Atrial Fibrillation in the Elderly: Balancing Stroke and Bleeding Risk

Atrial fibrillation is the most common sustained cardiac rhythm disturbance encountered in clinical practice, and its prevalence climbs steeply with age. In Australia, where more than 4.2 million people are aged 65 years and over, atrial fibrillation affects an estimated one in ten adults in this age bracket and accounts for roughly one in three ischaemic strokes. Survivors frequently face prolonged rehabilitation, loss of independence, and recurrent hospitalisation, placing pressure on families and on the public health system alike.

Balancing the protective effect of anticoagulation against the very real risk of bleeding is one of the central challenges in modern geriatric cardiology. For clinicians weighing whether to initiate, continue, or deprescribe therapy, the decision is rarely straightforward. This article examines current evidence, Australian prescribing frameworks, and practical strategies for tailoring stroke prevention to the realities of older patients across metropolitan, regional, and remote settings.

AF prevalence and stroke burden in older Australians

Data from the Australian Institute of Health and Welfare suggest that the number of Australians living with atrial fibrillation will more than double by 2050, driven largely by population ageing and improved survival from acute coronary events. The condition disproportionately affects people in their eighth and ninth decades, the very group in whom stroke confers the greatest disability and mortality. A first-ever ischaemic stroke in an 80-year-old carries a 30-day mortality approaching one in five, and survivors are more likely to be discharged to residential aged care than to return home.

The Australian Stroke Clinical Registry highlights ongoing gaps in secondary prevention, including under-prescription of anticoagulation in patients with known atrial fibrillation. Even when therapy is initiated, persistence at 12 months remains suboptimal, particularly among those who experience minor bleeding events or who lack structured follow-up. These patterns have motivated national quality-improvement initiatives such as the ANZACS-QI registry, which tracks antithrombotic prescribing after cardiac hospitalisation and feeds back comparative data to participating sites from Cairns to Canberra.

Quantifying thrombotic and bleeding risk

The CHA₂DS₂-VASc score remains the standard tool for stratifying stroke risk in non-valvular atrial fibrillation, with age 75 years or older contributing two points and age 65 to 74 contributing one. In Australian guidelines endorsed by the Cardiac Society of Australia and New Zealand, male patients with a score of one and female patients with a score of two are generally considered for anticoagulation. Yet the score treats age as a binary step rather than a continuum, and under-weights competing risks such as frailty, renal impairment, and cognitive decline.

Bleeding risk scores such as HAS-BLED were designed to flag modifiable hazards rather than to exclude patients from anticoagulation. A high HAS-BLED score should prompt clinicians to address uncontrolled hypertension, hazardous alcohol use, and concomitant antiplatelet therapy, not to withhold stroke prophylaxis altogether. Australian data from the Melbourne Collaborative Cohort and from several large tertiary centres have shown that bleeding risk in well-selected octogenarians is often lower than clinicians perceive, particularly when direct oral anticoagulants are used at appropriate renal doses.

Choosing oral anticoagulation in Australia

All four direct oral anticoagulants commonly used internationally are listed on the Pharmaceutical Benefits Scheme for non-valvular atrial fibrillation, with specific authority criteria relating to CHA₂DS₂-VASc score and creatinine clearance. This means that cost is rarely a barrier to initiation, although switching between agents because of formulary changes at the local hospital level can be disruptive. Apixaban and rivaroxaban dominate Australian prescribing, partly because once- or twice-daily dosing suits older patients who rely on Webster-pak medication organisers.

Warfarin retains a niche role in patients with mechanical heart valves, moderate-to-severe mitral stenosis, or advanced chronic kidney disease with estimated glomerular filtration rate below 30 mL/min/1.73 m². For these patients, point-of-care international normalised ratio monitoring through community pathology providers, including services that fly samples out from regional towns such as Broken Hill and Longreach, allows tight control without frequent travel. For the broader elderly population, however, direct oral anticoagulants offer a more predictable pharmacokinetic profile and fewer interactions with the typical Australian medication regimen, which often includes paracetamol, statins, and proton pump inhibitors.

