Effect of direct‐acting oral anticoagulants (DOACs) on bleeding and blood product usage in cardiac surgery compared to warfarin and controls. (3rd March 2020)

• Record Type: Journal Article
• Authors: Little, Christopher; Szydlo, Richard; Aw, TC; Laffan, Mike; Arachchillage, Deepa R. J.
• Is Part Of: British journal of haematology. Volume 190:Number 2(2020)
• Page Start: 284
• Page End: 293
• Publication Date: 2020-03-03
• Subjects: cardiac surgery -- direct‐acting oral anticoagulant -- warfarin -- bleeding -- thrombosis; Hematology -- Periodicals; Blood -- Diseases -- Periodicals; 616.15
• DOI: 10.1111/bjh.16521
• Languages: English
• ISSNs: 0007-1048

Abstract

Summary: In this retrospective, single‐centre, observational study, we assessed (i) use of anticoagulant and antiplatelet (AP) therapy, (ii) the duration of direct‐acting oral anticoagulant (DOAC) discontinuation, (iii) renal function and (iv) PT and APTT as predictors of bleeding and blood product usage; in adults (>18 years) undergoing major cardiac surgery from 01.01.2015 to 31.12.2018. Comparisons were made between each treatment group (warfarin, DOAC and DOAC + AP) and untreated controls, and between warfarin and DOAC. A total of 2928 patients were included for analysis. Median (range) of DOAC discontinuation prior to surgery was five days (1–22) for DOAC and five days (2–7) for DOAC + AP. There were no differences in bleeding between anticoagulant groups versus control, or DOAC versus warfarin. There were no differences in blood product use between DOAC and warfarin patients. The duration of DOAC discontinuation but not the creatinine clearance influenced bleeding and blood products use. Thrombosis occurred in 0·7% and 3·1% in controls and patients on warfarin respectively ( P = 0·099) with none among patients on DOAC or DOAC + AP. The PT/APTT had no predictive value. Median five‐day discontinuation of DOAC +/− AP irrespective of renal function prevents an increase in bleeding compared to patients on warfarin or controls with no increase in thrombosis.