HAS-BLED Bleeding Risk Score Calculator
An interactive implementation of the HAS-BLED bleeding risk score (Pisters, Lane and Lip): nine clinical factors scored 0 to 9 to estimate major bleeding risk in anticoagulated patients with atrial fibrillation.
Answer nine questions covering blood pressure control, kidney and liver function, prior stroke, bleeding history, INR stability, age, interacting drugs and alcohol to compute the HAS-BLED score from 0 to 9, with the corresponding published major bleeding rate. Use it to find and fix modifiable bleeding risks, not to deny anticoagulation on its own.
What is the HAS-BLED score?
The HAS-BLED score was developed to answer a question every clinician faces when starting anticoagulation for atrial fibrillation: how likely is this patient to bleed? Derived by Pisters, Lane and Lip from the Euro Heart Survey cohort and published in Chest in 2010, it condenses nine readily available clinical variables into a score from 0 to 9, each mapped to an observed rate of major bleeding per 100 patient-years. Its discrimination is modest, like most bleeding scores, but its real-world value lies less in the number and more in the structure: working through the nine items forces a systematic hunt for reversible bleeding risks.
The score's role in guidelines is often misunderstood. Because stroke and bleeding risk factors overlap heavily, withholding anticoagulation from everyone with a high HAS-BLED score would deny stroke prevention to the very patients who need it most. Both European and American guidance therefore position HAS-BLED as a tool for identifying and modifying bleeding risk factors and for flagging patients who need earlier review, not as a reason to withhold oral anticoagulation when it is otherwise indicated. Net clinical benefit analyses show that anticoagulation usually still favours the patient even at higher bleeding scores.
The nine factors and their definitions
Hypertension scores a point only when uncontrolled, defined in the original score as systolic pressure above 160 mm Hg, a stricter bar than the usual hypertension definition and a reminder that the item targets genuinely poor control. Abnormal renal function means dialysis dependence, prior kidney transplant, or serum creatinine above 200 micromol per litre (about 2.3 mg/dL). Abnormal liver function means cirrhosis or the combination of bilirubin more than twice the upper limit of normal with transaminases or alkaline phosphatase more than three times the upper limit. Prior stroke scores a point, as does any bleeding history or a predisposition such as anaemia.
Labile INR captures unstable anticoagulation control, conventionally time in therapeutic range below 60 percent, and applies only to patients on vitamin K antagonists; patients on direct oral anticoagulants score zero here. Elderly means age above 65. The drugs item covers concomitant antiplatelet agents and non-steroidal anti-inflammatory drugs, both of which raise bleeding risk without adding stroke protection in most atrial fibrillation patients. Alcohol excess is defined as eight or more standard drinks per week. Each item contributes exactly one point, keeping the score simple enough to compute at the bedside.
How to interpret the score
The published major bleeding rates per 100 patient-years are: 0 points, 1.13; 1 point, 1.02; 2 points, 1.88; 3 points, 3.74; 4 points, 8.70; and 5 or more points, 12.5. A score of 3 or higher is the conventional high-risk threshold, but the number should trigger action rather than inaction. For each point present, the clinician should ask whether it can be modified: can blood pressure be brought down, can the NSAID be stopped, can alcohol intake be reduced, can anaemia be investigated and treated, can anticoagulation control be tightened or the agent switched?
Context matters when reading the rate. A bleeding rate of 3.74 per 100 patient-years at a score of 3 sits alongside a stroke rate that, for the same patient, is often considerably higher without anticoagulation, and anticoagulation prevents about two-thirds of those strokes. The comparison that counts is therefore not bleeding risk versus zero but bleeding risk on treatment versus stroke risk off treatment. Documenting the score, the modifiable factors addressed, and the plan for follow-up turns HAS-BLED from a number into a management strategy.
HAS-BLED with direct oral anticoagulants
HAS-BLED was derived in the vitamin K antagonist era, and the labile INR item has no meaning for patients taking apixaban, rivaroxaban, dabigatran or edoxaban. In practice the score is still used as a structured checklist in direct oral anticoagulant users, with the labile item scored zero, and it retains value for flagging modifiable factors. Its absolute calibration is less certain in this population, so the bleeding rates should be read as approximate and clinical judgment should carry extra weight. Renal function deserves particular attention with direct agents, since most are renally cleared and dosing depends on creatinine clearance, age and weight.
