A Bayesian long-term survival model parametrized in the cured fraction

Area of Science:

• Biostatistics
• Survival Analysis
• Statistical Modeling

Background:

• Cure rate models are essential for analyzing data where a portion of the population is assumed to be immune to the event of interest.
• Existing models often have limitations in handling competing risks.
• There is a need for flexible models that can incorporate multiple event types and estimate cure fractions accurately.

Purpose of the Study:

• To propose and analyze a novel cure rate model incorporating competing risks.
• To utilize the negative binomial distribution to model the number of competing causes.
• To reparametrize the model using the cured fraction and link it to covariates via a logistic function.

Main Methods:

• Development of a cure rate model with competing risks using the negative binomial distribution.
• Reparametrization of the model through the cured fraction and logistic link function.
• Application of Bayesian inference using Markov chain Monte Carlo (MCMC) methods.

Main Results:

• The proposed model effectively integrates competing risks within a cure rate framework.
• The Bayesian approach provides a robust method for parameter estimation.
• The numerical example demonstrates the practical applicability of the model.

Conclusions:

• The developed cure rate model offers a flexible and comprehensive approach for survival data with competing risks.
• The Bayesian analysis facilitates reliable estimation of model parameters, including the cured fraction.
• This methodology can be valuable in various fields, including medicine and epidemiology, for analyzing long-term outcomes.