Saturday, November 2, 2001 1:30–2:13 pm Concurrent Session 5A: Lumbar Surgery— Decision Making and Outcomes: 1:30 Interspecialty differences in the use of lumbar spine surgery in the Medicare population

Abstract

Purpose of study: Prior research has documented marked geographic variation in spine surgery rates. However, little is known about how the local supply of neurosurgeons and orthopedic surgeons affects the rates of spine surgery.Methods used: Lumbar spine surgery was identified for all 65 years of age and older beneficiaries in fee-for-service Medicare in 1996 using Medicare Part A (hospital discharge) and Part B (physician services) data. Surgeon specialty was determined by linking physician identifiers to the physician file maintained by the Health Care Financing Administration. Using previously defined hospital referral regions, regional variation in population-based utilization rates, physician supply and specialty market share was evaluated using linear regression.of findings: Overall, 63% of the lumbar spine surgery among Medicare enrollees in 1996 was performed by neurosurgeons. A greater supply of neurosurgeons was associated with higher overall rates of spine surgery (r2=0.06, p<.001). The supply of orthopedic surgeons was not significantly associated with spine surgery rates (p=.14). The proportion of spine surgery performed by neurosurgeons varied from 19.5% in Akron, OH, to 99.4% in Rapid City, SD. Geographic regions with greater numbers of neurosurgeons and a greater proportion of spine surgeries performed by neurosurgeons had higher rates of lumbar discectomy (r2=0.09, p<.001) and lower rates of lumbar fusion (r2=0.09, p<.001). Rates of lumbar discectomy were 40% higher (0.91 vs. 0.65 per 1,000 enrollees; p<.001) in areas of high neurosurgical marketshare (greater than 75% of spine surgery performed by neurosurgeons vs. less than 50%) whereas rates of lumbar fusion were 38% lower (0.89 vs. 1.23 per 1,000 enrollees) in these areas. Orthopedic surgeons were more likely than neurosurgeons to perform fusions in association with lumbar discectomies (15% vs. 5%), lumbar decompressions (37% vs. 13%) and cervical spine procedures (75% vs. 59%).Relationship between findings and existing knowledge: There is considerable interspecialty variation in the surgical management of patients with lumbar spine conditions, which appears related to the known geographic variation in spine surgery rates.Overall significance of findings: Geographic areas with more neurosurgeons have higher overall rates of spine surgery. Patients cared for by neurosurgeons are more likely to undergo lumbar discectomy, and patients cared for by orthopedic surgeons are more likely to undergo fusion.Disclosures: No disclosures.Conflict of interest: No conflicts.