Correlation between Thrombin Generation, Standard Coagulation Assays, and Viscoelastic Assays for Hemostatic Assessment in Critically Ill Children.

Abstract

Accurate assessment of hemostatic function is essential to guide care in critically ill children with acute and acquired coagulopathies. Thrombin generation (TG) provides a global assessment of procoagulant and anticoagulant factors and is commonly used in hemostasis research laboratories. Our objective was to determine the correlation of clinically available hemostasis assays with TG in critically ill children. Children (<18 years old, >3 kg in weight) in the intensive care unit were enrolled from March 2016 to December 2019 in a prospective 2-center study. Coagulation tests were prothrombin time, activated thromboplastin time, anti-Xa assay, viscoelastic assays (thromboelastography [TEG], rotational thromboelastometry [ROTEM]), and TG (induced by 20 pM tissue factor in platelet poor plasma and reported as endogenous thrombin potential [ETP; nM*min]). Data are reported as median (interquartile range) or Spearman coefficient (ρ). Patients (n = 106, age 10.2 years [3.8-15.3]) were divided into 3 groups: (a) no anticoagulation (n = 46), (b) anticoagulation (unfractionated heparin) without extracorporeal life support (n = 34), or (c) with extracorporeal life support (n = 26). ETP was decreased in anticoagulated compared to non-anticoagulated patients (group 1: 902.4 [560.8-1234], group 2: 315.6 [0.0-962.2], group 3: 258.5 [0.0-716.6]; P < 0.0001). Across all patients, ETP correlated best with TEG kinetic time (TEG-K), in min (ρ = -0.639), followed by TEG reaction time, in min (ρ = -0.596). By group, ETP correlated best with international normalized ratio for group 1 (ρ = -0.469), TEG-K time for group 2 (ρ = -0.640), and anti-Xa for group 3 (ρ = -0.793). Standard and viscoelastic assays have varying correlation with TG in critically ill children. TEG-K time had the most consistent moderate correlation with ETP across all groups.