6.13 Test-Retest Reliability of the Repetitive Behavior Scale-Revised in Preschoolers With ASD and Related Neurodevelopmental Concerns

Restricted and repetitive behaviors (RRB), a core diagnostic feature of ASD, emerge in early childhood and continue over time with different trajectories across individuals. Importantly, RRB changes in preschoolers have been associated with longer-term outcomes. This underscores the relevance of valid and reliable RRB measures. The parent questionnaire Repetitive Behavior Scale-Revised (RBS-R) is an optimal candidate measure due to its established validity, scalability, interrater reliability, and wide use. However, studies on RBS-R test-retest reliability (TRT)—a critical requirement for accurate and reproducible assessments—are, to date, missing. Thus, we aim to gain knowledge on the stability of RBS-R scores and its variation by children’s characteristics in preschoolers with ASD and other neurodevelopmental disabilities by means of intraclass correlation (ICC) analyses. Analyses were conducted on RBS-R data of 129 children (Mage= 34 ± 7.8 months; 73% males; 93% ASD). Parent RBS-R were completed at 2 time points within 5 weeks (13 ± 8.8 days). TRT was indexed by ICC quantified via a one-way mixed-effects model for the total and the 6 subdomains scores: stereotyped, self-injurious, compulsive, routine, sameness, and restricted behaviors. To assess whether features known to be associated to RBB symptom affect TRT, we repeated ICC analyses in a subset of samples varying in developmental abilities, sex, and ASD severity. ICC was high for RBS-R total score (ICC(3,1) = 0.96; 95% CI, 0.95-0.97) and subdomains ranging from 0.89 to 0.94, for the stereotypic and restricted and the sameness subdomains, respectively. Although ICC remained overall high in subsets of children divided by sex at birth, symptom severity, and developmental stages, ICC was slightly lower in children with ASD and global developmental delay, particularly for the stereotype and ritualistic subdomains. Results from our analyses support the stability of RRB parent-ratings measured with the RBS-R. Given the crucial role that RRBs play in the early diagnosis of ASD in young children, validation of TRT on the RBS-R will prove important for longitudinal studies.