The utility of parent-report screening tools in differentiating autism versus attention-deficit/hyperactivity disorder in school-age children

Symptoms of autism spectrum disorder and attention-deficit/hyperactivity disorder often co-occur, challenging timely, and accurate diagnosis. We assessed the performance of three parent-report measures in discriminating autism spectrum disorder from attention-deficit/hyperactivity disorder without autism spectrum disorder (ADHDw/oASD) in school-age verbally fluent children. We examined the Autism Symptom Interview - School-Age and two widely used parent questionnaires: Social Responsiveness Scale - 2nd Edition and Social Communication Questionnaire - Lifetime. Receiver operating characteristic curves assessed each instrument's performance against the best-estimate clinician Diagnostic and Statistical Manual of Mental Disorders, 5th Edition diagnosis of autism spectrum disorder or of ADHDw/oASD (n = 74, n = 102, respectively; 6-11 years). These yielded moderate accuracies: area under the curve = 0.85, 0.79, and 0.78 for Social Communication Questionnaire - Lifetime, Autism Symptom Interview, and Social Responsiveness Scale - 2nd Edition, respectively. Area under the curve pairwise comparisons reached our statistical significance (p < 0.01) for the Social Communication Questionnaire - Lifetime versus the Social Responsiveness Scale - 2nd Edition. Within instruments, sensitivity and specificity varied across autism spectrum disorder cutoffs. Along with the between-instrument variability, this indicates that clinicians and researchers have valid options, depending on the settings and their goals. Comparing children correctly and incorrectly classified as autism spectrum disorder showed no differences in demographics, intellectual abilities, or in any specific clinical profile(s), except for the degree of parent concerns across autism spectrum disorder and comorbid psychopathology-related symptoms. Together, results suggest that complementing parent screeners with multiple sources may be needed to best differentiate school-age verbally fluent children with autism spectrum disorder versus ADHDw/oASD.