An observation-first strategy for liver injuries with "blush" on computed tomography is safe and effective.

Publication/Presentation Date



INTRODUCTION: The management of liver injuries in hemodynamically stable patients is variable and includes primary treatment strategies of observation (OBS), angiography (interventional radiology [IR]) with angioembolization (AE), or operative intervention (OR). We aimed to evaluate the management of patients with liver injuries with active extravasation on computed tomography (CT) imaging, hypothesizing that AE will have more complications without improving outcomes compared with OBS.

METHODS: This is a prospective, multicenter, observational study. Patients who underwent CT within 2 hours after arrival with extravasation (e.g., blush) on imaging were included. Exclusion criteria included cirrhosis, nontraumatic hemorrhage, transfers from outside facilities, and pregnancy. No hemodynamic exclusion criteria were used. The primary outcome was liver-specific complications. Secondary outcomes include length of stay and mortality. Angioembolization patients were compared with patients treated without AE. Propensity score matching was used to match based on penetrating mechanism, liver injury severity, arrival vital signs, and early transfusion.

RESULTS: Twenty-three centers enrolled 192 patients. Forty percent of patients (n = 77) were initially OBS. Eleven OBS patients (14%) failed nonoperative management and went to IR or OR. Sixty-one patients (32%) were managed with IR, and 42 (69%) of these had AE as an initial intervention. Fifty-four patients (28%) went to OR+/- IR. After propensity score matching (n = 34 per group), there was no difference in baseline characteristics between AE and OBS. The AE group experienced more complications with a higher rate of IR-placed drains for abscess or biloma (22% vs. 0%, p = 0.01) and an increased overall length of stay ( p = 0.01). No difference was noted in transfusions or mortality.

CONCLUSION: Observation is highly effective with few requiring additional interventions. Angioembolization was associated with higher rate of secondary drain placement for abscesses or biloma. Given this, a trial of OBS and avoidance of empiric AE may be warranted in hemodynamically stable, liver-injured patient with extravasation on CT.

LEVEL OF EVIDENCE: Therapeutic/Care Management; Level II.





First Page


Last Page





Medicine and Health Sciences




Department of Surgery

Document Type