Fact-checked by Grok 2 weeks ago

HAS-BLED

The HAS-BLED score is a clinical rule designed to estimate the one-year risk of major in patients with who are receiving oral anticoagulation therapy. Developed and validated using data from the Euro Heart Survey on involving over 3,900 patients, it offers a straightforward to quantify bleeding risk through seven readily assessable factors, aiding clinicians in balancing thrombotic and hemorrhagic risks. The acronym HAS-BLED represents the following components, each assigned one point unless otherwise specified: Hypertension (uncontrolled systolic blood pressure >160 mmHg), Abnormal renal function (chronic dialysis, renal transplantation, or serum creatinine ≥200 μmol/L [≥2.26 mg/dL]), Stroke (prior history), Bleeding history or predisposition (prior major bleeding or bleeding diathesis, anemia, etc.), Labile INR (unstable/high INRs or time in therapeutic range <60% for those on vitamin K antagonists), Elderly (age >65 years), and Drugs (concomitant antiplatelet agents or nonsteroidal anti-inflammatory drugs) or alcohol (≥8 units/week) concomitantly, with one point each for drugs and alcohol. Abnormal liver function (chronic hepatic disease [e.g., cirrhosis] or significant hepatic derangement [bilirubin >2× upper limit of normal, with AST/ALT/ALP >3× upper limit]) also contributes one point, allowing for a maximum total score of 9. A score of 0 corresponds to a low annual major bleeding risk of approximately 1%, increasing progressively to over 12% for scores of 5 or higher, with scores ≥3 indicating high risk and warranting closer , modifiable , and regular rather than withholding anticoagulation. The tool has been externally validated in multiple cohorts, including those on direct oral anticoagulants, and outperforms older scores like HEMORR₂HAGES and ATRIA in predicting events, leading to its endorsement in major guidelines such as those from the .

Introduction

Definition

The HAS-BLED score is a clinical tool designed to estimate the one-year risk of major events in patients with (AF) who are candidates for oral anticoagulation therapy. It was originally published in as a user-friendly alternative to existing bleeding risk models, derived from data in the Euro Heart Survey on AF involving nearly 4,000 patients. The acronym HAS-BLED expands to represent its seven key risk factors: Hypertension, Abnormal renal/liver function, Stroke, Bleeding history or predisposition, Labile international normalized ratio, Elderly (age >65 years), and Drugs or concomitantly. Each component contributes to a simple integer score that quantifies bleeding vulnerability, aiding clinicians in balancing thrombotic and hemorrhagic risks during treatment decisions. In this context, major bleeding is defined as any event involving , bleeding requiring hospitalization, a hemoglobin decrease greater than 2 g/dL, or the need for blood transfusion. This definition aligns with established criteria for clinically significant bleeds in anticoagulated populations, emphasizing outcomes that impact patient morbidity and management.

Development and History

The HAS-BLED score was developed in 2010 by a team of researchers including Ron Pisters, Deirdre A. Lane, Robby Nieuwlaat, Cees B. de Vos, .G.M. Crijns, and Gregory Y.H. Lip, primarily affiliated with institutions in the and the . The score was first introduced in the article titled "A user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with : the Euro Heart Survey," published in the journal Chest. The development drew from data in the Euro Heart Survey on (EHS-AF), a multinational conducted by the , which analyzed 3,978 patients with who had complete one-year follow-up data. This cohort allowed for the identification and weighting of key bleeding risk factors through univariate and multivariate analyses, enabling the creation of a simple predictive model. The EHS-AF dataset provided a real-world representation of diverse patients, many on oral anticoagulation, facilitating both derivation and internal validation of the score. The primary rationale for creating HAS-BLED was to address the absence of a validated, user-friendly tool for quantifying one-year major bleeding risk in patients receiving oral anticoagulation, surpassing informal or less structured clinical assessments previously used. Prior to its introduction, bleeding risks were often evaluated without standardized quantification, leading to inconsistent in antithrombotic therapy. HAS-BLED was designed to incorporate readily available clinical and laboratory factors, promoting practical application in routine care to balance thrombotic and bleeding risks. Since its 2010 publication, the HAS-BLED score has undergone no major revisions to its core components or methodology and continues to serve as a standard bleeding risk assessment tool in major guidelines. It remains widely recommended in clinical practice for management, with ongoing comparative studies evaluating its performance against newer scores as of 2025.

