Adverse prognosis of incidentally detected ambulatory atrial fibrillation. A cohort study

Abstract

It was the aim of this study to determine prognosis of incidentally detected ambulatory atrial fibrillation (IA-AF) and its response to antithrombotic therapy. We performed a cohort study of 5,555 patients with IA-AF (mean age 70.9 ± 10.1, 38.4% female) and 24,705 age- and gender-matched controls without AF followed three years using UK Clinical Practice Research Datalink. We measured incidence rates of stroke, all-cause mortality, myocardial infarction, major bleeding, and effect of antithrombotic therapy. Patients with IA-AF had mean CHA2DS2VASc score 2.5 ± 1.5, 73% with score ≥2. The stroke incidence rate (IR) was 19.4 (95% confidence interval 17.1 - 21.9)/1,000 person-years vs 8.4 (7.7 - 9.1) in controls (p<0.001), mortality 40.1 (36.8 - 43.6)/1,000 person-years vs 20.9 (19.8 - 22.0) in controls (p<0.001), and myocardial infarction 9.0 (7.5 - 10.8)/1,000 person-years vs 6.5 (5.9 - 7.2) in controls (p<0.001). IRs of all endpoints increased with age. Oral anticoagulant ± antiplatelet therapy received by 51.0% in year following IA-AF was associated with adjusted hazard ratio (HR) of 0.35 (0.17 - 0.71) for stroke, and 0.56 (0.36 - 0.85) for death compared to no therapy, while antiplatelet treatment was associated with a non-significant reduction of HR: 0.81 (0.51 - 1.29) for stroke, and 0.80 (0.55 - 1.15) for death, though both carried a similar small non-significant adjusted excess IR of major bleeding. In conclusion, asymptomatic AF detected incidentally is associated with a significant adverse effect on stroke and death, with reduction in both associated with oral anticoagulant but not antiplatelet treatment. This provides justification to assess cost-effectiveness of community screening to detect unknown AF.

Figure 1 Generation of incidentally detected ambulatory AF cohort

. AF: atrial fibrillation; CHF: congestive heart failure or use of cardiac glycosides; GP: general practitioner; MI: myocardial infarction; TIA: transient is-chaemic attack; OAC: oral anticoagulant, defined by prescription of vitamin K antagonist (VKA), read medical codes or mentioning in clinical notes of VKA use, or ≥2 INR tests in 60 days prior to index AF. *Study endpoint in the 14 days before the first AF record, within hospital stay, or within 7 days following AF/hospital discharge.

Figure 2 Crude and adjusted cumulative risk of stroke and all cause mortality in IA-AF and non-AF cohort (A), and crude and adjusted cumulative risk of major bleeding and MI in IA-AF and non-AF cohort (B)

. IA-AF: incidentally detected ambulatory AF. Adjusted for age, gender, smoking, hypertension, diabetes, previous TIA/stroke, coronary artery disease, peripheral artery disease, previous bleeding, cancer, AP therapy in previous year and Charlson index (0,1,2,3,4,5+)

Figure 3 Crude and adjusted cumulative incidence of stroke by antithrombotic treatment

. OAC: oral anticoagulant; AP: antiplatelets; First OAC and AP treatment episodes start in the first year after initial AF diagnosis. No treatment includes the person time prior to the first OAC/AP treatment episode or time until the end of observation (if patients remain un- treated). Cumulative incidence adjusted for age, gender, smoking, hypertension, diabetes, previous TIA/stroke, coronary artery disease, peripheral artery disease, previous major bleeding, cancer, AP therapy in previous year and Charlson index (0,1,2,3,4,5+).