Inter-rater Reliability and Misclassification of the ABCD(2) Score after Transient Ischemic Attack

Area of Science:

• Neurology
• Clinical Risk Assessment

Background:

• The ABCD(2) score aids clinicians in identifying high short-term stroke risk post-transient ischemic attack (TIA).
• Its widespread use in TIA risk stratification contrasts with limited data on inter-rater reliability.
• This study prospectively evaluates ABCD(2) score inter-rater reliability across different medical specialties.

Purpose of the Study:

• To assess the prospective inter-rater reliability of the ABCD(2) score in transient ischemic attack (TIA) patients.
• To compare ABCD(2) score assessments among raters from various specialties against a vascular neurology attending.
• To analyze agreement on total scores, risk categories, and individual components of the ABCD(2) score.

Main Methods:

• Prospective evaluation of emergency department patients with TIA within 48 hours of symptom onset.
• Inclusion criteria: Asymptomatic at enrollment, TIA defined as vascular-cause presumed focal cerebral/monocular symptoms <24 hours.
• Comparison of ABCD(2) scores assigned by internal medicine, emergency medicine, and neurology residents with a vascular neurology attending, using kappa statistics.

Main Results:

• Fair inter-rater reliability for ABCD(2) total score (κ=0.26) and risk category (κ=0.29) was observed.
• Raters agreed with the attending on risk category 67% of the time and total score 52% of the time.
• Component reliability varied: near perfect for age and diabetes, substantial for blood pressure, moderate for clinical features and duration.

Conclusions:

• The ABCD(2) score demonstrates only fair inter-rater reliability.
• Significant rater disagreement on ABCD(2) risk category occurred in nearly one-third of TIA patients.
• Findings suggest caution in relying solely on the ABCD(2) score for TIA risk stratification due to variability.