Components and Scoring

Risk Factors

The HAS-BLED score assesses bleeding risk in patients with , particularly those on oral anticoagulation, through an acronym representing eight key modifiable and non-modifiable risk factors. Each factor is assigned 1 point if present, yielding a total score from 0 to 9, though the categories for abnormal renal/liver function and drugs/alcohol can contribute up to 2 points each. These criteria were derived from the Euro Heart Survey on atrial fibrillation and validated in a of 3,978 patients. Hypertension (H) refers to uncontrolled high blood pressure, specifically a systolic blood pressure greater than 160 mm Hg. This threshold indicates poor blood pressure management, which can exacerbate vascular fragility and increase bleeding propensity in anticoagulated patients. Abnormal renal function (A) is defined by the presence of chronic dialysis, prior renal transplantation, or serum creatinine levels of 200 µmol/L (equivalent to ≥2.26 mg/dL) or higher. These indicators reflect significant impairment in kidney function, which can alter drug metabolism and heighten bleeding risks due to uremic platelet dysfunction. Abnormal liver function, also under this category, includes chronic hepatic disease such as cirrhosis or biochemical abnormalities like bilirubin more than twice the upper limit of normal, alongside aspartate aminotransferase (AST), alanine aminotransferase (ALT), or alkaline phosphatase (ALP) levels exceeding three times the upper limit of normal; each subcomponent scores 1 point, for a maximum of 2 in this category. Stroke (S) encompasses any prior history of , signifying underlying cerebrovascular vulnerability that may compound bleeding hazards during anticoagulation . Bleeding history or predisposition (B) involves a previous major bleeding event or conditions predisposing to hemorrhage, such as or other bleeding diatheses. Major bleeding is typically defined as , hospitalization, hemoglobin drop greater than 2 g/dL, or transfusion requirement, highlighting patients at elevated risk for recurrent events. Labile international normalized ratio (L) denotes unstable or elevated international normalized ratios (INRs) in patients on vitamin K antagonists, specifically when the time in therapeutic range is less than 60%. This instability reflects suboptimal anticoagulation control, which correlates with fluctuating and thrombotic risks. Elderly (E) applies to patients aged older than years, as advanced age is associated with frailty, reduced physiological reserve, and higher susceptibility to bleeding complications independent of other factors. Drugs (D) includes concomitant use of antiplatelet agents (e.g., aspirin or clopidogrel) or nonsteroidal drugs (NSAIDs), which can synergistically increase risk when combined with anticoagulants; this subcomponent scores 1 point. refers to excessive consumption, defined as 8 or more per week, which may impair liver function, , and medication adherence, also scoring 1 point for a maximum of 2 in this category.

Calculation

The HAS-BLED score is computed by summing one point for each applicable risk factor present in the patient, with the acronym representing hypertension (H), abnormal renal or liver function (A), stroke (S), bleeding history or predisposition (B), labile international normalized ratio (L), elderly age (E), and concomitant drugs or alcohol use (D). Abnormal renal or liver function contributes up to two points if both are present: one point for renal criteria (chronic dialysis, renal transplantation, or serum creatinine ≥ 200 µmol/L) and one for liver criteria (chronic hepatic disease or biochemical evidence such as bilirubin > 2 × upper limit of normal, or aspartate aminotransferase/alanine aminotransferase/alkaline phosphatase > 3 × upper limit of normal). Similarly, concomitant drugs or alcohol use can contribute up to two points: one for predisposing drugs (e.g., antiplatelet agents or nonsteroidal anti-inflammatory drugs) and one for excessive alcohol consumption (≥8 units per week). The remaining components each contribute a single point if present: uncontrolled hypertension (systolic > 160 mm Hg); prior ; history or predisposition (e.g., prior major or ); labile INR (time in therapeutic range < 60%); and age > 65 years. The total score ranges from 0 to 9 points, calculated as the simple sum of these assigned points (e.g., HAS-BLED = H + A_renal + A_liver + S + B + L + E + D_drugs + D_alcohol). For example, a with uncontrolled , age > 65 years, and excessive use would receive 1 point for H, 1 point for E, and 1 point for D (), resulting in a total score of 3.

