Patients with major bleeding
are commonly admitted to the ICU. A growing number are on either oral or parenteral anticoagulation
, but the impact of anticoagulation
on patient outcomes is unknown. We sought to examine this association between anticoagulation
therapy and mortality, as well as the independent effects of warfarin compared to direct oral anticoagulants.
Analysis of a prospectively collected registry (2011–2017) of consecutive ICU patients admitted with major bleeding
(as defined by International Society on Thrombosis and Haemostasis clinical criteria).
Two hospitals within a single tertiary care level hospital system.
We analyzed 1,598 patients identified with major bleeding
, of which 245 (15.3%) had been using anticoagulation
at the time of ICU admission. Of patients on anticoagulation
, 149 were using warfarin, and 60 were using a direct oral anticoagulant.
Measurements and Main Results:
The primary outcome, in-hospital mortality, was analyzed using a multivariable logistic regression model. Patients with anticoagulation
-associated major bleeding
had higher in-hospital mortality (adjusted odds ratio, 1.49; 95% CI, 1.16–1.92). Among survivors, anticoagulation
use was associated with longer median hospital length of stay, and higher mean costs
. No differences in hospital mortality were seen between warfarin- and direct oral anticoagulant-associated major bleeding
. Patients with warfarin-associated major bleeding
had longer median length of stay (11 vs 6 d; p
= 0.02), and higher total costs
than patients with direct oral anticoagulant-associated major bleeding
Among ICU patients admitted with major bleeding
, pre-admission anticoagulation
use was associated with increased hospital mortality, prolonged length of stay, and higher costs
among survivors. As compared to direct oral anticoagulants, patients with warfarin-associated major bleeding
had increased length of stay and costs
. These findings have important implications in the care of ICU patients with major bleeding