Detection of atrial fibrillation using an implantable loop recorder following cryptogenic stroke: implications for post-stroke electrocardiographic monitoring.

Abstract

Approximately 10-40% of strokes are cryptogenic (CS). Long-term electrocardiographic (ECG) monitoring has been recommended in these patients to search for atrial fibrillation (AF). An unresolved issue is whether ambulatory ECG (AECG) monitoring should be performed first, followed by an implantable loop recorder (ILR) if AECG monitoring is non-diagnostic, or whether long-term ECG monitoring should be initiated using ILRs from the onset. The purpose of this study was to assess, using an ILR, AF incidence in the first month after CS. We enrolled consecutive CS patients referred for an ILR. All patients were monitored via in-hospital continuous telemetry from admission until the ILR (Medtronic [Minneapolis, MN] LINQ™) was implanted. The duration and overall burden of all AF episodes ≥ 2min was determined. The cohort included 343 patients (68 ± 11years, CHA2DS2-VASc 3.5 ± 1.7). The time between stroke and ILR was 3.7 ± 1.5days. During the first 30days, only 18 (5%) patients had AF. All episodes were paroxysmal, lasting from 2min to 67h and 24min. The median AF burden was 0.85% (IQR 0.52, 10.75). During 1year of follow-up, 67 (21%) patients had AF. The likelihood of AF detection by an ILR in the first month post-CS is low. Thus, the diagnostic yield of 30days of AECG monitoring is likely to be limited. These data suggest a rationale for proceeding directly to ILR implantation prior to hospital discharge in CS patients, as many have AF detected during longer follow-up.