Fall risk and anticoagulation for atrial fibrillation in the elderly: A delicate balance

Author and Disclosure Information


Guidelines for managing atrial fibrillation recommend systemic anticoagulation for almost all patients age 65 and older, but in practice up to 50% of older patients do not receive maintenance anticoagulation therapy. The most common reason physicians cite for withholding anticoagulation in older patients with atrial fibrillation is a perception of a high risk of falling and associated bleeding, especially intracranial hemorrhage.


  • For most patients in this category, the benefits of anticoagulation outweigh the risks.
  • Although they are not perfect, scoring systems have been developed to predict the risk of stroke without anticoagulation and the risk of bleeding with anticoagulation.
  • The decision-making process is complex and should be shared with the patient and the patient’s family and caregivers.



An 86-year-old woman with hypertension, osteoporosis, and mild cognitive impairment presents with episodes of palpitations and heart “fluttering.” These episodes occur 1 to 2 times per week, last for up to several hours, and are associated with mild shortness of breath and reduced activity tolerance. She is widowed and lives in a retirement facility, but she is independent in activities of daily living. She has fallen twice in the past year without significant injury.

See related editorial

Physical examination is unremarkable. An electrocardiogram demonstrates sinus rhythm with left ventricular hypertrophy. A 30-day event monitor reveals several episodes of paroxysmal atrial fibrillation that correspond with her symptoms. A subsequent echocardiogram shows normal left ventricular systolic function, mild diastolic dysfunction, and no significant valvular abnormalities. Laboratory studies, including thyroid-stimulating hormone, are normal.

What is this patient’s risk of stroke? What is her risk of major bleeding from anticoagulation? How should fall risk be addressed in the decision-making process? What other factors should be considered?


The prevalence of atrial fibrillation increases with age, and nearly half of patients with atrial fibrillation in the United States are 75 or older.1 In addition, older age is an independent risk factor for stroke in patients with atrial fibrillation, and the proportion of strokes attributable to atrial fibrillation increases exponentially with age:

  • 1.5% at age 50 to 59
  • 2.8% at age 60 to 69
  • 9.9% at age 70 to 79
  • 23.5% at age 80 to 89.2

Numerous large randomized trials have shown that anticoagulation with warfarin reduces the risk of stroke by about two-thirds in patients with atrial fibrillation, and that this benefit extends to the elderly.

In the Birmingham Atrial Fibrillation Treatment of the Aged trial,3 973 patients at least 75 years old (mean age 81.5, 55% male) were randomized to receive either warfarin with a target international normalized ratio of 2.0 to 3.0 or aspirin 75 mg/day. Over an average follow-up of 2.7 years, the composite outcome of fatal or disabling stroke, arterial embolism, or intracranial hemorrhage occurred in 24 (4.9%) of the 488 patients in the warfarin group and 48 (9.9%) of the 485 patients in the aspirin group (absolute yearly risk reduction 2%, 95% confidence interval 0.7–3.2, number needed to treat 50 for 1 year). Importantly, the benefit of warfarin was similar in men and women, and in patients ages 75 to 79, 80 to 84, and 85 and older.

More recently, the oral anticoagulants dabigatran, rivaroxaban, apixaban, and edoxaban have been shown to be at least as effective as warfarin with respect to both stroke prevention and major bleeding complications, and subgroup analyses have confirmed similar outcomes in older and younger patients.4,5

But despite the proven value of anticoagulation for stroke prevention in older adults, only 40% to 60% of older patients who are suitable candidates for anticoagulation actually receive it.6 Moreover, the proportion of patients who are treated declines progressively with age. The most frequently cited reason for nontreatment is perception of a high risk of falls and associated concerns about bleeding, especially intracranial hemorrhage.7–10


Balancing the risk of stroke against the risk of bleeding related to falls is a commonly encountered conundrum in older patients with atrial fibrillation.

Stroke risk

The CHADS2 score was, until recently, the most widely used method for assessing stroke risk in patients with nonvalvular atrial fibrillation. CHADS2 assigns 1 point each for congestive heart failure, hypertension, age ≥ 75, and diabetes, and 2 points for prior stroke or transient ischemic attack (range 0–6 points). Annual stroke risk based on the CHADS2 score ranges from about 2% to about 18%
(Table 1).11

The CHA2DS2-VASc score,12 a modification of CHADS2, appears to assess the risk of stroke more accurately, especially at the lower end of the scale, and recent guidelines for managing atrial fibrillation recommend using the CHA2DS2-VASc algorithm.13 CHA2DS2-VASc is similar to CHADS2, except that it assigns 1 point for ages 65 to 74, 2 points for ages 75 and older, 1 point for vascular disease (coronary artery disease, peripheral arterial disease, aortic aneurysm), and 1 point for female sex (Table 1).11,12

For both CHADS2 and CHA2DS2-VASc, systemic anticoagulation is recommended for patients who have a score of 2 or higher. Our patient’s CHADS2 score is 2, and her CHA2DS2-VASc score is 4, corresponding to an annual estimated stroke risk of 4% with both scores (Table 1). Note, however, that the CHA2DS2-VASc score provides more information at the lower end of the spectrum.

Bleeding risk

Several scoring systems for assessing bleeding risk have also been developed (Table 2).14–16 Of these, the HAS-BLED score has come to be used more widely in recent years.

Perhaps not surprisingly, some of the same factors associated with risk of stroke also predict increased risk of bleeding (eg, older age, hypertension, prior stroke).14 Note, however, that history of falling or high risk of falling is only included in one of the bleeding risk models (HEMORR2HAGES).15

These tools are somewhat limited by their lack of consideration of concomitant antiplatelet therapy (only included in HAS-BLED) or history of bleeding (only ATRIA16 considers major and minor bleeding, HEMORR2HAGES does not specify bleeding severity, and HAS-BLED only considers major bleeding). The models also fail to include medications such as antibiotics or antiarrhythmic agents, which are commonly used by older patients with atrial fibrillation and may increase bleeding risk. In addition, all bleeding risk scores were developed for warfarin, and their applicability to patients treated with the newer oral anticoagulants has not been established.

At the time of presentation, our patient has a HAS-BLED score of 2 (1 point each for age and hypertension), placing her at intermediate risk of bleeding.14

Fear the clot, not the bleed

So how does one balance the risk of stroke vs the risk of bleeding? An adage from the early days of thrombolytic therapy for acute myocardial infarction was “fear the clot, not the bleed.” In other words, in the present context the consequences of a thrombus embolizing from the heart to the brain are likely to be more devastating and more permanent than the consequences of anticoagulation-associated hemorrhage.

Support for this view is underscored by a 2015 study by Lip et al,17 who examined stroke and bleeding risks and outcomes in a large real-world population of patients age 75 and older. The analysis included 819 patients ages 85 to 89 and 386 patients age 90 and older. The key finding was that the oldest patients derived the greatest net benefit from anticoagulation.

Moreover, the Canadian stroke registry of 3,197 patients, mean age 79, showed that advanced age was a more potent risk factor for ischemic stroke than it was for hemorrhagic stroke.18

Thus, the benefit from anticoagulation in patients with atrial fibrillation does not appear to have an upper age limit.

Next Article:

Cardiovascular disease: Innovations in devices and techniques

Related Articles