Hot Topics: HAS-BLED Tool – What is the Real Risk of Bleeding in Anticoagulation?

Gregory Y. H. Lip, M.D., F.A.C.C. (Disclosure)

September 17, 2012

Stroke prevention with appropriate use of antithrombotic therapy remains absolutely central to the overall management strategy of patients with atrial fibrillation (AF).

The first consideration is stroke risk assessment. Various risk factors have been used to derive stroke risk stratification schema, which have ‘artificially’ categorised patients into low, moderate and high risk stroke strata, so that the patients at highest risk can be identified for warfarin therapy. Many of these risk factors were derived from the non-warfarin arms of the historical trial cohorts, where only <10% of patients screened were randomised, and many risk factors were not systematically looked for, nor consistently defined. With the availability of the novel oral anticoagulants that are alternatives to warfarin, there is the need to be more inclusive of common stroke risk factors, to focus more on identification of ‘truly low risk patients’ with AF who do not need any antithrombotic therapy.

Indeed, the 2012 focused update to the European Society of Cardiology (ESC) guidelines recommends stroke risk assessment using the CHA2DS2-VASc score,¹ and strongly emphasises a clinical practice shift towards much more focus on defining the ‘truly low-risk’ patients with AF, instead of trying to identify ‘high-risk’ patients. These ‘truly low risk’ patients are those patients who fulfil the criteria of ‘age < 65 and lone AF (irrespective of gender) or CHA2DS2-VASc score=0’, who do not need any antithrombotic therapy).

The second aspect with regard to thromboprophylaxis is to assess bleeding risk.² The HAS-BLED score is the recommended score in the ESC and Canadian guidelines for this purpose.^1,3 HAS-BLED has been well validated,^4-6 and has been shown to outperform other risk scores (including HEMORR(2)HAGES and ATRIA) in predicting clinically relevant bleeding.^7-9 Indeed, limitations of some prior scores have previously been highlighted.¹⁰ Also, HAS-BLED has good predictive value for intracranial bleeding, whilst other scores (e.g. ATRIA) were not predictive.⁷ In the Swedish AF Cohort study, the rates of major bleeding (and intracranial bleeding) increased with increasing HAS-BLED score, but rates were fairly similar for warfarin and aspirin treated patients.¹¹

How to use HAS-BLED? A high HAS-BLED score (≥3) is indicative of the need for regular clinical review and followup, but should not be used per se as a reason for stopping oral anticoagulation.¹ Indeed, a high HAS-BLED score allows the clinician to ‘flag up’ patients at potential risk for serious bleeding in an informed manner, rather than relying on guesswork. The latter may be dangerous, as it has been shown that clinicians are poor in estimating bleeding risk.¹²

The HAS-BLED score also makes clinicians think about the potentially reversible risk factors for bleeding, e.g. uncontrolled blood pressure (the H in HAS-BLED), labile INRs if on warfarin (the L in HAS-BLED) and concomitant use of aspirin/NSAIDs (the D in HAS-BLED). The HAS-BLED score is also predictive of major bleeding in patients (both AF and non-AF) undergoing bridging therapy.¹³

Bleeding risk and stroke risk are closely related. Those patients with AF and a high HAS-BLED score derive a higher net clinical benefit from oral anticoagulation when balancing ischaemic stroke against intracranial bleeding.^14,15 This is irrespective of stroke risk strata, whether assessed by CHADS2 or CHA2DS2-VASc, with the exception of CHA2DS2-VASc score=0, where the net clinical benefit was negative reflecting the ‘truly low risk’ status of such patients that would result in a net disadvantage of warfarin therapy; of note, there was no stroke risk or HAS-BLED strata showing any positive net clinical benefit for aspirin.¹⁴

An illustrative application of the CHA2DS2-VASc and HAS-BLED scores to aid decision making has recently been published (Table 1).¹⁶