The variability of goniometric measurements in ambulatory children with spastic cerebral palsy.

This study assessed the variability of six goniometric measurements commonly used in the assessment of children with cerebral palsy (CP). Three experienced paediatric physiotherapists recorded three consecutive measurements of six joint ranges from 12 children with spastic CP. A fourth measurement was recorded 1 week later. The order of measurement with regards to both joint measurement and tester, was randomised to ensure a balanced experimental design. A blinded procedure was adopted with masked goniometers and a second assessor present to record the measurement. Results, analysed using generalizability theory, showed same day and different day measurement error of +/-10-14 degrees (95% confidence limits) for the foot/thigh angle, abduction and internal rotation of the hip. The source of this was child-assessor and residual error. For ankle dorsiflexion and the popliteal angles, same day measurement error was similar. However, for joint ranges recorded on different days, error ranged from +/-18-28 degrees. The major source of this was systematic differences in measurements recorded on different days. Results show a need for caution when using goniometric measurements in clinical decision making, particularly in the more tone dependent biarticular muscles after an initial assessment.