Predictors of prolonged opioid use following lumbar fusion

The United States has the highest rate of lumbar fusion in the world.1 The lumbar fusion rate increased 2.4 times between 1998 and 2008.2 Rates of opioid pain reliever (OPR) use at 24 months post-lumbar fusion have been reported in up to 30% to 67% of patients. Currently, the number of OPR-related deaths in the US exceeds deaths from motor vehicle accidents.3 The objective of this study is to assess predictors of prolonged opioid use after lumbar fusion. We designed a retrospective cohort of patients who underwent lumbar fusion in 2009. We used claims data from Clinformatics Data Mart. The cohort consisted of 630 non-workers compensation (WC) adults, age 23-62 years, who underwent lumbar fusion for degenerative indications and were enrolled in the commercial health plan for a minimum of 12 months prior to surgery and 24 months following lumbar fusion. Those excluded had ICD-9 codes indicating lumbar spine tumors, fracture, infection, inflammation, a major traumatic accident, or who were pregnant at the time of surgery. Variables indicating lumbar fusion type included posterior, anterior, circumferential (360), and outpatient minimally invasive (OPMI). The variables for indication included degenerative, post-laminectomy, and repeat fusion. In addition, we examined days of OPRs used in the year prior to fusion (divided into quartiles), Elixhauser Comorbidity Index, and diagnosis of smoking, depression, and obesity. The primary outcome variable was excessive OPR use (>364 days dispensed) in two years post-fusion. Logistic regression was used to assess the independent contributions of each of the aforementioned independent variables in predicting the binary outcome variable. Risk for having >364 days of OPRs dispensed post-lumbar fusion was observed for pre-fusion OPR use of 25-86 days dispensed odds ratio (OR) 3.5 (95% CI 1.6-8.3); pre-fusion OPR use of 87-266 days dispensed OR 12.4 (95% CI 5.6-27.2); pre-fusion OPR use (>266 days dispensed) OR 119.8 (95% CI 50.0-287.3); and smoking OR 1.74 (95% CI 1.04-2.92). There was no significant statistical association between fusion type or indication for lumbar fusion, with the binary outcome excessive post-fusion OPR use. These findings demonstrate need for rigorous development of evidence-based policy concerning OPR use following lumbar fusion.