Screening for Autism Spectrum Disorder with the SCQ and SRS: Variation Across Demographic and Developmental Factors

Friday, May 12, 2017: 12:00 PM-1:40 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
E. Moody1, N. M. Reyes2, C. Ledbetter3, L. D. Wiggins4, C. DiGuiseppi5, A. Alexander6, S. Jackson7, L. C. Lee8, S. E. Levy9 and S. Rosenberg10, (1)13121 E 17th Avenue, University of Colorado, Denver, Aurora, CO, (2)University of Colorado - Denver , Denver, CO, (3)School of Public Health, University of Colorado, Aurora, CO, (4)Centers for Disease Control and Prevention, Atlanta, GA, (5)University of Colorado - Denver, Aurora, CO, (6)University of Colorado, Aurora, CO, (7)CDC, Atanta, GA, (8)Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, (9)The Children's Hospital of Philadelphia, Philadelphia, PA, (10)University of Colorado, Aurroa, CO
Background: The American Academy of Pediatrics recommends that all children be screened for autism spectrum disorder (ASD) at an early age. Early identification of children with ASD is critical for referral to early intervention services, which improve outcomes and reduce long term care costs. To date, the performance of ASD screeners has not been comprehensively examined across demographic, behavioral, and developmental characteristics in young children.

Objectives: To determine how two widely used ASD screeners, the Social Communication Questionnaire (SCQ) and Social Responsiveness Scale (SRS), perform across a variety of demographic and child characteristics.

Methods: Data for this analysis come from The Study to Explore Early Development. SCQ and SRS were collected by phone interview and paper questionnaire, respectively. All parents completed the Child Behavior Checklist (CBCL) and all children completed the Mullen Scales of Early Learning (MSEL). The Autism Diagnostic Observation Schedule (ADOS) and Autism Diagnostic Interview–Revised (ADI-R) were administered to children who screened above 10 on the SCQ, or if the clinician observed behaviors sufficient to warrant an ASD evaluation. Following the evaluation, the clinician scored his or her global impression that the child has ASD using the Ohio State University Autism Rating Scale-4 (OARS-4), which served as our gold standard for all calculations. Additional demographic data were collected via a standardized parent interview. Sensitivity and specificity were calculated for the SCQ using cut-offs of 11, 13 and 15, and the SRS using a T score cut-off of 60 relative to the OARS-4. Sensitivity and specificity were then stratified by demographics (maternal race, ethnicity and education; household income; and MSEL and CBCL subscales.

Results: This analysis included 2317 children with a completed developmental evaluation, and OARS, CBCL, SCQ, and SRS assessments; 616 had ASD and 1701 were non-ASD, with 852 identified with developmental disability from educational and clinical settings and 849 population controls identified from state vital records. There was a significant difference in proportion of children with ASD and non-ASD who had below average cognitive functioning (74% versus 30% respectively). Overall, sensitivity and specificity were acceptable for the SCQ with a cut-off of 11 (0.87 and 0.81) and 13 (0.75 and 0.86), and for the SRS (0.85 and 0.77, respectively); however, specificity was considerably lower for African American and Hispanic mothers. Both the SCQ and SRS became less specific and more sensitive as maternal education and household income decreased. Specificity also decreased by as much as 50% when the child had below average scores on the MSEL or borderline to clinical scores on the CBCL.

Conclusions: These findings suggest that the SCQ and SRS cannot accurately differentiate many children with developmental delay from ASD within a diverse sample of children. Screening individuals from minority and lower socioeconomic backgrounds increases false positive rates. Refining these screeners to be more effective regardless of the child’s behavioral presentation and across cultures could help reduce the impact of false positives. When used clinically, additional testing with standardized instruments, such as the ADI-R and ADOS, is needed to inform the differential diagnosis.

See more of: Epidemiology
See more of: Epidemiology