What is the HAS‑BLED score, its components, scoring range, and how is it used to assess bleeding risk in atrial fibrillation patients on oral anticoagulation?

HAS-BLED Score: Components, Scoring, and Clinical Application

What is the HAS-BLED Score?

The HAS-BLED score is a validated bleeding risk assessment tool that should be calculated in all atrial fibrillation patients on oral anticoagulation to identify modifiable bleeding risk factors and guide monitoring intensity—not to withhold anticoagulation. 1

Score Components and Point Values

The HAS-BLED acronym represents nine clinical parameters, each worth 1 point (maximum score = 9): 2

• H - Hypertension (uncontrolled, systolic BP >160 mmHg) 1
• A - Abnormal renal function (dialysis, transplant, or creatinine >200 μmol/L) OR Abnormal liver function (cirrhosis or bilirubin >2x normal or AST/ALT/ALP >3x normal) 1, 2
• S - Stroke (prior history) 2
• B - Bleeding history or predisposition (prior major bleed, anemia, thrombocytopenia) 1, 2
• L - Labile INRs (time in therapeutic range <60% in VKA-treated patients) 1
• E - Elderly (age >65 years) 2
• D - Drugs (concomitant antiplatelet agents or NSAIDs) OR alcohol (≥8 drinks/week) 1, 2

Scoring Range and Risk Stratification

The score ranges from 0-9 points with corresponding annual major bleeding rates: 1

• Score 0: 0.59-1.13% annual bleeding risk 1
• Score 1: 1.02-1.51% annual bleeding risk 1
• Score 2: 1.88-3.20% annual bleeding risk 1
• Score 3: 3.74-19.51% annual bleeding risk 1
• Score ≥4: 8.70-21.43% annual bleeding risk 1

A score ≥3 defines "high bleeding risk" requiring more frequent monitoring and follow-up. 1

Clinical Application and Key Principles

Primary Use: Identify Modifiable Risk Factors

The HAS-BLED score should be used to address modifiable bleeding risk factors, NOT as a reason to withhold anticoagulation. 1 The 2018 CHEST guidelines provide a strong recommendation that bleeding risk assessment should focus on potentially modifiable factors at every patient contact. 1

Modifiable risk factors include: 1

• Uncontrolled blood pressure (treat to target <160 mmHg systolic)
• Labile INRs in VKA patients (improve anticoagulation control)
• Alcohol excess (counsel on reduction)
• Concomitant NSAIDs or aspirin (discontinue if not essential)
• Bleeding predisposition (treat peptic ulcers, optimize renal/liver function)

Monitoring Frequency Based on Score

Patients with HAS-BLED ≥3 warrant more frequent and regular clinical reviews. 1 The score is highly dynamic—follow-up or "delta HAS-BLED" scores are more predictive of major bleeding than baseline scores alone. 1

Validation and Performance

The HAS-BLED score demonstrates moderate predictive ability with C-statistics of approximately 0.63-0.72 across multiple populations. 1, 2, 3 Critically, HAS-BLED is the only bleeding score validated to predict intracranial bleeding specifically. 1

The score has been validated in: 1

• VKA-treated patients
• NOAC-treated patients
• Patients on no antithrombotic therapy
• Patients on antiplatelet therapy alone
• Different ethnic populations
• Both AF and non-AF cohorts

Comparison to Other Bleeding Scores

HAS-BLED performs better than HEMORR2HAGES and ATRIA scores for bleeding prediction, with superior sensitivity (62.8% vs 29.7% for ATRIA and 37.1% for ORBIT). 1, 4 The ORBIT score categorizes more patients as "low risk" (>83% vs 53% for HAS-BLED), potentially missing patients who require closer monitoring. 4

For VKA-treated patients specifically, HAS-BLED is preferred because it includes labile INR/time in therapeutic range, which ORBIT and ATRIA do not assess. 1

Critical Clinical Caveats

Never Withhold Anticoagulation Based on Score Alone

A high HAS-BLED score (≥3) is rarely a reason to avoid anticoagulation—the benefit of oral anticoagulation usually outweighs bleeding risk even in high-risk patients. 1 The net clinical benefit is often greatest in high bleeding risk patients because they typically also have high stroke risk. 5

Dynamic Reassessment is Mandatory

Bleeding risk assessment must be performed at every patient contact, as bleeding risk changes over time. 1 Static baseline scores are inferior to serial reassessment for predicting actual bleeding events. 1

Do Not Withhold Anticoagulation for Fall Risk

Do not withhold oral anticoagulation solely because the patient is at risk of falls. 1 This is explicitly stated in the 2018 CHEST guidelines as a strong recommendation. 1

Relationship to Thrombotic Risk

The HAS-BLED score also shows predictive value for cardiovascular events (HR 1.51) and all-cause mortality (HR 1.68), reflecting the relationship between thrombosis and bleeding risk factors. 5 However, it remains designed primarily for bleeding prediction, and specific stroke risk scores (CHA₂DS₂-VASc) should be used for thrombotic risk assessment. 5