Aun R. Shah, MD, MRCP1, Osama Hamid, MD, MRCPI2, Muhammad U. Butt, MD, MS3, Josue Davila, MD4, Erin Walsh, DO5, Sara Kamionkowski, DO5, Sara Ghoneim, MD3, Rubab Ali, MBBS6, Imad Asaad, MD5; 1University of Nebraska Medical Center, Omaha, NE; 2Cleveland Clinic Foundation, Cleveland, OH; 3Case Western Reserve University, Metrohealth Medical Center, Cleveland, OH; 4MetroHealth, Cleveland, OH; 5MetroHealth Medical Center, Cleveland, OH; 6Shifa College of Medicine, Islamabad, Islamabad, Pakistan
Introduction: The safety and efficacy of Direct Oral Anticoagulants (DOAC) when compared to warfarin or heparin products, has been established firmly over the past decade. However, the use of DOACs in patients with cirrhosis remains controversial, as the studies demonstrating safety of DOACs excluded these patients. This large population-based study tries to determine the predictors of Intracranial and Gastrointestinal bleeding in cirrhosis with use of Direct Oral Anticoagulants. Methods: We reviewed data from a large commercial database (Explorys IBM) that aggregates electronic health records from 26 large nationwide healthcare systems. Using systemized nomenclature of clinical medical terms (SNOMED CT), we identified adults with liver cirrhosis, intracranial and gastrointestinal bleeding. Risk factors known to be associated with bleeding such as alcohol abuse, NSAIDs, CKD and advanced age were collected. Bleeding risk with DOAC use was compared with coumadin. Univariable and multivariable logistic regression analyses were performed to investigate strongest predictors. Results: Out of 61.4 million active adult patients in the database, 468,580 patients (0.61%) had documented cirrhosis. 9.14 % and 3.14 % of the cirrhotic patients were on DOACs and coumadin, respectively. In cirrhotic patients, there was higher composite risk of ICH or GIB if they were on DOAC (35.47% vs 21.24 % OR 2.04 [2.01–2.07]), used NSAIDs (31.15% vs 23.63% OR 1.46 [1.44–1.48]), had CKD (32.61% vs 25.04 %, OR 1.45 [1.42–1.47]), or abused alcohol (36.22% vs 23.57%, OR 1.84 [1.81–1.87]). Risk of bleeding was higher with DOACs as compared to coumadin (35.47% vs 22.81%: p< 0.001). In Multivariable model CKD modified composite bleeding risk of ICH or GIB after adjustment for other risk factors. A cirrhotic patient on DOAC had higher risk of bleeding with CKD (OR 2.29 [(2.22-2.37)]) than without CKD (OR 1.66 [(1.63–1.69)]). This interaction was found to be statistically significant. Similar trend was noticed with use of coumadin. Discussion: The risk of major bleeding in patients with cirrhosis was significantly higher with use of DOACs as compared to warfarin. This risk was highest in patients with CKD. Selection of anticoagulation remains challenging in this population and there is a need for prospective studies to help determine safe and effective anticoagulation options for these patients.
Figure 1: Multivariable Model showing predictors of ICH or GIB
Table 1: Characteristics of cirrhotic patients
Disclosures: Aun Shah indicated no relevant financial relationships. Osama Hamid indicated no relevant financial relationships. Muhammad Butt indicated no relevant financial relationships. Josue Davila indicated no relevant financial relationships. Erin Walsh indicated no relevant financial relationships. Sara Kamionkowski indicated no relevant financial relationships. Sara Ghoneim indicated no relevant financial relationships. Rubab Ali indicated no relevant financial relationships. Imad Asaad indicated no relevant financial relationships.
