POSSUM and P-POSSUM overpredict mortality for carotid endarterectomy

Area of Science:

• Vascular Surgery
• Surgical Outcomes Research
• Health Services Research

Background:

• Carotid Endarterectomy (CEA) is a common procedure to prevent stroke.
• Accurate mortality prediction is crucial for surgical audit and patient risk stratification.
• Existing scoring systems like POSSUM and P-POSSUM are widely used but their applicability to specific procedures like CEA requires validation.

Purpose of the Study:

• To evaluate the predictive accuracy of the POSSUM and P-POSSUM scoring systems for 30-day mortality in patients undergoing CEA.
• To determine if these general surgical risk models are suitable for comparative audits of CEA procedures.

Main Methods:

• A retrospective and prospective study involving 499 CEAs performed between 1992 and 1999.
• Collection of patient physiological and operative parameters, and 30-day mortality data.
• Calculation of predicted mortality using POSSUM and P-POSSUM, followed by chi-squared analysis to compare observed versus predicted deaths.

Main Results:

• The observed 30-day mortality rate was 1.8% (9 out of 499 patients).
• POSSUM and P-POSSUM predicted significantly higher mortality rates (49 and 25, respectively) compared to the observed nine deaths.
• Statistical analysis revealed a significant lack of fit for both models (p<0.05), indicating poor predictive accuracy.

Conclusions:

• Both POSSUM and P-POSSUM demonstrate a tendency to overpredict mortality in patients undergoing CEA.
• These models are not suitable for comparative audit purposes in the context of CEA.
• Development of specific risk prediction models tailored for CEA procedures is recommended for improved accuracy.