Competing Risks Analysis of Kidney Transplant Waitlist Outcomes: Two Important Statistical Perspectives

Area of Science:

• Clinical Epidemiology
• Biostatistics
• Transplant Surgery

Background:

• Competing risks analysis is crucial in clinical epidemiology for understanding patient outcomes.
• Two primary goals include identifying etiologic associations and assessing differences in cause-specific patient risk.
• Kidney transplant waiting lists present a complex scenario with multiple distinct outcomes.

Purpose of the Study:

• To elucidate the goals of competing risks analysis in clinical epidemiology.
• To highlight the application of these models in deceased donor kidney transplant (DDKT) candidate analysis.
• To emphasize the importance of understanding cause-specific hazards as foundational for cumulative incidence functions (CIFs).

Main Methods:

• Utilizing cause-specific hazard models to understand associations between predictors and outcomes.
• Employing cause-specific subdistribution hazard models for analyzing differences in cause-specific patient risk.
• Discussing the estimation of cause-specific conditional failure probability for transplant waiting times.

Main Results:

• Cause-specific hazards are essential building blocks for cumulative incidence functions (CIFs).
• Subdistribution hazard ratios cannot be fully understood without knowledge of cause-specific hazards.
• The "one minus Kaplan-Meier" approach is inappropriate and produces biased estimates for competing risks.

Conclusions:

• Accurate competing risks analysis requires understanding both cause-specific hazards and appropriate CIF estimators.
• The analysis of kidney transplant candidates necessitates careful consideration of competing events.
• Clinicians must avoid biased estimation methods to ensure reliable patient risk assessment.