RSNA 2003 Scientific Papers > Reformatted Images of the Lumbar Spine from Abdominal ...
  Scientific Papers
  SESSION: Musculoskeletal (Spinal Trauma)

Reformatted Images of the Lumbar Spine from Abdominal Trauma Multi-Detector CT Protocols Can Replace Lumbar Spine Plain Films in Multiple Trauma Patients

  DATE: Tuesday, December 02 2003
  START TIME: 10:30 AM
  END TIME: 10:37 AM
  CODE: G22-721

Jeffrey Leichter MD
New Haven , CT
James Rhea MD
Thomas Ptak MD, PhD
Robert Novelline MD

Emergency Radiology
Abdomen, CT
Computed tomography (CT)
Spine, fractures
Emergency radiology

Purpose: The traditional workup of the abdominal trauma patient has been an abdominal trauma CT followed by a lumbar spine plain film (LSPF) to rule out lumbar spine fracture. Today, abdominal trauma multi-detector CT (ATMDCT) protocols permit excellent reformatted images of the lumbar spine. To determine if reformatted spine images could replace LSPF in trauma patients the following retrospective investigation was performed.

Methods and Materials: Between 1/1/01 and 12/31/01, 1042 patients were examined with ATMDCT at a level 1 trauma center. 172 of these patients were examined with LSPF. 45 were patients examined with both modalities, in which either showed a lumbar spine fracture. The findings at ATMDCT and LSPF were compared with each other. When there were discrepancies in interpretation of ATMDCT and LSPF the images were re-reviewed by all authors. Consensus reading was the gold standard for this investigation.

Results: 45 patients had fractures at ATMDCT, 28 (62.2%) of these patients had fractures at LSPF. ATMDCT showed a total of 85 fractures; LSPF showed 47 (55.3%) fractures. ATMDCT identified: 64 transverse process, 4 burst, 7 compression, 6 end plate, and 4 spinous process fractures. LSPF identified: 32 transverse process, 1 burst, 11 compression, and 3 end plate fractures. Fractures at ATMDCT overlooked by LSPF included: 32 transverse process, 3 burst, 3 superior end pate, and 4 spinous process fractures. LSPF misclassified the type of fracture in 3 patients. At consensus reading 4 compression fractures seen by LSPF were re-classified to 2 burst fractures and 2 end plate fractures. No fractures at LSPF were overlooked by ATMDCT. The sensitivity for detecting a patient with a fracture with LSPF as compared to ATMDCT is 62.2% (95% confidence interval: 46.5%-76.2%). The specificity for detecting a patient with a fracture using LSPF as compared to ATMDCT is 100% (95% confidence interval: 97.1%-100%). The positive predictive value of LSPF is 100%. The negative predictive value of LSPF is 88.2%.

Conclusion: ATMDCT images of the lumbar spine created from axial and coronal reformations of abdominal trauma multi-detector CT scans can replace lumbo-sacral spine films in the multiple trauma patient. Not only is ATMDCT more accurate in diagnosis but also radiation exposure in patients and length of imaging workup are both reduced.




Questions about this event email: