Board-level quick read · Cardiology
Atrial Fibrillation
ACC/AHA/ACCP/HRS 2023 (JACC; DOI 10.1016/j.jacc.2023.08.017) · reviewed 2026-07-21
Personal study digest, pitched for IM boards / a rising subspecialty fellow. Recommendations (COR/LOE) are the 2023 guideline’s. The “What’s changed since 2023” section is reviewer synthesis of newer evidence, each claim cited — not guideline text. Not a substitute for the full document.
In one line
Three jobs, every AF patient: modify risk factors, prevent stroke by estimated annual risk (CHA₂DS₂-VASc), and control rhythm early. The 2023 headlines: a new 4-stage classification, a push for early rhythm control, and catheter ablation upgraded to Class 1 first-line in selected patients (and in HFrEF).
Stage & classify (new in 2023)
Why it changed: the old duration-only labels emphasized AF only once diagnosed. The new stages frame AF as a progressive disease — so you act at the earliest stage (prevention, risk-factor control), not just when it’s established.
| Stage | What it means |
|---|---|
| 1 — At risk | Modifiable/nonmodifiable risk factors, no AF yet (HTN, obesity, DM, OSA, alcohol, etc.). |
| 2 — Pre-AF | Structural or electrical findings that predispose (atrial enlargement, frequent APBs, atrial flutter/other SVT). |
| 3 — AF | 3A paroxysmal · 3B persistent · 3C long-standing persistent · 3D successful ablation. |
| 4 — Permanent | Joint patient–clinician decision to stop rhythm-control attempts (a therapeutic choice, not a disease attribute). |
Duration definitions (unchanged): paroxysmal ends in ≤7 d; persistent >7 d; long-standing persistent >12 mo. Terms now abandoned: “valvular/nonvalvular,” “lone,” “chronic” AF.
Stroke prevention — the fellow’s core job
Who gets anticoagulated? Decide by estimated annual thromboembolic risk (use a validated score such as CHA₂DS₂-VASc), not by AF pattern or burden alone.
| Annual risk (CHA₂DS₂-VASc) | Action | COR |
|---|---|---|
| ≥2%/yr (≥2 men / ≥3 women) | Anticoagulation recommended | 1 |
| 1 to <2%/yr (1 men / 2 women) | Anticoagulation reasonable | 2a |
| Anticoagulant choice | DOAC over warfarin | 1 |
| Mod–severe rheumatic MS or mechanical valve | Warfarin (DOACs contraindicated) | 1 |
- LAAO (percutaneous): reasonable when there is a nonreversible contraindication to long-term OAC and CHA₂DS₂-VASc ≥2 2a (upgraded from 2019).
- Device-detected AHRE: episode ≥24 h + CHA₂DS₂-VASc ≥2 → anticoagulation reasonable 2a; 5 min–24 h + score ≥3 → may be reasonable 2b — inside shared decision-making (weigh duration × risk).
Rate vs. rhythm — and how to rate-control
- Target (no HF): resting HR <100–110 bpm, symptom-guided 2a (RACE II: lenient = strict for outcomes).
- First-line agents: beta blocker or non-DHP CCB (diltiazem/verapamil, only if EF >40%) 1. Digoxin as add-on / when the above fail 2a (keep level <1.2 ng/mL).
- HFrEF caution: do NOT give IV non-DHP CCB with moderate–severe LV dysfunction 3: Harm.
Rhythm control & ablation
- Early rhythm control: for AF diagnosed <1 year, rhythm control reduces hospitalization, stroke, and mortality 2a (EAST-AFNET 4: ~25% relative reduction in the CV composite).
- Reduced LV + persistent/high-burden AF: a trial of rhythm control to see whether AF is driving the cardiomyopathy 1.
- Catheter ablation — Class 1: (a) symptomatic AF after an antiarrhythmic failed/not tolerated/not preferred; (b) first-line in selected younger, low-comorbidity patients with symptomatic paroxysmal AF; (c) appropriate HFrEF patients (CASTLE-AF).
- Cardioversion anticoagulation: AF ≥48 h (or unknown) → 3 wks therapeutic AC or TEE to exclude thrombus before, and ≥4 wks after — regardless of stroke score (atrial stunning) 1.
Lifestyle & risk-factor modification (now a “pillar”)
- Weight loss ≥10% if overweight/obese (BMI >27) — reduces AF symptoms, burden, recurrence, progression 1.
- Exercise 210 min/wk moderate-to-vigorous 1; minimize/eliminate alcohol in patients pursuing rhythm control 1 (target ≤3 drinks/wk).
