Objective: To determine whether the complication or death rate from carotid endarterectomy can be predicted from hospital and physician structural variables, such as the hospital's teaching status or the number of endarterectomies done by the surgeon per year.
Design: Survey of medical records. After controlling for the severity of the patient's condition on the basis of data in the medical record at the time of the endarterectomy, regression analyses were used to predict the postoperative stroke, heart attack, and 30-day death rate as a function of patient, physician, and hospital characteristics.
Setting: Three geographic areas (states or large parts of states; average population, 3 million) in the United States.
Patients: Random sample of 1302 patients 65 years of age or older having carotid endarterectomy in 1981.
Intervention: Carotid endarterectomy.
Measurements and Main Results: Of 1302 patients, 11.3% had a postoperative stroke or heart attack or died within 30 days of the operation. Patient age, race, income, and gender; physician volume, board certification status, and age; and hospital size, for-profit status, ownership, and teaching status were not significantly related to the postoperative complication or death rate. If the surgeon was a graduate of a foreign, but not a Western European or Canadian, medical school, however, the average complication or death rate rose from 10.4% to 19.6% (P < 0.05).
Conclusions: The effectiveness of carotid endarterectomy depends heavily on its complication rate. Because complications after surgery cannot, in general, be predicted from structural variables, referring physicians cannot rely solely on the surgeon's experience and qualifications when recommending a carotid endarterectomy. The surgeon's and the hospital's actual postoperative complication and death rate should be considered.