Author + information
- Jonas B. Olesen, MD,
- Ron Pisters, MD,
- Vanessa Roldans, MD, PhD,
- Francisco Marin, MD, PhD and
- Deirdre A. Lane, PhD⁎ ()
- ↵⁎University of Birmingham, Centre for Cardiovascular Sciences, City Hospital, Birmingham B18 7QH, United Kingdom
Fang et al. (1) describe their bleeding risk scheme for anticoagulated patients with atrial fibrillation (AF), which includes 5 weighted risk factors: anemia, severe renal disease, age 75 years and older, previous hemorrhage, and diagnosed hypertension (1).
We welcome efforts to assess bleeding risk in AF patients requiring anticoagulation, but a number of major concerns regarding the derivation of this new schema may limit its clinical applicability. First, the ATRIA cohort consisted exclusively of anticoagulated AF patients (1), and rates of warfarin-associated hemorrhage may be underestimated due to the selected “warfarin-experienced” cohort and may not be applicable to patients initiating anticoagulation. Second, their previous study (2) reported that risk factors included in their new schema (previous intracranial/ gastrointestinal/other hemorrhage, renal insufficiency) were contraindications to warfarin (2). Patients with ≥1 of these risk factors were significantly less likely to receive warfarin (2) and consequently were excluded from current analyses (1). Third, several well-established major risk factors for bleeding, namely, recent (<90 days) hemorrhage (hazard ratio: 0.3; 95% confidence interval: 0.1 to 0.7), antiplatelet treatment (clopidogrel/ticlopidine), and reduced platelet count (3), had point estimates that appeared protective of future bleeding in their analyses (1).
Fourth, intensity of the anticoagulation effect (i.e., quality of INR control) plus treatment with aspirin and nonsteroidal anti-inflammatory drugs, which are important bleeding risk factors (3), were not included in the study (1). Fifth, no rationale was provided for only including risk factors with hazard ratios ≥1.5 in the initial model, resulting in previous stroke, a well-established risk factor for bleeding (3), being excluded.
Sixth, Fang et al. (1) included age 75 years and older in their bleeding prediction scheme, despite bleeding risk increasing from age 65 and older (3,4), and greater risk associated with age 65 and older found in initial analyses (1). Also noteworthy is the inclusion of diagnosed hypertension (1), despite evidence demonstrating that well-controlled hypertension is not significantly associated with bleeding in AF patients (3).
Finally, the lack of comparison of this new schema with one of the most contemporary bleeding prediction tools for AF currently available, HAS-BLED (4), is somewhat surprising, given the inclusion of HAS-BLED in recent European and Canadian AF guidelines. We note that many elements within the weighted (and, hence, more complicated) ATRIA score are already covered by the components within the more simple HAS-BLED score.
Please note: Dr. Olesen has received travel grants from AstraZeneca and Boehringer Ingelheim. Dr. Lane has received funding for research, educational symposia, and consultancy and lecturing fees from Boehringer Ingelheim, Bayer Pharma Schering, BMS/Pfizer, and AstraZeneca. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Fang M.C.,
- Go A.S.,
- Chang Y.,
- et al.
- Lip G.Y.H.,
- Andreotti F.,
- Fauchier L.,
- et al.