ESTIMASI PARAMETER MODEL REGRESI COX DENGAN METODEEFRON PARTIAL LIKELIHOOD PADA DATA PASIENENDOMETRIOSIS

Survival analysis is a statistical method used to analyze data relating to survival time, that is, from a starting point determined until an event occurs. In survival analysis, sometimes there are two or more individuals who have the same survival time (Ties). The purpose of this study was to form a proprtional hazard cox regression model with the method of estimating the partial likelihood efron parameters in survival data containing ties with the application of endometriosis patient data to find out the risk factors (Type of Operation, Fruit Consumption, Vegetable Consumption, Soy Consumption, Consumption of Fat, Time of Stay, Sports Time) , Gluten Consumption, Cheese Consumption, Yogurt Consumption, Milk Consumption, Kefir Consumption, Type of Endometriosis, Diet, and Rest) which influence the time of relapse of endometriosis patients after the first surgery. Endometriosis patient data obtained by conducting a survey that is distributing online questionnaires from August to December 2019 obtained as many as 40 patients who perform surgery. Based on the results of modeling that has been done with the help of R software, the factors that have a significant influence on relapse time are time of stay, sports time, cheese consumption, milk consumption, and the type of endometriosis with the estimated parameter results for each successive factor is 0.38, -0.42, -1.12, 0.47, 1.02. Sports time and cheese consumption can prolong recurrence time (negative parameter estimation) with the risk of recurrence respectively 44% and 67%. When staying up late, milk consumption, and types of endometriosis can accelerate the time of relapse (positive parameter estimation) with the risk of recurrence respectively 47%, 60%, and 177%.