Critical Clinical Needs In Classification and Prediction for Older Children, Adolescents and Adults with Autism

Autism is a complex disorder of brain connectivity and organization.  Many genes and their interactions with the environment appear to lead to abnormalities in early brain development and also dynamic aberrations of late brain development and maturation. The result is significant lifelong disability and impairment, often despite islets of significant ability. Major advances in the application of computer technology across all levels of analysis, from genes to brain and behavior, have begun to provide statistical tools that have potential to discern the “signature” of autism at a systems level from many interdisciplinary perspectives. When successfully translated into clinical and community care, biology-based classification and prediction have the potential for major beneficial impact on the lives of affected individuals, their families and communities. Autism research may be on the verge of such a major advance, but still has a long way to go to get there. In this talk I will discuss categorization and prediction, what they are, why they are important, and how they are used in clinical medicine.  I will discuss the pathway from conceptualization to use in the clinic, and present a scientific framework for the evaluation of new methods of classification and prediction that involves the assessment of their analytic validity, clinical validity, clinical utility, and consideration of their ethical and social implications.