A nonparametric statistical method that improves physician cost of care analysis

Area of Science:

• Health Services Research
• Biostatistics
• Health Economics

Background:

• Physician cost of care data analysis is crucial for healthcare management.
• Existing methods, like the observed-to-expected ratio, have limitations in stability and robustness.
• Accurate physician cost assessment impacts public reporting, pay-for-performance, and benefit design.

Purpose of the Study:

• To develop and evaluate a novel compositing method for physician cost of care data.
• To demonstrate improved performance compared to commonly used statistical tests.
• To enhance the reliability of physician cost assessment.

Main Methods:

• Utilized commercial preferred provider organization (PPO) claims data for internists.
• Developed a nonparametric composite performance metric incorporating risk adjustment via the Wilcoxon rank-sum (WRS) test.
• Compared the WRS algorithm against the parametric observed-to-expected ratio, assessing stability across outlier trimming methods and time periods.

Main Results:

• The WRS algorithm demonstrated significantly greater within-physician stability across various outlier trimming methods.
• The algorithm also exhibited significantly improved within-physician stability when analyzing physicians across different time periods.
• This indicates enhanced reliability in cost ratings.

Conclusions:

• The nonparametric WRS algorithm provides a more robust and stable methodology for evaluating physician cost of care.
• It surpasses the performance of traditional observed-to-expected ratio techniques.
• This improved methodology can enhance physician cost assessment for critical applications like public reporting and pay-for-performance initiatives.