Survival and hazard functions for progressive diseases using saddlepoint approximations

Area of Science:

• Biostatistics
• Mathematical Statistics

Background:

• Parametric survival analysis is crucial for modeling time-to-event data.
• Accurate computation of survival and hazard functions is essential for reliable analysis.
• Existing methods may be computationally intensive or complex for intricate models.

Purpose of the Study:

• To introduce and evaluate saddlepoint approximations for survival and hazard functions in parametric survival analysis.
• To demonstrate the computational efficiency and accuracy of these approximations.
• To highlight their utility in complex waiting time scenarios.

Main Methods:

• Utilizing the Lugannani and Rice saddlepoint approximation for survival function estimation.
• Employing numerical integration of saddlepoint density approximations.
• Approximating the hazard function using saddlepoint density and distribution functions.

Main Results:

• Saddlepoint approximations provide computationally fast and accurate results.
• The methods are relatively straightforward to implement.
• Demonstrated applicability to complex models involving convolutions of distributions.

Conclusions:

• Saddlepoint approximations are a valuable, yet underutilized, tool in parametric survival analysis.
• These methods enhance the feasibility of analyzing complex waiting time distributions.
• Further adoption can improve the efficiency and accuracy of survival data analysis.