Extended Cox Regression Model: the Choice of Timefunction
International Conference on Numerical Analysis and Applied Mathematics (ICNAAM), Rhodes, Greece, 19 - 25 September 2016, vol.1863, (Full Text)
- Publication Type: Conference Paper / Full Text
- Volume: 1863
- Doi Number: 10.1063/1.4992406
- City: Rhodes
- Country: Greece
- Hacettepe University Affiliated: Yes
Abstract
Cox regression model (CRM), which takes into account the effect of censored observations, is one the most applicative and usedmodels in survival analysis to evaluate the effects of covariates. Proportional hazard (PH), requires a constant hazard ratio over time, is the assumptionofCRM. Using extended CRM provides the test of including a time dependent covariate to assess the PH assumption or an alternative model in case of nonproportional hazards. In this study, the different types of real data sets are used to choose the time function and the differences between time functions are analyzed and discussed.