Reliability and validity of the gross motor function classification system for cerebral palsy.

PURPOSE The purposes of this study were to evaluate interrater reliability using videotapes and criterion-related and construct validity of the Gross Motor Function Classification System (GMFCS), aspects of reliability and validity not previously published. METHODS Two experienced pediatric physical therapists rated 30 videotapes of children with cerebral palsy (CP) or Down syndrome (DS) to test interrater reliability. Criterion-related validity was evaluated by comparing GMFCS levels with tests of motor and nonmotor development. Construct validity was assessed by comparing GMFCS trends over time in children with CP and DS. RESULTS Interrater reliability was 0.84. Correlation was higher between GMFCS level and tests of motor development than GMFCS level and tests of nonmotor development. The GMFCS level remained relatively stable in children with CP but tended to improve in children with DS. CONCLUSIONS This study extends reliability and validity of the GMFCS, supporting its use in clinical practice and research.