(087) Anticoagulant Deprescribing After An Emergency Department Visit Due To Gastrointestinal Bleeding Among Older Patients Using Multiple Medications

Background: Polypharmacy increases the risk of adverse drug events (ADE) in older adults. Anticoagulant use is common in this population and has been associated with drug interactions increasing the risk of ADEs, including gastrointestinal (GI) bleeding. A common recommendation after an ADE is the deprescribing of the drug causing it.

Objectives: To characterize the deprescribing events following Emergency Department (ED) visits for GI bleeding for an older population at higher risk of ADEs, using an anticoagulant and at least two other concomitant medications.

Methods: Using the MarketScan Medicare Supplement data between 2012 and 2020, we created a cohort including 8175 Medicare beneficiaries 65 years and older who had at least one ED visit due to GI bleeding and were using one anticoagulant and at least 2 concomitant medications. Deprescribing was considered present if an anticoagulant discontinuation, switching, or dose reduction was observed after the ED visit. Deprescribing events were characterized using descriptive statistics, and logistic regressions were performed to identify potential predictors of anticoagulant discontinuation, switching, or dose reduction, as well as anticoagulant deprescribing as a composite outcome.

Results: 57% of the cohort had a deprescribing event after the ED visit. Anticoagulant discontinuation was the most frequent among those who had a deprescribing event (84%). Having dementia, liver disease, and a higher Charlson comorbidity index were identified as potential predictors of anticoagulant deprescribing. Predictors for the individual deprescribing events were different for each case.

Conclusions: After an ED visit due to GI bleeding, more than 50% of those patients 65 years and older who are concomitant users of anticoagulants and at least two other medications had a deprescribing event. There are differences in the potential predictors for anticoagulant discontinuation, switching, dose reduction, and deprescribing as a composite outcome, indicating variations in real-world clinical practice and areas for further research.