Thromboelastography Reaction-Time Thresholds for Optimal Prediction of Coagulation Factor Deficiency in Trauma

Document Type

Journal Article

Publication Date

5-1-2020

Journal

Journal of the American College of Surgeons

Volume

230

Issue

5

DOI

10.1016/j.jamcollsurg.2020.01.033

Abstract

© 2020 American College of Surgeons Background: Coagulopathy is common in multitrauma patients and repletion of procoagulant factor deficiency with fresh frozen plasma (FFP) improves hemostasis. Optimal kaolin-thromboelastography thresholds for FFP transfusion in trauma patients have not been well established. Study Design: Adult trauma patients with an Injury Severity Score ≥15 were included in this retrospective observational cohort study. The primary end point was area under the receiver operating characteristic curve (AUROC) for reaction time (R-time) to detect procoagulant factor deficiency, as reflected by an elevated international normalized ratio (INR) or aPTT. Test characteristics for the optimal R-time threshold calculated in our study were compared against thresholds recommended by the American College of Surgeons for FFP transfusion. Results: Six hundred and ninety-four pairs of thromboelastography and conventional coagulation tests were performed in 550 patients, with 144 patients having additional pairs of tests after the first hour. The R-time was able to detect procoagulant factor deficiency (INR ≥1.5 AUROC 0.80; 95% CI, 0.75 to 0.85; aPTT ≥40 seconds AUROC 0.85; 95% 0.80 to 0.89) and severe procoagulant factor deficiency (INR ≥2.0 AUROC 0.82; 95% CI, 0.73 to 0.99; aPTT ≥60 seconds AUROC 0.89; 95% CI, 0.81 to 0.98) with good accuracy. Optimal thresholds to maximize sensitivity and specificity were 3.9 minutes for detection of INR ≥1.5, 4.1 minutes for detection of aPTT ≥40 seconds, 4.3 minutes for detection of INR ≥2.0, and 4.3 for detection of aPTT ≥60 seconds. Currently recommended R-time thresholds for FFP transfusion had 100% specificity for detecting procoagulant factor deficiency, but low sensitivity (3% to 7%). Conclusions: R-time can detect procoagulant factor deficiency in multitrauma patients with good accuracy, but currently recommended R-time thresholds are highly specific and not sensitive. Use of low-sensitivity thresholds might result in undertreatment of many patients with procoagulant factor deficiency.

Share

COinS