Incidence of atrial fibrillation in an opportunistic screening campaign in a general practitioner setting: a prospective population study

Abstract

Abstract Introduction The detection of atrial fibrillation (AF) largely depends on the method used, thus the reported true rate of incident AF is controversial. The detection of symptomatic AF is straight forward in most cases, while it is unclear which strategies provide the best results for the detection of silent AF. Traditional methods like pulse palpation have a low accuracy and pulse guided ECG confirmation is not deemed cost-effective. Purpose The aim of the study was to detect the incidence symptomatic AF, using ECG, and opportunistic screening of silent AF, using an AF-detection capable automatic sphygmomanometer, in patients ≥65 years old in the general practitioner (GP) setting. Methods This was a population-based prospective cohort study of unselected general population referred for routine visits in 93 randomly selected GPs. Patients of both sexes of ≥65 years without previous diagnosis of AF were considered for the study. Each patient if symptomatic was directly referred to perform an ECG, otherwise if asymptomatic, underwent blood pressure monitoring with an AF-detection capable automatic sphygmomanometer followed by an ECG in case the device signaled AF. Results The final population comprised of 14987 individuals of ≥65 years old. The follow up extended to 16838 patient-years. The overall incidence of AF was 2.25% patient-years (95% CI 2.03–2.48). AF incidence was significantly higher in male (male/female ratio 1.29), overweight (BMI: 28.2 kg/m2 vs 27.0 kg/m2), older (79.1 vs 75.4 years old) individuals, with 44.9% belonging to the ≥80-year-old group. Other risk factors for incident AF were higher systolic BP measurements, history of stroke/TIA, congestive heart failure and chronic kidney disease. On multivariate analysis, age (annual increment), mitral valve disease, previous stroke and the number of annual visits (>10), were independent predictors of incidental AF. About 25% of cases were silent AF (0.56% patient-years, 95% CI 0.46–0.69) and were identified using the AF-detection capable automatic sphygmomanometer. Independent predictors of silent AF were age (per year increment), overweight (BMI >28 kg/m2) and the number of visits (>10) in the GPs' office. Conclusions We found a higher than previously reported incidence of AF possibly due to capturing silent AF. This simple protocol, using an AF-capable automatic sphygmomanometer, might be feasible and easily implemented on the routine GP care where high rates of coverage can be achieved. Funding Acknowledgement Type of funding source: Public grant(s) – National budget only. Main funding source(s): Veneto Region Italy