A comparative study of mixture cure models with covariate


Oh, Yit Leng and Mohd Khalid, Zarina (2017) A comparative study of mixture cure models with covariate. AIP Conference Proceedings, 1842. 030011. ISSN 1551-7616

Full text not available from this repository.


In survival analysis, the survival time is assumed to follow a non-negative distribution, such as the exponential, Weibull, and log-normal distributions. In some cases, the survival time is influenced by some observed factors. The absence of these observed factors may cause an inaccurate estimation in the survival function. Therefore, a survival model which incorporates the influences of observed factors is more appropriate to be used in such cases. These observed factors are included in the survival model as covariates. Besides that, there are cases where a group of individuals who are cured, that is, not experiencing the event of interest. Ignoring the cure fraction may lead to overestimate in estimating the survival function. Thus, a mixture cure model is more suitable to be employed in modelling survival data with the presence of a cure fraction. In this study, three mixture cure survival models are used to analyse survival data with a covariate and a cure fraction. The first model includes covariate in the parameterization of the susceptible individuals survival function, the second model allows the cure fraction to depend on covariate, and the third model incorporates covariate in both cure fraction and survival function of susceptible individuals. This study aims to compare the performance of these models via a simulation approach. Therefore, in this study, survival data with varying sample sizes and cure fractions are simulated and the survival time is assumed to follow the Weibull distribution. The simulated data are then modelled using the three mixture cure survival models. The results show that the three mixture cure models are more appropriate to be used in modelling survival data with the presence of cure fraction and an observed factor.

Item Type: Article
Uncontrolled Keywords: Mathematical statistics, Survival analysis (Biometry)
Subjects: Q Science > QA Mathematics > QA273-280 Probabilities. Mathematical statistics
Divisions: Faculty of Business (FOB)
Depositing User: Ms Rosnani Abd Wahab
Date Deposited: 26 Oct 2020 15:49
Last Modified: 26 Oct 2020 15:49
URII: http://shdl.mmu.edu.my/id/eprint/7084


Downloads per month over past year

View ItemEdit (login required)