Therapy Management Software for Naturalistic Model-Based Behavioral Interventions -
DESCRIPTION (provided by applicant): This Phase I SBIR resubmission proposes to develop a prototype of a major new type of therapy management software for autism therapy provider groups, which will enable them to deliver emerging types of behavioral interventions that are designed for treating very young children with Autism Spectrum Disorder (ASD). Studies have shown that therapy is most effective when treatment begins at ages 2-4, while cognitive abilities are in earlier stages of development. However, traditional autism therapy (Applied Behavioral Analysis) is not well- suited for children under 5, as it requires adherence to a linear sequence of repetitive therapist-directed trials for 20 or more hours a week. Over the past decade, autism researchers have developed and studied new types of intensive interventions that are rigorous and data-driven, yet appropriate for very young children. The lack of tools to facilitate these types of therapies in a team environment is a significant obstacle to their widespread adoption by traditional ABA therapy groups. Existing therapy management systems do not accommodate the creation or importing of structured models associated with these interventions, and their therapy delivery and data collection interfaces are too rigid to support the naturalistic therapy techniques that these interventions require. The software proposed herein will consist of three major innovations: (1) a Model Manager, which will allow a director to import a structured model, tailor it to fit her own practice approach, and use it as a template system that governs the creation of curriculums within her organization, (2) a Curriculum Builder, that will allow program managers to construct and refine individual therapy curriculums based on the director's prescribed model, and (3) a Therapy Driver tablet application that facilitates naturalistic, child-
directed implementation of the prescribed curriculum, including an activity-based task prompter, a real-time learning step tuner, and an alert-driven therapy coverage indicator. This proposal will
allow us to complete the design and implementation of a fully operational prototype, with iterative feedback from autism intervention researchers at the University of Washington, and conduct a feasibility study by deploying Abacus to a large community therapy provider (Northwest Behavioral Associates) who will use it to deliver a naturalistic model- based therapy to young children while observers determine whether or not 14 key functions are achieved. At the end of Phase I, we will have developed a mature prototype of software with the potential to accelerate the adoption of emerging behavioral interventions. As such, it has the potential to be highly significant, as widespread adoption of new evidence-based therapies will improve long-term outcomes for many people with ASD, enabling them to live more independently, and saving society billions of dollars in reduced costs of care.
