Journal ID : TRKU-23-09-2020-11140
[This article belongs to Volume - 62, Issue - 09]
Total View : 409

Title : Mixture Cure Rate Modelling Approach to Cure Rate Estimation: A Mini Review

Abstract :

Cure fraction models are used when some units in the population of survival data are thought to survive the event of interest. Besides, modelling survival data with cure fraction provides a better fit. In this study, Mixture Cure Rate Modelling approach was briefly reviewed with some highlights on Non- Mixture and Defective Modelling approaches. Mixture Cure Rate Modelling assumes the population under study to be a mixture of susceptible and unsusceptible to the event of interest. Estimating cure fraction in the presence of either unobserved heterogeneity or zero-adjusted units were discussed. Mandatory assumption of the presence of cure fraction is a risk associated with this approach, which if wrongly done leads to misleading conclusion. The non-mixture modelling of cure fraction has a natural biological interpretation due to the process of its development. It is mostly associated with Bayesian statistical context. Estimating cure fraction using defective modelling approach makes no assumption of the presence of cure fraction; the cure fraction is estimated when the estimated value of the distribution shape parameter is negative. One of the shortcomings associated with this type of methodology is that, only few distributions can become defective. Zero-adjusted Frailty Mixture Cure Model is proposed

Full article