(103) Effectiveness and Safety of Oral Anticoagulants among Long-term Nursing Home Residents with atrial fibrillation (AF) and Moderate to Severe Impairment in Functional Status

Background:
Long-term nursing home residents with impaired functional status are at a high risk of falling and severe bleeding and their life expectancy is often limited, so the net clinical benefit of anticoagulation for stroke prevention for atrial fibrillation (AF) is unclear.

Objectives:
To compare the effectiveness and safety of oral anticoagulants (OACs) compared to no use in long-term nursing home residents with AF and impaired functional status.

Methods:
Using the US Medicare administrative claims data linked with the Minimum Data Set (MDS) from 2013 to 2020, we identified long-term nursing home residents with non-valvular AF and ≥3 items of impaired activities of daily living (ADL). We established three comparative new users (no use in the preceding year) vs. non-users of apixaban, rivaroxaban, or warfarin respectively, using incidence density sampling (index was the dispensing date of OAC for the users and the sampling date for the non-users). Primary outcomes include ischemic stroke, major bleeding, and death. We used 1:1 propensity score (PS) matching to balance 87 baseline covariates assessed in the 365 days prior to cohort entry. We used Cox proportional hazard models to estimate adjusted hazard ratios (aHRs) and generalized linear models to estimate adjusted incidence rate differences (aRDs).

Results:
Our study cohort included 3,731 warfarin initiators (median age 81.8 years, female 66.2%, median follow-up of 125 days), 7,239 apixaban initiators (median age 82.3 years, female 67.9%, median follow-up of 158 days), and 2,887 rivaroxaban initiators (mean age 82.1 years, female 68.3%, median follow-up of 135 days). Compared to OAC non-users, use of apixaban was associated with a lower risk of ischemic stroke (aHR, [95% confidence interval] 0.5 [0.4-0.7]; aRD, -11.1 [-30.1 to 8.0]), higher risk of major bleeding (aHR, 1.8 [1.5-2.3]; aRD, 19.1 [-4.2to 42.3]), and reduced mortality (aHR, 0.8 [0.7-0.8]; aRD, -116.2 [-198.5 to -33.8]). Similarly, rivaroxaban was associated with a lower risk of ischemic stroke (aHR, 0.5 [0.3-0.8]; aRD, -9.7 [-36.4 to 17.1]), higher risk of major bleeding (aHR, 2.8 [2.1-3.8]; aRD, 46.7 [5.3 to 88.1]), and reduced mortality (aHR, 0.7 [0.7-0.8]; aRD, -113.6 [-223.0 to -4.2]). And warfarin was associated with a lower risk of ischemic stroke (aHR, 0.4 [0.3-0.7]; aRD, -12.4 [-38.6 to 13.8]), higher risk of major bleeding (aHR, 2.2 [1.7-2.8]; aRD, 40.1 [-0.9 to 81.1]), and reduced mortality (aHR, 0.9 [0.8-1.0]; aRD, -29.6 [-143.2 to 84.0]).

Conclusions: Among the US long-term nursing home residents with AF and moderate to severe impairment functional status, use of apixaban, rivaroxaban, or warfarin versus non-use of OAC was associated with a favorable net clinical benefit (aRD) with the largest benefit observed in apixaban users.