Trends in anticoagulant use at nursing home admission and variation by frailty and chronic kidney disease among older adults with atrial fibrillation

Background — Atrial fibrillation (AF) is relatively common among nursing home residents, and decisions regarding anticoagulant therapy in this setting may be complicated by resident frailty and other factors.

Objectives — The aim of this study was to examine trends and correlates of oral anticoagulant use among newly admitted nursing home residents with AF following the approval of direct-acting oral anticoagulants (DOACs).

Methods — We conducted a retrospective cohort study of all adults aged > 65 years with AF who were newly admitted to nursing homes in Ontario, Canada, between 2011 and 2018 (N = 36,466). Health administrative databases were linked with comprehensive clinical assessment data captured shortly after admission, to ascertain resident characteristics. Trends in prevalence of anticoagulant use (any, warfarin, DOAC) at admission were captured with prescription claims and examined by frailty and chronic kidney disease (CKD). Log-binomial regression models estimated crude percentage changes in use over time and modified Poisson regression models assessed factors associated with anticoagulant use and type.

Results — The prevalence of anticoagulant use at admission increased from 41.1% in 2011/2012 to 58.0% in 2017/2018 (percentage increase = 41.1%, p < 0.001). Warfarin use declined (- 67.7%, p < 0.001), while DOAC use increased. Anticoagulant use was less likely among residents with a prior hospitalization for hemorrhagic stroke (adjusted risk ratio [aRR] 0.65, 95% confidence interval [CI] 0.60-0.70) or gastrointestinal bleed (aRR 0.80, 95% CI 0.78-0.83), liver disease (aRR 0.78, 95% CI 0.69-0.89), severe cognitive impairment (aRR 0.89, 95% CI 0.85-0.94), and non-steroidal anti-inflammatory drug (aRR 0.76, 95% CI 0.71-0.81) or antiplatelet (aRR 0.25, 95% CI 0.23-0.27) use, but more likely for those with a prior hospitalization for ischemic stroke or thromboembolism (aRR 1.30, 95% CI 1.27-1.33). CKD was associated with a reduced likelihood of DOAC versus warfarin use in both the early (aRR 0.62, 95% CI 0.54-0.71) and later years (aRR 0.79, 95% CI 0.76-0.83) of our study period. Frail residents were significantly less likely to receive an anticoagulant at admission, although this association was modest (aRR 0.95, 95% CI 0.92-0.98). Frailty was not associated with anticoagulant type.

Conclusions — While the proportion of residents with AF receiving oral anticoagulants at admission increased following the approval of DOACs, over 40% remained untreated. Among those treated, use of a DOAC increased, while warfarin use declined. The impact of these recent treatment patterns on the balance between benefit and harm among residents warrant further investigation.