Validation and determining an optimal cut-off score of the Infant Behavior Checklist for autism spectrum disorder.

Abstract

Given the escalating prevalence of autism spectrum disorder (ASD), the persistent paucity of child psychiatrists in Japan, and the need to prepare for unforeseen situations, such as the COVID-19 pandemic, it is essential to establish a reliable screening tool. This study aims to validate the Infant Behavior Checklist (IBC) developed by Japanese experts and determine its appropriate cut-off score. A total of 354 clinic-referred children, along with their caregivers, participated in this research. Clinical records, including diagnoses established through the sub-structured diagnostic interviews, and the IBC assessments, were subjected to rigorous statistical analysis. Among the 24 items, six failed to reach significance to differentiate ASD from non-ASD cases. The Cronbach's alpha coefficient for the IBC was calculated at 0.77. The IBC total score for ASD cases was significantly higher than that of non-ASD cases. With the chosen cut-off score, the IBC demonstrated an area under the ROC curve of 0.803, along with sensitivity, specificity, positive predictive value, and negative predictive value of 8.03, 0.79, 0.69, 0.34, and 0.94, respectively. The IBC exhibits satisfactory internal consistency and content and discriminant validity. The high sensitivity and other associated indices for the optimal cut-off score of the IBC affirm its validity as a screening instrument for ASD. Nevertheless, further investigations are warranted to ascertain the clinical utility of the IBC.