Interrater Reliability of Activity Questionnaires After an Intensive Motor-Skill Learning Intervention for Children With Cerebral Palsy

Abstract

ObjectiveTo investigate the reliability of parents-reported activity questionnaires after a motor-skill learning intervention for children with cerebral palsy (CP). We hypothesize that the intervention process might influence parental judgment. DesignDouble-blind randomized trial. SettingConventional therapy was delivered in the usual context while intensive intervention was provided at the Catholic University of Louvain. ParticipantsChildren with CP (N=41; age range 5-18y, Gross Motor Function Classification System I-IV) were randomized to a control group (CG) (n=21, 2 dropouts) receiving conventional therapy or an intervention group (IG) (n=20) receiving hand-arm bimanual intensive therapy—including lower extremities (HABIT-ILE). InterventionsConventional therapy (mostly neurodevelopmental) was delivered as ongoing treatment (1-5 times/wk). HABIT-ILE, based on motor-skill learning, was delivered over 2 weeks. All children were assessed at T1 (baseline), T2 (3wk after baseline) and T3 (4mo after baseline). Main Outcomes MeasuresABILHAND-Kids and ACTIVLIM-CP questionnaires rated by parents (perception) and 2 examiners (videotapes). ResultsAgreement (level/range) between examiners was systematically almost perfect (P≤.001). At baseline, moderate to almost perfect agreement (level/range) was observed between parents and examiners (P≤.001). At T2 and T3, a similar agreement (level/range) was observed for the CG. For the IG, a similar level of agreement was observed, but the range of agreement varied from poor to almost perfect (P≤.001), with parents estimating higher performance measures compared to examiners after intervention. Higher performance was associated with higher satisfaction scores of the child’s functional goals at T3. ConclusionParents and examiners have a similar perception of the child’s performance at baseline and during conventional therapy. Their perceptions are less congruent after a motor-skill learning intervention, probably due to the goal-oriented process of the intervention. Therefore, our results favor the use of blind observations of home-videotaped items after intensive motor-skill learning interventions.