IN BAD SHAPE: DOES CONTOUR OF HEMATOMA IN PELVIC FRACTURE PREDICT ARTERIAL EXTRAVASATION?

Objectives: Hemorrhage secondary to pelvic fractures is associated with significant morbidity and mortality. Optimal management continues to be debated but often includes interventional radiology (IR). We hypothesize that when pelvic hematomas are round or convex on CT, they are more likely to be associated with arterial hemorrhage, necessitating IR intervention and embolization.

Methods: All pelvic fractures from our trauma registry between 2017 and 2020 were reviewed. We then selected patients who underwent IR percutaneous pelvic angiography and had an initial CT scan. All CT scans were reviewed visually, by an emergency radiologist and trauma surgeon blinded to all clinical data and to each other’s review, and mathematically, via two methods to assess for convexity. These results were then correlated with the presence of active arterial hemorrhage at the time of angiography.

Results: We reviewed 31 patients who had pelvic fractures undergoing CT scan and pelvic angiography. Average age was 48 years, majority were male (72%) and half were Black. Average Injury Severity Score was 25. Over 75% required transfusions on presentation, as well as IR, surgery, and/or ICU care. 61% had arterial bleeding on angiography. There was excellent correlation with increasing percent area of the convex hull (a mathematical definition of convexity) with arterial bleeding. However, there was no correlation with the radiology or traumatology visual evaluation of convexity.

Conclusion: Mathematical measurement of hematoma convexity was able to predict arterial injury secondary to pelvic fractures better than visual assessment of this feature. Pelvic fracture hematoma shape may help inform our algorithms for managing unstable pelvic fractures and supports the development and incorporation of artificial intelligence tools into the interpretation of complex radiologic studies.