ORBIT Bleeding Risk Score

Learn the ORBIT bleeding score for anticoagulated AF: five items (age ≥75, anemia/Hb/Hct, bleeding history, eGFR <60, antiplatelet), 0–7 points, low/medium/high bands, derivation from ORBIT-AF, and how to use it with stroke-risk tools.

ORBIT Bleeding Risk Score — Atrial Fibrillation

Five-item score for major bleeding among patients on oral anticoagulation, derived from ORBIT-AF (O'Brien et al., Eur Heart J 2015). Intended for patients with AF already considered for or receiving OAC—not a substitute for integrated stroke/bleeding discussion.

Patient factors

Age (years)

+1 if age ≥75 years

Sex (for Hb/Hct thresholds)

MaleFemale

Hemoglobin (g/dL), optional

Hematocrit (%), optional

Reduced Hb/Hct or history of anemia (+2) — Check if documented anemia or values meet ORBIT thresholds (Hb <13 g/dL male / <12 g/dL female; Hct <40% male / <36% female), or if low Hb/Hct is entered above. Checking this applies +2 when labs are not used or are pending.Prior major or clinically relevant bleeding history (+2) — As captured in ORBIT-AF (e.g., gastrointestinal, intracranial, hemorrhagic stroke) at baseline assessment.Insufficient kidney function (+1) — eGFR <60 mL/min/1.73 m² (or CKD stage 3 or higher equivalent).Antiplatelet therapy (+1) — Aspirin, clopidogrel, prasugrel, ticagrelor, or similar (as in original registry definitions).

Disclaimer: The ORBIT score estimates population-level major bleeding rates among anticoagulated AF patients in ORBIT-AF and related analyses. It does not replace clinical judgment, does not address stroke risk (use CHA₂DS₂-VASc separately), and should not be the sole reason to start or stop anticoagulation. This tool is for education only.

ORBIT bleeding score — criteria explained

What is ORBIT?

The ORBIT bleeding risk score (Outcomes Registry for Better Informed Treatment) is a five-factor bedside tool derived from the ORBIT-AF registry to estimate risk of major bleeding in patients with atrial fibrillation who are taking oral anticoagulation. Point weights reflect the strength of association with bleeding in multivariable models (O'Brien et al., Eur Heart J 2015; PMID 26424865).

Components (sum all that apply)

• Older age: 75 years or older — +1
• Reduced hemoglobin / hematocrit / history of anemia: Hemoglobin <13 g/dL (men) or <12 g/dL (women), or hematocrit <40% (men) or <36% (women), or documented history of anemia as defined in the registry — +2
• Bleeding history: Prior bleeding events captured at baseline (e.g., gastrointestinal, intracranial, hemorrhagic stroke) — +2
• Insufficient kidney function: eGFR <60 mL/min/1.73 m² — +1
• Antiplatelet therapy: Aspirin, clopidogrel, prasugrel, ticagrelor, or related agents as used in the original analyses — +1

Maximum total score = 7. Major bleeding followed ISTH-oriented definitions in the primary publication (fatal, critical-site/symptomatic, Hb drop ≥2 g/dL or ≥2 units transfused, etc.).

Risk groups (ORBIT-AF, anticoagulated cohort)

Published categories compare observed major bleeding rates per 100 patient-years over follow-up:

Category	ORBIT score	Observed major bleeding (approx.)
Low	0–2	~2.4 per 100 patient-years (combined stratum)
Medium	3	~4.7 per 100 patient-years
High	≥4	~8.1 per 100 patient-years (combined stratum)

Bleeding rates for each integer score (0–7) are also reported in the primary paper and can be displayed in the calculator results. External validation (e.g., ROCKET-AF) showed ORBIT had favorable calibration vs some older scores.

Discrimination (context)

In ORBIT-AF, the five-item ORBIT score C-index for major bleeding was about 0.67 at two years—similar to the full regression model and competitive with other bleeding scores, with better calibration in some external comparisons. Use alongside stroke risk tools and individual patient factors.

Reference: O'Brien EC, Simon DN, Thomas LE, et al. The ORBIT bleeding score: a simple bedside score to assess bleeding risk in atrial fibrillation. Eur Heart J. 2015 Dec 7;36(46):3258-64. doi: 10.1093/eurheartj/ehv476. PMID: 26424865; PMCID: PMC4670965.

