Geographic variation in the use of breast-conserving treatment for breast cancer

In the past decade there has been an increase in the use of treatment designed to conserve the breast for women with breast cancer. The extent to which such treatment has been adopted in various regions of the country and whether characteristics of hospitals and patients predict its use are not known, however. We used national data on Medicare claims for inpatient care provided in 1986 to study 36,982 women 65 to 79 years of age, who had local or regional breast cancer and underwent either mastectomy or breast-conserving treatment (local excision, quadrantectomy, or subtotal mastectomy). Information about the hospitals at which these women were treated was obtained from an American Hospital Association survey. Of the 36,982 women, 12.1 percent had breast-conserving surgery and 87.9 percent had a mastectomy. The frequency of breast-conserving surgery ranged from 3.5 percent to 21.2 percent in various states. The highest rate of use was in the Middle Atlantic states (20.0 percent) and New England (17.2 percent), and the lowest was in the East South Central states (5.9 percent) and the West South Central states (7.3 percent). Breast-conserving treatment was used more often in urban than in rural areas, in teaching hospitals than in nonteaching hospitals, in large hospitals than in small hospitals, and in hospitals with on-site radiation therapy or geriatric services than in others. Most of the geographic variation persisted after adjustment for the characteristics of hospitals and patients for which data were available. There is substantial geographic variation in the use of breast-conserving surgery, which cannot be explained by differences in hospital characteristics. Hospital characteristics that were independently predictive of greater use of breast-conserving surgery were the size of the metropolitan area, the status of the institution as a teaching hospital, and the availability of radiation therapy and geriatric services. (N Engl J Med 1992;326:1102–7.).