1. Home>
2. Specialties>
3. Neurology>
4. Summary and Comment

Risk Models Help Predict Complications of Carotid Endarterectomy

Both general and surgery-specific risk indexes can be useful — but how should they be used in clinical practice?

Four risk indexes are widely used for predicting cardiac risk in patients undergoing noncardiac surgery: the ASA index, Goldman index, Detsky index, and Revised Cardiac Risk (RCR) index. To assess the value of risk indexes specifically for carotid endarterectomy (CEA), these authors analyzed how well various indexes predicted outcomes in 1998 patients who underwent single, unilateral CEA. In addition to the four general cardiac-risk indexes, the researchers used two CEA-specific indexes that predict the risk for the combined endpoint of death, nonfatal stroke, and cardiac, noncardiac, minor neurologic, and wound complications: the Halm score (previously formulated based on the data used in this study) and the Tu score (a similar formula derived from another patient group). The authors employed a technique from signal-detection theory that uses the area under the receiver operating characteristics (AROC) curve to test the predictive ability of each model.

All risk models except the Tu score were equivalently predictive of cardiac and noncardiac medical complications (AROC range, 0.58–0.68). The generic RCR index and the CEA-specific Halm score also were predictive of the combined endpoint of all-cause death and stroke (AROCs with RCR and Halm, 0.61 and 0.72, respectively) as well as of minor neurologic and wound complications of CEA (AROC range, 0.56–0.61).

Comment: These findings show that several indexes designed to assess cardiac risk also are predictive of noncardiac risks. Moreover, the generic RCR index is predictive of CEA-specific complications such as stroke and minor neurologic deficits. Although the predictive abilities of the scores in this study were comparable to their predictive abilities in other validation studies for death and stroke risk and for medical complications, their prognostic abilities were modest and only slightly better than chance (AROC=0.50). Therefore, one should use caution in applying any such index. Although they are useful guides, the indexes should not replace good clinical judgment.

— Arthur Day, MD, and Rose Du, MD, PhD

Dr. Day is Director, Division of Cerebrovascular Surgery, and Director, Cerebrovascular Center, Brigham and Women’s Hospital, Boston. Dr. Du is a Fellow in Cerebrovascular and Skull Base Surgery, Department of Neurosurgery, Brigham and Women’s Hospital, Boston.

Published in Journal Watch Neurology September 12, 2006

Citation(s):

Press MJ et al. Predicting medical and surgical complications of carotid endarterectomy: Comparing the risk indexes. Arch Intern Med 2006 Apr 24; 166:914-20.

• Original article (Subscription may be required)
• Medline abstract (Free)