Intra-rater and inter-rater reliability, minimal detectable change, and construct validity of the Edinburgh Visual Gait Score in children with cerebral palsy

Background: The Edinburgh Visual Gait Score (EVGS) has been used for observational gait assessment in children with cerebral palsy (CP). However, the measurement error of the EVGS and its detailed relationship with gross motor function remain unclear. Research questions: This study aimed to confirm the intra-rater and inter-rater reliability as well as the minimal detectable change (MDC) values for the EVGS with the use of the video analysis software and examine the relationship between the EVGS and the Gross Motor Function Measure 66 (GMFM-66) with regard to construct validity. Methods: This cross-sectional study was conducted for 62 children (mean age 11.3 +/- 3.9 years) with spastic CP at Gross Motor Function Classification System (GMFCS) level I (32 children), II (25 children) or III (5 children). Three raters independently scored the EVGS using Kinovea video analysis software. The intra-rater and inter rater reliability were calculated using intra-class correlation coefficients (ICC2,1), and the MDC90 was calculated using standard error of measurement. The construct validity was examined by correlating the EVGS with the GMFM-66. Results: The EVGS showed good or excellent reliability within each rater (ICC2,1 = 0.90-0.97) and between raters (ICC2,1 = 0.91). The MDC90 of the EVGS ranged from 3.6 to 6.0. There was a significant correlation between the EVGS and the GMFM-66 (r = 0.69 to 0.73, p < 0.001). Significance: The intra-rater and inter-rater reliability of the EVGS is sufficient for observational gait assessment. The high correlation between the EVGS and the GMFM-66 supports its construct validity. The authors propose an MDC of 6.0 for the EVGS. These results can help in the application of EVGS to children with CP at GMFCS level I-II with mild to moderate gait pathology, as there were few children with CP at GMFCS level III.