NPC Archive Item: Warfarin more effective than aspirin for stroke risk in AF but what about the risk of bleeding? – The BAFTA study

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
MeReC Rapid Review

NPC Logo

Warfarin versus aspirin for stroke prevention in an elderly community population with atrial fibrillation (the Birmingham Atrial Fibrillation Treatment of the Aged Study, BAFTA): a randomised controlled trial

What is the story here?

Treating patients with AF with warfarin compared to aspirin has been shown in randomised controlled trials (RCTs) to be highly effective at reducing the incidence of stroke, but this benefit is at the cost of a higher risk of major bleeding. NICE AF guidance recommends that the physician must balance the risks and benefits for the individual when deciding between warfarin or aspirin treatment. However, increasing age is an independent risk factor for serious haemorrhage.

Older patients are under-represented in AF clinical trials. The Birmingham Atrial Fibrillation Treatment of the Aged Study, (BAFTA), was a prospective randomised open-label trial with blind assessment of endpoints which attempted to clarify the risks and benefits associated with warfarin treatment in an elderly population.

What were the details of this study?

BAFTA randomised 973 community patients aged 75 years and over (mean age 81.5 years) with AF in England and Wales to warfarin (target INR 2-3) or aspirin 75mg daily. Follow-up was for a mean of 2.7 years. Inclusion was restricted to patients for whom there was clinical uncertainty as to whether warfarin or aspirin should be used. Approximately 72% had a CHADS2 score of 1-2, so only 28% were at moderate to high risk with a CHADS2 score of 3-6 (see CKS on AF for  more information on CHADS2 scores).

The primary endpoint was fatal or non-fatal stroke (ischaemic or haemorrhagic), other intracranial haemorrhage, or clinically significant arterial embolism. Secondary outcomes were major extracranial haemorrhage, other admissions to hospital for haemorrhage, hospital admission or death as a result of non-stroke vascular event, and all cause mortality.

What were the results?

Results for the primary endpoint support the current evidence that warfarin is more effective than aspirin in reducing stroke in patients with AF (relative risk (RR) 0.48, 95% CI 0.28 to 0.80, p= 0.003; absolute risk reduction (ARR) 2%, 95% CI 0.7 to 3.2). The number needed to treat (NNT) for one year to prevent one event is 50. The primary event rate and ARR observed was lower than in other AF trials comparing warfarin and aspirin, and the authors suggest several explanations for this. For example, the patients in BAFTA might have had a lower baseline cardiovascualr risk, than other populations studied.

The study was underpowered to detect differences in secondary outcomes. However, the observed bleeding risk from warfarin was no different from aspirin. The yearly risk of extracranial haemorrhage did not differ between warfarin and aspirin (RR 0.87, 0.43 to 1.73; ARR 0.2%, 95% CI -0.7 to 1.2).

The authors suggested several reasons for this, including that the INR in the BAFTA trial (2-3) was much lower than other AF trials using warfarin (which have been up to 4.5). They note that the observed beneficial effects of stroke prevention were of the same magnitude as the trials in which the target INR was higher.

So what?

The BAFTA trial adds to the evidence that warfarin is an effective intervention at reducing the risk of stroke compared to aspirin in patients with AF. It adds reassurance that warfarin with a target INR of 2-3 can be used safely in an elderly population with a low risk of bleeding apart from age. Nearly 80% of elderly patients with AF that were identified for this study were excluded from taking part and therefore these results may not apply to the general elderly population with AF. Clinicians should continue to weigh up the risks and benefits of warfarin vs. aspirin in each individual with AF using the NICE algorithm.

For more information on the diagnosis, assessment and management of AF, visit the AF section of NPC