Abstract 15518: What Monitoring Duration is Needed to Rule Out High Atrial Fibrillation Burden?

Introduction: Atrial fibrillation (AF) burden contributes to stroke risk. AF detection rates increase with monitoring duration, but the relationship between time to the first AF episode and AF burden has not been described. We aimed to identify AF free monitoring durations after which a given AF burden can be ruled out. Methods: We included all subjects without permanent AF (n=13,106, median age 65 years (IQR 51-74 years), 40.1% male) monitored for ≥14 days in 2021 in the USA with PocketECG (MEDICALgorithmics), a full disclosure mobile cardiac telemetry device. We graphed 14-day AF burden (% time in AF/total time) by time in days before detection of prespecified AF durations (<30s (micro-AF), ≥30s-6 minutes, ≥6 minutes-1 hour and ≥1 hour in mutually exclusive strata) and calculated the 95th percentile of 14-day AF burden by time to first AF episode of different durations. Confidence intervals were estimated using bootstrapping with 1000 replications. Results: There were 5,605 patients with micro-AF episodes, 1,788 patients with 30s-6min AF episodes, 748 patients with 6 min-1 hour AF episodes and 227 patients with ≥1 hour AF episodes (Fig 1 a-d). The overall AF burden was higher in subjects with longer episodes, but decreased by time to first episode for all episode durations, Figure 1e-h. After 3 days of monitoring the 95th percentile for overall AF burden was 0.03% (95%CI 0.02-0.03%) in patients without micro-AF, 0.1% (95%CI 0.08-0.11% in patients without ≥30s AF, 0.4% (95%CI 0.3-0.4%) without ≥6-minute AF episodes, and 1.5% (95%CI 1.2-1.8%) without ≥1 hour episodes of AF. Conclusions: The probability of a high AF burden decreases as AF-free monitoring time increases, and AF free monitoring time can be used to inform the need for longer monitoring. When no AF episodes lasting ≥1 hour have been detected after 3 days of monitoring the probability of an AF burden >2% can be effectively ruled out.