Computing Thromboembolic Risk Per Year in Atrial Fibrillation
Thromboembolic risk per year is computed using the CHA₂DS₂-VASc scoring system, which assigns points to clinical risk factors and translates directly to an annualized stroke/thromboembolism rate ranging from 0% (score 0) to >8% (score ≥6) per year. 1
The CHA₂DS₂-VASc Scoring System
The calculation assigns points as follows 1:
- Congestive heart failure: 1 point
- Hypertension: 1 point
- Age ≥75 years: 2 points
- Diabetes mellitus: 1 point
- Stroke/TIA/thromboembolism (prior): 2 points
- Vascular disease (prior MI, peripheral artery disease, aortic plaque): 1 point
- Age 65-74 years: 1 point
- Sex category (female): 1 point
The maximum possible score is 9 points 1.
Annualized Risk Rates by Score
The adjusted annual thromboembolic event rates correspond directly to the CHA₂DS₂-VASc score 1:
- Score 0: 0% per year 1
- Score 1: 1.3% per year (range 0.6-1.4% across validation cohorts) 1
- Score 2: 2.2% per year 2
- Score 3: 3.2% per year 2
- Score 4: 4.0% per year 2
- Score ≥6: >8% per year 1
Important Caveats About Score 1
Not all risk factors within a CHA₂DS₂-VASc score of 1 carry equal risk. 3 Among patients with a score of 1, age 65-74 years confers the highest thromboembolic risk (hazard ratios 1.9-3.9), while vascular disease tends to show lower event rates 3. Female gender alone (without other risk factors) should not count toward risk stratification, as it only increases risk when combined with age ≥75 years 1.
Alternative Risk Computation Methods
By AF Type
AF pattern modifies baseline thromboembolic risk 1:
- Paroxysmal AF: 1.50% per year (95% CI: 1.23-1.76%) 1
- Non-paroxysmal AF: 2.17% per year (95% CI: 1.81-2.53%) 1
- Relative risk increase: Non-paroxysmal AF carries 1.355-fold higher risk (95% CI: 1.169-1.571, P<0.001) 1
Historical CHADS₂ Score (Less Preferred)
The older CHADS₂ system assigns 1:
- Congestive heart failure: 1 point
- Hypertension: 1 point
- Age ≥75 years: 1 point
- Diabetes: 1 point
- Stroke/TIA (prior): 2 points
The CHADS₂ score has inferior predictive accuracy (c-statistic 0.68) compared to CHA₂DS₂-VASc (c-statistic 0.82) 1, and fails to identify many intermediate-risk patients who would benefit from anticoagulation 2.
Risk in Untreated Patients
For patients with AF not receiving anticoagulation, baseline stroke rates vary by population 1:
- General AF population: 5% per year average, increasing from 1.5% (age 50-59) to 23.5% (age 80-89) 1
- Primary prevention cohorts: 2.0-2.7% per year 1
- With aspirin therapy alone: 2.2-2.7% per year (only 19% risk reduction vs placebo) 1
Clinical Application Algorithm
Calculate CHA₂DS₂-VASc score by summing all applicable risk factor points 1
Determine annualized risk using the score-to-risk conversion above 1
Consider AF pattern as a risk modifier: multiply baseline risk by 1.35 if non-paroxysmal AF 1
Account for individual risk factor hierarchy in score 1 patients: prioritize age 65-74 as highest risk 3
Compare against bleeding risk using HAS-BLED score to determine net clinical benefit of anticoagulation 1
The threshold for anticoagulation consideration is ≥1% annual thromboembolic risk, which corresponds to CHA₂DS₂-VASc ≥1 in males or ≥2 in females 1. Warfarin reduces this risk by approximately 61% (event rate 1.66% per year on warfarin), while NOACs show similar or superior efficacy 1.