Resonant acoustic rheometry for assessing plasma coagulation in bleeding patients

Abstract

Disordered hemostasis associated with life-threatening hemorrhage commonly afflicts patients in the emergency department, critical care unit, and perioperative settings. Rapid and sensitive hemostasis phenotyping is needed to guide administration of blood components and hemostatic adjuncts to reverse aberrant hemostasis. Here, we report the use of resonant acoustic rheometry (RAR), a technique that quantifies the viscoelastic properties of soft biomaterials, for assessing plasma coagulation in a cohort of 38 bleeding patients admitted to the hospital. RAR captured the dynamic characteristics of plasma coagulation that were dependent on coagulation activators or reagent conditions. RAR coagulation parameters correlated with TEG reaction time and TEG functional fibrinogen, especially when stratified by comorbidities. A quadratic classifier trained on selective RAR parameters predicted transfusion of fresh frozen plasma and cryoprecipitate with modest to high overall accuracy. While these results demonstrate the feasibility of RAR for plasma coagulation and utility of a machine learning model, the relative small number of patients, especially the small number of patients who received transfusion, is a limitation of this study. Further studies are need to test a larger number of patients to further validate the capability of RAR as a cost-effective and sensitive hemostasis assay to obtain quantitative data to guide clinical-decision making in managing severely hemorrhaging patients.