Background and purpose

Atrial fibrillation (AF) increases the risk of ischemic stroke and systemic embolism. Oral anticoagulation (OAC) reduces stroke risk substantially in eligible patients, but clinicians and patients must weigh that benefit against the risk of major bleeding. Several bleeding-risk scores exist; many were derived in selected cohorts, require data that are not always available, or have shown uneven calibration when applied outside their original populations.

The ORBIT bleeding score was developed from the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF), a large, community-based U.S. registry of patients with electrocardiographically confirmed AF. The goal was a simple, bedside-friendly score using routinely collected clinical information to estimate risk of major bleeding among patients already on OAC, with emphasis on practical usability and reasonable performance in external populations.

Derivation cohort and outcome

The primary analysis focused on ORBIT-AF participants who were taking OAC at baseline and had follow-up data. Major bleeding was defined using criteria aligned with accepted clinical trial and registry standards for “major bleed”: this includes fatal bleeding; symptomatic bleeding in critical sites (for example intracranial, intraocular, retroperitoneal, or other specified serious locations); or bleeding associated with a substantial hemoglobin decrease or need for transfusion, as operationalized in the published work.

Over a median follow-up near two years, major bleeding occurred at a clinically meaningful rate in the anticoagulated cohort, supporting the need for tools that stratify bleeding risk without over-relying on specialized laboratory or procedural data.

Model development approach

Investigators began with a broad set of candidate predictors collected at the baseline registry visit. A multivariable time-to-event model (Cox proportional hazards) evaluated independent associations with time to major bleeding, accounting for site clustering as described in the primary publication. Continuous predictors were examined for non-linear relationships, with spline-based handling where appropriate.

From the full model, the authors retained the five strongest predictors (by contribution to the model) to create a transparent integer score. Points were assigned according to the relative strength of each predictor’s association with bleeding, yielding a compact scale that preserves much of the discrimination of the fuller model while remaining easy to compute in clinic.

The five ORBIT components

Each component below contributes points that are summed to produce the total ORBIT score (range 0–7 in the published integer scoring scheme).

1. Older age (≥75 years) — 1 point

Age is a consistent marker of bleeding risk in anticoagulated populations, reflecting vascular fragility, comorbidity burden, falls risk, renal physiology changes, and polypharmacy. In ORBIT, “older age” is operationalized as 75 years or older at the time of assessment.

2. Reduced hemoglobin, reduced hematocrit, or history of anemia — 2 points

This item captures both current laboratory evidence of anemia or low red-cell indices and documented history of anemia as recorded in the registry definitions. Threshold examples used in the score’s specification include hemoglobin below sex-specific cutoffs (commonly cited as <13 g/dL in men and <12 g/dL in women) and hematocrit below sex-specific cutoffs (commonly cited as <40% in men and <36% in women), in addition to clinician-documented anemia when those thresholds are not met at the moment of assessment.

This component receives two points, reflecting its strong association with future bleeding in the derivation analyses. Clinically, anemia may indicate occult blood loss, marrow stress, nutritional deficiency, chronic inflammation, or prior bleed-related consumption—each of which can modify both risk and the margin of safety with antithrombotic therapy.

3. Bleeding history — 2 points

Prior bleeding—particularly gastrointestinal bleeding, intracranial hemorrhage, or hemorrhagic stroke documented at baseline—marks patients whose hemostatic reserve, mucosal integrity, or vascular vulnerability has already been stress-tested. In registry-based scores, “bleeding history” is defined by what was reliably captured at enrollment; in practice, clinicians should interpret this item using the same spirit: clinically significant prior bleeding events that inform future risk, not trivial epistaxis without medical evaluation.

Like anemia, this item carries two points, underscoring its prognostic weight in the ORBIT derivation.

4. Insufficient kidney function — 1 point

Kidney dysfunction alters drug exposure for many anticoagulants, influences platelet function and endothelial biology, and often coexists with hypertension and vascular disease. The ORBIT score assigns one point for estimated glomerular filtration rate (eGFR) below 60 mL/min/1.73 m², consistent with stage 3 chronic kidney disease or worse by conventional eGFR staging.

When applying this item, use the same creatinine-based eGFR estimate your institution standardizes (for example CKD-EPI), recognize instability during acute illness, and remember that dosing adjustments for direct oral anticoagulants are drug-specific and not implied by the score itself.

