New scheme for estimating bleeding risk in afibbers

MAASTRICHT, THE NETHERLANDS. Anticoagulation with drugs such as warfarin (Coumadin) and dabigatran (Pradaxa) and antiplatelet therapy with aspirin or clopidogrel treads a fine line between benefit and risk. While anticoagulation and, to a lesser extent, antiplatelet therapy can reduce the risk of ischemic stroke in atrial fibrillation (AF) patients with coexisting risk factors, these therapies also significantly increase the risk of hemorrhagic stroke and major internal bleeding. While there now are two commonly used schemes (CHADS2 and CHA2DS2-VASc) for predicting stroke risk, there is no universally accepted scheme for predicting risk of major bleeding.

A team of researchers from the University of Maastricht and the University of Birmingham has now developed a simple, quite accurate bleeding risk score called HAS-BLED where the letter in the acronym and their assigned risk scores are as follows:

  • H = Hypertension - 1 point
  • A = Abnormal kidney and liver function - 1 point each
  • S = Stroke (previous ischemic) - 1 point
  • B = Bleeding (previous event/events) - 1 point
  • L = Labile INRs (difficulty maintaining stable INR) - 1 point
  • E = Elderly - 1 point
  • D = Drug or alcohol use - 1 point each

The research team applied the HAS-BLED risk score to a group of 3,456 patients with AF without structural heart disease (non-valvular AF). The average age of the group was 67 years and 39% were women. At discharge from hospital, 52% of patients were prescribed an anticoagulant (most likely warfarin), 12.8% were prescribed anticoagulant + aspirin and/or clopidogrel, 24% received antiplatelet therapy (aspirin or clopidogrel) on its own, and the remaining 10.2% received no antithrombotic therapy. The most common reason for prescribing therapy was age over 65 years, although the researchers point out that the biological age of an elderly patient is probably more relevant to bleeding risk than is the chronological age.

During a 1-year follow-up, 52 patients (1.56%) experienced a major bleeding event (requiring hospitalization and/or blood transfusion). The annual risk (%/year) of a bleeding event increased with increasing HAS-BLED score as shown below.

Bleeds/Patient-year, %

The overall annual bleeding rate was highest for patients treated with anticoagulants (1.75%/year) followed by those receiving no antithrombotic treatment (1.42%/year), and those on antiplatelet therapy alone (0.97%/year). Anticoagulation and antiplatelet therapy is not recommended for afibbers with a CHADS2 score of 0 and thus the HAS-BLED score is not really relevant here. However, in the case of a CHADS2 score of 1, the researchers suggest that the HAS-BLED score must exceed 2 in order for the risk of anticoagulation to offset its benefits. For a CHADS2 score of 2 or higher, they suggest that the risk of bleeding outweighs the potential benefits of anticoagulation if the HAS-BLED score exceeds the CHADS2 score. The researchers acknowledge the limitation of not including INR variation in their evaluation.

Pisters, R, et al. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation. Chest, Vol. 138, No. 5, November 2010, pp. 1093-1100

Editorís comment: Anticoagulation and antiplatelet therapy is a double-edged sword, in that benefits and risks must be carefully weighed for the individual patient before being prescribed. Until now, only schemes dealing with stroke risk have been employed with no or little attention paid to bleeding risk. Hopefully, this will change with the widespread use of the HAS-BLED scheme.