Whole‐body computed tomography in the initial assessment of trauma patients: Is there optimal criteria for patient selection?

Abstract

To describe the use of whole-body computed tomography (WBCT) at this Major Trauma Centre; to determine independent predictors of multi-region injury; and to evaluate the accuracy of the decision to perform WBCT in detecting multi-region injury. A prospective cohort study was performed at a single Major Trauma Centre in New South Wales, Australia. All adult patients who triggered trauma team activation and required an initial CT scan were studied. Primary outcome was the presence of multi-region injury. Logistic regression with stepwise selection was used to derive a prediction model for the need for WBCT based on our primary outcome. Receiver operator characteristic (ROC) analysis was used to compare the accuracy of the derived model and the clinical decision to perform WBCT. Six hundred and sixty patients were studied. WBCT scanning rate was 9.3% of all trauma activations. Of the patients who underwent WBCT, 31/98 (32.0%) had multi-region injury compared with 31/562 (5.5%) who underwent selective CT scanning (P < 0.001). Predictors of multi-region injuries were GCS <9 (OR 3.0, 95% CI 1.3-7.0, P = 0.01), full trauma activation (OR 2.9, 95% CI 1.5-5.3, P = 0.001), fall >5 m (OR 4.8, 95% CI 1.8-13.4, P = 0.003) and pedal cyclist (OR 3.0, 95% CI 1.2-7.5, P = 0.02). Area under ROC curve for the clinical decision to perform WBCT was 0.70 (95% CI 0.63-0.76) compared with 0.74 (95% CI 0.67-0.80) for the prediction model. The decision to perform WBCT scans in trauma should be at the discretion of the treating clinician. Applying a prediction rule would increase the number of WBCT scans performed without improving overall accuracy.