Analysis of Health Care Billing via Quantile Variable Selection Models

Area of Science:

• Health Economics
• Medical Data Analytics
• Statistical Modeling

Background:

• Billions lost annually to healthcare fraud, particularly Medicare overpayments.
• Existing analytical methods struggle with complex billing data distributions.
• Understanding provider billing behavior is crucial for policy makers.

Purpose of the Study:

• To apply a quantile regression framework for medical billing analysis.
• To identify factors influencing provider billing aggressiveness at various distribution levels.
• To enhance fraud detection strategies using provider billing characteristics.

Main Methods:

• Utilized quantile regression for analyzing medical billing data.
• Implemented a variable selection approach to identify key billing drivers.
• Focused analysis on mammography procedures to demonstrate the method.

Main Results:

• The proposed method effectively captures varying impacts of variables across different billing aggressiveness quantiles.
• Identified specific factors influencing conservative and aggressive billing behaviors.
• Demonstrated the utility of the approach in a real-world medical procedure context.

Conclusions:

• Quantile regression offers a robust framework for analyzing asymmetric medical billing data.
• The method provides nuanced insights into provider billing patterns, aiding fraud detection.
• Recommendations are offered for improving fraud detection by understanding billing aggressiveness.