Risk-Adjusted Surgical Learning Curve Assessment Using Comparative Probability Metrics.

Area of Science:

• Medical Statistics
• Surgical Education
• Health Informatics

Background:

• Surgical learning curves assess trainee proficiency.
• Cumulative sum (CUSUM) methods are common but have limitations.
• CUSUM methods rely on fixed thresholds and lack interpretability.

Purpose of the Study:

• Introduce a risk-adjusted surgical learning curve assessment (SLCA) method.
• Develop a novel approach for evaluating surgical trainee progress.
• Address limitations of existing CUSUM-based learning curve techniques.

Main Methods:

• Proposed a risk-adjusted SLCA method using estimation, not hypothesis testing.
• Utilized Weibull distribution for right-skewed outcomes like surgery durations.
• Employed weighted estimating equations, prioritizing recent performance data.

Main Results:

• The SLCA method provides enhanced interpretability and deeper insights.
• It avoids reliance on difficult-to-determine external performance levels.
• The approach emphasizes clinical equivalence and noninferiority.

Conclusions:

• The proposed SLCA method offers a more insightful and practical approach to surgical learning curve assessment.
• This method is particularly suitable for skewed outcome data.
• SLCA enhances the evaluation of surgical trainee proficiency and performance improvement.