Racial Differences in Perioperative Opioid Utilization in Lumbar Decompression and Fusion Surgery for Symptomatic Lumbar Stenosis or Spondylolisthesis

Study Design: Retrospective cohort study. Objectives: To assess for racial differences in opioid utilization prior to and after lumbar fusion surgery for patients with lumbar stenosis or spondylolisthesis. Methods: Clinical records from patients with lumbar stenosis or spondylolisthesis undergoing primary <3-level lumbar fusion from 2007 to 2016 were gathered from a comprehensive insurance database. Records were queried by International Classification of Diseases diagnosis/procedure codes and insurance-specific generic drug codes. Opioid use 6 months prior, through 2 years after surgery was assessed. Multivariate regression analysis was employed to investigate independent predictors of opioid use following lumbar fusion. Results: A total of 13 257 patients underwent <3-level posterior lumbar fusion. The cohort racial distribution was as follows: 80.9% white, 7.0% black, 1.0% Hispanic, 0.2% Asian, 0.2% North American Native, 0.8% "Other," and 9.8% "Unknown." Overall, 57.8% patients utilized opioid medications prior to index surgery. When normalized by the number opiate users, all racial cohort saw a reduction in pills disbursed and dollars billed following surgery. Preoperatively, Hispanics had the largest average pills dispensed (222.8 pills/patient) and highest average amount billed ($74.67/patient) for opioid medications. The black cohort had the greatest proportion of patients utilizing preoperative opioids (61.8%), postoperative opioids (87.1%), and long-term opioid utilization (72.7%), defined as use >1 year after index operation. Multivariate logistic regression analysis indicated Asian patients (OR 0.422, 95% CI 0.191-0.991) were less likely to use opioids following lumbar fusion. Conclusions: Racial differences exist in perioperative opioid utilization for patients undergoing lumbar fusion surgery for spinal stenosis or spondylolisthesis. Future studies are needed corroborate our findings.