Clinical Predictors of Risk for Atrial Fibrillation

Objective: To create a risk score using clinical factors to determine whom to screen and monitor for atrial fibrillation (AF). Patients and Methods: The AF risk score was developed based on the summed odds ratios (ORs) for AF development of 7 accepted clinical risk factors. The AF risk score is intended to assess the risk of AF similar to how the CHA2DS2-VASc score assesses stroke risk. Seven validated risk factors for AF were used to develop the AF risk score: age, coronary artery disease, diabetes mellitus, sex, heart failure, hypertension, and valvular disease. The AF risk score was tested within a random population sample of the Intermountain Healthcare outpatient database. Outcomes were stratified by AF risk score for OR and Kaplan-Meier analysis. Results: A total of 100,000 patient records with an index follow-up from January 1, 2002, through December 31, 2007, were selected and followed up for the development of AF through the time of this analysis, May 13, 2013, through September 6, 2013. Mean SD follow-up time was 3106 819 days. The ORs of subsequent AF diagnosis of patients with AF risk scores of 1, 2, 3, 4, and 5 or higher were 3.05, 12.9, 22.8, 34.0, and 48.0, respectively. The area under the curve statistic for the AF risk score was 0.812 (95% CI, 0.805-0.820). Conclusion: We developed a simple AF risk score made up of common clinical factors that may be useful to possibly select patients for long-term monitoring for AF detection. a 2014 Mayo Foundation for Medical Education and Research. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). n Mayo Clin Proc. 2014;89(11):1498-1505 A trial fibrillation (AF) affects 2.3 million Americans, is associated with an increased risk of stroke, and often occurs with other comorbidities, such as congestive heart failure. Although anticoagulation can reduce the risk of stroke, anticoagulation can only be initiated if the diagnosis of AF is made. Because AF is often asymptomatic, it is frequently undiagnosed, and patients therefore do not undergo anticoagulation. As a consequence, these patients may be exposed to a higher risk of stroke. This potential adverse risk is highlighted in a study that found that subclinical AF accounted for approximately 23% of cryptogenic strokes. In addition, addressing AF may also positively affect other associated comorbidities. Atrial fibrillation, especially if paroxysmal and asymptomatic, may bemissed during clinical evaluations, electrocardiography, and periodic ambulatory telemetry monitoring. Screening for AF after an ischemic stroke has the anticipated Mayo Clin Proc. n November 2014;89( n a 2014 Mayo Foundation for Medical Education and Research. T nses/by-nc-nd/3.0/). benefit of identifying 4.4 new cases of AF for every 100 patients monitored. With long-term continuous monitoring, as is available with implantable devices, the diagnostic yield of previously undetected AF after a stroke increases significantly. Expansion of long-term monitoring to detect AF before a stroke occurs in large populations is not likely to be cost-effective or time effective, unless high-risk features for AF genesis can be determined to improve selection criteria. In addition, data from the Asymptomatic Atrial Fibrillation and Stroke Evaluation in Pacemaker Patients and the Atrial Fibrillation Reduction Atrial Pacing Trial (ASSERT) suggest that strokes oftenoccur independently of AF episodes, and as such remote monitoring as a means to detect AF early and reduce events by starting anticoagulation may be insufficient. As such, we sought to create a simple, readily accessible AF risk score according to general clinical markers to determine which patients are at highest risk of AF and should be considered for long-term 11):1498-1505 n http://dx.doi.org/10.1016/j.mayocp.2014.08.016 his is an open access article under the CC BY-NC-ND license CLINICAL ATRIAL FIBRILLATION PREDICTORS monitoring to improve earlydiagnosis and initiate anticoagulation strategies.