Balancing Immediate Surgical Risks with Deferred Benefits in Preventive Surgery.

Area of Science:

• Medical Statistics
• Surgical Outcomes Research

Background:

• Preventive surgeries like carotid endarterectomy (CEA) carry immediate risks, but traditional analyses (Kaplan-Meier, Cox models) inadequately address competing risks and event timing.
• Assessing the risk-benefit trade-off in preventive surgery is crucial for patient outcomes.

Purpose of the Study:

• To evaluate alternative methodologies, specifically competing risk analysis and restricted mean survival time (RMST).
• To re-analyze data from a landmark CEA trial (ACST-1) using these advanced statistical approaches.

Main Methods:

• Reconstructed individual patient data from the ACST-1 trial (n=3,120) comparing immediate vs. deferred CEA for asymptomatic carotid stenosis.
• Utilized the Aalen-Johansen estimator for competing risks and RMST to quantify stroke-free days and life years gained/lost.
• Mortality data were modeled from a comparable study, ensuring reconstructed data accuracy within 1% of original parameters.

Main Results:

• Competing risk analysis reduced 10-year stroke risk estimates by 20% and lowered the estimated benefit of immediate CEA from 6.1% to 5.6%.
• No significant difference in stroke-free days between arms at 10 years (40 days, 95% CI -52 to 129).
• Immediate CEA showed a small net disadvantage, ranging from 1 to 16 life days lost over 10 years, considering procedural deaths versus prevented stroke deaths.

Conclusions:

• Competing risk analysis and RMST provide a more comprehensive evaluation of risk-benefit trade-offs in preventive surgery compared to traditional methods.
• These advanced analytical frameworks can alter study conclusions and facilitate better intertrial comparisons and clinical decision-making.