2024 ACC Expert Consensus Decision Pathway on Practical Approaches for Arrhythmia Monitoring After Stroke

Authors, Journal, Affiliations, Type, DOI

• Writing Committee: Michael T. Spooner, Steven R. Messé, Seemant Chaturvedi, Monika M. Do, Ty J. Gluckman, Janet K. Han, Andrea M. Russo, Sherry J. Saxonhouse, Newton B. Wiggins
• Journal: JACC, Volume 85, No. 6, February 18, 2025: pages 657–681
• Organization: American College of Cardiology Solution Set Oversight Committee
• Type: Expert Consensus Decision Pathway (ECDP)
• Approved: December 2024
• DOI: https://doi.org/10.1016/j.jacc.2024.10.100

Overview

This 2024 ACC Expert Consensus Decision Pathway establishes the first comprehensive, structured clinical framework for arrhythmia (primarily AF) monitoring after ischemic stroke. It stratifies patients into three populations based on stroke etiology, defines monitoring strategies and duration for each, and reviews the range of medical-grade and consumer monitoring technologies. The key anticoagulation trigger after AF detection is ≥5 minutes duration combined with CHA₂DS₂-VASc ≥3 or equivalent stroke risk; AF <5 minutes should not trigger anticoagulation. Risk scoring tools (CHASE-LESS, AS5F performing best) help identify which patients warrant long-term implantable monitoring. While prolonged monitoring reliably detects more AF, definitive RCT evidence that this translates to reduced stroke outcomes remains pending.

Keywords

Arrhythmia monitoring, atrial fibrillation, stroke, cryptogenic stroke, ESUS, embolic stroke of undetermined source, implantable cardiac monitor, cardiac monitoring, anticoagulation, CHA₂DS₂-VASc, CHASE-LESS, wearable devices, consumer monitoring, secondary stroke prevention

Key Takeaways

1. Introduction and Background

• Stroke is a leading cause of death and disability worldwide; AF causes approximately 1 in 7 strokes and raises stroke risk ~5-fold
• AF is often asymptomatic; traditional brief ECG methods frequently miss transient AF episodes
• Longer monitoring duration detects more AF, but AF detected long after stroke is less likely to be the proximate cause of that event
• Wide variability in practice exists; this ECDP provides guidance stratified by stroke etiology

2. Definitions

• AF burden: Time spent in AF during a monitoring period (absolute duration of longest paroxysm or proportional duration relative to total monitoring)
• AF detected after a stroke: Detection of AF post-stroke/TIA; associated with lower prevalence of risk factors and lower risk of recurrent stroke than AF detected prior to stroke
• Cryptogenic stroke (CS): Ischemic stroke without identified etiology despite thorough assessment including head/neck arterial imaging, echocardiography, extended cardiac rhythm monitoring, and key labs
• ESUS: Non-lacunar cryptogenic ischemic stroke without proximal arterial stenosis or cardioembolic source; ~17% of all ischemic strokes; recurrence 4–5%/year
• Subclinical AF: Asymptomatic AF detected by intracardiac, implantable, or wearable monitors; confirmed by intracardiac electrogram or ECG review
• Medical-grade monitor: Regulatory-cleared, prescription-required device reimbursable for cardiac rhythm diagnosis
• Consumer-grade monitor: Off-the-shelf device; may have FDA clearance but purchased without prescription; output self-reported to purchaser

3. Pathway — Three Patient Populations

3.1 Stroke of Presumed Cardiac Origin (Already on Anticoagulation)

• Conditions requiring long-term anticoagulation: AF, LA/LV thrombus, post-ablation LA appendage isolation, rheumatic mitral stenosis, mechanical valves, cardiac amyloidosis, LV non-compaction, severely reduced EF (especially with anterior akinesis/dyskinesis), LVAD
• Role for monitoring is LIMITED — the anticoagulation indication already exists; monitoring is only warranted if there is a possibility of stopping anticoagulation (e.g., postoperative AF, reversible AF cause, LV thrombus with transient wall motion abnormality) or if other treatment decisions depend on arrhythmia detection

3.2 Stroke from Presumed Small- or Large-Vessel Disease

• Short-term monitoring (≤7 days): AF identified in 2.2% (large-vessel) and 2.4% (small-vessel) — lower than CS (9.2%)
• STROKE-AF trial: ICM vs usual care in 496 patients; AF detected in 12.1% (ICM) vs 1.8% (usual care); median time to first AF episode 99 days; heart failure (HR 5.06) and LA enlargement (HR 3.32) were strongest predictors
• Post-hoc analysis: AF rates similar in large-vessel (11.7%) vs small-vessel disease (12.6%); patients with HF and/or LA enlargement had 23.4% AF detection at 12 months vs 5.0% without
• Cost-effectiveness: ICM equated to $37,760/QALY vs SoC overall; improved to $22,016/QALY when risk criteria selected highest-risk patients
• Recommendation: 2–4 weeks monitoring is reasonable; ICM may be considered initially or in select patients with higher AF risk (HF history, LA enlargement, elevated CHASE-LESS score) if external monitoring is unrevealing