Falls, frailty, and cognitive impairment

The widely held belief that a fall automatically contraindicates anticoagulation is not supported by contemporary evidence. Modelling studies suggest that an older patient would need to fall roughly 295 times per year for the intracranial haemorrhage risk to outweigh the stroke-prevention benefit of warfarin, and the threshold is even higher for direct oral anticoagulants. Australian falls-prevention programmes such as Stay On Your Feet, funded through state health departments, complement anticoagulation by addressing the underlying hazards rather than prompting therapeutic nihilism.

Frailty and cognitive impairment introduce distinct challenges. Patients with early dementia may forget doses or take double doses, while those with advanced cognitive decline may resist medication altogether. Involving a spouse, adult child, or community nurse in medication administration is often the simplest safeguard, and dose administration aids prepared by the local pharmacist are rebatable under Medicare for residents of eligible aged-care facilities. Where behavioural and psychological symptoms of dementia necessitate sedating agents, clinicians should review the anticholinergic burden and avoid combinations that increase sedation-related falls.

Geographic, rural, and cultural realities

Australians living outside major metropolitan centres face structural barriers to specialist cardiac care. Cardiac electrophysiologists are concentrated in capital cities, and patients in the Pilbara, the Top End, or western New South Wales may travel more than 500 kilometres for a single outpatient review. The so-called tyranny of distance that shapes much of rural Australian life means that some patients simply cannot attend a tertiary clinic without giving up a working week. Telehealth consultations, funded through Medicare since the pandemic-era expansion of item numbers, have narrowed this gap, but access to procedural services such as catheter ablation remains uneven. For frail elderly patients, the trade-off between stroke prevention through anticoagulation and the morbidity of long-haul travel usually favours medical therapy.

Aboriginal and Torres Strait Islander peoples experience stroke at younger ages and at higher rates than the broader Australian population, and atrial fibrillation is increasingly recognised as a contributor to this disparity. Culturally safe care, delivered in partnership with Aboriginal Community Controlled Health Organisations, can improve anticoagulation uptake and persistence. Having a yarn with an Aboriginal Health Worker or community elder, rather than relying solely on a clinician-led consent process, builds the trust needed for long-term therapy. Simple adaptations such as explaining CHA₂DS₂-VASc in plain language, involving family in decision-making, and avoiding jargon-heavy handouts make a measurable difference in adherence and persistence.

Shared decision-making and integrated care

Guidelines from the National Heart Foundation of Australia and the Cardiac Society of Australia and New Zealand emphasise shared decision-making as a core competency in atrial fibrillation management. For elderly patients, this conversation should explicitly address the patient's own priorities: avoiding a disabling stroke, preserving independence, minimising tablets, or reducing clinic visits. Decision aids developed by Australian researchers, including the AF Shared Decision Making tool piloted in Adelaide and Perth, help structure these discussions and document patient preferences.

Integrated atrial fibrillation clinics, often led by nurse practitioners in collaboration with cardiologists, general practitioners, and pharmacists, have demonstrated improved guideline adherence and reduced hospital readmissions in Victorian and New South Wales health services. Routine review of renal function, weight, and bleeding symptoms at six- to twelve-month intervals allows timely dose adjustment and reinforces adherence. Embedding these reviews within existing aged-care or chronic-disease management plans reduces duplication and is generally well accepted by patients and their families. Clinicians seeking to update their protocols will find original research, reviews, and themed collections on the Journal of Arrhythmia website.

Stroke prevention in elderly patients with atrial fibrillation is rarely a question of whether to anticoagulate, but rather which agent, at what dose, and with what support. Australian clinicians are fortunate to work within a system that subsidises evidence-based therapy and increasingly recognises the value of multidisciplinary review. For practising cardiologists, geriatricians, general physicians, and trainees who manage this complex population, ongoing engagement with the contemporary literature remains essential. Researchers and clinicians are encouraged to submit original work, suggest themed issues, or correspond with the editorial team through the contact page.