Pelvic trauma imaging: a blinded comparison of computed tomography and roentgenograms.

To determine the sensitivity for detecting pelvic pathology and instability, roentgenograms and computed tomographic (CT) scans from 59 patients with pelvic injuries that had been admitted to a Level I trauma center were randomly reviewed by a orthopedic surgeon blinded to the study. Normal control roentgenograms and CT scans were included to decrease observer bias. The anteroposterior (AP) roentgenogram detected 66% of all pelvic injuries, 78% of those involving the anterior ring, and 53% of those involving the posterior pelvic ring. The trauma CT scan, 10-mm axial images of the abdomen and pelvis, detected 86% of all pelvic injuries, and 78% of anterior ring and 93% of posterior ring injuries. The sensitivity for detecting pelvic instability from one plain film AP pelvis roentgenogram taken in a trauma room setting was 74%. Inlet and outlet views were 75% sensitive. Trauma CT scans were 93% sensitive and high-definition pelvic CT scans (5-mm pelvic cuts) yielded 100% sensitivity. The mechanism of injury could be ascertained with 73% sensitivity by plain films and with 79% sensitivity by inlet and outlet views; trauma CT scans were 96% and high-definition pelvic CT scans were 100% sensitive. When combined, the AP pelvis roentgenograms and trauma CT scans identified 96% of the injured structures and were 100% sensitive in determining injury force patterns and instability. The data suggested that a good quality AP pelvis roentgenogram in conjunction with a complete trauma CT scan of the abdomen/pelvis should identify both the injury mechanism and pelvic instability with a high degree of sensitivity.