Adaptação e validação de instrumento de avaliação dinâmica das habilidades motoras da fala

Abstract

The characterization and diagnosis of childhood apraxia of speech (CAS) are often discussed in literature. The gold-standard for the diagnosis of this disorder is still the experts’ opinion since the evaluation instruments are limited and not all have established psychometric proprieties. An instrument that evaluates CAS is the Dynamic Evaluation of Speech Motor Skills (DEMSS), which is validated and used for speech pathologists in United State of America. The present research aimed to adapt for Brazilian Portuguese and search for reliability and validity evidence for the DEMSS. The adaptation process involved five stages: 1) translation and retranslation of the instrument; 2) New items (words) selection; 3) analysis by experts judges (speech therapists and linguist); 4) Analysis by non-specialist judges; 5) application of the instrument on a pilot sample. For reliability and criterion validity studies the sample was composed of 136 children between 3:0 and 6:7 years old (114 with typical language acquisition and 22 with disorders of speech sounds). Reliability was assessed by internal consistency (Cronbach's alpha), test-retest and inter e intraobserver analyses. Test-retest reliability was calculated using the Pearson correlation coefficient and the reliability inter and intra-observer Kendall and Kappa analysis were used. Comparisons of sociodemographic variables (type of speech development, sex, type of school and age) and performance on DEMSS-BR constituted the criterion validation data. Student t test for independent samples and univariate analysis were used. The test d Cohen was used to calculate the effect size between the significant results. Data were analyzed using SPSS software version 22, considered significant when p ≤ 0:05. The DEMSS translation was performed by two independent translators. After, the instrument was retranslated and it was sent to one of the authors of the original instrument. Two Speech therapists experts selected 269 words which 94 were considered adequate by experts’ judges. Finally, non-experts judges and pilot sample judged the 94 chosen words and 44 were selected for the instrument. The DEMSS-BR reliability study showed adequate accuracy and stability (> 0.8). The criterion validity showed that DEMSS-BR is able to discriminate children with impaired speech of those with typical development. In addition, girls performed better on the instrument. Type of school and age variables do not appear to influence the test results. DEMSS-BR came to fill a clinical and scientific gap in the Brazilian reality of quantitative instruments, whereas there are few instruments which intend to evaluate severe speech sounds disorder in children. The instrument demonstrated content validity, criterion and presented stable, consistent and accurate measurements, as demand a good test.