Clinical Application

Interpretation of Scores

The HAS-BLED score categorizes with into levels for major bleeding over one year, primarily based on validation in the Euro Heart Survey cohort. Scores of 0-1 indicate low , with an annual major bleeding rate of approximately 1% (0.6% for score 0 and 1.5% for score 1). A score of 2 denotes moderate , associated with an annual major bleeding rate of around 2-3%. Scores of 3 or higher signify high , with annual major bleeding rates ranging from 3% to 9% or more, escalating with each additional point (e.g., up to 8.9% for scores ≥5 in broader validations). In the original validation study, the risk of major demonstrated a significant increasing trend with higher HAS-BLED scores (p < 0.001), with each one-point increment linked to a hazard ratio of approximately 1.47 for bleeding events. This predictive pattern holds across subgroups, including those on antiplatelet therapy alone (C-statistic 0.91) or no antithrombotic therapy (C-statistic 0.85), underscoring the score's utility in estimating relative risk escalation. A high HAS-BLED score (≥3) signals the need for caution in anticoagulation management but does not contraindicate therapy, as the net clinical benefit often favors continuation after balancing stroke risk. Instead, it prompts proactive addressing of modifiable risk factors, such as optimizing blood pressure control, stabilizing INR, minimizing concomitant drugs or alcohol use, to mitigate bleeding potential while maintaining protective anticoagulation. Regular reassessment of the score is recommended to reflect changes in patient status.
HAS-BLED ScoreRisk LevelApproximate Annual Major Bleeding Rate (%)
0-1Low~1
2Moderate~2-3
≥3High3-9+ (increases per point)

Use in Practice

The HAS-BLED score is recommended by the (ESC) in its 2024 guidelines for the management of (AF) as a formal tool to assess bleeding risk in patients considered for oral anticoagulation with (VKAs) or (DOACs). Similarly, the 2023 (ACC)/ (AHA)/American College of Chest Physicians (ACCP)/ (HRS) guideline endorses HAS-BLED for evaluating bleeding risk as part of shared decision-making (SDM) in AF patients initiating anticoagulation therapy. In clinical practice, HAS-BLED is integrated with the CHA2DS2-VASc score to balance thrombotic stroke risk against bleeding risk before commencing anticoagulation in AF patients. This dual assessment ensures that anticoagulation is prescribed when stroke prevention benefits outweigh bleeding hazards, particularly in patients with elevated scores indicating high stroke risk. Practical application involves calculating the HAS-BLED score at baseline prior to starting anticoagulation to identify at-risk patients and guide initial decisions. Reassessment is advised periodically or when clinical factors change, such as the introduction of new medications like antiplatelets or NSAIDs, to dynamically adjust bleeding risk estimates and therapy. For instance, if a patient's score increases due to a new comorbidity, clinicians may opt for closer monitoring or alternative dosing strategies. The score highlights modifiable risk factors, enabling targeted interventions to mitigate bleeding potential without necessarily discontinuing anticoagulation. These include optimizing blood pressure control to address uncontrolled hypertension (a key component), avoiding concomitant use of unnecessary antiplatelet agents, and counseling on alcohol limitation to reduce excessive intake. Pharmacist-led or multidisciplinary interventions focusing on these elements have shown feasibility in reducing overall bleeding risk in anticoagulated AF cohorts. HAS-BLED is preferred over informal or subjective bleeding risk assessments, as it provides a structured, evidence-based framework to flag high-risk individuals and inform SDM, especially in elderly patients or those with multiple comorbidities where bleeding concerns may deter anticoagulation. In such scenarios, the score facilitates discussions on patient preferences, lifestyle modifications, and potential dose adjustments to promote adherence while minimizing harm.

