Carotid plaque characterization by duplex scanning: observer error may undermine current clinical trials

Area of Science:

• Medical imaging
• Cardiovascular research
• Ultrasound technology

Background:

• Current clinical studies rely on subjective ultrasound imaging for carotid plaque characterization.
• Limited data exists on the intraobserver and interobserver variability of these subjective classifications.

Purpose of the Study:

• To assess the intraobserver and interobserver variability in grading carotid plaque morphology from ultrasound images.
• To compare the reproducibility of different classification schemes and image storage methods.

Main Methods:

• Plaque morphology was graded using two classification schemes (2-category and 4-category) from both digital and hard copy ultrasound images.
• Four observers assessed interobserver agreement, and within-observer agreement was evaluated over six months.
• Accuracy was determined by comparing ultrasound grading with histological findings.

Main Results:

• Moderate to good agreement was observed for full-color digital image analyses (kappa values ranging from 0.49 to 0.62).
• Hard copy image analysis yielded lower agreement values compared to digital.
• More experienced observers demonstrated higher within-observer agreement and better correlation with histology.

Conclusions:

• Consistent and reproducible grading of ultrasound images for carotid plaque morphology is challenging, even among experienced observers.
• Variability in subjective characterization may impact clinical trial reliability.
• A simpler 2-category grading system for digital images can enhance interobserver agreement, but regular audits are recommended.