Is it possible to use transaminases for deciding on surgical or non-operative treatment for blunt liver trauma?

Abstract

We aimed to research the relation of transaminase levels in blunt liver trauma (BLT) with the intensity of the trauma and the use of transaminase levels for deciding on surgical or non-operative treatment. In all, 44 patients with BLT diagnosed by computerized tomography (CT) were involved in this retrospective study. By testing the correlation of the transaminase levels and the grade of liver injury with receiver operator characteristics (ROC), area under the curve (AUC) was calculated; besides, the sensitivity, specificity, and cut-off values of transaminases were calculated separately for the grades. Moreover, same method was repeated for the surgically and non-operatively treated patients. Cut-off value was assessed for surgical and non-operative treatments. The efficiency of transaminases in deciding non-operative treatment was compared with that of other methods using ROC test applied on focused abdominal sonography in trauma (FAST), hemodynamic instability, blood replacement rate, aspartate aminotransferase (AST), and alanine aminotransferase (ALT). It was observed that the AUC, sensitivity, and specificity increased correspondingly with the grade rise of transaminase levels in BLT. In the selection of non-operative treatment/surgery, following values have been confirmed: AUC for AST: 0.851 (sensitivity: 86 %, specificity: 73 %, cut-off value: 498 U/L), AUC for ALT: 0.880 (sensitivity: 86 %, specificity: 81 %, cut-off value: 498 U/L), AUC for replacement: 0.948 (sensitivity: 86 %, specificity: 94 %), AUC for hemodynamic instability: 0.902 (sensitivity: 86 %, specificity: 94 %), and AUC for FAST: 0.642 (sensitivity: 57 %, specificity: 75 %). It was found that in BLT, transaminases can predict the injury rating with higher accuracy as the grade rises, and they outrival FAST in terms of determining the need for laparotomy.