The belief that cirrhotic patients are “auto-anticoagulated” often results in anticoagulation therapy being withheld in these patients. We aimed to understand patterns of use of anticoagulation and to determine the risk of bleeding complications in cirrhotic patients.
Methods
We retrospectively analyzed 320 cirrhotic patients treated with anticoagulation therapy from July 15, 2014 to January 30, 2018. We performed bivariate and multivariate analyses to identify risk factors for clinically relevant bleeding. We conducted a separate analysis using propensity score matching to compare bleeding rates of a noncirrhotic cohort group on anticoagulation to anticoagulated patients with cirrhosis.
Results
Nonalcoholic steatohepatitis (47%) was the most common cause of cirrhosis, and 49% were classified as Child-Pugh class B, a mean model for end-stage liver disease score of 14 and Charlson comorbidity index of 7. Anticoagulation was initiated for atrial fibrillation/atrial flutter in 56% of patients; warfarin was used in 57% of patients and concomitant use of antiplatelet therapy in 25%. Bleeding occurred in 18%, with upper gastrointestinal bleeding (53%) being the most common source. In the propensity-matched cohort, bleeding rates were higher in cirrhotics than in control patients who were matched for baseline characteristics. In multivariate analysis of the cirrhotic patients, the presence of esophageal varices was associated with higher odds of clinically relevant bleeding.
Conclusion
Anticoagulated cirrhotic patients who have esophageal varices are at an increased risk of bleeding. We recommend that patients with cirrhosis and esophageal varices who require anticoagulation have their varices managed carefully prior to initiation of anticoagulation.
To read this article in its entirety please visit our website.
-Roula Sasso, MD, Don C. Rockey, MD
This article originally appeared in the June 2019 issue of The American Journal of Medicine.
