Tradução e avaliação das propriedades psicométricas do teste challenge para crianças e adolescentes com paralisia cerebral

Abstract

The Challenge was developed to assess gross motor function of children and adolescents with Cerebral Palsy in the Gross Motor Function Classification System (GMFCS) levels I and II, since other standardized tests were limited in evaluating this population. The instrument’s items are commonly performed by children over five years old and youths during sports and recreation. The aims of this study were: 1) describe the Challenge’s translation to Portuguese and cultural adaptation process, and 2) evaluate its psychometric proprieties including face validity, reliability and responsiveness. The translation was proceeded in four stages: translation, synthesis, back translation and review of the translated version. Face validity was assessed by a 10 Brazilian Physical Therapists that judged the usefulness of each test item. For the reliability test, 50 children and adolescents with cerebral palsy were evaluated in two assessments (seven to 10 days of interval), by two independent assessors. The interrater reliability analysis was performed using independent t-test and the test-retest, using paired t-test. In both reliability tests, the Bland-Altman analysis and its limits of agreement complement the results, presenting the Challenge’s score variation. In addition, the indexes Intraclass Correlation Coefficient (ICC), Weighted Kappa (kw), Standard Error of Measurement (SEM), and Coefficient of Variation (CV) were used. Responsiveness was tested in 30 of the participants assessed during the reliability test. These participants underwent to physical therapy treatment weekly and, after three months of the reliability test, they were evaluated again through the Challenge test. Paired t-test was used for the responsiveness analysis. In order to interpret the changes after three months, effect size measures and minimal detectable change (MDC) and minimal clinically important difference (MCID) indexes were estimated. For MCID estimation, a Global Rating Scale was used as anchor. In this rating scale, parents and caregivers judged the participant’s changes in gross motor function execution after three months (worse, no changes, better). The Challenge translated version had minor cultural adaptations for better understanding of north American metric units, symbols and colloquial expressions in order to ensure semantic, idiomatic and experimental equivalences. This version’s face validity evaluation was good for all test items. The translated version total scores were adequate for inter-rater reliability (ICC: 0.98, SEM: 2.07, and CV: 6.03%) and test-retest reliability (ICC: 0.97, SEM: 3.05 and CV:8.50%). In the test-retest reliability, a statistic significant difference was detected between the assessments (1.70 points, p=0.01). Bland-Altman’s limits of agreement were -6.27 to 5.43 points (inter-rater) e -6.69 to 10.09 points (test-retest). The test items reliability indexes varied from moderate to almost-perfect (kw: 0.32 a 0.96). The Challenge was capable to detect statistically significant changes after three months of treatment (p=0.01), with 8.42 and 5.11 of MDC and MCID, respectively. The Challenge translated version has relevant items for gross motor function evaluation of children and adolescents with CP. This version showed face validity and reliability proprieties similar to the original version. After three months of physical therapy treatment, changes over 8.42 points are not attributed to measurement error and are considered relevant by parents and caregivers.