Development of a scale for identifying autism spectrum disorder during early childhood

CONTEXT: Assessment forms an important part of the early intervention in autism spectrum disorders (ASD). However, there is not much choice with regard to tools and methods for early identification and assessment of ASD. AIMS: The aim is to develop a scale for identifying ASD during infancy and early childhood by mapping empirically supported skill behaviors and excess behaviors. SETTINGS AND DESIGN: The study was conducted in community-based organisations and the local communities. Moreover, it employed a post facto research design, with survey method. MATERIALS AND METHODS: Following the standard procedures to pool items and standardization, a scale was developed and administered to three groups of children (N = 190) including children with autism (n = 100), global developmental delay suggestive of mental retardation (n = 40) and neurotypicals (n = 50) in the age range of 1.5–6 years. The ICD-10 Classification of Mental and Behaviooural Disorders and Baby and Infant Screen for Children with aUtIsm Traits (BISCUIT) were the gold standards to diagnose ASD. STATISTICAL ANALYSIS USED: Descriptive statistics (frequency, percentages, median, and quartile deviation), Pearson's correlation, Cronbach's α, factor analysis, and binary logistic regression analysis with receiver operating characteristic curve were performed. RESULTS: The new scale also demonstrated a high diagnostic efficiency by yielding a hit rate of 0.89, specificity of 0.90, and sensitivity of 0.88. CONCLUSIONS: The new scale can be used for early identification of ASD in the Indian population, though further validation with large population is required.