Contrast extravasation and hematoma volume as predictors of the need for embolization in patients with pelvic fractures

Abstract

Angioembolization is effective and safe in patients with pelvic bone fractures with arterial bleeding. However, there is still no consensus regarding the indications for angiography after pelvic fractures. This study investigated predictors of embolization in patients with pelvic bone fractures through a comparative analysis of patients who had extravasation on angiography. From January 2009 to December 2021, 1431 patients with pelvic fractures were admitted to a single trauma center. After the application of exclusion criteria, 949 patients were enrolled in the study. We divided the patients into two groups to identify predictors of patients with arterial bleeding: the therapeutic embolization (TE) group (n = 149) versus the non-TE group (n = 800). Vital signs and laboratory data were significantly worse in the TE group, except for the Glasgow Coma Scale. When the fracture patterns were compared, the vertical shearing type and the anteroposterior compression type were significantly more common in the TE group. Multivariate logistic regression analysis identified five predictors of therapeutic embolization: systolic blood pressure <90 mmHg in the emergency department (odds ratio (OR) = 2.63; 95% credible intervals (CI) = 1.52-4.53; p = 0.001), combined injury to the abdomen (Abbreviated Injury Scale >= 3) (OR = 3.94; 95% CI = 2.23-6.97; p < 0.001), contrast extravasation on enhanced computed tomography (OR = 30.41; 95% CI = 16.08-57.52; p < 0.001), sacroiliac joint disruption (OR = 2.40; 95% CI = 1.35-4.28; p = 0.003), and hematoma volume >25 mL (OR = 3.79; 95% CI = 2.06-6.98; p < 0.001). Systolic blood pressure less than 90 mmHg, contrast extravasation on enhanced computed tomography, sacroiliac joint disruption, and pelvic hematoma were significant predictors of embolization in pelvic fracture patients. Trauma surgeons' clinical decision-making for patients with pelvic bone fractures should consider clinical features and radiologic findings.