Improved reliability of the NIH Stroke Scale using video training. NINDS TPA Stroke Study Group

Area of Science:

• Neurology
• Clinical Trials
• Medical Education

Background:

• Multicenter stroke trials rely on clinical rating scales.
• Lack of standardized investigator training hinders consistent scale use.
• Need for reliable assessment in acute stroke therapy studies.

Purpose of the Study:

• To train and certify investigators in using the National Institutes of Health Stroke Scale (NIHSS).
• To ensure comparable patient ratings among examiners in acute stroke therapy trials.

Main Methods:

• Developed a two-camera videotape method for optimal presentation of examination findings.
• Assessed training effectiveness by having investigators evaluate videotaped patients.
• Calculated interobserver agreement using a devised scoring system.

Main Results:

• Trained and certified 162 investigators.
• Achieved moderate to excellent agreement on most NIHSS items (kappa > 0.60).
• Identified facial paresis and ataxia as items needing revision (kappa < 0.40).
• Video training improved performance compared to previous studies.
• Motor rating of unaffected limbs did not impact score reliability.

Conclusions:

• Video training and certification provide a practical, effective method for standardizing scale use.
• A two-camera setup is crucial for accurate clinical finding presentation.
• This approach is adaptable for multi-center trials with numerous investigators.