Intended for healthcare professionals

CCBYNC Open access

Rapid response to:


Risk of stroke and transient ischaemic attack in patients with a diagnosis of resolved atrial fibrillation: retrospective cohort studies

BMJ 2018; 361 doi: (Published 09 May 2018) Cite this as: BMJ 2018;361:k1717

Opinion: The importance of asking the right research question

Rapid Response:

Reasons why guideline recommendations for anticoagulants should not be extended to patients with ‘resolved atrial fibrillation’

There are at least two reasons why the results of this study should not be used for broadening guideline recommendations for prescribing anticoagulants to patients with ‘resolved atrial fibrillation (AF)’:

1. In keeping with comments from responder 2 (Dr Wicke), Adderley et al have not defined ‘resolved AF’ as applied in their study. They simply say they used the ‘atrial fibrillation resolved’ clinical code in the ‘THIN’ database of electronic primary care records from UK general practices. Therefore, the target population of their recommendations regarding anticoagulation has not been defined for generalized use. Their study sample is likely to include representation from highly heterogeneous populations. Clear and generalizable definitions are also missing for the comparator groups (‘unresolved’ and ‘no history of’ AF) and for stroke and TIA (used for patient selection and outcome comparisons). Clinical, brain imaging and electrocardiograph (ECG) results (including number and temporal relationships of investigations) are required for these definitions.

2. The population best represented by the ‘resolved AF’ code is highly likely to be significantly different compared to the population who received an overall stroke prevention from anticoagulation in the randomised trials comparing vitamin K antagonists, antiplatelet therapy, placebo and direct oral anticoagulants. 1-4 We know this because the stroke and TIA rate was significantly lower in their patients with ‘resolved AF’ compared to those with ‘AF’. In contrast, the great majority of patients in randomised anticoagulation trials had recently documented ‘non-valvular or ‘non-rheumatic’ AF with ECG confirming AF at least 1-3 times within the day to 24 months before trial enrolment.1-4 In addition, in many of these randomised anticoagulation trials patients with ‘transient’ or ‘reversible causes’ of AF (such as cardiac surgery or hyperthyroidism) or undergoing cardioversion were excluded.1-4 One cannot assume that other populations of patients with AF will achieve an overall benefit from anticoagulation. Atrial fibrillation is a spectrum of disorder and anticoagulation is not without risk. Appropriately designed studies are required before extending guideline recommendations for anticoagulation.

1. Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Annal Int Med 2007;146:857-67.
2. Connolly SJ, Eikelboom J, Joyner C, et al. Apixaban in patients with atrial fibrillation. New Eng J Med 2011;364:806-17.
3. Patel MR, Mahaffey KW, Garg J, et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. New Eng J Med 2011;365:883-91.
4. Granger CB, Alexander JH, McMurray JJ, et al. Apixaban versus warfarin in patients with atrial fibrillation. New Eng J Med 2011;365:981-92.

Competing interests: No competing interests

22 May 2018
Anne L Abbott
Monash Univeristy, Melbourne, Australia
Monash Univeristy, Melbourne, Australia