Fast interpretation of thromboelastometry in non-cardiac surgery: reliability in patients with hypo-, normo-, and hypercoagulability

Abstract

Background: Conventional coagulation test are not useful to guide haemostatic therapy in severe bleeding due to their long turn-around time. In contrast, early variables assessed by point-of-care thromboelastometry (ROTEM(®)) are available within 10-20 min and increasingly used to guide haemostatic therapy in liver transplantation and severe trauma. However, the reliability of early ROTEM(®) variables to predict maximum clot firmness (MCF) in non-cardiac surgery patients with subnormal, normal, and supranormal MCF has not yet been evaluated.

Methods: Retrospective data of 14,162 ROTEM(®) assays (3939 EXTEM(®), 3654 INTEM(®), 3287 FIBTEM(®), and 3282 APTEM(®) assays) of patients undergoing non-cardiac surgery were analysed. ROTEM(®) variables [clotting time (CT), clot formation time (CFT), α-angle, A5, A10, and A15] were related to MCF by linear or non-linear regression, as appropriate. The Bland-Altman analyses to assess the bias between early ROTEM(®) variables and MCF and receiver operating characteristics (ROC) were also performed.

Results: Taking the best and worst correlation coefficients for each assay type, CT (r=0.18-0.49) showed the worst correlation to MCF. In contrast, α-angle (r=0.85-0.88) and CFT (r=0.89-0.92) demonstrated good but non-linear correlation with MCF. The best and linear correlations were found for A5 (r=0.93-0.95), A10 (r=0.96), and A15 (r=0.97-0.98). ROC analyses provided excellent area under the curve (AUC) values for A5, A10, and A15 (AUC=0.962-0.985).

Conclusions: Early values of clot firmness allow for fast and reliable prediction of ROTEM(®) MCF in non-cardiac patients with subnormal, normal, and supranormal MCF values and therefore can be used to guide haemostatic therapy in severe bleeding.