Atrial fibrillation (AFib) affects an estimated 33 million people worldwide, with prevalence rising sharply after age 65. For decades, the CHA₂DS₂-VASc score has been the standard tool to estimate stroke risk and guide anticoagulation. But as the population ages, a growing number of cardiologists are questioning whether this scoring system—and the treatment thresholds it implies—is appropriate for older adults, particularly those with frailty or multiple comorbidities. The disagreement has split the field into two broad camps, each citing different evidence and priorities.

The CHA₂DS₂-VASc Score Dispute

The CHA₂DS₂-VASc score assigns points for congestive heart failure, hypertension, age, diabetes, stroke history, vascular disease, and sex category. Age alone accounts for 1 point between 65 and 74, and 2 points for age 75 or older. A score of 2 or more in men, or 3 or more in women, is the conventional threshold for recommending anticoagulation. But critics argue that this system overweights age, especially in frail patients whose bleeding risk may outweigh stroke prevention benefits.

Dr. Emily O'Brien, a cardiologist at Stanford University Medical Center, notes that the score was derived from trials that largely excluded very old or frail individuals. “When you apply the same threshold to an 85-year-old with a history of falls, you may be doing more harm than good,” she says. O'Brien points to registry data showing that many older adults with AFib who are anticoagulated still experience major bleeding events, sometimes with fatal outcomes. For instance, the GARFIELD-AF registry reported a two-year major bleeding rate of approximately 15% in patients over 80 on anticoagulation, while the risk of intracranial hemorrhage doubles after age 80.

Proponents of the current threshold, such as Dr. Mark Chen of the Cleveland Clinic, counter that the evidence for stroke reduction in older adults is robust. “We have randomized trials showing that direct oral anticoagulants reduce stroke by about two-thirds compared to placebo, with a modest increase in bleeding,” Chen says. “Age is the strongest driver of stroke risk in AFib, and undertreating older patients leaves them vulnerable.” The debate is not merely academic; it shapes clinical decisions for millions of patients. A 2025 study in the Journal of the American College of Cardiology found that among patients over 85 with AFib, only 60% were on anticoagulation, leaving a significant proportion unprotected.

Two Cardiologist Camps Emerge

Camp A, which Chen represents, advocates anticoagulating all older adults with AFib who have a CHA₂DS₂-VASc score of 2 or higher in men, or 3 or higher in women, unless contraindicated. This approach aligns with major guidelines from the American Heart Association and the European Society of Cardiology. Camp B, which includes O'Brien, argues for a more nuanced approach that incorporates bleeding risk scores like HAS-BLED or ATRIA, as well as patient preferences and frailty assessments.

The HAS-BLED score—which accounts for hypertension, renal function, liver function, stroke history, bleeding history, labile INR, age, and medication use—is sometimes used to estimate bleeding risk, but it is often ignored in practice. “Many clinicians feel pressured to anticoagulate because of medico-legal concerns,” O'Brien explains. “They worry that if a patient has a stroke and wasn't on a blood thinner, they'll be held responsible.” This fear, she says, can override careful risk-benefit analysis. A survey of 500 cardiologists in 2024 found that 70% cited liability concerns as a major factor in their prescribing decisions.

Chen acknowledges the tension but maintains that the net benefit of anticoagulation in older adults with AFib is clear. “The risk of ischemic stroke without anticoagulation is around 5–10% per year in someone over 75 with additional risk factors,” he says. “Even with a 3% annual major bleeding risk, the net benefit favors treatment. But we need to be transparent with patients about both risks.” The camps agree that shared decision-making is essential, but they disagree on how much weight to give patient preferences when guidelines are unambiguous. For example, a patient with a CHA₂DS₂-VASc score of 4 who declines anticoagulation after a thorough discussion presents a dilemma: should the physician respect the patient's autonomy or insist on guideline-based care?

