Applicability of the Autism Spectrum Screening Questionnaire Parent Form to 5-Year-Old Children

Background: The study of High-Functioning Autism Spectrum Disorder(HF-ASD) has reported that secondary maladjustment such as school refusal in school age is one of the serious problems of them. The Autism Spectrum Screening Questionnaire(ASSQ; Ehlers, Gillberg, & Wing, 1999) was developed to screen school children for HF-ASD. Although the ASSQ consists of only 27 items, it has shown both validity and reliability, with good sensitivity and specificity in clinical settings(Ehlers et al., 1999). The ASSQ could be appropriate for use as a population screening, as it short, it’s easy for parents and teachers to complete. Although the ASSQ has been psychometrically tested in only 7-16-year-old children, an application for younger children than 7-year-old is also expected to make early detection of HF-ASD and prevent secondary maladjustment of them. From the clinical necessity, the ASSQ has been used in younger children than 7-year-old in clinical setting without standalization data(Kopp & Gillberg, 2011), therefore earlier application possibility of the ASSQ should be examined urgently.

Objectives: The purpose of our study was twofold: first, to study psychometric properties of the ASSQ before school age, and second, to evaluate the sensitivity, specificity, positive predict value(PPV) and negative predict value(NPV) on the ASSQ.

Methods: This study was conducted as a part of a larger study-the Hirosaki Five years check-up(HFC) study-assessing mental health among children in Hirosaki(n=2571). We sent ASSQ and other questionnaires to the parents of all children aged 5 in this city. The response rate was 74.6%(n=1919 boy=1002). In the HFC study, children were screened by multi-aspects in their difficulties; communication, behavior, motor skill, daily adaptation of children and their parent's stress. If children had showed higher score than cut-off scores in at least one screening, they were invited to additional assessments and the diagnostic interview by child psychiatrists. The children who met the DSM-5 criteria for a diagnosis of ASD were diagnosed as ASD(n=55, boy=35).

Results: The internal consistency for the entire ASSQ was good, with Cronbach’s alpha of .853. The results of confirmatory factor analysis confirmed that the factor structure was the same for 5-year-old as it was for school-aged children(RMSEA=.050). The receiver operating characteristic(ROC) analyses were performed to assess the discriminant power of the ASSQ in distinguishing ASD from non-ASD cases, and it demonstrated favorable accuracy of identification of ASD(AUC=.91:95%CI=.85-.95). The present study suggested that using cut-off score of > 8 provided the most efficient screening with sensitivity of .85 and specificity of .90. For the cut-off of 9, PPV was .22, NPV.99.

Conclusions: The current study demonstrates the possibility that the ASSQ could identify ASD in the 5-year-old children with a certain degree of accuracy. Although NPV was excellent, showing that only few children fell below the cut-off scores, PPV was low. This problem has been also pointed out by the study of the school age(Mattila et al., 2009). Therefore, it should be emphasized that the ASSQ is a screening instrument, not a diagnostic instrument; all screening-positive children have to undergo diagnostic examinations.