Abstract
BACKGROUND: The objective was to assess the predictive performance of different intravascular contrast extravasation (ICE) characteristics for need for pelvic transarterial embolization (TAE) to determine the risk factors of false positives. METHODS: A retrospective study was performed in our trauma center between 2010 and 2015. All severe trauma patients with pelvic fracture were included. Pelvic ICE characteristics on computed tomography (CT) scan were studied: arterial (aSICE), portal surface (pSICE), and extension (exSICE) anatomic relationships. The overall predictive performance of ICE surfaces for pelvic TAE was analyzed using receiver operating characteristic curves. The analysis focused on risk factors for false positives. RESULTS: Among 311 severe trauma patients with pelvic ring fracture (mean age, 42 \textpm 19 years; mean Injury Severity Score, 27 \textpm 19), 94 (30%) had at least one pelvic ICE on the initial CT scan. Patients requiring pelvic TAE had significantly larger aSICE and pSICE than others (p = 0.001 and p = 0.035, respectively). The overall ability of ICE surfaces to predict pelvic TAE was modest (aSICE area under the receiver operating characteristic curve, 0.76 [95% confidence interval, 0.64-0.90]; p = 0.011) or nonsignificant (pSICE and exSICE). The high-sensitivity threshold was defined as aSICE 20 mm or more. Using this threshold, 76% of patients were false positives. Risk factors for false positives were admission systolic blood pressure of 90 mm Hg or greater (63% vs 20%; p = 0.03) and low transfusion needs (63% vs 10%; p = 0.009), extravasation in contact with complex bone fracture (78% vs 30%; p = 0.008), or the absence of a direct relationship between extravasation and a large retroperitoneal hematoma (100% vs 38%; p \textless 0.001). CONCLUSION: A significant pelvic ICE during the arterial phase does not guarantee the need for pelvic TAE. Three quarters of patients with aSICE of 20 mm or more did not need pelvic TAE. Several complementary CT scan criteria will help to identify this risk of false positives to determine adequate hemostatic pelvic procedures. LEVEL OF EVIDENCE: Therapeutic study, level IV.