Clinical background
CHA₂DS₂-VASc is well established, but it requires the clinician to remember to score age at two thresholds, sex, and six different risk factors with different weights, and then to translate the sum into a treatment recommendation whose thresholds rest partly on absolute-risk considerations and have shifted between guideline versions. In primary care and the emergency department, where most patients with atrial fibrillation are managed, this complexity leads to both undertreatment and overtreatment.
CHADS-65 was developed as a direct response to this. It rests on the observation that age is the single strongest risk marker for stroke in atrial fibrillation, that the absolute risk rises markedly around the age of 65, and that the remaining risk factors in CHA₂DS₂-VASc serve mainly to identify younger patients in whom age alone is not sufficient. The result is a decision tree in which every patient either should or should not receive an oral anticoagulant, without an intermediate category in which the clinician must interpret a score [1,2].
The algorithm was adopted in the 2020 comprehensive guideline update from the Canadian Cardiovascular Society (CCS) and the Canadian Heart Rhythm Society (CHRS) as the simplified first-line approach to routine stroke prevention in patients with non-valvular atrial fibrillation [1].
Applying CHADS-65
CHADS-65 is a decision algorithm, not a summed risk score. Unlike CHA₂DS₂-VASc, the tool does not give a number to be interpreted against a table, but a direct recommendation on anticoagulation.
The decision logic, formally stated:
where a means previous stroke/TIA/systemic embolism, hypertension, heart failure (or a reduced ejection fraction) or diabetes.
The variables assessed are therefore:
- Age: a single threshold at 65 years
- Previous stroke/TIA/systemic embolism: binary, present or not
- Hypertension: binary
- Heart failure or reduced EF: binary, combined into one variable
- Diabetes: binary
- Coronary artery disease or peripheral arterial disease: binary, combined into one variable
The derivation rests on concepts introduced in connection with the development of the 2018 CCS guidelines and further developed up to the comprehensive 2020 update [1,3]. Unlike CHA₂DS₂-VASc, which was derived from a cohort of 108,000 non-anticoagulated patients in the Swedish registry, CHADS-65 was not produced through a traditional derivation study in a specific cohort with a specified outcome rate. It is a logical simplification of CHA₂DS₂-VASc, based on the fact that age ≥65 years has such strong predictive power that it alone justifies anticoagulation, and that each of the remaining CHADS₂ risk factors (stroke/TIA, hypertension, heart failure, diabetes) carries enough risk to justify anticoagulation in younger patients as well [2,3].
Interpretation in practice
The algorithm has three outcomes and three actions:
| Outcome by CHADS-65 | Clinical action |
|---|---|
| Age ≥65 years | Start an oral anticoagulant irrespective of other risk factors |
| Age <65 years with a CHADS₂ risk factor (stroke/TIA, hypertension, heart failure, diabetes) | Start an oral anticoagulant |
| Age <65 years with coronary or peripheral arterial disease | Start an oral anticoagulant, not an antiplatelet agent |
| No risk factor and age <65 years | No antithrombotic therapy |
Crucial and often misunderstood is the third row: in patients with coronary or peripheral arterial disease without CHADS₂ risk factors, the recommendation is an oral anticoagulant alone. The CCS states explicitly that an antiplatelet agent should not be started for the atrial fibrillation alone. If the patient is already on an antiplatelet agent for a separate cardiological indication, that treatment continues, but neither is an OAC added to it nor is it combined with an antiplatelet agent for the purpose of reducing the embolic risk of the atrial fibrillation.
Patients who fall into the "no antithrombotic therapy" category should be reassessed at every new contact, particularly on reaching the age of 65, but also on the onset of hypertension, heart failure or diabetes.
Validation and performance
CHADS-65 has not undergone external validation in randomised trials with stroke as the outcome. Its performance is judged on the basis that the algorithm is a collinear transformation of CHA₂DS₂-VASc with thresholds that approximate those at which CHA₂DS₂-VASc gives an OAC recommendation [2].
A comparison of the CCS/CHRS, ESC and AHA/ACC/HRS guidelines shows that CHADS-65 classifies fewer patients as low risk than the European reading of CHA₂DS₂-VASc, in which the ESC 2020, for example, indicates that men with a CHA₂DS₂-VASc of 1 should generally not be anticoagulated. CHADS-65 takes a more conservative position: age is never a non-qualifying risk factor in patients aged 65 and over [3].
Discrimination for stroke in external cohorts, measured as a c-statistic, has not been published specifically for CHADS-65, and would not differ appreciably from CHA₂DS₂-VASc since the same variables underlie both. Calibration is likewise absent: the CCS gives no absolute risk percentage, and the tool is not intended to predict an individual risk figure.
Limitations
CHADS-65 applies to non-valvular atrial fibrillation. With a mechanical valve prosthesis, mitral stenosis with a haemodynamically significant flow disturbance or other valvular disease carrying a high embolic risk, anticoagulation is indicated irrespective of the algorithm's output, and CHADS-65 must not be used to justify withholding it.
The algorithm does not include bleeding risk. The patient's bleeding risk should always be assessed before anticoagulation is started, and the drug class and dose adapted to renal function, liver function and interactions. In patients with coronary artery disease already on an antiplatelet agent, the indication extends to combination therapy, and here bleeding risk must be weighed against ischaemic risk according to current cardiological practice.
CHADS-65 has not been specifically validated for:
- Patients with hyperthyroidism-related cardiovascular disease
- Patients under 50 years, in whom the baseline risk is very low even with risk factors
- Patients with hypertrophic cardiomyopathy, in whom the embolic risk is regarded as valve-related or valve-like and is therefore handled separately
- Dialysis-dependent renal failure, in which the evidence for OAC is uncertain and guidelines differ
A common error is to apply CHADS-65 as a score and give, for example, a 67-year-old woman with hypertension "2 points" and thereby an increased risk class. The algorithm does not give points; it gives a decision.
Place in current practice
Contemporary European and North American practice follows CHA₂DS₂-VASc, and both the ESC 2020 and the ACC/AHA/HRS 2023 prescribe risk stratification by sex, in which men with a CHA₂DS₂-VASc ≥2 and women with ≥3 should be anticoagulated, with grey zones below. CHADS-65 is a Canadian algorithm and is not part of standard practice elsewhere; it would give anticoagulation to patients who under the ESC algorithm fall into the grey zone, particularly 65-year-old men without other risk factors (CHA₂DS₂-VASc 2 by age, a grey zone under the ESC). Outside Canada, CHADS-65 is chiefly an educational complement, or a tool in primary care where the full application of CHA₂DS₂-VASc is not assured.
References
- Andrade JG et al. The 2020 Canadian Cardiovascular Society/Canadian Heart Rhythm Society Comprehensive Guidelines for the Management of Atrial Fibrillation. Can J Cardiol 2020. PMID: 33191198
- Nattel S, Lip GYH. Guideline Implications of Prothrombotic State Assessment in Low-Risk Atrial Fibrillation Patients: Consistency With CHA₂DS₂-VASc and Support for CHADS-65. Can J Cardiol 2019. PMID: 31030855
- Cheung CC et al. Management of Atrial Fibrillation in 2021: An Updated Comparison of the Current CCS/CHRS, ESC, and AHA/ACC/HRS Guidelines. Can J Cardiol 2021. PMID: 34186113
