AskMed

HAS-BLED score interpretation: what each score from 0 to 9 means

HAS-BLED score interpretation in short: the score counts nine bleeding risk factors in atrial fibrillation, so it runs from 0 to 9. A score of 0 to 2 is below the high-risk cut-off, and 3 or more is high bleeding risk, which European guidelines treat as a reason to correct the risk factors that can change and to review the patient more often, not as a reason on its own to withhold anticoagulation. Here is what each score means, with the bleed rates from the original study.

Open the HAS-BLED calculator

What a HAS-BLED score means

A HAS-BLED score is the number of bleeding risk factors a patient with atrial fibrillation has, out of nine, so the HAS-BLED score range is 0 to 9. Pisters, Lip and colleagues built the score from 3,978 patients in the Euro Heart Survey and published it in Chest in 2010 to estimate the risk of a major bleed within 1 year. Major bleeding meant bleeding that was intracranial, led to hospital admission, caused a fall in hemoglobin or needed a transfusion. Each letter scores 1 point: uncontrolled Hypertension, Abnormal kidney and Abnormal liver function (1 point each), previous Stroke, Bleeding history or predisposition, Labile INR, Elderly (over 65), and Drugs and alcohol (1 point each). In the 3,071 patients the paper scored, major bleeding rose from about 1 per 100 patient-years at scores of 0 and 1 to 8.70 at a score of 4 and 12.50 at a score of 5. The 2020 ESC guideline calls a score of 3 or more high bleeding risk.

Major bleeds per 100 patient-years at each HAS-BLED score in the Euro Heart Survey (Pisters et al., 2010), from 798, 1,286, 744, 187, 46 and 8 patients. Grey bars are below the cut-off of 3, blue bars at or above it. Only 2 patients scored 6 or more, so there is no rate for 6 to 9. The HAS-BLED calculator shows the same figures.

HAS-BLED score of 0, 1 or 2: below the high-risk cut-off

A HAS-BLED score of 0, 1 or 2 means the patient sits below the high-risk cut-off of 3 used by the 2020 ESC atrial fibrillation guideline. Most patients in the Euro Heart Survey fell here: 798 scored 0, 1,286 scored 1 and 744 scored 2, together 2,828 of the 3,071 patients scored (92%). Their major bleeding rates were 1.13, 1.02 and 1.88 per 100 patient-years. The rate at a score of 0 came out slightly higher than at 1 because each figure rests on few events, 9 and 13 bleeds, so scores of 0 and 1 are best read as the same level, about 1 major bleed per 100 patient-years. A score of 2 is nearly double that. A low score is not zero risk, and it can change quickly: a single new item, such as starting an antiplatelet agent or a nonsteroidal anti-inflammatory drug, moves a patient from 2 to 3. At a score of 2, check which of the two items could still change.

HAS-BLED score of 3 or more: high bleeding risk

A HAS-BLED score of 3 or more is what the 2020 ESC atrial fibrillation guideline calls high bleeding risk. In the Euro Heart Survey, only 243 of the 3,071 patients scored (7.9%) reached 3 or more, and the bleed rate climbed steeply across them: 3.74 major bleeds per 100 patient-years at a score of 3 (7 bleeds in 187 patients), 8.70 at 4 (4 bleeds in 46 patients) and 12.50 at 5 (1 bleed in 8 patients). Just 2 patients scored 6 or more and neither bled, so the paper gives no rate for scores of 6 to 9, and the HAS-BLED calculator says so rather than guessing. A score of 3 roughly doubles the rate at 2, and a score of 4 more than doubles it again, but the uncertainty around the higher rates is wide because they rest on 12 bleeds in all. In the ESC reading, a high score asks for the changeable factors to be corrected and for earlier, more frequent review.