- Treat hypertension, diabetes, OSA, and stop smoking — comprehensive guideline-directed LRFM 1.
What’s changed since 2023
Reviewer synthesis of evidence published after (or alongside) the guideline — each claim cited. The 2023 ACC/AHA/ACCP/HRS document is the current U.S. guideline; ESC issued its own AF guideline in 2024.
- Device-detected subclinical AF is NOT clinical AF for anticoagulation. The 2023 AHRE recs landed just as two RCTs reported: ARTESIA (apixaban) cut stroke but raised major bleeding, and NOAH-AFNET 6 (edoxaban) showed no significant stroke reduction with more bleeding/death NEJM 2023; 10.1056/NEJMoa2303062. Pooled, OAC lowers ischemic stroke modestly but increases major bleeding — individualize by episode duration + risk Circulation 2023; 10.1161/CIRCULATIONAHA.123.067512.
- Factor XI inhibitors did NOT arrive — the lead compound failed. OCEANIC-AF was stopped early: asundexian was inferior to apixaban (stroke/systemic embolism 1.3% vs 0.4%; HR 3.79, 95% CI 2.46–5.83) despite less bleeding. DOACs remain standard; abelacimab trials ongoing NEJM 2024; 10.1056/NEJMoa2407105.
- LAAO is moving beyond “OAC-contraindicated only.” OPTION: after AF ablation, Watchman FLX was noninferior to continued OAC for death/stroke/systemic embolism and caused less non-procedural bleeding — an option even in patients who could take OAC NEJM 2025; 10.1056/NEJMoa2408308.
- ESC 2024 diverges on stroke scoring. The 2024 ESC guideline adopted CHA₂DS₂-VA (drops the sex-category point) and the AF-CARE framework; anticoagulate at CHA₂DS₂-VA ≥2 (I), consider at 1 (2a). ACC/AHA still use CHA₂DS₂-VASc with sex — expect a one-point difference in women Eur Heart J 2024; 10.1093/eurheartj/ehae176.
- Pulsed-field ablation (PFA) is now real-world. ADVENT — first RCT of PFA vs conventional thermal ablation for paroxysmal AF — showed PFA noninferior for efficacy and safety; several PFA systems gained FDA approval 2023–24, postdating the guideline NEJM 2024; PMID 38305503.
- Still current: early rhythm control (EAST-AFNET 4) and Class 1 first-line ablation for paroxysmal AF and for HFrEF are strongly reaffirmed; weight loss ≥10% and comprehensive risk-factor control remain foundational.
Recency-sensitive — confirm before quoting as current
- OPTION published 2025 (post my training cutoff; web-verified) — LAAO after ablation. NEJM 2025
- 2024 ESC AF guideline and the shift to CHA₂DS₂-VA (drops sex) — a live divergence from ACC/AHA; web-verified. Eur Heart J 2024
- PFA FDA approvals (e.g., Farapulse, PulseSelect) 2023–2024; web-verified.
Anki cards minted this run (4)
- 2023 4-stage AF classification — Stage 1 at-risk / Stage 2 pre-AF (pearl carries 3A–D and permanent).
- Cardioversion of AF ≥48 h — 3 wks AC (or TEE) before, ≥4 wks after (all patients).
- Device-detected/subclinical AF ≠ clinical AF for AC — ARTESIA / NOAH-AFNET 6.
- OCEANIC-AF — asundexian inferior to apixaban; factor XI inhibitors not ready.
Held (already in deck or capped): EAST-AFNET 4 / early rhythm control (already carded), catheter-ablation first-line (already carded), lenient rate-control target <110 (RACE II), DOAC-over-warfarin-except-MS/mechanical-valve, and weight-loss ≥10% — deduped or below the 4-card cap.
Sources: 2023 ACC/AHA/ACCP/HRS AF Guideline (JACC; DOI 10.1016/j.jacc.2023.08.017) — local verbatim capture · ARTESIA + NOAH-AFNET 6 pooled (Circulation 2023; 10.1161/CIRCULATIONAHA.123.067512) · NOAH-AFNET 6 (NEJM 2023; 10.1056/NEJMoa2303062) · OCEANIC-AF / asundexian (NEJM 2024; 10.1056/NEJMoa2407105) · OPTION (NEJM 2025; 10.1056/NEJMoa2408308) · 2024 ESC AF Guideline (Eur Heart J 2024; 10.1093/eurheartj/ehae176) · ADVENT / PFA (NEJM 2024; PMID 38305503) · EAST-AFNET 4 (NEJM 2020), CASTLE-AF (NEJM 2018), RACE II — as cited in the guideline.