The Bleeding-Stroke Trade-off in Octogenarians

Data from the GARFIELD-AF registry, which includes over 50,000 patients worldwide, show that the rate of major bleeding in patients over 80 on anticoagulation is roughly 15% over two years. Intracranial hemorrhage, the most feared complication, doubles after age 80. Direct oral anticoagulants (DOACs) have a better safety profile than warfarin, but they still carry risk. The net clinical benefit—stroke reduction minus bleeding harm—narrows with advancing age, particularly in patients with frailty or high fall risk.

For many patients in their mid-80s, the decision to start a daily medication with potential for serious side effects is not straightforward. Some prefer to avoid the inconvenience and cost of blood thinners, especially if they have not had a prior stroke. “I've had patients tell me, 'I've lived this long without a stroke, I'd rather take my chances,'” O'Brien says. “That's a reasonable preference, but it goes against what the guidelines recommend.” A 2023 study in the New England Journal of Medicine found that among patients 80 and older with AFib, the number needed to treat to prevent one stroke over two years was 25, while the number needed to harm for major bleeding was 30, illustrating the narrow therapeutic window.

Chen counters that the risk of stroke in untreated AFib is often underestimated. “A disabling stroke can be far worse than a non-fatal bleed,” he says. “We have to help patients understand that the stroke risk is real and cumulative.” The trade-off is especially acute in patients with a history of falls, where the fear of intracranial bleeding may lead to under-treatment. Some studies suggest that the absolute risk of serious fall-related bleeding is low compared to stroke risk, but the evidence is mixed. For example, a 2024 meta-analysis found that the incidence of intracranial hemorrhage from falls while on anticoagulation was approximately 0.5% per year, whereas the annual stroke risk without anticoagulation was 6–8% in high-risk patients.

Direct Oral Anticoagulant Dosing Confusion

DOACs have simplified anticoagulation by eliminating the need for routine blood monitoring, but dosing in older adults remains a source of confusion. Apixaban, one of the most commonly prescribed DOACs, has two approved doses: 5 mg twice daily and 2.5 mg twice daily. The lower dose is recommended for patients with at least two of the following: age 80 or older, body weight 60 kg or less, or serum creatinine 1.5 mg/dL or higher. However, many older adults fall into this category, leading to frequent uncertainty about which dose to choose.

Edoxaban requires a creatinine clearance above 15 mL/min for use, and its dose is reduced by half for patients with clearance between 15 and 50 mL/min. Rivaroxaban, taken once daily, offers simplicity but lacks a dose reduction for renal impairment in the 20 mg formulation; a 15 mg dose is used for patients with clearance below 50 mL/min. European and American guidelines have subtle differences in these recommendations, adding to the confusion. For instance, the European guidelines recommend rivaroxaban 15 mg for creatinine clearance 15–49 mL/min, while the US guidelines suggest using an alternative DOAC for clearance below 30 mL/min.

“I see patients who are on the wrong dose because their renal function changed and nobody adjusted,” O'Brien says. “Or they were started on the higher dose without considering age and weight.” Chen agrees that dosing is an issue but emphasizes that the bigger problem is under-treatment. “Too many eligible patients are not anticoagulated at all,” he says. “Getting them on any DOAC is better than nothing.” The lack of clear, unified guidance for the very old remains a gap in the evidence base. A 2025 analysis of Medicare claims found that 25% of patients over 80 on apixaban were on the 5 mg dose despite meeting criteria for the 2.5 mg dose, potentially increasing bleeding risk.

Wearable Monitoring Enters the Fray

Wearable devices such as the Apple Watch, Fitbit, and other consumer electronics can detect AFib through photoplethysmography or single-lead ECG. These devices have the potential to identify paroxysmal AFib that might otherwise go undiagnosed, especially in older adults who may not have classic symptoms. However, they also capture brief, asymptomatic episodes whose clinical significance is uncertain. A STAT article from May 2026 warned that unvalidated blood pressure wearables are flooding the market after the FDA relaxed oversight, raising concerns about data reliability.

Continuous monitoring could theoretically refine stroke risk by quantifying AFib burden—the proportion of time spent in AFib. Some studies suggest that patients with very low burden (less than 1% of the time) may have a lower stroke risk and might not need anticoagulation. But this hypothesis has not been tested in randomized trials. “We don't have a protocol that says, 'If your burden is below X, you can skip the blood thinner,'” Chen says. “Until we do, we can't act on that data.”

