Screening also included an assessment of the sociodemographic profile of the participants using a socioeconomic status (SES) pro forma and a behavioral checklist.[,] There were separate SES pro formas (modified Kuppuswamy for the urban area and modified Prasad scale for the rural and tribal areas) for the urban, rural, and tribal areas. The ISAA has been developed and validated by the National Institute of Mentally Handicapped (NIMH), Ministry of Social Justice and Empowerment, Government of India, for diagnosing and measuring the severity of autism in 2009. This scale is based on childhood-autism rating scale (CARS) and has 40 items divided under six domains – social relationship and reciprocity; emotional responsiveness; speech, language, and communication; behavior patterns; sensory aspects; and cognitive component. The items are rated from 1 to 5 with an increased score indicating increased severity of the problem. A score of <70 indicates no autism, 70-106 mild autism, 107-153 moderate autism, and >153 severe autism. The screening instrument was administered by investigators trained in the diagnosis of autism.


A training workshop in the assessment of “autism” for investigators and clinical psychologists was specially designed. The training workshop comprised theory (lectures) and clinical posting (in pediatrics). The training program was conducted by a neurologist, psychiatrist, pediatrician, and three public health specialists. The training workshop was followed up by field testing as detailed above.

B. Evaluation phase: After the screening phase, all the subjects who were scoring above 70 on the ISAA were considered suspected cases of autism and were evaluated by a clinical psychologist (VK) and a public health specialist (SKR). The evaluation included an account of the prenatal conditions, birth history, developmental and medical histories, findings from earlier evaluations (including history of hearing impairment), and intellectual and behavioral functioning. The clinical evaluation also included observing the child for a few minutes. The screening scores on the scale were also reconsidered and altered (wherever required). Further, 10% of all the children who scored less and were thus classified as nonsuspects for autism were also evaluated.