Accuracy of Telehealth-Administered Measures to Screen Language in Spanish-Speaking Preschoolers

Introduction: There is a critical need for telehealth language screening measures for use with Spanish-speaking children because of the shortage of bilingual providers and the current lack of psychometrically sound measures that can be administered via telehealth. The purpose of the current study was to describe the classification accuracy of individual telehealth language screening measures as well as the accuracy of combinations of measures used with Spanish-speaking preschoolers from rural and underserved areas of the country. Materials and Methods: This study applied a hybrid telehealth approach that implemented synchronous videoconferencing, videocasting, and traditional pen and paper measures. Screening measures included a processing efficiency measure (Spanish nonword repetition [NWR]), language sampling, and a developmental language questionnaire. Eighty-two mostly Spanish-speaking preschool-age children and their parents participated. Thirty-four children had language impairment (LI), and 48 had typical language development. Results: Although many of the individual measures were significantly associated with standardized language scores (r=0.27-0.55), not one of the measures had classification values of 0.8 or higher, which is recommended when screening for LI. However, when NWR scores were combined with language sample or parent survey measures, promising classification accuracy values that approached or were higher than 0.8 were obtained. Conclusions: This research provides preliminary evidence showing the effectiveness of a hybrid telehealth model in screening the language development of Spanish-speaking children. A processing efficiency measure, NWR, combined with a parent survey or language sample measure can provide informative and accurate diagnostic information when screening Spanish-speaking preschool-age children for LI.