Duplex ultrasonography versus arteriography in the diagnosis of arterial injury: an experimental study.

Duplex ultrasonography (DUS) and arteriography (ART) were evaluated using an experimental model of arterial trauma in order to determine the accuracy of DUS compared with ART and to define the characteristic ultrasonographic features of arterial injuries. Occlusions (n = 19), blunt injuries (n = 24), lacerations (n = 25), arteriovenous fistulae (n = 13), or no injuries (sham, n = 19) were surgically created in the femoral and carotid arteries of 25 dogs. Following closure of the incisions, DUS was performed and interpreted by a staff vascular surgeon without knowledge of the presence or type of injury. Biplane selective ART was evaluated by an independent staff radiologist. Although DUS and ART were equally accurate in evaluating arterial injuries, DUS was more sensitive (90.1% +/- 3.3% versus 80.2% +/- 4.4%, p = 0.002) and better at identifying lacerated arteries (p = 0.01). However, ART had greater specificity (94.7% +/- 5.1% versus 68.4% +/- 10.7%, p = 0.04) and was more accurate for identifying normal arteries (p = 0.04). The validity of DUS increased in the latter half of the study, thus demonstrating a learning curve. Duplex ultrasonography was a more sensitive screening modality than ART for evaluating arterial injuries in our experimental model, thereby supporting its use in clinical trials to evaluate its accuracy, reliability, and cost effectiveness in the trauma setting. Furthermore, management decisions can be based on the specific type of injury, pathologic condition of the arterial wall, and hemodynamic factors identified by DUS. Clinically occult arterial injuries can be followed by repeat DUS to define the natural history of these injuries.