Pelvic angioembolization: how urgently needed?

PURPOSE: Angioembolization (AE) has been questioned as first-line modality for hemorrhage control of pelvic fracture (PF)-associated bleeding due to its potential inconsistent timely availability. We aimed to describe the patterns of AE use with hemostatic resuscitation and hypothesized that time to AE improved during the study period.

METHODS: A Level-1 trauma center's prospective PF database was analyzed. All consecutive PFs referred to angiography between 01/01/2009 and 12/31/2018 were included. All suspected pelvic hemorrhage was managed with AE; pelvic packing was not performed. Demographics, injury/shock severity, 24-h transfusion data, time to AE and mortality were recorded. Data are presented as median (IQR).

RESULTS: During the 10-year study period, 1270 PF patients were treated. Thirty-six (2.8%) [75% male, 49 (33;65) years, ISS 36 (24;43), base deficit 3.65 (5.9;0.6), transfusions 4(2;7)] had AE. The indication for AE was clinical suspicion (CS) of pelvic bleeding [CS 24(67%)] or arterial blush on CT [CT 12 (33%)]. Median time to AE was 141 min for CS, and 223 min for CT, with no change over the study period. Patients with CS had a higher ISS, worse base deficit, greater transfusion requirements and faster time to AE. Five patients (14%) died. There were no deaths attributed to exsanguination.

CONCLUSIONS: Time to AE did not improve. Patients referred from CT are physiologically different from CS and should be analyzed accordingly, with CS resulting in faster time to AE in sicker patients. Contemporary resuscitation challenges the need for hyperacute AE as no patients exsanguinated despite time to AE of more than 2 h.