Outcome measures

The MAS will be used to assess the patient’s ability to perform functional activities rather than just cooperative motor patterns. The scale has eight areas of motor function, for which each item is scored from 0 to 6; the scale is highly reliable, with an inter-rater correlation of 0.95 and a test–retest correlation of 0.98.³² The Spearman correlation coefficient between the total MAS score and the FMA after stroke reflects good validity (r=0.96, excluding the general tonus item) and scores range from 0 to 48.³³ This study focused on the overall function of patients, therefore, we choose the total score as the main outcome measures.

The FMA is the most widely accepted movement function scale in patients with stroke, with a reliability of at least 0.95.³⁴ The simplified FMA uses a 100-point motor domain and has been demonstrated to have excellent reliability and construct validity.³⁵ The simplified FMA has 50 evaluation items, divided into three levels (0, 1 or 2 points). A total score of fewer than 50 points indicates that the patient has severe motor dysfunction; a score between 50 and 84 points indicates obvious motor dysfunction; a score between 85 and 95 is classified as moderate motor dysfunction; a score between 96 and 99 is classified as mild motor dysfunction.

Upper limb functioning will be assessed with the ARAT scale, which has 19 items divided into four-arm movement tests: grasp, grip, pinch and gross. Performance on each item is rated on an ordinal scale that ranges from 0 to 3 points.³⁶ The Spearman’s rank correlation coefficients was excellent with the FMA (r=0.77–0.87),³⁷ and intraclass correlation coefficients (ICCs) for inter-rater and intrarater reliability ranged from 0.92 to 0.97.³⁸

The seven-level, modified Rankin Scale (mRS) measures neurological recovery in patients after stroke³⁹ and can distinguish effective from ineffective trials. Inter-observer variability, using pooled reliability, yield a weighted kappa of 0.90.⁴⁰ Its concurrent validity is demonstrated by strong correlations with measures of infarct volumes as well as with the BI (r=−0.89) in acute stroke.⁴¹

The MBI will be used to assess the degree of patient independence. The MBI is a 10-item, 100 total points scale, and each item has five levels, with each item weighted differently. It has been widely used in China and has good reliability (ICC=0.866–0.997) and high criterion-related validity (r=0.816–1.000) with the BI.⁴²

The degree of neurological deficit will be assessed by the NIHSS. For all 11 parameters, a value of 0 is normal, and the top score is 42 points.⁴³ The total score, determined by neurologists and trained nurses, has a high level of agreement (ICC=0.92–0.96).⁴⁴

Except for NIHSS and mRS, for which lower scores indicate better function, all of the scales show that the patient’s function is better if the scores are higher. All assessments will be performed by two blinded assessors (prioritize that the same assessor for one patient), who were trained and qualified before a formal assessment and testing for Cohen’s kappa (κ≥0.75).