To Treat or Not to Treat: Comparing Oral Anticoagulant Outcomes Among U.S. Nursing Home Residents with Atrial Fibrillation

OBJECTIVES:

Evidence on the effectiveness and safety of oral anticoagulants (OACs) in older nursing home residents with atrial fibrillation (AF) is limited, as this population is excluded from most clinical trials.

METHODS:

Using Medicare administrative data and Minimum Data Set (MDS) 3.0 assessments, we conducted a retrospective cohort study of US nursing home residents with AF between July 1, 2011 and December 31, 2016. The user group consisted of residents who initiated OACs on or after July 1, 2011 (n=48,093). The primary effectiveness outcome was ischemic stroke or systemic embolism while the primary safety outcome was intracranial hemorrhage or extracranial bleeding. Mortality was a secondary outcome and competing risk for the primary outcomes. Cox proportional hazard models estimated crude and multivariable adjusted hazard ratios (aHRs).

RESULTS:

The median age was 86 years and nearly all (99.6%) of the study population had CHA₂DS₂-VASc risk scores for ischemic stroke of 2 or higher. Among OAC users and non-users, the rates of ischemic stroke or systemic embolism were 1.1% and 2.5% (per person-year) while the rates of intracranial or extracranial hemorrhage were 2.7% and 2.2%, respectively. The all-cause death rate was 42.0% and 91.3% per person-year among OAC users and non-users. In Cox regression models, OAC use was associated with a lower risk of ischemic stroke or systemic embolism (aHR: 0.65; 95% Confidence Interval (CI): 0.59- 0.71), a higher risk of intracranial or extracranial hemorrhage (aHR: 1.93; 95% CI: 1.82- 2.06) and a lower risk of death (aHR: 0.72; 95% CI: 0.71- 0.73).

CONCLUSIONS:

The lower risk of ischemic events and death, but higher risk of bleeding among OAC users, suggests the potential net benefit of OAC treatment in the nursing home setting. Estimating a resident's risk for both safety and effectiveness outcomes and weighing resident and caregivers' preferences would strengthen shared decision-making processes regarding OAC use.