Limitations
Like CHA2DS2-VASc, HAS-BLED has only modest discriminative ability, with c-statistics in the 0.6 range, and it reduces continuous variables to binary items, so a systolic pressure of 161 counts the same as 200. The bleeding rates come from a European cohort from over a decade ago, and absolute rates shift with contemporary practice, including wider use of direct oral anticoagulants and proton pump inhibitor co-prescription. The score does not capture frailty, fall risk, thrombocytopenia, or genetic bleeding tendencies. It should be recalculated when the clinical situation changes, for example after a bleed, a new interacting prescription, or a decline in renal function, rather than treated as a one-time label.
References
- Pisters R, Lane DA, Nieuwlaat R, de Vos CB, Crijns HJGM, Lip GYH. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: the Euro Heart Survey. Chest. 2010;138(5):1093-1100.
- Lip GYH, Frison L, Halperin JL, Lane DA. Comparative validation of a novel risk score for predicting bleeding risk in anticoagulated patients with atrial fibrillation: the HAS-BLED score. J Am Coll Cardiol. 2011;57(2):173-180.
- Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation. Eur Heart J. 2021;42(5):373-498.
Key takeaways
- HAS-BLED sums nine factors, uncontrolled hypertension, abnormal renal or liver function, prior stroke, bleeding history, labile INR, age over 65, bleeding-risk drugs and alcohol excess, into a score from 0 to 9.
- Published major bleeding rates run from about 1.1 per 100 patient-years at 0 points to 12.5 at 5 or more points; 3 or higher is conventionally high risk.
- A high score should prompt correction of modifiable factors and closer follow-up, not automatic withholding of anticoagulation.
- Stroke risk and bleeding risk usually rise together; net clinical benefit of anticoagulation generally persists even at higher bleeding scores.
- The labile INR item applies only to warfarin users; score it as absent for patients on direct oral anticoagulants.
- Reassess the score when the clinical picture changes, after bleeding, new interacting drugs, or worsening renal function.
Frequently asked questions
What is the HAS-BLED score used for?
HAS-BLED estimates the risk of major bleeding in patients with atrial fibrillation who are receiving oral anticoagulation. It sums one point each for uncontrolled hypertension, abnormal renal function, abnormal liver function, prior stroke, bleeding history or predisposition, labile INR, elderly age over 65, drugs that increase bleeding, and alcohol excess, for a total of 0 to 9. Its main clinical role is to flag modifiable bleeding risk factors, not to deny anticoagulation.
At what HAS-BLED score is bleeding risk considered high?
A score of 3 or more is conventionally considered high bleeding risk, corresponding to about 3.74 major bleeds per 100 patient-years in the original validation cohort. Scores of 0 to 2 indicate low to moderate risk. A high score should prompt closer review and correction of reversible factors rather than automatic withholding of anticoagulation.
Should a high HAS-BLED score stop anticoagulation?
No. Guidelines are explicit that a high HAS-BLED score alone should not lead to withholding oral anticoagulation when it is otherwise indicated, because the patients at highest bleeding risk are usually also at highest stroke risk, and the net clinical benefit of anticoagulation generally persists. Instead, the score should trigger action on modifiable factors: control blood pressure, stabilise anticoagulation intensity, review interacting medicines and alcohol, and correct anaemia.
Does HAS-BLED apply to patients on direct oral anticoagulants?
HAS-BLED was derived and validated in patients taking vitamin K antagonists, where labile INR is a meaningful variable. It is still widely used as a structured bleeding risk checklist for patients on direct oral anticoagulants, with the labile INR item scored as absent, but its calibration in that population is less certain and clinical judgment carries more weight.
What counts as labile INR in HAS-BLED?
Labile INR means unstable or highly variable INR values, commonly operationalised as time in therapeutic range below 60 percent. It applies only to patients taking a vitamin K antagonist such as warfarin; for patients on direct oral anticoagulants, which do not require INR monitoring, this item scores zero.
What is the difference between HAS-BLED and other bleeding scores?
HAS-BLED is the simplest and most widely validated bleeding score in atrial fibrillation, using nine bedside clinical items. Alternatives such as ATRIA, ORBIT and the ABC score add laboratory variables or biomarkers and may discriminate slightly better in some cohorts, but HAS-BLED remains the guideline-referenced standard because of its simplicity and its explicit focus on modifiable factors.