beginning of therapy on each patient until the event occured, when the patients enter the stage of AIDS (CD4 values less than 200 cells/mm3). The data used are secondary data obtained from the Polyclinic of HIV in the Dharmais’s hospital, Jakarta. The Data population are all patients who get the therapy in this Polyclinic on the period of 2005 to 2011. Those 902 patient with ART who have been qualified by WHO criteria (CD4 count ≤ 350 cells/mm3 on beginning). The results of logistic regression analysis are the factors that significantly affect the survival time of patients under level of alpha which is 10% are HIV transmission, sex and work status. While the results survival analysis using cox proportional hazard regression models provide information that all of the independent variables included in the model are significant. Plot log-minus-log survival data shows each stratum of each independent variable (covariates) is parallel with the time, it means the covariates are proportional, so hazard assumption of proportional are fulfilled.