S Sachse1, M Saracino, W von Suchodoletz. 1. Institut für Kinder- und Jugendpsychiatrie und Psychotherapie, Ludwig-Maximilians-Universität München, Germany. Steffi.Sachse@lrz.uni-muenchen.de
Abstract
BACKGROUND: Early identification of developmental language disorders is a precondition for early intervention and therefore of high clinical relevance. The ELFRA-1 was constructed to identify children at risk for language disorders in the context of the routine examination of children at the age of 12 months. The aim of the study was to determine the predictive validity of this screening tool. SUBJECTS AND METHOD: 121 parents completed the questionnaire for their 12 months old children. One year later the language development was assessed by means of the ELFRA-2, the German version of the CDI for two year old children. RESULTS: 63% of the children were correctly classified. The sensitivity amounted to 52% and the specificity to 65%. The RATZ-index (relative improvement of the hit rate in comparison to the random hit rate) was 23%, which has to be considered as unsatisfying. Improvement of language abilities were less likely in boys compared to girls and in cases of low educational background of the mother. CONCLUSIONS: The prognostic validity of the ELFRA-1 is insufficient and too low to identify children at risk for later language impairment. The screening overlooked too many late talkers and classified too many normally developed children as language impaired. The ELFRA-1 can not be recommended as a useful method for the early identification of language impaired children at the age of 12 months.
BACKGROUND: Early identification of developmental language disorders is a precondition for early intervention and therefore of high clinical relevance. The ELFRA-1 was constructed to identify children at risk for language disorders in the context of the routine examination of children at the age of 12 months. The aim of the study was to determine the predictive validity of this screening tool. SUBJECTS AND METHOD: 121 parents completed the questionnaire for their 12 months old children. One year later the language development was assessed by means of the ELFRA-2, the German version of the CDI for two year old children. RESULTS: 63% of the children were correctly classified. The sensitivity amounted to 52% and the specificity to 65%. The RATZ-index (relative improvement of the hit rate in comparison to the random hit rate) was 23%, which has to be considered as unsatisfying. Improvement of language abilities were less likely in boys compared to girls and in cases of low educational background of the mother. CONCLUSIONS: The prognostic validity of the ELFRA-1 is insufficient and too low to identify children at risk for later language impairment. The screening overlooked too many late talkers and classified too many normally developed children as language impaired. The ELFRA-1 can not be recommended as a useful method for the early identification of language impairedchildren at the age of 12 months.