5. Treatment with antiplatelet agents — 1 point

Concomitant antiplatelet therapy increases bleeding risk in anticoagulated AF patients, especially when combined therapy is used for longer durations or in patients with additional risk factors. The ORBIT item includes commonly used agents such as aspirin and P2Y12 inhibitors (for example clopidogrel, prasugrel, ticagrelor), consistent with the antithrombotic combinations captured in the registry analyses.

This item is worth one point. It should be interpreted as a marker of added pharmacologic bleeding pressure; decisions about whether combination therapy is necessary (for example recent coronary stenting) remain separate from the score’s arithmetic.

How to calculate and interpret the total score

Add the points from all applicable items. The maximum total is 7 when every item is present. The published work emphasizes interpretive bands that map to observed bleeding rates in the anticoagulated ORBIT-AF cohort:

Risk group (published categorization)	ORBIT score	Interpretive anchor
Low	0–2	Lowest published risk stratum as a group; bleeding is not absent, but modeled rates are lower than higher strata.
Medium	3	Intermediate stratum with higher observed bleeding than the low group.
High	≥4	Highest published risk grouping; modeled major bleeding rates rise substantially versus lower scores.

Many publications also report bleeding rates per 100 patient-years for each integer score (0 through 7), illustrating a graded increase in observed events as the score rises. This granularity can help patients understand that risk is continuous even though clinical decisions often use broader bands.

Performance characteristics (what the score can and cannot claim)

In ORBIT-AF, the five-item ORBIT score demonstrated discrimination for major bleeding broadly similar to a fuller multivariable model, with a reported C-index in the range often summarized as about 0.67 for two-year major bleeding (as presented in the primary report). Discrimination quantifies how well the score ranks patients who bleed versus those who do not; it does not, by itself, guarantee perfect probability estimates in every new population.

An important motivation for ORBIT was calibration—how closely predicted bleeding frequencies match observed frequencies when the tool is transported to new datasets. In external evaluation contexts described in the primary literature, ORBIT showed more favorable calibration patterns than some older bleeding scores in certain comparisons, particularly when examined alongside widely used alternatives. This property matters for shared decision-making because well-calibrated tools reduce systematic over- or under-estimation of risk across risk strata.

The score is not a diagnostic test for bleeding diathesis, does not replace medication-specific contraindication review, and does not incorporate every potential modifier (for example fall risk, alcohol use, or procedural planning) unless those factors are indirectly reflected in the five items.

Relationship to stroke risk assessment

Bleeding risk should not be interpreted in isolation. For nonvalvular AF, stroke/embolism risk is commonly estimated with tools such as CHA₂DS₂-VASc (or related schemas), while bleeding tools inform the safety margin. A patient may have a high ORBIT score yet still have compelling indications for anticoagulation when stroke risk is high; conversely, a low ORBIT score does not remove bleeding risk or eliminate the need for monitoring.

Practical anticoagulation decisions should integrate stroke risk, bleeding risk, values and preferences, renal function, drug interactions, adherence supports, and follow-up plans (including INR monitoring for vitamin K antagonists or appropriate laboratory follow-up for direct oral anticoagulants per guideline and label).

How ORBIT compares conceptually to other bleeding scores

Scores such as HAS-BLED and ATRIA remain widely referenced. ORBIT differs in derivation population (large U.S. outpatient registry of contemporary AF), endpoint definition handling, and the specific five-item structure. Clinicians may encounter situations where scores disagree because they weight different domains. In those cases, the disagreement itself is clinically informative: it prompts review of modifiable factors (blood pressure control, avoidable antiplatelet exposure, NSAID use, alcohol, fall mitigation, anemia work-up) rather than automatic cessation of anticoagulation.

Limitations and prudent use

Population transportability: Performance may shift in patients markedly different from registry enrollees (for example critically ill inpatients, perioperative contexts, or populations with very different comorbidity profiles).
OAC era and regimen mix: Derivation included predominantly vitamin K antagonist users with a smaller fraction on early direct oral anticoagulant use; contemporary practice mixes may change absolute bleeding rates even if ranking remains useful.
Data integrity: Score accuracy depends on faithful application of each item’s definition (especially anemia thresholds and “bleeding history” documentation).
Not a substitute for judgment: The ORBIT score supports risk communication and systems-level triage; it should not be the sole determinant of therapy, nor used to label patients as “unsafe” for anticoagulation without individualized reasoning.

Educational content for clinical context only; not individualized medical advice.