Validation and Evidence

Clinical Studies

The HAS-BLED score was originally derived from the Euro Heart Survey on atrial fibrillation, a multinational cohort study involving 3,978 patients with atrial fibrillation conducted between 2003 and 2004. In this derivation, the score demonstrated good discrimination for one-year major bleeding risk through internal validation, with annual bleeding rates increasing progressively with higher scores, supporting its utility in a real-world population of patients receiving antithrombotic therapy. External validations have confirmed the score's applicability across diverse cohorts and anticoagulation regimens. In the RE-LY trial, a randomized controlled trial of 18,113 patients with atrial fibrillation comparing dabigatran to warfarin, post-hoc analyses applied HAS-BLED to the full cohort, including dabigatran and warfarin arms, affirming its relevance in the direct oral anticoagulant (DOAC) era. Similarly, a population-based study using Danish nationwide registries evaluated HAS-BLED in 57,930 patients with atrial fibrillation initiating non-vitamin K antagonist oral anticoagulants (NOACs), demonstrating its ability to predict major bleeding events independently of anticoagulation type. Recent studies from 2023 to 2025 have extended validations to contemporary and diverse populations, particularly in the post-DOAC era. A 2024 prospective European registry of 2,834 atrial fibrillation patients on DOACs reported consistent but modest predictive performance for major bleeding, highlighting HAS-BLED's ongoing relevance despite evolving treatment landscapes. In Asia, a Japanese multicenter registry involving 2,101 patients with atrial fibrillation validated HAS-BLED, showing its applicability in East Asian cohorts on DOACs with predictions aligning with observed bleeding incidences, though performance remained modest in this demographic. Key findings across these studies indicate that HAS-BLED predicts bleeding risk independently of anticoagulation type, with each one-point increase associated with a hazard ratio of approximately 1.2 for major bleeding events. Meta-analyses pooling data from over 100,000 patients, including one encompassing 18 studies and 321,888 individuals with atrial fibrillation, have affirmed the score's utility in identifying high-risk patients across vitamin K antagonists and DOACs, underscoring its broad clinical value despite moderate overall discrimination.

Performance Metrics

The HAS-BLED score demonstrates moderate discriminatory ability for predicting major bleeding events in patients with atrial fibrillation, with c-indices (equivalent to area under the curve, AUC) typically ranging from 0.60 to 0.70 across validation studies. In the original derivation cohort, the score achieved a c-index of 0.65 for major bleeding. A 2022 meta-analysis of real-world cohorts reported a pooled c-index of 0.63 (95% CI: 0.60–0.66), confirming consistent moderate performance. Calibration of the HAS-BLED score, which assesses agreement between observed and predicted bleeding rates, is generally good for low- to moderate-risk patients (scores up to 5), where predicted risks closely match event rates. However, underestimation occurs at higher scores (≥5), particularly in high-risk subgroups with multiple comorbidities, leading to discrepancies in observed versus predicted events. Net reclassification improvement (NRI) analyses indicate that HAS-BLED provides modest gains in risk stratification compared to no formal assessment, with positive NRI values over older scores like HEMORR2HAGES and ATRIA. When compared to contemporary scores, NRI is similar or slightly lower, as seen in evaluations where ORBIT showed marginal improvements over HAS-BLED for major bleeding reclassification. At the common high-risk cutoff of ≥3, HAS-BLED exhibits balanced sensitivity and specificity for major bleeding, with sensitivity approximately 50–60% and specificity 70–80% in anticoagulated cohorts. This threshold effectively identifies patients warranting closer monitoring, though it misses some events in lower-score groups. Recent 2025 analyses in direct oral anticoagulant (DOAC) cohorts report a c-index of 0.651 for HAS-BLED in predicting major bleeding, slightly lower than ORBIT's 0.664, underscoring its continued relevance despite minor differences in newer populations.

