ABC | Volume 111, Nº2, August 2018

Original Article Cryptogenic Acute Ischemic Stroke: Assessment of the Performance of a New Continuous Long-Term Monitoring System in the Detection of Atrial Fibrillation Rogerio Ferreira Sampaio, Isabel Cristina Gomes, Eduardo Back Sternick Faculdade Ciências Médicas de Minas Gerais, Belo Horizonte, MG - Brazil Mailing address: Eduardo Back Sternick • Alameda do Morro 85, Condominio Olympus, Torre 4, Apto 1900, Vila da Serra. Postal Code 34006-083, Vila da Serra, Nova Lima, MG – Brazil E-mail: eduardosternick@gmail.com, eduardo.sternick@cienciasmedicasmg.edu.br Manuscript received June 16, 2017; revised manuscript January 23, 2018; accepted April 11, 2018 DOI: 10.5935/abc.20180112 Abstract Background: Long-term monitoring has been advocated to enhance the detection of atrial fibrillation (AF) in patients with stroke. Objective: To evaluate the performance of a new ambulatory monitoring system with mobile data transmission (PoIP) compared with 24-hour Holter. We also aimed to evaluate the incidence of arrhythmias in patients with and without stroke or transient ischemic attack. Methods: Consecutive patients with and without stroke or TIA, without AF, were matched by propensity score. Participants underwent 24-hour Holter and 7-day PoIP monitoring. Results: We selected 52 of 84 patients (26 with stroke or TIA and 26 controls). Connection and recording times were 156.5 ± 22.5 and 148.8 ± 20.8 hours, with a signal loss of 6,8% and 11,4%, respectively. Connection time was longer in ambulatory (164.3±15.8 h) than in hospitalized patients (148.8±25.6 h) (p=0.02), while recording time did not differ between them (153.7 ± 16.9 and 143.0 ± 23.3 h). AF episodes were detected in 1 patient with stroke by Holter, and in 7 individuals (1 control and 6 strokes) by PoIP. There was no difference in the incidence of arrhythmias between the groups. Conclusions: Holter and PoIP performed equally well in the first 24 hours. Data transmission loss (4.5%) occurred by a mismatch between signal transmission (2.5G) and signal reception (3G) protocols in cell phone towers (3G). The incidence of arrhythmias was not different between stroke/TIA and control groups. (Arq Bras Cardiol. 2018; 111(2):122-131) Keywords: Atrial Fibrillation; Stroke; Electrocardiography, Ambulatory; Cell Phone; Ischemic Attack, Transient. Introduction Atrial fibrillation (AF) is the main predictive factor of stroke. 1 Many studies have suggested that frequent short runs of atrial tachycardia (AT) or supraventricular extrasystoles (SVES) may yield early left atrial remodeling and predict AF and increased risk for stroke. 2-4 The risk for stroke is independent of clinical presentations of AF and recent studies have shown that in up to 30% of the cases, arrhythmia is diagnosed before, during or following an ischemic event. 5 The diagnosis of AF requires documentation, and the detection of paroxysmal AFmay be challenging. 6 By convention, the diagnosis of AF requires a minimumduration of 30 seconds. 7 The prognostic value of short episodes of AF is still debatable, and some authors have suggested that their occurrence may not be a benign condition. 8 Detection of paroxysmal AF has been performed by different monitoring techniques, and the importance of its early detection is due to the fact that the prompt initiation of anticoagulation significantly reduces the risk of stroke recurrence by up to 40%. 8-10 The American Heart Association and the Stroke Association recommend a long-term electrocardiographic monitoring of 30 days for the diagnosis of AF in post-cryptogenic stroke (class IIa; level of evidence C). Further evidence in support of this recommendation and for the establishment of the role of short AF episodes is still needed. 11,12 The aim of this study was to evaluate the performance of a new ambulatory electrocardiographic monitoring system using cell phone transmission in the diagnosis of AF during the acute phase of stroke or transient ischemic attack (TIA) and compared it with 24-hour Holter, and to evaluate the incidence and the type of supraventricular arrhythmias in patients with and without stroke/TIA in its acute phase. 13 Methods Subjects : patients with recent (less than 15 days of the event) stroke/TIA were enrolled based on clinical and imaging findings. Stroke was classified as cryptogenic based on the Trial of Org 10172 in Acute Stroke Treatment (TOAST). 14 Ambulatory patients without stroke/TIA, but with risk factors for these events (control group) were also included, and both groups had normal sinus rhythm at electrocardiography (ECG) and no history of AF or atrial flutter (AFL). 122