RACE Score Parameters
The RACE (Rapid Arterial oCclusion Evaluation) scale consists of 5 clinical parameters scored 0-9 total: facial palsy (0-2 points), arm motor function (0-2 points), leg motor function (0-2 points), head/gaze deviation (0-1 point), and aphasia or agnosia (0-2 points). 1
Individual Parameter Scoring
Facial Palsy (0-2 points)
- Scored from 0 to 2 based on the degree of facial weakness 1
- This parameter contributes significantly to LVO detection, and simplifying it reduces correct classification by 9% 2
Arm Motor Function (0-2 points)
- Evaluates upper extremity motor strength on a 0-2 scale 1
- Assesses motor deficit severity in the affected arm 3
Leg Motor Function (0-2 points)
- Evaluates lower extremity motor strength on a 0-2 scale 1
- Measures motor deficit severity in the affected leg 3
Head/Gaze Deviation (0-1 point)
- Binary assessment (present or absent) worth 0-1 point 1
- Removing this parameter decreases correct classification by 9% 2
Aphasia or Agnosia (0-2 points)
- Cortical sign assessment scored 0-2 points 1
- Evaluates language dysfunction or inability to recognize objects 3
- Removing this parameter reduces correct classification by 4.5% 2
Clinical Performance Characteristics
The optimal threshold for LVO detection is RACE ≥5, which demonstrates sensitivity of 66-85%, specificity of 68-79%, positive predictive value of 29-67%, and negative predictive value of 80-94%. 1, 4, 5, 3
The RACE scale was specifically designed from NIHSS items with the highest predictive value for large vessel occlusion, making it practical for prehospital use by emergency medical technicians 1. The scale shows strong correlation with NIHSS (r=0.76) and comparable discriminatory capacity (AUC 0.72-0.82 for RACE versus 0.85 for NIHSS) 1, 4, 3.
Performance varies by time window: RACE ≥5 shows higher sensitivity for LVO detection in patients presenting within 6 hours (sensitivity 0.74, specificity 0.79) compared to the 6-24 hour window (sensitivity 0.65, specificity 0.80). 3
Common Pitfalls
The original 5-item RACE scale should not be simplified, as any modification reduces predictive accuracy 2. While false positive rates can reach 61%, the majority of these patients (approximately 43%) still require comprehensive stroke center care for conditions like non-LVO strokes, intracranial hemorrhages, or other neurologic emergencies requiring specialized intervention 5.
The RACE scale performs optimally when used at thresholds of ≥3 to ≥5 depending on the clinical setting, with ≥5 providing the best balance of sensitivity and specificity in most validation studies 6, 1, 4.