Applying autism screening research to real-world scenarios: a commentary on Sheldrick et al. (2023)

Abstract

Early identification of autism spectrum disorder (ASD) continues to be a challenge despite universal screening efforts. One explanation is that screening tools have lower sensitivity and specificity than initial studies report when accounting for incomplete follow-up for all children screened. Sheldrick and colleagues used statistical modeling to demonstrate the impact on sensitivity and specificity when assumptions about the diagnostic outcome of children who do not pursue diagnostic evaluation are altered. Crucially, the work of Sheldrick et al. serves as a reminder that autism screening in primary care is just one component of the clinical assessment and should not be conflated with a diagnostic evaluation. Thus, lack of follow-up after a positive screen is a feature, not only a bug when using a screen in a clinical setting. Engaging families in shared decision-making around screening may help encourage follow-up, and thus, screening tool psychometric performance.