3.3 Ischemic Stroke with Unclear Source (Cryptogenic/ESUS)

• AF identified in up to 30% of CS patients on long-term monitoring
• Key caveats: (1) unclear minimal AF burden required to increase stroke risk; (2) brief AF episodes (<5–6 min) have unknown stroke risk; (3) temporal discordance between AF and stroke observed; (4) subclinical AF ≥5 min occurs at similar rates in older adults with and without stroke history
• ESUS anticoagulation trials (NAVIGATE-ESUS, RE-SPECT ESUS, ARCADIA, ATTICUS) all failed to show anticoagulant superior to antiplatelet — unselected anticoagulation is not beneficial; AF identification is therefore needed to select patients for anticoagulation
• 2021 AHA/ASA guideline: long-term cardiac monitoring Class 2a/B in CS (with or without ESUS)
• 2023 ACC/AHA/ACCP/HRS AF guideline: ICM for stroke/TIA of undetermined cause reasonable (Class 2a/B); OAC for AHRE ≥24h + CHA₂DS₂-VASc ≥2 (Class 2a/A); OAC for AHRE 5 min–24h + CHA₂DS₂-VASc ≥3 may be reasonable (Class 2b/B); OAC not recommended for <5 min (Class 3/B)
• Recommendation: Prolonged cardiac monitoring (2–4 weeks) should be offered if candidate for anticoagulation; 24–48h monitoring has limited yield; ICM in select higher-risk patients if initial external monitoring unrevealing; anticoagulation for AF ≥5 min + CHA₂DS₂-VASc ≥3 or equivalent; avoid anticoagulation for AF <5 min without other indications

4. Post-Stroke AF Risk Scoring

• Taiwanese Stroke Registry validation: AS5F and CHASE-LESS showed highest correlation with future AF (C-statistics 0.730 and 0.741 respectively)
• Machine learning algorithm: C-statistic 0.77 for poststroke AF prediction
• ECG/echo markers of atrial cardiomyopathy (ARCADIA criteria): P-wave terminal force in V₁ >5,000 mcV·ms, NT-proBNP >250 pg/mL, or LA diameter index ≥3 cm/m²

5. Medical-Grade Monitoring Devices

5.1 PPG-Based Devices

• Optical sensors detect blood volume changes; software analyzes pulse wave variation
• Subject to motion artifact; provide intermittent (not continuous) monitoring; do not generate ECG
• Clinician overread and ECG confirmation required
• Only FDA-cleared and CE-marked PPG medical-grade AF detection app: FibriCheck (Belgium) — 96% sensitivity, 97% specificity; used in >6,000 European patients; requires prescription in the US; not specifically studied post-stroke

5.2 ECG-Based Wearable Devices

• Single-use, self-adhesive ambulatory ECG patches; up to 30 days continuous monitoring
• Mobile cardiac telemetry options offer near real-time detection
• Key trials proving superiority over short-term monitoring:
  • EMBRACE: 30-day monitor vs 24h Holter — AF detection 16.1% vs 3.2% (P<0.001)
  • Find-AF RANDOMISED: 10-day ECG vs 24h Holter — AF detection 14% vs 5% (P=0.002)
  • EPACS: 14-day patch vs 24h Holter — AF detection 16.3% vs 2.1% (P<0.02)
  • SCREEN-AF: 14-day patch vs SoC in age ≥75 + CHA₂DS₂-VASc ≥2 — AF detection 5.3% vs 0.5% (P<0.001)
• Single-use 14-day patch (iRhythm Zio): highest diagnostic yield, lowest need for repeat testing vs Holter/event recorders/MCT

5.3 Implantable Cardiac Monitors (ICMs)

• Battery life 3–5.5 years; continuous single-lead ECG; minimally invasive outpatient implantation
• AF detection yield: ~12–16% at 1 year, up to 22.8% at 24 months, 28.5% at 36 months
• False-positive rates up to 55% (noise, PACs/PVCs); AI algorithms improved PPV from 53.9% to 74.5–98.5%
• Remote reprogramming reduces provider alarms significantly (0.13→0.03/day for Medtronic LINQ II; 0.15→0.01/day for Boston Scientific LUX-Dx)
• Key ICM trials:
  • CRYSTAL-AF (2014): CS patients; ICM vs SoC — AF detection 8.9% vs 1.4% at 6 months (HR 6.4; P<0.001); 30.0% vs 3.0% at 36 months (HR 8.8; P<0.001); 74% AF episodes asymptomatic
  • PER DIEM (2021): 300 ischemic stroke patients; ICM vs external loop recorder — AF detection 15.3% vs 4.7% at 12 months (P=0.003)
  • STROKE-AF: ICM vs usual care in large/small-vessel stroke — AF detection 12.1% vs 1.8% (HR 7.4; P<0.001)
  • LOOP trial: 6,004 patients aged >70 with AF risk factors (no prior AF); ICM vs usual care — 3× more AF diagnosed (31.8% vs 12.2%); 3× more anticoagulation; NO reduction in stroke/SE (HR 0.80; P=0.11); per-protocol analysis with ≥3yr follow-up: significant reduction (HR 0.75; P=0.047)