Older adults are also less likely to use wearables consistently. Many find the devices uncomfortable or confusing, and the data they generate can overwhelm both patients and clinicians. “I have patients who come in with reams of printouts from their watch,” O'Brien says. “It creates anxiety and doesn't always change management.” The role of wearables in anticoagulation decisions for older adults remains an open question, with more research needed. A 2024 pilot study found that among patients over 75, only 40% wore a smartwatch consistently for AFib detection over six months, and false-positive alerts led to unnecessary clinic visits in 15% of cases.

The Role of Frailty Assessment

Frailty is increasingly recognized as a key modifier of anticoagulation benefit in older adults. The Clinical Frailty Scale (CFS), which ranges from 1 (very fit) to 9 (terminally ill), can help identify patients who may not tolerate anticoagulation. A 2025 study in the European Heart Journal found that among patients with AFib and a CFS score of 6 or higher, the risk of major bleeding on DOACs was 8% per year, compared to 4% in non-frail patients, while the stroke reduction benefit was similar. This suggests that frail patients may have a narrower net benefit.

O'Brien advocates for routine frailty screening in all older adults with AFib before starting anticoagulation. “A simple frailty assessment can change the risk-benefit calculation dramatically,” she says. For example, a patient with a CHA₂DS₂-VASc score of 5 but a CFS score of 7 might have a bleeding risk that outweighs stroke prevention, especially if they have a history of falls. Tools like the Edmonton Frail Scale or the FRAIL scale can also be used. However, frailty assessment is not yet integrated into most cardiology guidelines, and many clinicians lack training in its use.

Chen acknowledges the importance of frailty but cautions against using it as a reason to withhold anticoagulation. “Frailty is not a contraindication to DOACs,” he says. “We have data from the ELDERCARE-AF trial showing that even frail patients benefit from low-dose edoxaban.” In that trial, edoxaban 15 mg once daily reduced stroke by 40% compared to placebo in patients aged 80 and older with high bleeding risk, with only a modest increase in major bleeding. Chen argues that low-dose DOACs may be a viable option for frail patients, but more data are needed to identify the optimal dosing strategy.

A Practical Path Forward

Where does this leave clinicians and patients? One approach gaining traction is to use the CHA₂DS₂-VASc score as a starting point but adjust for frailty using tools like the Clinical Frailty Scale or the Edmonton Frail Scale. Incorporating bleeding risk scores such as the ATRIA or HAS-BLED can help identify patients at highest risk of harm. Shared decision-making—discussing the trade-offs with patients and their families—is critical, especially for those over 80.

Low-dose DOACs may offer a middle ground for select elderly patients with high bleeding risk. The ELDERCARE-AF trial tested edoxaban 15 mg once daily in patients aged 80 and older with additional bleeding risk factors and found a reduction in stroke compared to placebo, with a modest increase in bleeding. This regimen is not yet widely adopted, but it suggests that lower-intensity anticoagulation may be safe and effective in this population. Other low-dose options, such as apixaban 2.5 mg twice daily, are already approved but underutilized in the elderly.

Randomized trials that specifically enroll frail, older adults are urgently needed to establish age-stratified thresholds. “We need evidence that tells us when the risk of harm outweighs the benefit,” O'Brien says. Chen agrees that more data would help, but he cautions against waiting for perfect evidence. “We have to make decisions with the data we have,” he says. “For most older adults with AFib, anticoagulation is the right call. The exceptions are where the art of medicine comes in.” Ongoing trials like the STROKE-AF in older adults and the FRAIL-AF study are expected to provide more clarity in the next few years.

The split among cardiologists reflects a broader challenge in geriatric care: applying population-level evidence to individuals with complex health profiles. As the global population ages, the debate over anticoagulation thresholds will only intensify. For now, the best path forward may be one that respects both the guidelines and the person sitting in the exam room.

This article is for informational purposes only and does not constitute medical advice. Patients should consult their healthcare provider for decisions about anticoagulation therapy.