Relationship of thromboelastography and conventional clotting test values with severe bleeding in critically ill...

This study aimed to ascertain the associations of thromboelastography (TEG^® ) and standard laboratory test (SLTs) values with the presence of bleeding in critically ill patients with known coagulopathy.

Three groups of coagulopathic patients with (a) hepatic failure, (b) postoperative period after prolonged cardiac surgery, and (c) complex abdominal surgery with sepsis were prospectively included in this study. On intensive care unit (ICU) admission, patients were stratified into two groups according to whether they had major bleeding (MB) (evident overt bleeding, important bleeding apparent on imaging studies, and/or need for moderate-massive blood transfusion and hemodynamic instability). Blood samples were drawn for the SLTs (international normalized ratio [INR], activated partial thromboplastin time [aPTT], platelet count, and fibrinogen level [Clauss]) and TEG whole blood coagulation assays. Receiver operating characteristic (ROC) curves were generated to determine the efficiency of TEG and SLTs for detecting bleeding. The correlations between SLTs and TEG parameters with similar coagulation profiles were evaluated by Spearman rank-order analysis.

Eighty-three patients were included, and bleeding was confirmed in 45 (54%). The fibrinogen level demonstrated the best accuracy for detecting bleeding with an area under the curve and 95% confidence intervals [AUC (95% CI)] of 0.74 (0.63-0.85) with the best cutoff value of ≤ 2 g/L. Regarding TEG-MA, the AUC (CI) obtained with the optimal cutoff value of ≤ 51 mm was 0.68 (0.56-0.80).

Both conventional clotting tests and TEG values were poorly associated with bleeding in this critically ill cohort of patients with coagulopathy.