Abstract
AbstractBackgroundExtracorporeal cardiopulmonary resuscitation (eCPR) improves outcomes after out of hospital cardiac arrest (OHCA). However, bleeding and thrombosis are common complications. The post cardiac arrest syndrome including trauma and altered hemostasis creates challenges when balancing the risk of bleeding versus thrombosis. We aimed to describe the incidence and predictors of bleeding and thrombosis and their association with in-hospital mortality.MethodsConsecutive patients presenting with refractory VT/VF OHCA between December 2015 and March 2022 who met criteria for eCPR initiation at our center were included. Patients were anticoagulated with unfractionated heparin while on ECMO. Major bleeding was defined by the extracorporeal life support organization’s (ELSO) criteria as clinically significant bleed associated with transfusion of ≥2 units of red blood cells in 24h or bleeding in a critical area. Thrombosis was defined by clinical signs and symptoms and or imaging evidence of thrombi. Adjusted analyses were done to seek out risk factors for bleeding and thrombosis and evaluate their association with mortality.ResultsMajor bleeding occurred in 135/200 patients (67.5%) with traumatic bleeding from CPR in 73/200 (36.5%). Baseline demographics and arrest characteristics were similar between groups. In multivariable regression analysis only fibrinogen was independently associated with bleeding (aHR 0.98 per every 10mg/dl rise, 95% CI: 0.96-0.99). Patients who died had a higher rate of bleeds per day (0.21 vs 0.03,p<0.001) though bleeding was not significantly associated with in-hospital death (aHR 0.81, 95% CI: 0.55-1.19). A thrombotic event occurred in 23.5% (47/200). Venous thromboembolism (VTE) occurred in 11% (22/200) and arterial thrombi in 15.5% (31/200). Clinical characteristics were comparable between groups. In adjusted analyses, antithrombin III level on admission (aHR 0.97, 95% CI: 0.94-0.99) and use of dual anti-platelet therapy (DAPT) (aHR 0.31, 95% CI: 0.11-0.86) were associated with the incidence of thrombosis. Thrombosis was not associated with in-hospital death (aHR 0.65, 95% CI: 0.42- 1.03).ConclusionBleeding is a frequent complication of eCPR that is associated with decreased fibrinogen levels on admission. Thrombosis is less common occurring in 24% of the cohort. Neither bleeding nor thrombosis was significantly associated with in-hospital mortality.Clinical PerspectiveWhat is new?In this large single center study with a protocolized approach to anticoagulation for patients with VT/VF OHCA bleeding as defined by ELSO major bleeding was common occurring in 68% of the cohort while thrombosis was less frequent occurring in 24% with the more than half of the events being arterial thrombi.Bleeding events are multifactorial including trauma from prolonged CPR, access site bleeding, and mucosal bleeding.Neither bleeding nor thrombosis is associated with overall in-hospital mortality.What are the Clinical implications?The high incidence of bleeding and low incidence of thrombosis with the current approach to anticoagulation which often involves antithrombotic therapy reinforces the need for a better method to risk stratify patients to better tailor anticoagulation strategies.The lack of association of bleeding and thrombotic events with mortality support a less aggressive anticoagulation strategy to minimize bleeding.
Publisher
Cold Spring Harbor Laboratory