Kaplan—meier, marginal or conditional probability curves in summarizing competing risks failure time data?

Key Points

• To evaluate the suitability of alternative descriptive estimators and two-sample test statistics for summarizing failure time data in the presence of competing risks.
• Assessed the descriptive validity of the standard Kaplan-Meier estimator in competing risk settings.
• Examined marginal probability and conditional probability estimators as alternative summarization techniques.
• Presented two-sample test statistics designed for comparative analysis of competing risk failure time data.
• Demonstrated that the standard Kaplan-Meier estimator is frequently unsuitable for summarizing survival data when competing risks are present.
• Showed that marginal and conditional probability estimators provide viable descriptive alternatives for summarizing failure time curves and calculating two-sample comparisons.