Development of Screening Tool for Autism in Two-Year-Olds- Taiwan Version

Background: The STAT is a twelve-item interactive screening measure for autism, which has been investigated and demonstrated strong psychometric properties and shows promising utility as a Level 2 screening measures for autism in the Western society. A modified T-STAT developed from the original STAT was used in the study. Objectives: The purpose of current study was to examine whether the T-STAT is a useful screening tool for the young children with autism in Taiwan. Methods: The participants were thirty-seven 29-month-olds (range = 24-36 months) children with autism, CA-matched eleven children with PDDNOS and twenty-four children with non-ASD. All of participants were recruited from a local hospital in southern Taiwan and assessed and diagnosed by senior psychiatrists. ADOS was administered independently by the first author. For developing new T-STAT, three items were substituted and held the nine items from the original STAT. Results: Using the same algorithm and cutoff of original STAT, yields T-STAT’s sensitivity and specificity. 1. Based on the autism and non-ASDs diagnosis from ADOS, the sensitivity and specificity are .92 and .83, respectively. Based on the autism and non-ASDs diagnosis from clinical judgment, the sensitivity and specificity are .87 and .70, respectively. 2. Based on the ASDs and non-ASDs diagnosis from ADOS, the sensitivity and specificity are .83 and .83, respectively. Based on the ASDs and non-ASDs diagnosis from clinical judgment, the sensitivity and specificity are .87 and .70, respectively. Comparing the three areas of T-STAT between autism/ASDs and non-ASDs groups by ADOS classification, there were significant differences in areas of play, requesting and directing attention, but not in the area of imitation. Conclusions: Results of this study demonstrated that the T-STAT is quite good tool to differentiate the groups between autism/ASDs and non-ASDs.