Validation of four scales for the acute stage of stroke

OBJECTIVES: To validate whether a simplified scale for the acute stage of stroke--the Scandinavian Stroke Supervision scale--is sufficient for monitoring symptom progression of prognostic importance. DESIGN: The capacity of the scale was compared to that of the Mathew, Toronto and Fugl-Meyer stroke scales and the Barthel ADL index. SETTING: The stroke unit of Danderyd Hospital, which cares for a defined population. SUBJECTS: Fifty noncomatose patients with objectively recorded symptoms at entry were examined over a period of 5 days by one physician (250 ratings/scale). The last 10 patients were also investigated by another doctor in the same manner (50 ratings/scale) and by nurses (50 ratings) the Scandinavian scale only. The amount of time required by each scale was tested in another 10 patients by two nurses (100 ratings/scale). INTERVENTIONS: Routine investigation and treatment. MAIN OUTCOME MEASURES: The reliability, validity, time requirement and correlation of the scales were evaluated, as was their ability to reveal progress of symptoms. RESULTS: All scales were highly significantly correlated. The interobserver agreement was excellent between the physicians but not as good between the physician and the nurses. The prediction of the outcome at discharge and after 3 months was very satisfactory for all of the scales with regard to the whole groups of patients, but none of them could predict the outcome for an individual with certainty. The Scandinavian Stroke Supervision scale was least time consuming, and had the fewest uncertainties expressed by the testers. This scale recorded the progression of slightly fewer symptoms, but did record those with a more serious impact on patient outcome. CONCLUSION: The time-saving, simplified Scandinavian scale was as reliable and as valid as the other scales, and detected deterioration that was important for patient outcome as satisfactorily as the other scales.