Reevaluation of Hepatic Angioembolization for Trauma in Stable Patients: Weighing the Risk.

Abstract

Angioembolization (AE) is recommended for extravasation from liver injury on CT. Data supporting AE are limited to retrospective series that have found low mortality but high morbidity. These studies did not focus on stable patients. We hypothesized that AE is associated with increased complications without improving mortality in stable patients. We queried the 2016 Trauma Quality Improvement Project database for patients with grade III or higher liver injury (Organ Injury Score ≥ 3), blunt mechanism, with stable vitals (systolic blood pressure ≥ 90 mmHg and heart rate of 50 to 110 beats/min). Exclusion criteria were nonhepatic intra-abdominal or pelvic injury (Organ Injury Score ≥ 3), laparotomy less than 6 hours, and AE implementation more than 24 hours. Patients were matched 1:2 (AE to non-AE) on age, sex, Injury Severity Score, liver Organ Injury Score, arrival systolic blood pressure and heart rate, and transfusion in the first 4 hours using propensity score logistic modeling. Primary outcomes were in-hospital mortality, length of stay, transfusion, hepatic resection, interventional radiology drainage, and endoscopic procedure. There were 1,939 patients who met criteria, with 116 (6%) undergoing hepatic AE. Median time to embolization was 3.3 hours. After successfully matching on all variables, groups did not differ with respect to mortality (5.4% vs 3.2%; p = 0.5, AE vs non-AE, respectively) or transfusion at 4 to 24 hours (4.4% vs 7.5%; p = 0.4). A larger percentage of the AE group underwent interventional radiology drainage (13.3% vs 2.2%; p < 0.001), with more ICU days (4 vs 3 days; p = 0.005) and longer length of stay (10 vs 6 days; p < 0.001). Hepatic AE was associated with increased morbidity without improving mortality, suggesting the benefits of AE do not outweigh the risks in stable liver injury. Observing these patients is likely a more prudent approach.