[Severe blunt liver injury: reappraisal of criteria for nonoperative management decision. Experience with 88 cases]

Abstract

Aim of the study: To analyze the predictive value of computed tomography (CT) and initial physiologic and laboratory data findings in the immediate operative (OP) or non-operative (NOP) management of blunt liver injury (BL).

Methods: Eighty-eight BL, grade III (51), grade IV (28) and nine grade V (9), aged 26.2 years (16-75) were identified. Hemoperitoneum on CT, hemodynamic status, physiologic and laboratory data <24 hours or preoperative (transfusion, vascular filling) and follow-up >48 hours were analyzed.

Results: Data of 71/88 (80%) NOP and 17/88 (20%) OP patients were reviewed. A secondary laparotomy or laparoscopy was necessary in 11/71 TNO. Six OP (35%) and 1 NOP patients died. Blood units transfused were 1.33 (0-10) vs 5.9 (0-22) and vascular filling 1.45 (0.5-5.5) vs 3.6L (2-12) (P<10(-6), P<4.10(-3) respectively). NOP patients had less severe hemoperitoneum (31 vs 94%, P<10(-5)) and hemodynamic instability (8.5 vs 94%, P<10(-4)). But, there was an overlap of values of blood units transfused, amount of vascular filling and initial haemoglobin levels between NOP and OP patients and among CT grades of liver injury. No cut-off values could be determined: 33% NOP received >4 blood units and >3 L vascular filling; 30% had severe hemoperitoneum. In OP group 23.5% patients had lower values and no severe hemoperitoneum.

Conclusion: In the management of BL, vascular filling and blood transfusion increased with the grade of CT liver injury and were globally more elevated in the operative group but did not individually correlate with hemodynamic stability and did not authorize, by themselves, to decide between operative versus non-operative management.