Anticoagulation and stroke in hospitalized patients newly diagnosed with secondary atrial fibrillation: a population-based cohort study

Abstract Background Atrial fibrillation (AF) can be triggered by acute precipitants, with reversion to sinus rhythm after the stressor resolves. This has been described as "secondary AF", to differentiate it from AF without acute provocation ("primary AF"). Objective To describe anticoagulation patterns and the risk of stroke following hospitalization with a new diagnosis of secondary AF, with comparison to a first hospitalization for primary AF. Methods We created a cohort of adults aged ≥66 years who were discharged alive from hospital after a first diagnosis of AF between April 1, 2013, and March 31, 2019. The main exposure was "secondary" AF, with "primary" AF serving as the comparator group. This was based on a validated approach utilizing the AF discharge diagnosis type. Patients were followed for 1 year. We used drug dispensation records to determine the proportion of people anticoagulated after discharge from hospital. We also identified hospitalizations for stroke. We used the cumulative incidence function to estimate the risk of stroke while not anticoagulated, while censoring on dispensation of anticoagulation and treating death as a competing risk. Inverse probability of censoring weights were used to reduce bias from informative censoring. Cause-specific hazards regression was used to estimate the hazard ratio (HR) for stroke associated with secondary AF (relative to primary AF) while accounting for differences in baseline characteristics and time-varying anticoagulation status. Results We identified 13,011 people with secondary AF and 11,065 with primary AF. People with primary AF were older (mean 79.5 yrs vs. 77.1 yrs for secondary AF), more likely to be female (58.6% vs. 42.3% for secondary AF), and more likely to have prior HF. Secondary AF was positively associated with prior bleeding, diabetes, ischemic heart disease, and peripheral vascular disease. Mortality was higher for people with secondary vs. primary AF: 1.7% (95%CI 1.5%-2.0%) vs. 0.7% (95%CI 0.6%-0.9%) respectively at 7 days, 4.2% (95%CI 3.9%-4.6%) vs. 2.5% (95%CI 2.2%-2.8%) at 30 days, and 16.1% (95%CI 15.5%-16.7%) vs. 14.6% (13.9%-15.2%) at 1-year post-discharge. The proportion of people who were dispensed anticoagulants within 7 days of discharge was lower for secondary than primary AF (26.9% vs. 68%), and this continued to be lower at one year (41.7% vs. 82.3%, p<0.001 for both comparisons). At 1 year, the risk of stroke in people who were not anticoagulated was 2.2% (95% CI 1.6-2.8%) after primary AF and 1.2% (95% CI 1.0-1.4%) after secondary AF (p< 0.001). Relative to patients with primary AF, the hazard of stroke (adjusted for baseline characteristics and time-varying anticoagulation status) was lower in patients with secondary AF (HR 0.74; CI 0.57-0.97; p= 0.03). Conclusion Secondary AF is associated with higher mortality, less anticoagulation, and a lower risk/ hazard of stroke than primary AF. These data can inform post-discharge care in people with secondary AF.Proportion anticoagulated over timeRisk of stroke without anticoagulation