Reliability and measurement error of multi-segment trunk kinematics and kinetics during cerebral palsy gait

Children with cerebral palsy (CP) have been shown to have altered trunk movements during gait resulting in increased loads at the lower lumbar spine. Detailed assessment is possible using 3D gait analysis. However, reliability and quantification of measurement error have not been established. The aim of this study was to evaluate test-retest reliability of thorax and lumbar segment kinematics and L5/S1 kinetics during gait in children with CP. Eight children with CP participated in this study with repeat assessments conducted within 1 week. Reliability was assessed using the one-way random ICC, standard error of measurement and an examination of extrinsic-to-intrinsic variability. Thorax kinematics demonstrated mixed level of reliability with SEM values ranging from 5.94^o to 1.15^o. Lumbar kinematics demonstrated poor-to-good reliability with the largest SEM values for peak lumbar flexion at 4.14^o. L5/S1 moment values demonstrated only poor to good test-retest reliability while L5/S1 reactive forces demonstrated poor to excellent test-retest reliability.This study provides estimates of reliability and change needed to exceed measurement error. While reliability was mixed and some measures for thorax movement were above 5^o, stated as a measure of acceptable error, the results of this study support the use of these measures in children with CP.