Correspondence: national survey of patients with AF in the acute medical unit: a day in the life survey

Br J Cardiol 2013;20:160 Leave a comment
Click any image to enlarge

Correspondence from the world of cardiology

Anticoagulation - BJC Learning programme
For healthcare professionals only

National survey of patients with AF in the acute medical unit: a day in the life survey

Dear Sirs,

The first national survey examining the management of atrial fibrillation (AF) within acute medical units up and down the country has just been published in the British Journal of Cardiology.1 Essentially it seems to show that secondary care is just as bad as primary care in initiating warfarin for AF patients. This group of patients is five times more likely to have a thromboembolic cerebrovascular accident than matched populations in sinus rhythm and yet doctors are ineffective at influencing change. This study took place over a 24-hour period and involved 31 hospitals and 149 patients.

The results showed that 80% of patients with a significant risk of stroke (CHA2DS2-VASc score of over 2) did not receive anticoagulation. Stroke risk scores were documented in less than a quarter of patients and the same fraction received digoxin as monotherapy. The rationale for this treatment decision was only documented in a quarter of cases and was not consistent with national standards. Although the study was small, it does strongly suggest that both general practitioners (GPs) and hospital doctors are struggling to anticoagulate AF patients and are also selecting the wrong drugs for AF rate management.

I have spent several years trying to understand why we as GPs are performing below par in anticoagulation2 and now it seems that hospital doctors are just the same. Of course it takes time to explain and organise warfarinisation and patients are naturally apprehensive. When we do have the discussion, it is easy to persuade patients to take aspirin – it reduces stroke risk by 20%. In contrast, it’s a different issue to persuade them to take warfarin, which reduces risk by nearly 70%. I believe we as GPs encourage the British compromise of aspirin because we are unduly and disproportionately influenced by the risk of bleeds.

It is a paradox that the BAFTA (Birmingham Atrial Fibrillation Treatment of the Aged) study showed that in the elderly, aspirin was actually more dangerous than warfarin in causing both intracerebral and extracerebral bleeds! Nevertheless we tend to feel if we prescribe warfarin then we take on enduring risk both of bleeding and prescribing/compliance errors, and so we push the discussion away from the real data. Patients and GPs are therefore complicit in not preventing 15,000 AF strokes a year – and now hospital doctors are embroiled as well!The risk of stroke needs to be balanced with the risk of major haemorrhage. Despite this, the study showed only 36% (n=54) of patients with a CHA2DS2-VASc score >2 had a HAS-BLED score of 3 (8.7 major bleeds per 100 patient-years). In addition, 24% (n=15) of patients at risk of major haemorrhage (HAS-BLED score of 3) were discharged on anticoagulation therapy. Four of this group were concomitantly discharged on antiplatelet therapy.

Decades have gone by with new tools and education to try to influence the prescription of more anticoagulants in AF, but performance has not improved significantly. Surely the time has now come for a more radical and aggressive way forward for stroke prevention in AF. The mantra needs to be that every patient with AF should be on warfarin unless a consultant cardiologist or stroke physician feels this is clinically inappropriate. If the patient declines – as is their informed right – then a letter needs to be in the electronic notes signed by the patient absolving the medical attendants of clinical responsibility and explaining the risks. 

This letter will confirm informed decision making and make both patients and their physicians think more about the real risks – and could have some significant impact on stroke prevention. This is a large clinical performance gap that needs to be closed, and patently it will not happen without a more assertive approach.

Conflict of interest

None declared.

Dr John Havard
General Practitioner

Saxmundham Health, Lambsale Meadow, Saxmundham, IP17 1AS


1. Soong J, Balasanthiran A, MacLeod DC, Bell D. National survey of patients with AF in the acute medical unit: a day in the life survey. Br J Cardiol 2013;20:106.

2. Havard J. Why are we so bad in primary care at initiating warfarin in atrial fibrillation patients? Br J Cardiol 2009;16:237–40. 

The authors’ reply

We are grateful to Dr Havard for his detailed response to our article, highlighting in particular the recurring issue of anticoagulation in AF. 

It should be appreciated that this study focused on acute management strategies. Outwith the context of DC cardioversion during the index episode, the initiation of anticoagulation for the majority of AF patients is a deferred process involving the patient and their GP, supported by secondary care, commonly in relation to echocardiographic assessment.

However, Dr Havard appropriately draws attention to the low rate (<25%) of risk scoring in our study, and it might reasonably be proposed that all unanticoagulated AF patients discharged from acute medical units should have both CHA2DS2-VASc and HASBLED scores documented to guide decision making.

Barriers to good practice need to be studied. Interventions must be locally relevant and embedded within existing systems to be sustainable. Quality improvement methodology may provide the necessary framework for “real world” investigation and implementation of best practice.

Dr John Soong
RCP Clinical Quality Improvement Research Fellow

Chelsea & Westminster Hospital NHS Foundation Trust, 369 Fulham Road, London, SW10 9NH