Limitations and Comparisons

Limitations

Despite its widespread use, the HAS-BLED score exhibits modest discriminatory ability, with c-indices typically ranging from 0.60 to 0.70 across various cohorts, which restricts its precision for individual patient predictions and makes it more suitable for population-level risk assessment rather than personalized clinical decisions. For instance, in a 2025 comparative analysis of bleeding risk scores in atrial fibrillation patients, the HAS-BLED achieved a c-index of 0.651 for major bleeding, underscoring its limited capacity to distinguish high- from low-risk individuals effectively. The score's development from the Euro Heart Survey cohort, comprising primarily European patients, has led to underrepresentation of non-Caucasian populations, with limited validation in diverse ethnic groups or low-income settings where bleeding risks may differ due to genetic, socioeconomic, or environmental factors. Studies in Asian cohorts, for example, have shown variable performance, with scores often lower than in European populations, potentially underestimating or overestimating risks in these groups due to differences in body size, comorbidities, or anticoagulant pharmacokinetics.00216-X/fulltext) Validation efforts in non-European settings remain sparse, highlighting a need for broader applicability testing. As a static tool, HAS-BLED does not incorporate dynamic changes in patient status, such as fluctuating renal function or evolving comorbidities, nor does it address direct oral anticoagulant (DOAC)-specific risks, given its derivation from warfarin-treated patients in the pre-DOAC era. This limitation is particularly evident in scenarios like progressive renal decline on DOACs, where bleeding risk may escalate without corresponding adjustments in the score. Consequently, it may fail to capture time-varying factors that influence bleeding events over the course of therapy. The inclusion of modifiable risk factors, such as uncontrolled hypertension, labile international normalized ratio, and concomitant antiplatelet use, can lead to overestimation of bleeding risk if these elements are successfully managed during follow-up, as the score does not account for such interventions. Guidelines emphasize periodic reassessment to mitigate this, but the static framework inherently limits its responsiveness to therapeutic optimizations. In the context of 2025 advancements, emerging machine learning models have demonstrated superior performance over in predicting bleeding risks within certain cohorts, achieving higher c-indices and better calibration through integration of complex, non-linear interactions among variables. Nevertheless, retains endorsement in major guidelines, such as the 2024 European Society of Cardiology recommendations, owing to its simplicity, ease of calculation without specialized tools, and established role in identifying addressable risks.

Comparison with Other Scores

The HAS-BLED score is frequently compared to the ORBIT score, which incorporates factors such as anemia (defined by hemoglobin <10 g/dL) and treatment with antiplatelet or nonsteroidal anti-inflammatory drugs, in addition to age ≥75 years, bleeding history, and abnormal renal or liver function. In patients with atrial fibrillation using direct oral anticoagulants (DOACs), recent analyses indicate that ORBIT demonstrates modestly superior discrimination for major bleeding, with c-indices ranging from 0.66 to 0.70 compared to HAS-BLED's 0.65, alongside improvements in net reclassification. However, both scores exhibit only moderate overall predictive performance (c-index <0.7), with no clear superiority in calibration for clinically relevant non-major bleeding.00526-4/fulltext) In contrast to HAS-BLED, the ATRIA score includes a broader set of variables, such as , estimated glomerular filtration rate <30 mL/min, prior severe or recent bleeding, hypertension requiring hospitalization, diabetes, and age ≥85 years, making it more complex with weighted points up to 15. Performance comparisons reveal similar discriminatory ability between ATRIA and HAS-BLED for major bleeding in anticoagulated atrial fibrillation patients (c-indices around 0.60-0.65), though HAS-BLED often shows better calibration and net reclassification for clinically relevant bleeding events due to its simplicity. ATRIA's inclusion of bleed severity and renal thresholds provides nuanced risk stratification but at the cost of increased computational burden in clinical settings. The 2023 DOAC score, developed specifically for bleeding risk in DOAC-treated atrial fibrillation patients, integrates factors like age >75 years, anemia, and prior bleeding, assigning points for DOAC-specific elements such as >1.5 mg/dL and treatment with antiplatelet agents. In validations, the DOAC score has outperformed HAS-BLED, achieving a c-index of 0.68 versus 0.65 for major bleeding prediction, particularly in diverse cohorts including East Asian populations. This edge stems from its tailoring to DOAC , though HAS-BLED remains competitive in identifying very low-risk patients.00277-2/fulltext) HAS-BLED complements the CHA2DS2-VASc score for risk assessment but shares overlaps in components like age and , necessitating integrated use to balance thrombotic and risks in management. This synergy allows clinicians to weigh net clinical benefit, as HAS-BLED addresses modifiable factors absent in CHA2DS2-VASc. Despite these alternatives, HAS-BLED is preferred in major guidelines for its simplicity and established validation across antagonists and DOACs, with scores ≥3 prompting modifiable risk interventions. Newer tools like AF-BLEED are emerging but lack widespread adoption as of 2025, as they have not yet demonstrated consistent superiority in large-scale prospective data.

