Abstract
OBJECTIVES: In middle- and lower-income countries, the scarcity of psychiatric expertise and IQ testing should be considered when developing and testing assessment tools. This study aims to assess the psychometric properties of a modified Arabic version of the Childhood Autism Rating Scale 2 Standard Form. METHODS: A diagnostic accuracy study, including 286 children, was conducted from December 1, 2023, to August 1, 2024, at Baghdad's National Centre and Al-Subtain Academy for Autism. Confirmatory factor analysis was utilized for structural validity, standardized alpha and the average inter-item correlation for reliability, and ROC curve analysis for diagnostic accuracy. RESULTS: Good fit was demonstrated for the 1-, 2-, and 3-factor solutions. In all models, 'activity levels' had the lowest loadings and higher scores in children with ADHD (p-value = 0.0002). Excellent internal consistency was shown with a standardised alpha of 0.90 (> 0.7) and an average inter-item correlation of 0.394 (> 0.15). At the optimal cut-off (≥ 28.5), the scale had a sensitivity of 82.6%, a specificity of 93.4%, and would increase the child's odds of having ASD by 12.5. CONCLUSION: The modified scale has good psychometric properties and might be utilised to adjust the psychiatric evaluation of children suspected of having a neurodevelopmental disorder.