Are plain radiographs of the spine necessary during evaluation after blunt trauma? Accuracy of screening torso computed tomography in thoracic/lumbar spine fracture diagnosis

Background: Fracture of the thoracolumbar (TL) spine is reported in 8 to 15% of victims of blunt trauma. Current screening of these patients is done with conventional radiography. This may require repeated sets of films and take hours to days. It is imperative that these patients get timely, accurate evaluation to allow for treatment planning and early mobilization; alternatives to plain films would aid in this. The objective of this study is to determine whether the data obtained from admission chest/abdomen/pelvis (CAP) computed tomography (CT) scans after blunt trauma has utility in thoracolumbar spine evaluation. Methods: The records of all patients admitted to a Level I trauma center over a 2-month period who underwent CAP CT were reviewed for the presence of TL spine fracture, time to completion of plain film evaluation, and clinical course. Admission CT scans were reviewed by an attending radiologist who was blinded to any previously diagnosed spine fractures. The two tests were compared for diagnostic accuracy and their discriminatory ability was compared using receiver operating characteristic (ROC) curves. Significance was defined as p < 0.05. Results: In all, 103 patients were admitted from January 1, 2003 to February 28, 2003 and underwent CAP CT scan as part of their initial trauma evaluation. Of these, 26 (25%) had thoracolumbar fractures. Seven (27%) thoracolumbar fractures were not seen on plain radiographs taken during the trauma evaluation. Average time until plain film completion in this group was 8 hours (range, 44 minutes to 38 hours). All 26 (100%) patients with fractures, however, were diagnosed on CT scan performed shortly after admission. Of the remaining 77 patients, two (2.6%) were falsely read as positive for fracture on CT. Sensitivity and specificity of CT scan for thoracolumbar fracture were excellent at 100% and 97%, respectively, with a negative predictive value of 100%. Plain radiographs were 73% sensitive, 100% specific, and had a negative predictive value of 92%. Area under the ROC curve for CT was 0.98, but for plain film was 0.86 (p < 0.02). Conclusion: Admission CAP CT obtained as part of the routine trauma evaluation in these high-risk patients is more sensitive than plain radiographs for evaluation of the TL spine after blunt trauma. In addition, CAP CT can be performed faster. Omission of plain radiographs will expedite accurate evaluation allowing earlier treatment and mobilization.