HAS-BLED scoreMeaningEuro Heart Survey rate
0–2Below the high-risk cut-off1.02 to 1.88 per 100 patient-years
3High bleeding risk3.74 (187 patients)
4High bleeding risk8.70 (46 patients)
5High bleeding risk12.50 (8 patients)
6–9High bleeding riskNo rate: 2 patients, no bleeds

What guidelines say about a high HAS-BLED score

Guidelines read a high HAS-BLED score as a reason for closer care, not as a reason on its own to withhold anticoagulation. The 2020 European Society of Cardiology (ESC) atrial fibrillation guideline says: “For a formal risk-score-based assessment of bleeding risk, the HAS-BLED score should be considered to help address modifiable bleeding risk factors, and to identify patients at high risk of bleeding (HAS-BLED score ≥3) for early and more frequent clinical review and follow-up.” The same guideline says that estimated bleeding risk, in the absence of an absolute contraindication, should not by itself decide whether to use oral anticoagulation for stroke prevention. The 2024 ESC guideline goes further. It recommends assessing and managing modifiable bleeding risk factors in every patient eligible for oral anticoagulation, advises against using bleeding risk scores to decide on starting or withdrawing anticoagulation, and no longer names HAS-BLED as its score of choice. Stroke risk is weighed separately, with CHA₂DS₂-VA in the 2024 guideline.

Stroke risk can be scored with the CHA₂DS₂-VASc calculator. The UK NICE guideline uses the ORBIT score for bleeding instead; how the two compare is covered in HAS-BLED vs ORBIT.

Which HAS-BLED items are modifiable

Four of the nine HAS-BLED items can usually be changed, which is why the 2020 ESC guideline uses the score to help address modifiable bleeding risk factors, not only to rank patients. Uncontrolled hypertension, a systolic blood pressure over 160 mmHg, can be brought under control. Labile INR, meaning unstable or high INRs or a time in therapeutic range under 60%, can improve with better INR control. Drugs predisposing to bleeding, antiplatelet agents and nonsteroidal anti-inflammatory drugs, can be reviewed and stopped when they are not needed. Alcohol excess, 8 or more units a week, can be cut down. Each of these that is fixed takes 1 point off the score. Two items are partly changeable: the bleeding predisposition in the B item includes anemia, which may be treatable, and kidney or liver function can sometimes improve. Age over 65, a previous stroke, a past major bleed, dialysis, a kidney transplant and cirrhosis cannot be undone, so those points stay.

HAS-BLED itemCan it change?
H: systolic over 160 mmHgYes
L: labile INRYes
D: antiplatelets or NSAIDsYes
D: alcohol excessYes
B: anemiaSometimes
A: kidney functionSometimes, not dialysis or a transplant
A: liver functionSometimes, not cirrhosis
B: past major bleedNo
S: previous strokeNo
E: age over 65No

Common HAS-BLED scoring mistakes

Most HAS-BLED scoring mistakes come from reading a letter loosely instead of using the definition from the 2010 paper, and each one shifts the score by a point, which matters most around the cut-off of 3. The H item is uncontrolled hypertension, a systolic pressure over 160 mmHg, not a diagnosis of hypertension: a patient whose treated blood pressure is controlled scores 0, unlike in CHA₂DS₂-VASc, where a history of hypertension counts. The A and D letters each carry up to 2 points, because kidney and liver function are scored separately and so are drugs and alcohol. The kidney item needs chronic dialysis, a kidney transplant or a serum creatinine of 200 µmol/L (2.26 mg/dL) or more; an eGFR under 60 alone does not count. The E item is age over 65, so a patient of exactly 65 scores 0. These errors push scores both up and down, so the score is only as good as the chart review behind it.

  • Counting known hypertension. Only uncontrolled pressure, systolic over 160 mmHg, scores.
  • Scoring A or D once. Abnormal kidney and liver function are 1 point each; so are drugs and alcohol.
  • Using eGFR for the kidney item. HAS-BLED uses dialysis, transplant or creatinine 200 µmol/L or more.
  • Ticking labile INR without INR records. The item rests on unstable or high INRs or a time in therapeutic range under 60%, so it needs a monitored INR history.
  • Scoring once and never again. Blood pressure, drugs, alcohol and kidney function change, and so does the score.
  • Reading 3 as a stop sign. Neither ESC guideline uses a high score alone to withhold anticoagulation.

Worked example on the HAS-BLED calculator

The HAS-BLED calculator opens on a worked example with two boxes ticked: uncontrolled hypertension, systolic over 160 mmHg, and age over 65. It shows a score of 2 out of 9, the label “Below the high-risk cut-off of 3”, and the note “Euro Heart Survey: 1.88 major bleeds per 100 patient-years at a score of 2.” Now take a 72-year-old with atrial fibrillation whose systolic pressure is 172 mmHg, who takes a nonsteroidal anti-inflammatory drug for knee pain and drinks more than the eight units of alcohol a week that HAS-BLED counts as excess. Ticking drugs and alcohol as well gives 4 points: the calculator switches to “High bleeding risk (3 or more)” and shows 8.70 major bleeds per 100 patient-years at a score of 4. Three of those four points can change. If the blood pressure is controlled, the anti-inflammatory drug is stopped and alcohol drops below eight units a week, only age is left: a score of 1, with 1.02 major bleeds per 100 patient-years.

