Exponentiated Weibull regression for time-to-event data.

Area of Science:

• Statistics
• Survival Analysis
• Biostatistics

Background:

• Traditional models like Weibull, log-logistic, and log-normal distributions have limitations in capturing diverse hazard rate shapes (monotone, unimodal).
• The exponentiated Weibull distribution offers greater flexibility, accommodating monotone, unimodal, and bathtub-shaped hazard rates.

Purpose of the Study:

• To extend the utility of the exponentiated Weibull distribution to regression modeling in survival analysis.
• To demonstrate that the exponentiated Weibull distribution is closed under the accelerated failure time (AFT) family.

Main Methods:

• Formulation of an accelerated failure time regression model using the exponentiated Weibull distribution.
• Development of large sample theory for statistical inference within the proposed regression framework.
• Description of a Bayesian approach for parameter estimation and inference.

Main Results:

• The exponentiated Weibull distribution is shown to be closed under the AFT family.
• A novel regression model based on the exponentiated Weibull distribution is developed.
• Comparative studies using real and simulated data indicate the model's effectiveness.

Conclusions:

• The proposed exponentiated Weibull regression model provides a valuable tool for analyzing time-to-event data with complex hazard functions.
• This flexible modeling approach can improve the description and understanding of various time-to-event phenomena.