Model Regresi Cox Weibull Dengan Metode Penaksiran Parameter Efron Partial Likelihood

Studi Kasus : Lama Perawatan Pasien Penderita Tuberkulosis Di Puskesmas Loa Ipuh Tenggarong Tahun 2017

Abstract

Survival analysis is an analysis that involves statistical tests to analyze data on the time or length of time until the occurrence of a particular event. One regression model for time duration data modeling is the Cox PH regression model. Cox PH regression applied to time duration data with Weibull distribution is called Cox PH Weibull regression. The purpose of this study is to obtain a Cox PH Weibull regression model and determine the factors that influence the length of treatment for tuberculosis patients. The parameter estimation in this study is Efron partial likelihood method. The Efron partial likelihood method is suitable for estimating Cox PH regression parameters to data containing ties. Based on the results of parameter estimation the best model is obtained by using  AIC crtiteria. Based on the partial test,  age is factor that influence to the length of treatment. The results of the study show that every increasing age patient one year, then length of treatment until the patient recovered, will increas 1,024 times.