Readmission rates after decompression surgery in patients with lumbar spinal stenosis among Medicare beneficiaries

Abstract

Study design: Retrospective observational cohort analysis of administrative claims.

Objective: Estimate readmission rates after spine stenosis decompression surgery in a 5% randomly selected sample of Medicare beneficiaries.

Summary of background data: Operative management of lumbar spinal stenosis has significant and measurable benefits compared with nonoperative care. Revision rates for lumbar decompression with and without fusion have been reported with significant variability. An understanding of readmission and reoperation rates informs decisions regarding the cost-effective management of lumbar stenosis.

Methods: Patients were identified in 2005-2009 Medicare claims who had both a procedure code for decompression (03.09), and a diagnosis of lumbar spinal stenosis (724.02). Patients diagnosed with spondylolisthesis, and those receiving revision surgery or fusion of more than 3 segments were excluded. Kaplan-Meier product limit method was used to estimate univariate rates of readmission for fusion, decompression, or injection and Cox proportional hazards to examine whether fusion decreased the likelihood of readmission.

Results: The overall 1-year readmission rate was slightly higher in patients undergoing fusion with decompression (9.7%) than patients who underwent decompression alone (7.2%, P = 0.03). Rates at 2 years were 14.6% and 12.5%, respectively. Patients receiving decompression with fusion were slightly younger and more likely female. Procedures performed during readmission were similar for the fusion and no fusion cohorts with 56% receiving fusion, 23% decompression, and 22% injection for pain management. Of the patients who were not readmitted, more than 25% of patients received outpatient injections for pain management during the 3-month quarter of their surgery and approximately 20% in the subsequent quarter.

Conclusion: Readmission rates for spinal stenosis decompression were approximately 8% to 10% per year. Fusion at the index procedure did not protect against subsequent readmission. Large databases can inform choice of surgical options by focusing examination on indications for surgery and reasons for readmission. Fusion along with decompression does not seem to impact readmission rates.