Current multivariate risk scores in patients undergoing non-cardiac surgery

Area of Science:

• Cardiology
• Anesthesiology
• Health Services Research

Background:

• Numerous indexes predict perioperative cardiovascular risk, with the Revised Cardiac Risk Index (RCRI) being widely used since 1999.
• The RCRI identifies risk based on ischemic heart disease, cardiovascular disease history, heart failure, diabetes, and renal failure.
• While the RCRI has a high negative predictive value, its accuracy is suboptimal in many clinical settings.

Purpose of the Study:

• To evaluate the performance of various perioperative cardiovascular risk prediction indexes.
• To compare the accuracy of newer indexes, such as the Orion score, against established ones like the RCRI.

Main Methods:

• Review of existing cardiovascular risk indexes including RCRI, Gupta index, VISION, POSSUM, POSPOM, and the Orion score.
• Analysis of validation studies and cohort data, including a 9000-patient cohort for the Orion score.
• Comparison of accuracy using Area Under the Receiver Operating Characteristic (AUROC) curves.

Main Results:

• The Orion score, derived from 9000 patients, demonstrated higher accuracy (AUROC=0.872) compared to the RCRI (AUROC=0.807) in both derivation and validation cohorts.
• Other indexes like the Gupta index and universal surgical risk tools have shown good performance for mortality and morbidity.
• Despite European guidelines recommending risk stratification, clinical adoption of these indexes remains low.

Conclusions:

• The Orion score offers improved accuracy for predicting major adverse cardiovascular events in the perioperative period compared to the RCRI.
• Despite evidence supporting their utility, the clinical implementation of perioperative cardiovascular risk indexes requires further attention.
• Enhanced risk stratification tools are crucial for optimizing patient care in non-cardiac surgery.