Monitoring surgical performance using risk-adjusted cumulative sum charts

Area of Science:

• Medical Statistics
• Health Services Research
• Surgical Quality Improvement

Background:

• The cumulative sum (CUSUM) procedure is a standard graphical tool for quality monitoring.
• CUSUM charts are increasingly used to monitor surgical outcomes and detect changes in failure rates.
• A key limitation of standard CUSUM is its susceptibility to confounding by changes in patient case-mix, such as an increase in high-risk patients.

Purpose of the Study:

• To develop and describe a novel CUSUM procedure that accounts for pre-operative patient risk.
• To provide a more accurate method for detecting true changes in surgical performance, independent of referral patterns.
• To enhance the reliability of CUSUM charts in clinical settings with variable patient populations.

Main Methods:

• A new CUSUM procedure was developed using a likelihood-based scoring method.
• The method incorporates individual patient pre-operative risk assessment into the CUSUM calculation.
• The adjusted CUSUM procedure is designed to differentiate between changes in surgical performance and changes in patient risk.

Main Results:

• The proposed CUSUM procedure effectively adjusts for patient pre-operative risk of surgical failure.
• This adjustment allows for more precise identification of shifts in actual surgical performance.
• The method is robust to variations in patient case-mix over time.

Conclusions:

• The risk-adjusted CUSUM procedure offers a significant improvement for monitoring surgical quality.
• It provides a more reliable signal of changes in surgical performance in diverse patient populations.
• This methodology is particularly valuable for healthcare settings experiencing fluctuating patient risk profiles.