Evaluation of possible traumatic thoracic aortic injury using aortography and CT

To identify and compare the methods of evaluation for suspected traumatic aortic rupture, 1,031 trauma charts from 1983-1989 were reviewed. Fifty-nine patients were evaluated for possible aortic injury. Patients who died before completion of the CT or aortogram were excluded. Widening of the mediastinum on chest x ray was the most frequent indication for follow-up studies. Twenty-five had a CT of the aortic arch alone. No study showed disruption. There were no false negative studies. Thirty patients had only aortography. Twenty-four were read as normal (one false negative). Six were read as positive (one false positive). In four, both studies were performed (CT/aortography--TP/TP, TN/TN, TP/FN, FP/FP). (FP = False Positive, TP = True Positive, FN = False Negative, TN = True Negative.) Six received surgical repair of the aortic injury (one death). In this experience, CT was used successfully as a screening tool for aortic disruption. It was highly sensitive in recognizing aortic injuries when present (100% vs. 75% for aortography) and in most cases did not require aortographic verification. False positive rates were comparable (CT = 3.8%, aortography = 7.7%). Specificity was also comparable (CT = 96%, aortography = 92%). Overall, four aortograms were inaccurate while only one CT was inaccurate. We recommend the use of CT for the evaluation of widened mediastinum in the stable patient.