DOI: 10.1097/SA.0000000000000140
,
Issn Print: 0039-6206
Publication Date: 2015/06/01
Facebook
Twitter
LinkedIn
Email
 |
Checking for direct PDF access through Ovid | |
Excerpt
Of 401 patients who had LT, 239 undergoing LDLT were included. Anesthesia was performed based on the institution’s standard protocol. Transfusions were administered to maintain prothrombin time of less than 2.0 international normalized ratio (INR), Fib concentration greater than 100 mg/dL, and PLT count greater than 30,000/μL. During LDLT, coagulation assays and ROTEM test were performed using blood sample obtained 1 hour after induction of general anesthesia, 1 hour after surgical incision, 30 minutes after hepatectomy, and 30 minutes after graft reperfusion and after hepatic artery anastomosis. The study included 1139 EXTEM, 1182 INTEM, and 1125 FIBTEM tests. ROTEM variables were clotting time, clot formation time, maximum clot firmness (MCF), α angle, and A5 and A10. The correlations between A5 and MCF index, PLT count, and Fib concentration were calculated. Receiver operating characteristic analysis with area under the curve (AUC) was used to determine A5 thresholds predictive of PLT counts of less than 30,000/μL, less than 50,000/μL, and Fib concentrations less than 100 mg/dL. All statistical analyses were done with MedCalc statistical software (MedCalc Software, Mariakerke, Belgium). A 2-tailed P < 0.05 indicated statistical significance.
During LDLT, the median PLT count was 47,000/μL (interquartile range [IQR], 32,000–65,000/μL); the median Fib concentration was 100 mg/dL (IQR, 77–137 mg/dL), and the median prothrombin time was 1.8 INR (IQR, 1.6–2.2 INR). Among the ROTEM parameters, A5EXTEM (r = 0.961, P < 0.0001) and A10EXTEM (r = 0.975, P < 0.0001) were highly correlated with MCFEXTEM and with MCFINTEM (r = 0.950 and r = 0.965, respectively; P < 0.0001 for both). A5FIBTEM (r = 0.908, P < 0.001) and A10FIBTEM (r = 0.951, P < 0.0001) showed high correlation with MCFFIBTEM. Cutoff values of A5EXTEM and A10EXTEM predicting PLT count of less than 30,000/μL were 15 and 22 mm, respectively (sensitivities 86% and 85%; specificities 77% and 78%, respectively). Cutoff values to predict PLT count of less than 50,000/μL were 19 and 27 mm, respectively (sensitivities 82% and 82%; specificities 78% and 77%, respectively). Areas under the curve for A5EXTEM and A10EXTEM were 0.90 and 0.898, respectively, and 0.871 and 0.857, respectively, for PLT count of less than 30,000/μL and PLT count of less than 50,000/μL, respectively. Cutoff values of A5INTEM and A10INTEM were 16 and 22 mm, respectively (sensitivities 82% and 85%; specificities 86% and 83%, respectively), for predicting PLT count of less than 30,000/μL and 19 and 27 mm for predicting PLT count of less than 50,000/μL (sensitivities 82% and 79%; specificities 77% and 79%, respectively). Areas under the curve of A5INTEM and A10INTEM were 0.914 and 0.915, respectively, for PLT count of less than 30,000/μL and 0.873 and 0.869, respectively, for PLT count of less than 50,000/μL (all P > 0.0001). Receiver operating characteristic curve analysis found that A5FIBTEM and A10FIBTEM predicting Fib concentration of less than 100/mg/dL were 4 and 5 mm, respectively. The AUCs of A5FIBTEM and A10FIBTEM for Fib concentration of less than 100 mg/dL were 0.86 and 0.87, respectively.
