SCREENINGScreening for Autism Spectrum Disorders With the Social Communication QuestionnaireEAVES, LINDA C. Ph.D.1; WINGERT, HEATHER D. Ph.D.1; HO, HELENA H. M.D.2; MICKELSON, ELIZABETH C.R. M.D.2Author Information 1Department of Psychology, Sunny Hill Health Centre for Children 2Division of Developmental Pediatrics, Sunny Hill Health Centre for Children, Department of Pediatrics, University of British Columbia Faculty of Medicine, Vancouver, British Columbia, Canada Received August 2005; accepted February 2006. Address for reprints: Psychology Department, Sunny Hill Health Centre for Children, 3644 Slocan Street, Vancouver, British Columbia, Canada V5M-3E8; e-mail: email@example.com. Journal of Developmental & Behavioral Pediatrics: April 2006 - Volume 27 - Issue 2 - p S95-S103 Buy Abstract ABSTRACT. The Social Communication Questionnaire (SCQ) is a parent report screening measure for autism spectrum disorders (ASDs) based on the Autism Diagnostic Interview-Revised (ADI-R). To examine its validity in a young sample, the SCQ was given to parents of 151 children at a mean age of 5 years, before assessment in tertiary autism or preschool clinics. Overall sensitivity was .71, the same for both clinics, but specificity was better for the preschool clinic (.62) than for the autism clinic (.53) reflecting fewer false-positives in the former. The "hit rate" was 65% with 28% of the children with autism missed by the SCQ at a cutoff score of 15 (false-negatives) and 38% of the nonautistic misidentified as having an ASD (false-positives). Item validity analysis, contrary to what was previously published, indicated that only 15 or 46% of the items distinguished between children with and without ASD in this much younger sample. False-negatives were somewhat higher functioning. The SCQ would seem to be a useful tool for identifying young children in need of further assessment and assisting in routing them to the appropriate clinic, especially if used in conjunction with a screening by a community professional. There remain questions about the "best" cutoff score to use and whether a shorter version, based on the items that distinguished autistic from nonautistic, would be more reliable and valid with younger children. Furthermore, it may be that an adjusted score is required when parents omit items or with nonverbal children who cannot be scored on some of the items. © 2006 Lippincott Williams & Wilkins, Inc.