Postoperative ileus risk after posterior thoracolumbar fusion performed with total intravenous anesthesia versus inhaled anesthesia.

Abstract

There has been an increase in the use of total intravenous anesthesia (TIVA) for intraoperative neuromonitoring during thoracolumbar posterior spinal fusion (PSF). Although prior studies have identified risk factors for postoperative ileus (PI) after PSF, to the authors' knowledge, PI rates in patients receiving inhaled anesthetic versus TIVA have not been evaluated. In this study the authors analyzed whether TIVA is associated with greater risk of PI in PSF patients. In this retrospective single-institution cohort study, all patients undergoing PSF at the authors' tertiary academic institution from May 2014 to December 2020 were included. Patients undergoing anterior/lateral approaches or who had concurrent abdominal procedures unrelated to ileus in the same admission were excluded. PI was defined using radiographic and/or clinical diagnoses (postoperative radiographs, abdominal CT, and/or ICD-9 or -10 codes) and was confirmed via chart review. The use of TIVA or inhaled anesthetic was captured from the anesthesia record; patients were excluded if they were missing anesthesia technique data. Postoperative occurrence of PI was compared between patients who had TIVA or inhaled anesthetics while controlling for collected demographic, clinical, and surgical variables. Of the 2819 patients meeting inclusion criteria, 283 (10.0%) had PI (mean ± SD age 59.3 ± 15.8 years; 155 [54.8%] male). The mean patient length of stay was 7.7 ± 5.0 days, which was significantly longer than that of patients without PI (4.9 ± 3.9 days, p < 0.001). Patients with PI had more levels fused (46% of PI patients with ≥ 5 levels fused vs 25% of non-PI patients, p < 0.001) and longer operations (6.0 ± 2.2 vs 5.4 ± 1.9 hours, p < 0.001). TIVA patients were more likely than inhalation-only patients to experience PI, but this finding did not reach significance on univariate analysis (11.0% PI rate vs 8.9%, p = 0.06). After propensity matching 125 non-PI patients and 50 PI patients by age, sex, operative time, and number of levels fused, there was a significant difference in intraoperative opiate dosing between TIVA and inhalational patients (275.7 ± 187.5 intravenous morphine milligram equivalents vs 120.9 ± 155.5, p < 0.001). On multivariate analysis of PI outcome, TIVA was an independently significant predictor (OR 1.45, p = 0.02), as was anesthesia time (OR per hour increase: 1.09, p = 0.03) and ≥ 8 levels fused (OR 1.86, p = 0.01). In a large cohort of PSF patients, TIVA was associated with a higher rate of PI compared with inhaled anesthetic. This effect is likely due to higher intraoperative opiate use in these patients.