Effects of shorter activated coagulation time on hemorrhage during venoarterial extracorporeal membrane oxygenation

抄録

Background: Extracorporeal membrane oxygenation (ECMO) has been used for decades, but optimal anticoagulation control remains unknown. This study aimed to compare shorter target activated coagulation time (ACT) criteria with the usual target ACT criteria in terms of complications.

Methods: We retrospectively identified patients who received ECMO between 1 January 2013 and 31 December 2018 in an acute tertiary care hospital. Patients were divided into two groups: (I) those whose target ACT was 160–180 sec and (II) those whose target ACT was 180–220 sec. Cox proportional hazard models and Fine–Gray models adjusted for propensity score to account for the competing risk of death were used to compare the incidence of hemorrhage during ECMO between the groups.

Results: We identified 74 patients, 25 of whom were managed with target ACT 160–180 sec, and 49 of whom were managed with target ACT 180–220 sec. In crude analysis, the proportions of patients with hemorrhage in the under 180-sec group were significantly more than those in the over 180-sec group [60.0% (15/25) vs. 28.6% (14/49), p=0.009]. Shorter target ACT was not associated with hemorrhage during ECMO in either Cox regression (hazard ratio, 1.67; 95% confidence interval, 0.59–4.80; p=0.336) or Fine–Gray model (hazard ratio, 1.58; 95% confidence interval, 0.64–3.91; p=0.324).

Conclusions: The shorter ACT target was not associated with improved hemorrhage and inappropriate coagulation than the usual target ACT criteria. This study is the first to compare the ACT target of patients with ECMO.