Evaluation of the Reliability of the Challenge when used to Measure Advanced Motor Skills of Children with Cerebral Palsy

Aims: The Challenge was designed as an extension to the GMFM-66 to assess advanced motor skills of children with cerebral palsy (CP) who walk/run independently. This study evaluated the Challenge's inter-rater and test-retest reliability. Methods: Thirty children with CP (GMFCS level I [n = 24] and II [n = 6]) completed the Challenge, with re-testing one to two weeks later. Seven physiotherapist assessors passed the Challenge criterion test pre-administration. A single assessor administered and scored test and retest sessions (test-retest reliability). A second assessor independently scored one of these sessions (inter-rater reliability). Results: Inter-rater reliability was excellent (ICC = 0.97, 95%CI 0.94-0.99, CoV < 10%), with no bias (Bland-Altman plot). Test-retest ICC was excellent (ICC = 0.94, 95% CI 0.88-0.97. CoV < 10%, and Minimum Detectable Change (MDC90) was 4.47 points. Many participants indicated practising at home pre-retest session. Conclusions: There was strong rating consistency between assessors. While test-retest ICC estimates were also high, Challenge scores were higher at retest. The MDC90 was still in a range (>4.5 points) that seems clinically viable for change detection. Test-retest reliability could be reassessed with children instructed not to practice between assessments to determine the extent to which between-session practice influenced scores.