A predictive model of motor deficit in civilian ballistic lumbosacral spine injuries

Spine J. 2024 Mar;24(3):446-453. doi: 10.1016/j.spinee.2023.11.011. Epub 2023 Nov 21.

Abstract

Background context: Civilian gunshot wounds to the spine are an increasingly common injury in the USA. A majority of the available research is focused on a military population suffering high energy missile injury. Minimal research has focused on civilian ballistic injuries to the lumbosacral spine as the available studies focus on the entire spine due to limited numbers.

Purpose: Characterize ballistic injuries to the lumbosacral spine and develop a model to predict the presence of neurological deficit based upon a patients presenting fracture morphology.

Study design: Retrospective chart review.

Patient sample: One hundred forty-eight consecutive patients that were presented to an urban level 1 trauma center with ballistic injures to the spine involving the levels L1-S2.

Outcomes measures: Neurological status at presentation and final follow up using the American Spinal Injury Association (ASIA) Impairment Scale.

Methods: IRB approval was obtained, and retrospective chart review was performed. Extracted data included patient demographics, neurological status on presentation and final follow-up, fracture morphology, assessment of stability, other associated injuries, and surgical procedures performed. Proportional analysis was performed to characterize the fractures and their associated neurological injuries. Chi-square testing was done to identify fracture characteristics associated with neurologic injury. A multiple logistical regression was performed using fracture characteristics highly associated with neurological deficit to develop a model to predict neurologic deficit. The model was then validated with a receiver operator curve.

Results: Of the 148 patients, 14 patients underwent spinal surgery with the most common indication being decompression and foreign body removal. There was a high incidence of intra-abdominal injury (73.6%). Fractures were characterized by level, affected vertebral component, and spinal canal involvement. Neurological injury was classified using the ASIA scale at presentation and final follow up. Odds-ratios of vertebral fracture characteristics showed neurologic deficit was highly associated with pedicle fractures (OR=9.07 [4.14-21.54] - 95% CI), lamina fractures (OR=6.42 [3.16-13.62] - 95% CI), facet fractures (OR=5.95 [2.90-12.79] - 95% CI), intra-canal bone (OR=12.79 [5.98-29.05] - 95% CI), and an intra-canal trajectory (OR=1078 [4.48-28.98]) - 95% CI. Multiple logistic regression was performed to construct a predictive model of neurologic deficit which showed that intra-canal trajectory, pedicle fracture and facet fracture are associated with neurologic deficit. An ROC curve was made with an area of 0.849 ([0.7853 to 0.9128 - 95% CI], p<.0001) demonstrating a good model fit.

Conclusions: Ballistic injuries to the lumbosacral spine involve complex injury patterns in an often polytraumatized patient. Fractures involving the pedicle, lamina, and facet are highly associated with neurologic injury, as is a trans-canal trajectory. A patients fracture morphology can be used to predict if a neurologic deficit is present.

Keywords: Ballistic; Fracture; Gunshot wound; Lumbar; Modeling; Sacral; Trauma.

MeSH terms

  • Humans
  • Retrospective Studies
  • Spinal Fractures* / surgery
  • Spine
  • Wounds, Gunshot* / complications
  • Wounds, Gunshot* / epidemiology
  • Wounds, Gunshot* / surgery