References

  1. [1]
    A novel user-friendly score (HAS-BLED) to assess 1-year risk of ...
    Conclusion: This simple, novel bleeding risk score (HAS-BLED) provides a practical tool to assess the individual bleeding risk of real-world patients with AF, ...
  2. [2]
    The HAS-BLED (Hypertension, Abnormal Renal/Liver Function ...
    We recently derived and validated the HAS-BLED bleeding risk schema for AF (also called the Birmingham AF bleeding schema: Hypertension, Abnormal Renal/Liver ...
  3. [3]
    Performance of the HAS‐BLED, ORBIT, and ATRIA Bleeding Risk ...
    Nov 29, 2022 · Patients with HAS‐BLED ≥3, ATRIA ≥5, or ORBIT ≥4 were classified as at high bleeding risk. Receiver operating characteristic analysis for each ...
  4. [4]
    2024 ESC Guidelines for the management of atrial fibrillation
    Aug 30, 2024 · ESC Clinical Practice Guidelines aim to present all the relevant evidence to help physicians weigh the benefits and risks of a particular ...
  5. [5]
    A comparative analysis of HAS-BLED, ORBIT, DOAC, and AF ...
    Aug 20, 2025 · A HAS-BLED score of ≥3 indicates a high bleeding risk. The ORBIT score [5] includes data on age ≥75 years, anemia, bleeding history, chronic ...
  6. [6]
    2023 ACC/AHA/ACCP/HRS Guideline for the Diagnosis and ...
    The “2023 ACC/AHA/ACCP/HRS Guideline for the Diagnosis and Management of Atrial Fibrillation” provides recommendations to guide clinicians in the treatment ...
  7. [7]
    Use of the CHA2DS2-VASc and HAS-BLED Scores to Aid Decision ...
    Aug 14, 2012 · The HAS-BLED scores range from 0 to 9, with scores of ≥3 indicating high risk of bleeding, for which caution and regular review of the patient ...Total Citations317 · Use Of Chads-Vasc To Assess... · Bleeding Risk With...<|control11|><|separator|>
  8. [8]
    HAS-BLED Tool – What is the Real Risk of Bleeding in ...
    Sep 17, 2012 · 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.
  9. [9]
    Regular reassessment of bleeding risk using the HAS-BLED score ...
    Nov 25, 2020 · The HAS-BLED score was introduced to draw attention to modifiable bleeding risk factors and to identify patients with atrial fibrillation (AF) ...Missing: interpretation levels
  10. [10]
    Performance of HAS-BLED and DOAC scores to predict major ...
    Jul 4, 2024 · Both HAS-BLED and DOAC scores showed a significant association with the risk of major bleeding events, with only modest predictive ability. •.
  11. [11]
    Pharmacist Interventions to Reduce Modifiable Bleeding Risk ... - NIH
    The authors identified modifiable bleeding risk factors using the HAS-BLED score and evaluated pharmacist interventions to correct these factors.Missing: reassessment | Show results with:reassessment
  12. [12]
    Development and Validation of the DOAC Score: A Novel Bleeding ...
    Sep 19, 2023 · The score had superior performance to the HAS-BLED score in RE-LY (C statistic, 0.73 versus 0.60; P for difference <0.001) and among 12 296 ...
  13. [13]
  14. [14]
    Bleeding risk in East Asian patients with atrial fibrillation
    The DOAC score outperformed established bleeding risk scores, including HAS-BLED (C-index 0.523), Anticoagulation and Risk Factors in Atrial Fibrillation (C- ...
  15. [15]
    Predictive value of bleeding risk scores in elderly patients with atrial ...
    Sep 28, 2023 · The HASBLED bleeding risk score is based on 3,978 patients from the Euro Heart Survey registry, with a mean age of 65 years. ATRIA and ORBIT ...
  16. [16]
    Risk Stratification for Bleeding Complications in Patients With ...
    Mar 7, 2018 · The HAS‐BLED score has good predictive validity for bleeding risks in patients with VTE. The addition of cancer as an independent bleeding risk factor merits ...
  17. [17]
    Accuracy of HAS‐BLED and other bleeding risk assessment tools in ...
    This network meta‐analysis investigated the sensitivity and specificity of HAS‐BLED, and other bleeding risk assessment tools, to predict major bleeding events ...
  18. [18]
    Diagnostic Accuracy of the HAS-BLED Bleeding Score in VKA- or ...
