Significance of 'blush' on computed tomography scan in children with liver injury

Abstract

Background/purpose: The aim of this study was to determine if the presence of "blush" (an indication of active bleeding) on abdominal CT in children with blunt liver injury adversely affected their clinical outcome as has been reported in adults.

Methods: The authors reviewed the records of 105 children ages 1 to 16 years with blunt liver injury seen on admission IV contrast CT seen over a 6-year period. Demographic characteristics measured were age, mechanism of injury, and injury severity score (ISS). Clinical outcomes included ICU stay, hospital length of stay (LOS), transfusion requirement (milliliters per kilogram), operations performed, and mortality rate. CT scans were evaluated retrospectively by a radiologist blinded to prior reports, for a "blush" and grade of liver injury. No patient underwent arterial embolization. The authors eliminated children with grade I-II injuries (30 patients), because only one had a blush, and analyzed the 75 patients with severe liver injuries (grades III-V). Those patients without a blush (n = 53) seen on CT were the control group, whereas patients with a blush (n = 22) were the study group. Data were analyzed using the Fisher's Exact and Mann-Whitney U test. The level of significance was set at.05.

Results: Patients with a blush had a significantly larger transfusion requirement (17.3 +/- 30.5 mL/kg v 5.0 +/- 10.9 mL/kg; P =.02) and mortality rate (23% v 4%; P =.02), but the ISS also was significantly greater (25.8 +/- 14.5 v 17.5 +/- 12.2; P =.019). All other data were similar between the 2 groups.

Conclusions: Children with a blush seen on abdominal CT after blunt liver injury have higher transfusion requirements and greater risk of mortality than those without blush. Mortality is primarily related to the severity of their other injuries.