The Weibull-Exponentiated Exponential Cure Fraction Model for Right Censored Survival Data with Applications to Cancer Data

Abstract

The cure fraction model also known as the long-term survival model is used in fitting data from a population with two different types of individuals: individuals who experienced the event of interest (susceptible) and individuals who will never experience the event of interest (non-susceptible). The present paper introduced a cure fraction model considering the Weibull exponentiated exponential distribution that will be used in modeling such type of information. The parameters of the model were estimated via the maximum likelihood procedure (MLE) under the assumption of right censoring. Furthermore, the statistical properties of the model were studied comprehensively. Simulation studies and medical data sets were used to demonstrate the applicability of the proposed methodology. Bias and standard error were used as discrimination criteria in the simulation study while Akaike Information criteria (AIC), Bayesian Information Criteria (BIC), and Consistent Akaike Information criteria (CAIC) were used as discrimination criteria in real-life applications. Results from the applications showed that the Weibull exponentiated exponential non-mixture cure fraction model is a strong competitor.