    Nov 22, 2021 · The HAS-BLED score has moderate predictive abilities for bleeding risks in patients with AF regardless of type of oral anticoagulants.
  19. [19]
    Diagnostic Accuracy of the HAS-BLED Bleeding Score in VKA- or ...
    Chang et al. (5) performed a network meta-analysis of 18 studies involving 321,888 patients and found that the HAS-BLED score was the most balanced bleeding ...Missing: 100000 | Show results with:100000
  20. [20]
    Predictive Performance of HAS-BLED Score in Patients with Atrial ...
    Aug 29, 2025 · For example, results from the RE-LY trial revealed a C-statistic of 0.62 for HAS-BLED score in predicting major bleeding in an anticoagulant- ...
  21. [21]
    The HAS‐BLED Score for Predicting Major Bleeding Risk in ... - NIH
    The HAS‐BLED score originates from a 2010 European Heart Survey database, in which a score of ≥3 points is considered to be an indicator of high‐risk bleeding.<|control11|><|separator|>
  22. [22]
    A comparative analysis of HAS-BLED, ORBIT, DOAC, and AF ...
    Aug 20, 2025 · The ORBIT score showed a slight improvement in net reclassification improvement for major bleeding compared with HAS-BLED, with a modest ...
  23. [23]
    External Validation of the ORBIT Bleeding Score and the HAS-BLED ...
    Oct 1, 2019 · The ORBIT bleeding and the HAS-BLED scores both had modest discrimination ability to identify those who had bled versus who had not (C index ...
  24. [24]
    Comparative validation of HAS-BLED, GARFIELD-AF and ORBIT ...
    We found no advantage of the ORBIT over HAS-BLED score for bleeding risk prediction, even in direct oral anticoagulant users.Missing: Danish | Show results with:Danish
  25. [25]
    Filling an Important Knowledge Gap: The DOAC Score | Circulation
    Aug 25, 2023 · The current leading bleeding risk score is the HAS-BLED score. ... Scores range from 0 to 9, with scores of ≥3 indicating a high risk of bleeding.
  26. [26]
    Bleeding Risk of Direct Oral Anticoagulants in Patients With Heart ...
    Feb 5, 2021 · Renal function decline resulted in increased bleeding in patients with DOACs. DOAC dose adjustment was often indicated, associated with ...
  27. [27]
    Optimal Management of Anticoagulation Therapy in Asian Patients ...
    Jul 21, 2021 · Moreover, the HAS-BLED score should be reassessed at periodic intervals to address potentially modifiable bleeding risk factors because bleeding ...
  28. [28]
    Machine Learning Predicts Bleeding Risk in Atrial Fibrillation ...
    Feb 25, 2025 · This study compares machine learning (ML) models with conventional bleeding risk ... ML models outperformed HAS-BLED, ATRIA, and ORBIT in 1-year ...
  29. [29]
    Based Bleeding Risk Predictions in Atrial Fibrillation Patients on ...
    May 27, 2024 · Our findings demonstrate the superior performance of ML models compared to conventional bleeding risk scores and identify novel risk factors.
  30. [30]
    The 2024 European Society of Cardiology Atrial Fibrillation Guidelines
    The omission of the HAS-BLED score, a simple and validated bleeding risk ... Notably, US guidelines have retained the HAS-BLED score alongside renal ...
  31. [31]
    Comparison of HAS-BLED with other risk models for... - Medicine
    Jun 19, 2020 · The HAS-BLED, HEMORR2HAGES, ATRIA, and ORBIT scores stratified AF patients into low, moderate, and high bleeding risk categories (Table 1).
  32. [32]
    Performance of the HEMORR 2 HAGES, ATRIA, and HAS-BLED ...
    Aug 1, 2012 · HAS-BLED was the only score that demonstrated a significant predictive performance for intracranial hemorrhage (c-index: 0.75; p = 0.03). An ...Abstract · The Hemorrhages Bleeding... · DiscussionMissing: original paper
  33. [33]
    Development and Validation of the DOAC Score: A Novel Bleeding ...
    The DOAC Score had superior prediction of major bleeding risk (C Statistic: 0.67 [95% CI: 0.64-0.69]) than the HAS-BLED score (C Statistic: 0.63 [0.61-0.65], P ...
  34. [34]
    2024 ESC Guidelines for Management of Atrial Fibrillation: Key Points
    Sep 17, 2024 · For a formal risk-score-based assessment of bleeding risk, the HAS-BLED score should be considered to help address modifiable bleeding risk ...