Introduction: The G.A.I.T. scale was developed to evaluate gait and identify deviations from normal gait post-stroke. Purpose: The aim was to assess the reliability of the Slovenian translation of the G.A.I.T. scale for intra-rater and inter-rater reliability, as well as concurrent validity with the walking speed test. Additionally, internal consistency and criterion validity of the scale were evaluated. The minimal detectable change was also calculated. Methods: The study included 30 subjects in the chronic phase post-stroke. The translation followed guidelines for cross-cultural adaptation. The interval between assessments for intra-rater reliability was seven days. Inter-rater reliability was assessed between two raters. Concurrent validity of the G.A.I.T. scale was determined using the 10-meter walk test at a comfortable and fast walking speed. Internal consistency was calculated for the entire scale and its subscales. Criterion validity was calculated for scale items that include degrees of movement in their scoring. Kinovea software was used for the objective measurement of range of motion. Results: The reliability of intra-rater for the entire scale was excellent (ICC = 0.974), while item reliability ranged from low to excellent (ICC = 0.279 – 1.00). Inter-rater reliability for the entire scale was high (ICC = 0.850) and ranged from low to excellent (ICC = -0.147 – 0.935) for individual items. The G.A.I.T. scale outcomes were highly positively correlated with the 10-meter walk test at a comfortable (ρ = 0.867) and fast walking speed (ρ = 0.765). Internal consistency for the entire scale was moderate (α = 0.897), while consistency of subscales ranged from low to high (α = 0.243 – 0.816), low for the stance and swing phase subscale, and high for the swing phase. Criterion validity for items 2, 5, 6, 11 – 14, 19, 20, 26, and 27 was moderate to excellent (ρ = 0.554 – 1.00). The minimal detectable change for the scale was 8.78 points. Discussion and conclusion: Our study demonstrated excellent intra-rater reliability and high inter-rater reliability for the G.A.I.T. scale, confirming its utility in clinical practice. Item reliability was low to excellent for both intra-rater and inter-rater assessments. Additionally, very high concurrent validity with walking speed was found. Criterion validity ranged from moderate to excellent. Internal consistency for the entire scale was moderate, while subscale consistency ranged from low to moderate. Future research should include patients with lower FAC scores and those in various stages post-stroke.
|