Related Experiment Videos
Interconversion of stroke scales. Implications for therapeutic trials
K W Muir1, D G Grosset, K R Lees
1University Department of Medicine and Therapeutics, Gardiner Institute, Western Infirmary, Glasgow, Scotland.
Stroke
|July 1, 1994
Summary
Stroke scale conversion is reliable between the Canadian Neurological Scale and Middle Cerebral Artery Neurological Score. However, the National Institutes of Health stroke scale is not reliably convertible, especially for patients with dysphasia and total anterior circulation strokes.
Area of Science:
- Neurology
- Clinical Trials
- Biostatistics
Background:
- Stroke scales are crucial for measuring stroke severity in clinical trials.
- Existing scales are used for trial entry, patient group comparison, and as secondary endpoints.
- The validity of using stroke scale scores in meta-analysis remains unproven due to potential interconversion issues.
Purpose of the Study:
- To determine if stroke severity scores from different scales can be reliably interconverted.
- To assess the feasibility of using various stroke scales interchangeably in research and clinical settings.
Main Methods:
- A single observer assessed 433 stroke patients using the Canadian Neurological Scale, Middle Cerebral Artery Neurological Score, and National Institutes of Health Stroke Scale.
- Linear regression models were developed to predict scores between scales using training and testing datasets.
- Strokes were classified using the Oxfordshire Community Stroke Project criteria to analyze subtype-specific interconversion accuracy.
Main Results:
- Conversion between the Canadian Neurological Scale and Middle Cerebral Artery Neurological Score demonstrated high reliability (R2 = 94.7%) with acceptable prediction errors.
- Conversion from the National Institutes of Health Stroke Scale yielded lower reliability (R2 = 87.5%-89.0%) and significantly greater prediction errors.
- Interconversion accuracy was poorest for patients with dysphasia and total anterior circulation strokes, even after adjusting the National Institutes of Health scale's motor component.
Conclusions:
- The Canadian Neurological Scale and Middle Cerebral Artery Neurological Score are reliably convertible.
- The National Institutes of Health Stroke Scale is not reliably predictive of scores on the other two scales.
- Interconversion of stroke scales is challenging, particularly for specific stroke subtypes, questioning the validity of meta-analyses using averaged scores from different scales.