5.4 ICMs and Anticoagulation Outcomes

• Meta-analysis: long-term rhythm monitoring associated with lower recurrent stroke in observational studies (RR 0.29) but NOT in RCTs (RR 0.72)
• LOOP trial: despite 3× more AF diagnosis and anticoagulation, stroke rates not significantly different — may be underpowered due to higher-than-expected AF detection in control group
• ARTESIA trial (meta-analysis with NOAH-AFNET 6): OAC reduces stroke/SE (RR 0.68; CI 0.50–0.92) in device-detected AF, even without prior stroke
• Ongoing trials: SAFFO (secondary prevention), Find-AF2, ARTESIA sub-analysis in prior stroke patients

6. Consumer Monitoring Devices

6.1 Available Devices

• Smartwatches with FDA 510(k) clearance for AF detection: Apple Watch (Series 4–9), Fitbit Sense, Samsung Galaxy Watch 2/3, Google Verily, Withings ScanWatch
• Technology: PPG (passive, intermittent sampling every ~2 hours) + ECG (spot-check, active user-initiated)
• Handheld ECG: KardiaMobile (AliveCor, US); Zenicor (Europe) — for occasional rhythm checks

6.2 Key Limitations

• No continuous monitoring; require charging; not worn 24h/day
• PPG algorithms only detect AF ≥30 min episodes; cannot detect flutter (regular rhythm)
• Real-world sensitivity/specificity lower than manufacturer reports when devices compared head-to-head
• No prospective studies demonstrating clinical outcome improvement post-stroke
• ECG outputs can be suboptimal; require physician interpretation; false alarms cause anxiety and unnecessary testing
• Healthcare disparities: cost and cognitive demands limit access

6.3 When Consumer Devices May Be Helpful

• Patients who cannot tolerate external monitors and refuse ICM
• Limited access to medical-grade monitors
• Long-term supplementary surveillance after medical-grade monitoring complete
• Arrhythmia trigger identification and risk-factor modification support

6.4 EHR Integration

• Requires 3-step pipeline: consumer monitor → network interface → cloud analytics platform → EHR via API
• 21st Century Cures Act mandates FHIR-based interoperability for third-party apps
• Current reality: most devices rely on patient-printed PDF uploads; seamless transmission limited

Limitations of the Document

• No randomized trial data confirming that extended AF monitoring post-stroke reduces recurrent stroke events (RCT evidence remains absent; observational data supports benefit)
• Many ESUS anticoagulation trials excluded patients with known AF and required only ≤24h monitoring for enrollment, limiting generalizability
• AF detected post-stroke has lower stroke risk than pre-existing AF; causal relationship between post-stroke AF detection and prior stroke is uncertain
• Consumer device performance studies predominantly conducted in patients aged 40–50s; performance in older post-stroke populations uncertain
• LOOP trial underpowered due to unexpectedly high AF detection in control group; whether routine ICM prevents stroke in non-cryptogenic stroke populations remains unproven
• Machine learning–based AF detection improves ICM PPV significantly, but false-positive rates remain clinically significant
• Cost-effectiveness of ICM varies greatly depending on patient risk strata; routine use in low-risk patients not cost-effective

Key Concepts Mentioned

• concepts/Poststroke-AF-Monitoring — central framework of this source; stratified three-population approach
• concepts/Subclinical-AF — detailed anticoagulation thresholds for device-detected AF (5 min/24h/CHA₂DS₂-VASc)
• concepts/Atrial-Cardiomyopathy — ARCADIA trial criteria for atrial cardiomyopathy as AF risk predictor post-stroke
• concepts/Inter-Atrial-Block — P-wave markers as atrial cardiomyopathy indicators predictive of post-stroke AF
• concepts/CHA2DS2-VA — primary stroke risk score applied in post-stroke anticoagulation thresholds

Key Entities Mentioned

• entities/Atrial-Fibrillation — primary arrhythmia of concern; cause of ~1 in 7 strokes

Wiki Pages Updated

• Created wiki/sources/arrhythmia-monitoring-stroke-acc-2024.md (this file)
• Created wiki/concepts/Poststroke-AF-Monitoring.md
• Updated wiki/concepts/Subclinical-AF.md — added post-stroke AF monitoring section and LOOP/CRYSTAL-AF context
• Updated wiki/wikiindex.md
• Updated wiki/sourceindex.md
• Updated log.md