Preoperative epidural steroid injections are not associated with increased rates of infection and dural tear in lumbar spine surgery

Authors:
Publisher website
Google Scholar
Abstract
Purpose The study objectives were to use a large national claims data resource to examine rates of preoperative epidural steroid injections (ESI) in lumbar spine surgery and determine whether preoperative ESI or the timing of preoperative ESI is associated with rates of postoperative complications and reoperations. Methods A retrospective longitudinal analysis of patients undergoing lumbar spine surgery for disc herniation and/or spinal stenosis was undertaken using the MarketScan® databases from 2007–2015. Propensity-score matched cohorts were constructed to compare rates of complications and reoperations in patients with and without preoperative ESI. Results Within the year prior to surgery, 120,898 (46.4%) patients had a lumber ESI. The median time between ESI and surgery was 10 weeks. 23.1% of patients having preoperative ESI had more than one level injected, and 66.5% had more than one preoperative ESI treatment. Patients with chronic pain were considerably more likely to have an ESI prior to their surgery [OR 1.62 (1.54, 1.69), p < 0.001]. Patients having preoperative ESI within in close proximity to surgery did not have increased rates of infection, dural tear, neurological complications, or surgical complications; however, they did experience higher rates of reoperations and readmissions than those with no preoperative ESI (p < 0.001). Conclusion Half of patients undergoing lumbar spine surgery for stenosis and/or herniation had a preoperative ESI. These were not associated with an increased risk for postoperative complications, even when the ESI was given in close proximity to surgery. Patients with preoperative ESI were more likely to have readmissions and reoperations following surgery.
Funding Information
  • National Center for Advancing Translational Sciences (UL1 TR001085)

This publication has 26 references indexed in Scilit: