Carotid endarterectomy in a community hospital: a change in physicians' practice patterns.
A total of 243 consecutive carotid endarterectomies (CEA) performed at Providence Medical Center in Portland, Oregon, were retrospectively reviewed over a 22-month period. Of these, 137 patients (56%) underwent CEA for asymptomatic disease, 52 (37%) of whom had stenotic lesions of 79% or less. There were 6 deaths (3%) and 12 strokes (5%). Four strokes were in asymptomatic patients. These data prompted development of criteria for CEA: (1) hemisphere-specific transient ischemic attacks, reversible ischemic neurologic deficits, or amaurosis fugax with an appropriate carotid lesion; (2) completed stroke with major recovery and significant carotid stenosis; (3) asymptomatic lesion with greater than 80% stenosis (D+) either by carotid arteriogram or non-invasive lab evaluation; and (4) other indications only with a supporting second opinion from a disinterested vascular surgeon, neurosurgeon or neurologist. A prospective review followed institution of the guidelines. In 21 months, 148 operations were performed, a 36% reduction over the initial study period. Of these, 46 (31%) were for asymptomatic lesions. Two patients (4%) did not fulfill the guideline criteria. There were six strokes (4%) and no deaths. The reduction of CEAs appears to be related to a significant decrease in "inappropriate" operations being performed. Surgeons' familiarity with the data rather than external pressures seems to be the major factor in changing practice patterns. The decrease in stroke/death rate is not statistically significant.