Prevalence of "Flat-Line" Thromboelastography During Extracorporeal Membrane Oxygenation for Respiratory Failure in Adults

We evaluated the prevalence of a thromboelastography reaction time (R time) > 90 min ("flat-line") reversible with heparinase during extracorporeal membrane oxygenation (ECMO). We evaluated the association between "flat-line" thromboelastography, other coagulation tests, and risk of bleeding during ECMO. Thirty-two consecutive patients on ECMO were included. Anticoagulation was provided by continuous infusion of unfractionated heparin to maintain an activated partial thromboplastin time (aPTT) ratio between 1.5 and 2.0. Activated clotting times (ACTs) thromboelastography without and with heparinase were measured. Occurrence of bleeding was recorded. Median heparin infusion rate was 16 (12-20) IU/kg/h, aPTT ratio was 1.67 (1.48-1.96), and ACT was 173 (161-184) sec. One hundred forty-five (46%) of 316 paired thromboelastography samples were "flat lines" all reversed with heparinase. Patients with "flat-line" thromboelastography received more heparin (p = 0.001) but had similar platelet count (p = 0.164) and fibrinogen level (p = 0.952) than those without. Activated partial thromboplastin time, ACT, and R time without heparinase weakly correlated between each other (Spearman correlation <= 0.36) with poor agreement (Cohen's kappa <= 0.10). Major bleeding occurred in seven (22%) patients. Bleeding during ECMO was not predicted by any of the used test. In conclusion, adjusting heparin infusion to maintain aPTT ratio between 1.5 and 2.0 frequently resulted in "flat-line" thromboelastography.