Thromboelastography (TEG®) compared to conventional coagulation tests in surgical patients - a laboratory evaluation

Background. Several methods exist for evaluation of hypocoagulation in patients with perioperative bleeding, e.g. thromboelastography (TEG (R)) and conventional methods (platelet count, aPTT, INR and fibrinogen). Considering the vast experience of conventional methods it is important to investigate how well the methods correspond. Methods. Sixty surgical patients were included prospectively and blood samples were taken perioperatively. TEG (R) and conventional parameters were analyzed simultaneously. An assessment of coagulopathy, based on a synthesis of the conventional methods, was done by two experienced coagulation specialists, blinded from the results of TEG (R) and from the results of each other. Hypocoagulation, defined by TEG (R) parameters; reaction time (R-time), angle, maximal amplitude (MA) and fibrinolysis, was evaluated according to a commonly used algorithm. Results. To detect a platelet count below 150 x 10(9) L-1, the sensitivity of TEG was 17% (95% CI, 7-36%) with angle and 25% (95% CI, 11-45%) with MA. The sensitivity to detect fibrinogen below 2 g/L was 11% (95% CI, 3-29%) with angle and 21% with MA (95% CI, 8-43%). To detect aPTT more than 40 s and INR more than 1.2 with R-time, the sensitivity was 19% (95% CI, 8-37%) and 0% (95% CI, 0-69%) respectively. The agreement of the evaluator's assessments of hypocoagulation was 100%, but the agreement with the overall TEG (R) analysis was poor with a sensitivity of 33% and a specificity of 95%. Conclusion. The agreement between conventional laboratory tests and TEG is poor, but it remains uncertain which type of coagulation tests that best reflects the actual bleeding risk.