#atrial-fibrillation #stroke-prevention #anticoagulation #risk-stratification

CHA2DS2-VA Score

Definition

CHA2DS2-VA is the updated thromboembolic risk stratification score for atrial fibrillation introduced in the 2024 ESC AF Guidelines. It is derived from the prior CHA2DS2-VASc score by removing the sex category ('Sc'), reflecting evidence that female sex alone does not independently increase AF-related stroke risk. It is used to guide decisions on initiation of oral anticoagulation.

Key Concepts

Score Components

C — Congestive heart failure: 1 point
H — Hypertension: 1 point
A2 — Age ≥75 years: 2 points
D — Diabetes mellitus: 1 point
S2 — Prior stroke/TIA/arterial thromboembolism: 2 points
V — Vascular disease (prior MI, PAD, or aortic plaque): 1 point
A — Age 65–74 years: 1 point
Maximum score: 9 points (sources/AF-ESC-2024, rating: very high)

Clinical Decision Thresholds

Score ≥2: OAC is recommended (Class I/C) (sources/AF-ESC-2024)
Score = 1: OAC should be considered (Class IIa/C) (sources/AF-ESC-2024)
- This threshold now has randomized evidence (SINGLE-AF, 2026). The Class IIa recommendation rested on observational data alone, which had been conflicting. SINGLE-AF (n=1,803; DOAC vs no anticoagulation; 24 months; 18 South Korean centres) enrolled exactly this population — CHA₂DS₂-VASc 1 in men / 2 in women, equivalently CHA₂DS₂-VA = 1 in both sexes — and found the composite of stroke, systemic embolism, major bleeding or CV death reduced from 1.5% to 0.5% (difference −1.0 percentage points, 95% CI −2.0 to −0.1; P=0.03; HR 0.31, 0.10–0.94), with stroke 0.3% vs 1.1% and major bleeding essentially equal (0.3% vs 0.5%). Rated high rather than very high: the result rests on 17 events with a CI bound touching null, the primary end point was amended mid-trial, and the trial was open-label and confined to one East Asian country. (sources/single-af-nejm-2026, rating: high)
Score = 0: OAC is generally not recommended; no treatment for stroke prevention
Reassessment at periodic intervals recommended as risk factors may change over time (Class I/B) (sources/AF-ESC-2024)

Key Caveats

Temporal pattern of AF (paroxysmal, persistent, permanent) is not relevant to OAC decisions — thromboembolic risk is present regardless of AF type (Class III — do not withhold OAC based on AF type) (sources/AF-ESC-2024)
OAC recommended regardless of CHA2DS2-VA score in AF patients with hypertrophic cardiomyopathy or cardiac amyloidosis (Class I/B) (sources/AF-ESC-2024)
Device-detected subclinical AF with high CHA2DS2-VA score: DOAC may be considered (Class IIb/B), balancing stroke reduction (ARTESiA: HR 0.63 for stroke/SE) against bleeding increase (HR 1.36 for major bleeding) (sources/AF-ESC-2024)

Anticoagulant Choice

DOACs (apixaban, dabigatran, edoxaban, rivaroxaban) preferred over VKAs for eligible patients (Class I/A) (sources/AF-ESC-2024)
VKAs remain indicated for mechanical heart valves and moderate-to-severe mitral stenosis (sources/AF-ESC-2024)
Reduced DOAC doses should only be used if specific dose-reduction criteria are met — avoiding underdosing is critical (Class III against underdosing) (sources/AF-ESC-2024)

Contradictions / Open Questions

Stroke occurs in very young AF with score = 0: In a prospective cohort of 122 patients with AF before age 40 without structural heart disease, 5 patients (4.1%) suffered a stroke despite all having CHA₂DS₂-VA = 0 and none being anticoagulated — 4 events at first AF presentation. This observational finding challenges the assumption that score 0 confers negligible thromboembolic risk in very young AF patients, though the mechanism (LA thrombus vs. paradoxical embolism vs. other) was not characterized. Current guidelines provide no anticoagulation guidance for this phenotype. (sources/eoaf-riskfactor-ehj-2026, rating: medium)
Evidence base for anticoagulation was generated using CHA2DS2-VASc, not CHA2DS2-VA; the practical impact of removing sex category is a modest refinement (sources/AF-ESC-2024)
Score 1 no longer lacks RCT evidence; score 0 still does. ESC 2024 noted that patients with CHA₂DS₂-VA 0 or 1 were largely excluded from major RCTs, so evidence for OAC in this group was limited. SINGLE-AF (2026) has supplied randomized evidence at score 1 — composite 0.5% vs 1.5%; HR 0.31 (0.10–0.94) — though on 17 events with wide confidence intervals, and its authors call for confirmatory trials or pooled analyses before the magnitude is accepted. Score 0 remains untested by any randomized trial, and the recommendation against anticoagulation there is unchanged. The boundary between "consider" at 1 and "do not treat" at 0 is therefore now supported on one side by a single fragile trial and on the other by no trial at all. (sources/AF-ESC-2024; sources/single-af-nejm-2026, rating: high)
Optimal approach for patients with device-detected subclinical AF (duration, burden threshold for anticoagulation) remains undefined (sources/AF-ESC-2024)
ESC vs. AHA sex category disagreement: The 2024 ESC guideline uses CHA₂DS₂-VA (sex removed, max score 9), whereas the 2023 AHA guideline retains CHA₂DS₂-VASc (sex category retained, max score 9). The AHA rationale: female sex remains a risk modifier that interacts with other risk factors, particularly in patients aged ≥65 years. The ESC rationale: female sex alone does not independently increase stroke risk in the absence of other risk factors. The two scores produce discordant recommendations in women with score 0 (ESC: no OAC; AHA: may recommend OAC if CHA₂DS₂-VASc = 1 in the female patient, score = 0 by CHA₂DS₂-VA but 1 by CHA₂DS₂-VASc). (sources/AF-AHA-2023, sources/AF-ESC-2024)
AHA device-detected AHRE thresholds differ from ESC: AHA uses duration-stratified thresholds (≥24h = Class IIa; 5min–24h + score ≥3 = Class IIb; <5min = no OAC), while ESC applies a single Class IIb/B recommendation for high-risk patients with device-detected AF regardless of duration (sources/AF-AHA-2023, sources/AF-ESC-2024)