At a score of 5 the calculator adds that the 12.50 rate comes from only 8 patients, and from 6 upwards it says too few Euro Heart Survey patients scored that high to give a rate. That is the HAS-BLED score explained in one patient: the meaning of a score depends on which items produce it, because only some of them can change. Try the same steps in the HAS-BLED calculator.

Limits of HAS-BLED score interpretation

The HAS-BLED score predicts bleeding only moderately, so a score describes a group's average risk, not one patient's outcome. In the 2010 derivation, the risk factors had a C statistic of 0.72 overall, and HAS-BLED gave similar values in subgroups, rising to 0.91 in patients on antiplatelet agents alone and 0.85 in those on no antithrombotic therapy. That derivation rested on just 53 major bleeds, and patients with mitral stenosis or valve surgery were left out. External tests found weaker discrimination. In 7,329 anticoagulated patients from the SPORTIF III and V trials, Lip and colleagues found HAS-BLED the best of the schemes tested, with bleeding rising step by step with the score, though C statistics in the whole cohort ranged from 0.50 to 0.67. In 182,678 patients in the Swedish Atrial Fibrillation cohort, HAS-BLED and HEMORR₂HAGES both had C statistics of about 0.6. A 2021 meta-analysis of 39 studies by Gao and colleagues pooled 0.63, the same with vitamin K antagonists and direct oral anticoagulants.

Questions

What does a HAS-BLED score of 2 mean?

A score of 2 is the top of the range below the high-risk cut-off of 3. In the Euro Heart Survey, patients scoring 2 had 1.88 major bleeds per 100 patient-years, against 1.13 at 0 and 1.02 at 1.

What does a HAS-BLED score of 3 mean?

A score of 3 is high bleeding risk in the 2020 ESC guideline, with 3.74 major bleeds per 100 patient-years in the Euro Heart Survey. The guideline uses it to flag patients for earlier and more frequent review and for correction of modifiable risk factors.

Does a high HAS-BLED score mean no anticoagulation?

No. The 2020 ESC guideline says bleeding risk alone, without an absolute contraindication, should not decide against oral anticoagulation, and the 2024 guideline advises against using bleeding scores to decide on starting or stopping it.

Is HAS-BLED still in the 2024 ESC guideline?

The 2024 ESC guideline does not name HAS-BLED as its bleeding score. It asks for modifiable bleeding risk factors to be assessed and managed in every patient eligible for oral anticoagulation, and HAS-BLED's items remain a quick checklist for that.

Can AskMed explain a HAS-BLED score with sources?

Yes. AskMed answers from guidelines and journals, with a citation after every claim, and does not give doses. It is on the App Store and Google Play, or use it on the web.

Sources

  1. 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. doi:10.1378/chest.10-0134
  2. 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. doi:10.1016/j.jacc.2010.09.024
  3. Friberg L, Rosenqvist M, Lip GYH. Evaluation of risk stratification schemes for ischaemic stroke and bleeding in 182 678 patients with atrial fibrillation: the Swedish Atrial Fibrillation cohort study. Eur Heart J. 2012;33(12):1500–1510. doi:10.1093/eurheartj/ehr488
  4. Gao X, Cai X, Yang Y, Zhou Y, Zhu W. Diagnostic accuracy of the HAS-BLED bleeding score in VKA- or DOAC-treated patients with atrial fibrillation: a systematic review and meta-analysis. Front Cardiovasc Med. 2021;8:757087. doi:10.3389/fcvm.2021.757087
  5. Hindricks G, Potpara T, Dagres N, et al. 2020 ESC guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with EACTS. Eur Heart J. 2021;42(5):373–498. doi:10.1093/eurheartj/ehaa612
  6. Van Gelder IC, Rienstra M, Bunting KV, et al. 2024 ESC guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Eur Heart J. 2024;45(36):3314–3414. doi:10.1093/eurheartj/ehae176

Cited medical answers
you can trust.

AskMed on iPhone, iPad, Android and the web.