Flexible Parametric Survival Cure Rate Models for Pulmonary Tuberculosis Data

Abstract

This article mainly aims to compare Flexible parametric cure rate models using relative survival function and to predict cure fraction for tuberculosis (TB) data. In survival analysis, the Cox proportional-hazards model of time-to-event data is effective, but still there may be some benefits of using parametric models than non-parametric or semi-parametric models. Sometimes, it happens that a certain fraction of the data corresponds to subjects who are never involved in the event when assessing time-to-event data. Survival models that take this characteristic into account are typically referred to as cure rate models. Hence, in this article the parametric cure model to time-to-event (sputum conversion) on pulmonary TB data with the survival time distribution such as Weibull, Gamma, Exponential and Lognormal is developed. The objective of this article is to compare cure rate models to find the best model fitting survival time using the relative survival function and to predict cure fraction of TB data. The data were analyzed using “R-4.0.2” and STATA 15.0.0 statistical tools.