Penerapan Regresi Weibull pada Data Pasien Data Pasien Penderita Kanker Serviks RSUD Kota Makassar Tahun 2017-2019

Abstract

In survival analysis, the commonly used method is cox proportional hazard regression, but if the data to be studied meet the assumptions for Weibull regression, Weibull regression analysis will provide better results. Weibull regression is a regression model developed from the Weibull distribution of 2 parameters, namely scale parameters and form parameters that can be expressed in regression parameters. Weibull regression models include the Weibull survival regression model, the Weibull hazard regression model and the mean model. The purpose of this study was to determine the shape of the model between the condition of cervical cancer patients and survival time using the Weibull regression model and to find out what factors affect survival time until cervical cancer patients are declared cured. Parameter estimation is done using Maximum Likelihood Estimation (MLE) but the assessment results are not closed form so they are overcome by Newton-Raphson iteration. The Weibull regression model was applied to cervical cancer patient data at RSUD Kota Makassar in 2017-2019. Based on the research conducted, it can be concluded that the factors that affect the cure of cervical cancer patients, namely the stage with the interpretation of cervical cancer patients in stage 2, have a risk of experiencing a failure rate of 4.4309 times that of cervical cancer patients in stage 1. While cervical cancer patients in stage 3 have a risk of failure rate of 8.4554 times that of cervical cancer patients in stage 1.