Prospective validation of computed tomographic screening of the thoracolumbar spine in trauma

J Trauma. 2003 Aug;55(2):228-34; discussion 234-5. doi: 10.1097/01.TA.0000076622.19246.CF.


Objective: Concern for thoracolumbar spine (TLS) injuries after major trauma mandates immobilization pending radiographic evaluation. Current protocols use standard posteroanterior and lateral radiographs of the thoracolumbar spine (XR/TLS), but many patients also undergo abdominal or thoracic computed tomographic (CT) scanning. We sought to evaluate whether helical truncal CT scanning performed to evaluate visceral trauma images the spine as well as dedicated XR/TLS.

Methods: We prospectively studied 222 consecutive patients sustaining high-risk trauma requiring TLS screening because of clinical findings or altered mentation. The chest, abdomen, and pelvis were imaged with one intravenous contrast infusion. All patients had CT scan of the chest, abdomen, and pelvis (CT/CAP) and XR/TLS. Initial radiologic diagnoses were compared with the discharge diagnosis of acute fractures confirmed by thin-cut CT scan and/or clinical examination of the patient when alert.

Results: Of 222 patients studied, 215 were fully evaluated. Thirty-six (17%) had acute TLS fractures. The accuracy of CT/CAP for TLS fractures was 99% (95% confidence interval [CI], 96-100%). The accuracy of XR/TLS was 87% (95% CI, 82-92%). Sensitivity, specificity, and positive and negative predictive values were better for CT/CAP than for XR/TLS. CT/CAP found acute fractures XR/TLS missed, and correctly classified old fractures XR/TLS read as "possibly" acute. The total XR/TLS misclassification rate was 12.6% (95% CI, 8.4-19%); for CT/CAP it was 1.4% (95% CI, 0.3-3.3%). No fractures were missed by CT/CAP. No unstable fracture was missed by either technique.

Conclusion: CT/CAP diagnoses TLS fractures more accurately than XR/TLS. Neither misses unstable fractures, but CT scanning finds small fractures that benefit by treatment and identifies chronic disease better. CT screening is far faster and shortens time to removal of spine precautions. CT scan-based diagnosis does not result in greater radiation exposure and improves resource use. Screening the TLS on truncal helical CT scanning performed for the evaluation of visceral injuries is more accurate than TLS imaging by standard radiography. CT/CAP should replace plain radiographs in high-risk trauma patients who require screening.

Publication types

  • Comparative Study
  • Validation Study

MeSH terms

  • Adult
  • Female
  • Glasgow Coma Scale
  • Humans
  • Injury Severity Score
  • Lumbar Vertebrae / diagnostic imaging*
  • Lumbar Vertebrae / injuries*
  • Male
  • Predictive Value of Tests
  • Prospective Studies
  • Radiography*
  • Reproducibility of Results*
  • Sensitivity and Specificity
  • Spinal Injuries / diagnostic imaging*
  • Thoracic Vertebrae / diagnostic imaging*
  • Thoracic Vertebrae / injuries*
  • Tomography, Spiral Computed*
  • Viscera / diagnostic imaging
  • Viscera / injuries