Parametric versus non-parametric methods for estimating cure rates based on censored survival data.

Area of Science:

• Biostatistics
• Survival Analysis
• Medical Statistics

Background:

• The Kaplan-Meier curve is a standard method for estimating survival rates.
• Accurate estimation of cure rates is crucial for assessing long-term patient outcomes.
• Data recording errors can significantly impact survival analysis results.

Purpose of the Study:

• To investigate the impact of incorrectly recorded early failure times on Kaplan-Meier cure rate estimates.
• To introduce a novel parametric approach for cure rate estimation.
• To discuss the implications of data recording errors in survival analysis.

Main Methods:

• Analysis of the effect of early failure time misrecording on Kaplan-Meier curve plateau.
• Development and application of a parametric model based on the Gompertz distribution for cure rate estimation.

Main Results:

• Correction of early failure times leads to a counter-intuitive decrease in the estimated cure rate via the Kaplan-Meier method.
• The proposed Gompertz distribution-based parametric approach offers an alternative for cure rate estimation.

Conclusions:

• Data accuracy is paramount in survival analysis; early failure time errors can distort cure rate estimations.
• The Gompertz distribution provides a viable parametric framework for robust cure rate estimation, particularly when data integrity is a concern.