Inter-rater reliability of the diagnosis of vascular cognitive impairment at a memory clinic

Area of Science:

• Neurology
• Cognitive Science
• Medical Diagnostics

Background:

• Traditional consensus criteria for vascular dementia (VaD) are evolving towards data-driven approaches.
• Accurate diagnosis of vascular cognitive impairment (VCI) is crucial for patient management and research.
• Empirically-derived criteria offer a potentially more robust diagnostic framework.

Purpose of the Study:

• To evaluate the inter-rater reliability of novel, empirically-derived criteria for diagnosing vascular cognitive impairment (VCI).
• To assess the consistency of VCI classification among physicians in a memory clinic setting.

Main Methods:

• A study involving 36 patients randomly selected from a Memory Disability Clinic.
• Independent chart reviews and diagnoses by four physicians using new VCI criteria.
• Statistical analysis using intraclass correlation coefficient (ICC) to determine reliability.

Main Results:

• A high average ICC of 0.88 (95% CI = 0.80-0.93) indicated significant inter-rater reliability for VCI diagnosis.
• Discrepancies primarily occurred within VCI subtypes or between Alzheimer's disease (AD) and VCI diagnoses.
• Most diagnostic disagreements involved only a single rater.

Conclusions:

• The new empirically-derived criteria for VCI demonstrate high reliability among physicians.
• These criteria can be reliably employed in clinical practice, particularly in memory clinics.
• This reliability supports the construct validity of the VCI criteria in the absence of biological markers.