Modeling traumatic brain injury lifetime data: Improved estimators for the Generalized Gamma distribution under small...

Area of Science:

• Medical Statistics
• Trauma Research
• Survival Analysis

Background:

• Traumatic brain injury (TBI) from traffic accidents presents challenges in survival rate modeling due to limited patient data.
• Accurate statistical modeling is crucial for understanding TBI outcomes and improving patient care.

Purpose of the Study:

• To develop and compare statistical estimation methods for modeling TBI patient survival rates, especially with small sample sizes.
• To identify the most efficient estimation method for TBI data analysis.

Main Methods:

• Utilized the Generalized Gamma distribution for flexible modeling of TBI data.
• Compared various statistical estimation methods through extensive numerical simulations.
• Employed the Simulated Annealing technique to address numerical optimization challenges.

Main Results:

• Penalized maximum likelihood estimators demonstrated the smallest mean square errors and biases.
• These estimators proved most efficient for small sample sizes in TBI survival analysis.
• The Generalized Gamma distribution showed utility across different risk levels in real-world TBI datasets.

Conclusions:

• Penalized maximum likelihood estimation is recommended for TBI survival modeling with limited data.
• The Generalized Gamma distribution offers a robust and simple alternative for analyzing traffic accident TBI.
• The findings support improved statistical approaches for trauma research.