Using the ages and stages questionnaire in the general population as a measure for identifying children not at risk of a neurodevelopmental disorder

BACKGROUND: Early detection of neurodevelopmental disorders (NDDs) enables access to early interventions for children. We assess the Ages and Stages Questionnaire (ASQ)’s ability to identify children with a NDD in population data. METHOD: Children 4 to 5 years old in the National Longitudinal Survey of Children and Youth (NLSCY) from cycles 5 to 8 were included. The sensitivity, specificity, positive and negative predictive values were calculated for the ASQ at 24, 27, 30, 33, 36 and 42 months. Fixed effects regression analyses assessed longitudinal associations between domain scores and child age. RESULTS: Specificity for the ASQ was high with 1SD or 2SD cutoffs, indicating good accuracy in detecting children who will not develop a NDD, however the sensitivity varied over time points and cut-offs. Sensitivity for the 1 SD cutoff at 24 months was above the recommended value of 70% for screening. Differences in ASQ domains scores between children with and without NDD increases with age. CONCLUSIONS: The high specificity and negative predictive values of the ASQ support its use in identifying children who are not at the risk of developing a NDD. The capacity of the ASQ to identify children with a NDD in the general population is limited except for the ASQ-24 months with 1SD and can be used to identify children at risk of NDD. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s12887-018-1105-z) contains supplementary material, which is available to authorized users.