Assessment of interobserver variability in a Dutch multicenter study on acute ischemic stroke.

Area of Science:

• Neurology
• Clinical Trials
• Biostatistics

Background:

• Quantitative patient data is crucial in multicenter clinical studies.
• Assessing interobserver agreement is vital when multiple centers are involved.

Purpose of the Study:

• To evaluate interobserver agreement in acute ischemic stroke assessments.
• To identify factors contributing to variability in clinical data collection.

Main Methods:

• Conducted a multicenter study involving twelve acute ischemic stroke patients.
• Four investigators assessed patients using the Mathew scale.
• Calculated interobserver agreement using kappa statistics.

Main Results:

• Significant interobserver variability was observed in the assessments.
• Agreement was notably lower for items relying on subjective patient information.
• Potential time-dependent changes in subjective information may impact agreement.

Conclusions:

• Interobserver variability poses a challenge in multicenter stroke studies.
• Assessment scales should be designed to minimize subjective items.
• Avoiding items with low interobserver agreement is recommended for scale development.