Thursday, September 27, 2018 8:30 AM–9:30 AM Best Papers: 106. Do preoperative epidural steroid injections increase the risk of infection after lumbar spine surgery?

Abstract

BACKGROUND CONTEXT Lumbar radiculopathy occurs in up to 5% of the population accounting for significant morbidity. Lumbar epidural corticosteroid injections (ESI) may provide diagnostic and therapeutic benefit in the nonoperative management of radiculopathy. However, concern exists regarding whether preoperative ESI's might predispose surgical patients to an increased risk of postoperative infection. PURPOSE To determine an association and temporal relationship between preoperative ESI and postoperative infection in a large population of patients undergoing lumbar surgery. STUDY DESIGN/SETTING Retrospective review of all patients who underwent elective lumbar surgery for radiculopathy or stenosis at a single institution between 2000 and 2017 with at least 90 days follow-up. PATIENT SAMPLE All patients who underwent lumbar surgery were identified in our institutional database by ICD and CPT code. Those who underwent preoperative lumbar ESI were also identified by CPT code. Patients were categorized as no ESI, preoperative ESI less than 30 days, 30-90 days, and greater than 90 days before surgery. Pre- and postoperative clinical data was available in a single electronic medical record. OUTCOME MEASURES The primary outcome measure was postoperative infection requiring reoperation within 90 days of index procedure. Postoperative infections were identified by ICD code and confirmed by operative report for each patient. METHODS Of the patients who underwent lumbar spine surgery, those who received preoperative lumbar ESI were identified. Postoperative infection rates were determined by ICD code and operative report. All demographic information including age, sex, body mass index (BMI) and Charlson Comorbidity Index (CCI) was determined. Comparison and regression analysis was used to determine an association between preoperative ESI exposure, demographics, comorbidities, and postoperative infection. RESULTS A total of 15,011 patients who underwent elective lumbar surgery were included. Of these, 197 (1.3%) were diagnosed with postoperative infection. Of the identified patients, 595 had lumbar ESI within 30 days of surgery, 1709 within 30-90 days of surgery, and 2036 greater than 90 days before surgery. Overall, there was a marginally higher rate of infection among the cohort of patients who had a preoperative ESI that did not reach statistical significance (1.54% vs. 1.22%, p=.118). Among the temporal subgroups, there was no significant difference in postoperative infection rates between those patients who had ESI within 30 days of surgery (12/595, 2.0%, p=.089) or those who had ESI 30-90 days before surgery (15/1709, 0.9%, p=.23) compared to those with no preoperative ESI (130/10541, 1.2%). Those patients who underwent ESI greater than 90 days before surgery had a higher rate of infection (40/2036, 2.0%) compared to those with no ESI (p=.007). Sub analysis of those patients who underwent ESI greater than 90 days before surgery demonstrated that patients with a postoperative infection had a higher BMI (p CONCLUSIONS This retrospective single institution review of over 15,000 patients undergoing lumbar spine surgery demonstrated that increased body mass index and Charlson Comorbidity Index are stronger predictors of postoperative infection after lumbar spine surgery than having received a preoperative ESI. Although the data did not demonstrate a temporal relationship between preoperative lumbar epidural steroid injection and risk of postoperative infection, there was a trend toward a higher infection rate among all patients who underwent preoperative ESI when compared with those who did not. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.