Incidence, impact, and risk factors of adverse events in thoracic and lumbar spine fractures: An ambispective cohort analysis of 390 patients

R. Andrew Glennie, Tamir Ailon, Kyun Yang, Juliet Batke, Charles G. Fisher, Marcel F. Dvorak, Alexander R. Vaccaro, Michael G. Fehlings, Paul Arnold, James S. Harrop, John T. Street

Research output: Contribution to journalArticlepeer-review


Background context Adverse events (AEs) in thoracic and lumbar spine fractures are common, but little is known about the type of AEs that are specific to this population. Furthermore, very little is known about the incidence and clinical impact of these AEs on patients in the presence of traumatic spinal cord injury and whether they are treated operatively or nonoperatively. Purpose The purpose of this study was to determine primarily the incidence of AEs in patients with thoracic or lumbar spine fractures treated both operatively and nonoperatively and their impact on length of stay (LOS) and secondarily the difference in the incidence of AEs in both neurologically intact and compromised patients. Study design/setting This is an ambispective cohort study at a quaternary referral center. Patient sample Patients admitted at our institution with thoracic or lumbar fractures from January 2009 to December 2013 were identified. Patients with full Spine Adverse Events Severity System (SAVES) data were included. Outcome measures Number and type of AEs collected from SAVES were assessed. Impact of AE on acute LOS was also determined. Methods Data on intraoperative, preoperative, and postoperative AEs were prospectively collected using the SAVES data collection. Logistic regression was used to model the likelihood of experiencing at least one AE based on the patient characteristics. The impact of the total number of AEs experienced by a patient and that of each of the most common AEs on LOS was determined using Poisson regression. Results Three hundred and ninety patients were included in the final analysis. Two hundred and seventy-six patients (70.8%) were treated operatively. One hundred and forty patients (36%) experienced neurologic deficit as a result of their initial injury. Adverse events occurred 56% of the time in the operatively treated patients and only 13% of the time in the nonoperative group. The presence of neurologic deficit increased the risk of AEs especially in high thoracic (T1-T6) trauma increasing the odds of having an AE by 12.1 (p<.0001). The most common AEs were urinary tract infections (19.7%), neuropathic pain (12.3%), pneumonias (11.8%), delirium (10.5%), and ileus (6.2%). Length of hospital stay increased significantly with pneumonia (p<.0001) and delirium (p=.0001). Conclusions The presence of neurologic injury and the need for operative fixation of thoracic or lumbar injuries lead to a greater risk of AEs. Only pneumonia and delirium consistently increase LOS.

Original languageEnglish (US)
Pages (from-to)629-637
Number of pages9
JournalSpine Journal
Issue number4
StatePublished - Apr 1 2015
Externally publishedYes


  • Adverse event
  • Lumbar
  • Operative
  • Spine
  • Thoracic
  • Traumatic spinal cord injury

ASJC Scopus subject areas

  • Surgery
  • Orthopedics and Sports Medicine
  • Clinical Neurology


Dive into the research topics of 'Incidence, impact, and risk factors of adverse events in thoracic and lumbar spine fractures: An ambispective cohort analysis of 390 patients'. Together they form a unique fingerprint.

Cite this