Risk of Ischemic Stroke in Asymptomatic Atrial Fibrillation Incidentally Detected in Primary Care Compared with Other Clinical Presentations

Background It is uncertain whether stroke risk of asymptomatic ambulatory atrial fibrillation (AA-AF) incidentally detected in primary care is comparable with other clinical AF presentations in primary care or hospital. Methods The stoke risk of 22,035 patients with incident nonvalvular AF from the United Kingdom primary care Clinical Practice Research Datalink with linkage to hospitalization and mortality data was compared with 23,605 controls without AF (age- and sex-matched 5:1 to 5,409 AA-AF patients). Incident AF included 5,913 with symptomatic ambulatory AF (SA-AF); 4,989 with primary and 5,724 with nonprimary hospital AF discharge diagnosis (PH-AF and non-PH-AF); and 5,409 with AA-AF. Ischemic stroke adjusted subhazard ratios (aSHRs) within 3 years of AA-AF were compared with SA-AF, PH-AF, non-PH-AF, and no AF, accounting for mortality as competing risk and adjusted for ischemic stroke risk factors. Results There were 1,026 ischemic strokes in 49,544 person-years in patients with incident AF (crude incidence rate: 2.1 ischemic strokes/100 person-years). Ischemic stroke aSHR over 3 years showed no differences between AA-AF and SA-AF, PH-AF, and non-PH-AF groups (aSHR: 0.87-1.01 vs. AA-AF). All AF groups showed a significantly higher aSHR compared with no AF. Conclusion Ischemic stroke risk in patients with AA-AF incidentally detected in primary care is far from benign, and not less than incident AF presenting clinically in general practice or hospital. This provides justification for identification of previously undetected AF, e.g., by opportunistic screening, and subsequent stroke prevention with thromboprophylaxis, to reduce the approximately 10% of ischemic strokes related to unrecognized AF.