Should a single haemoglobin (Hb) measurement or an average of multiple measurements be used to assess risk in patients with haemoglobin levels close to the cut-off for low haemoglobin on the ORBIT (Ongoing Risk Blood Identification Tool) score?

ORBIT Score Hemoglobin Assessment: Single Measurement vs. Average

For patients with hemoglobin levels near the ORBIT score cut-off, use a single hemoglobin measurement rather than an average, as the ORBIT score was derived and validated using single point-in-time laboratory values. 1, 2

Evidence Supporting Single Measurement Approach

The ORBIT bleeding risk score was specifically developed and validated using single hemoglobin measurements from routine clinical practice, not averaged values. 2 The original ORBIT derivation study analyzed data from 7,411 patients in the ORBIT-AF registry, where hemoglobin/hematocrit/anemia status was assessed as a binary variable (present or absent) based on standard laboratory values at enrollment. 2

The ORBIT score assigns 2 points for "reduced hemoglobin/hematocrit/history of anemia," making it the highest-weighted single factor in the score. 1, 2 This binary classification system was designed for practical bedside use with readily available single laboratory values, not serial measurements. 2

Why Averaging Is Not Recommended

High Hemoglobin Variability in Clinical Practice

Hemoglobin levels demonstrate substantial within-patient variability over time. 1 In chronic kidney disease populations, only 30% of patients at any single time point have hemoglobin levels within a narrow target range of 11.0-12.0 g/dL, even when specifically targeted to that range. 1 A single sampling in a single patient cannot be expected to lie within a narrow hemoglobin range, but this inherent variability is already accounted for in the ORBIT score's validation. 1

Score Validation Used Point-in-Time Values

The ORBIT score demonstrated good discrimination (C-index 0.67) and superior calibration compared to HAS-BLED and ATRIA scores when validated in the ROCKET-AF trial population—all using single hemoglobin measurements. 2 Averaging values would fundamentally alter the score's performance characteristics from its validated form. 2

Practical Application Algorithm

When assessing bleeding risk with ORBIT score:

• Use the most recent hemoglobin value available from routine laboratory testing 1, 2
• If hemoglobin is borderline (within 0.5 g/dL of the anemia cut-off), consider repeating the measurement within 1-2 weeks to confirm the value, then use the most recent single measurement for scoring 3
• Do not calculate an average of multiple hemoglobin values, as this approach was not used in score derivation or validation 2
• For patients on erythropoietin therapy, hemoglobin should be measured every 1-2 weeks, and the most recent stable value should be used for ORBIT scoring 3

Important Caveats and Pitfalls

Context-Specific Hemoglobin Interpretation

Avoid using hemoglobin values during acute bleeding or immediately post-transfusion, as these do not reflect the patient's baseline bleeding risk. 1, 4 In acute bleeding scenarios, initial hemoglobin values may not reflect true blood loss due to fluid shifts and resuscitation. 4 Wait until hemodynamic stability is achieved before calculating ORBIT score. 4

Laboratory Method Considerations

Use hemoglobin rather than hematocrit when possible, as hematocrit can falsely increase by 2-4% with prolonged sample storage and is affected by hyperglycemia, while hemoglobin remains stable. 3 The ORBIT score specifically includes "reduced hemoglobin/hematocrit/history of anemia" as a composite variable, but hemoglobin is the preferred measurement. 1, 2

Anemia Definition Variability

The ORBIT score does not specify exact hemoglobin cut-offs for "reduced hemoglobin/anemia." 1, 2 In clinical practice, use standard WHO definitions (hemoglobin <13 g/dL for men, <12 g/dL for women) or document clinical anemia based on laboratory reference ranges. 5 Some studies suggest lower cut-offs may be appropriate (9.65 g/dL for children, 10.81 g/dL for women based on healthy population fifth percentiles), but the ORBIT score was validated using conventional definitions. 5

Serial Monitoring vs. Risk Scoring

While serial hemoglobin measurements are essential for monitoring disease progression or treatment response (particularly in chronic kidney disease, trauma, or acute bleeding), 3, 4 risk score calculation should use single point-in-time values as validated. 2 The distinction is critical: serial monitoring informs clinical management, but risk stratification tools must be applied as validated. 2, 6

Comparative Performance Context

Recent real-world data from 57,930 atrial fibrillation patients on non-vitamin K antagonist oral anticoagulants showed that ORBIT, HAS-BLED, and ATRIA scores all had modest predictive ability (C-statistics approximately 0.59). 7 ORBIT categorized more patients as low risk (>83%) compared to HAS-BLED (53%), with lower sensitivity (37.1% vs. 62.8%). 7 However, a 2022 European registry analysis found that ORBIT showed poorer calibration compared to HAS-BLED in contemporary NOAC users and did not provide reclassification improvement. 6

Despite these limitations, the ORBIT score remains a validated tool when applied as designed—using single hemoglobin measurements at the time of risk assessment. 1, 2, 6, 7