Relation between surgeons' practice volumes and geographic variation in the rate of carotid endarterectomy.

We examined the relation between the number of operative procedures carried out by individual surgeons and the variation in the rate of carotid endarterectomy among Medicare beneficiaries in areas of high, average, and low use of the procedure in 1981. Rates ranged from 48 per 100,000 in the low-use area to 178 per 100,000 in the high-use area. Two variables accounted for most of the differences in the rates: the number of surgeons performing the procedure and the number of endarterectomies performed by surgeons with high practice volumes. Twice as many surgeons in the high-use area and 25 percent more in the average-use area performed carotid endarterectomy as compared with those in the low-use area. If the average number of cases per surgeon had been the same, the differences in the number of surgeons would have accounted for 36 percent and 15 percent, respectively, of the differences in use. Surgeons who performed 15 or more carotid endarterectomies during the year accounted for most of the variation in the rates. These high-volume surgeons represented 15 percent and 17 percent of the surgeons in the areas of high and average use, respectively, as compared with 4 percent of those in the low-use area. They accounted for 60 and 77 percent, respectively, of the additional endarterectomies. Three fourths of the surgeons performing carotid endarterectomies carried out fewer than 10, and 24 percent did only 1. We conclude that most of the geographic variation in the rate of carotid endarterectomy is caused by a few surgeons in high-use areas who perform large numbers of operations.