Is the use of pan-computed tomography for blunt trauma justified? A prospective evaluation

Many trauma centers use the pan-computed tomography (CT) scan (head, neck, chest, and abdomen/pelvis) for the evaluation of blunt trauma. This prospective observational study was undertaken to determine whether a more selective approach could be justified.

We evaluated injuries in blunt trauma victims receiving a pan-CT scan at a level I trauma center. The primary outcome was injury needing immediate intervention. Secondary outcome was any injury. The perceived need for each scan was independently recorded by the emergency medicine and trauma surgery service before patients went to CT. A scan was unsupported if at least one of the physicians deemed it unnecessary.

Between July, 1, 2007, and December, 28, 2007, 284 blunt trauma patients (average Injury Severity Score = 11) underwent a pan-CT after the survey form was completed. A total of 311 CT scans were judged to be unnecessary in 143 patients (27%), including scans of the head (62), neck (50), chest (116), and abdomen/pelvis (83). Of the 284 patients, 48 (17%) had injuries on 52 unsupported CT scans. An immediate intervention was required in 2 of the 48 patients (4%). Injuries that would have been missed included 5 of 62 unsupported head scans (8%), 2 of 50 neck scans (4%), 33 of 116 chest scans (28%), and 12 of 83 abdominal scans (14%). These missed injuries represent 5 of the 61 patients with closed head injuries (8%) in the series, 2 of the 23 with C-spine injuries (9%), 33 of the 112 with chest injuries (29%), and 12 of the 86 with abdominal injuries (14%). In 19 patients, none of the four CT scans was supported; nine of these had an injury identified, and six were admitted to the hospital (1 to the intensive care unit). Injuries that would have been missed included intraventricular and intracerebral hemorrhage (4), subarachnoid hemorrhage (2), cerebral contusion (1), C1 fracture (1), spinous and transverse process fractures (3), vertebral fracture (6), lung lacerations (1), lung contusions (14), small pneumothoraces (7), grade II-III liver and splenic lacerations (6), and perinephric or mesenteric hematomas (2).

In this small sample, physicians were willing to omit 27% of scans. If this was done, two injuries requiring immediate actions would have been missed initially, and other potentially important injuries would have been missed in 17% of patients.