The autism therapy industry will first see AI applied en masse to business-focused functions. Not far beyond that more advanced trend, providers should expect AI to become an increasingly common part of their clinical work.
Several industry leaders told Autism Business News that clinicians will never be replaced by AI. However, many see it eventually becoming an essential tool to perfect treatment plan development and tracking, refine clinical decision making, lessen administrative burdens, increase access to care and improve the accuracy of diagnostics.
“I think clinical efficiency is going to be huge,” Brett Blevins, CEO and founder of Commonwealth Autism Care, told ABN while attending the 2024 Autism Investor Summit in Los Angeles. “When you look at the practical application of AI and what we do on a day-to-day basis, I think clinical efficiency is probably at the top of the list.”
Commonwealth Autism Care offers in-center, in-home and in-school services, including applied behavior analysis, and operates 7 locations in Georgia, Kentucky, North Carolina and Virginia, according to its website.
Several powerful forces strain autism therapy providers individually and collectively to be as efficient as possible. The widespread adoption of AI in business and society generally, coupled with the industry’s needs, makes the adoption of AI an inevitability, if not a necessity.
AI’s most common use case in autism therapy is making it easier to collect, summarize and analyze data. Registered behavior technicians (RBTs) and board-certified behavior analysts (BCBAs) are obligated to collect data — observations of their patients, documentation of goal progress and tracking of specific interventions — for clinical and reimbursement purposes. A speculative example could be video- or audio-enabled, AI-powered devices that have the potential to observe sessions and generate notes and other insights for clinicians to review and approve.
This could alleviate the administrative burden and make the clinical burden on BCBAs easier to manage. BCBAs are tasked with developing, delivering, overseeing and reassessing treatment plans for those with autism. They also oversee RBTs that deliver interventions according to the treatment plan. On top of that, they often have several administrative duties to attend to.
The already short supply of BCBAs is worsened by the multitudinous daily burdens they face, making assistance for BCBAs a crucial need for autism therapy organizations to address.
“Think about what we do now: You make the treatment plan, provide the services, collect the data and then a BCBA is evaluating all that data to make determinations about if and what they need to do differently,” Rob Marsh, CEO of Chatsworth, California-based autism therapy provider 360 Behavioral Health, told ABN. “A lot of that can be automated. … I have to believe that you could have an AI looking at all of the variables that a BCBA would never have the time to look at in the amount of time they have to review a case or do a progress report to find the things that really work to motivate the client.”
360 Behavioral Health operates 21 locations in California and is working on two locations in Nebraska.
Eventually, the increased amount of hopefully better clinical data enabled by AI could deepen understanding of the care provided at the patient, individual and industry levels. AI-backed systems could improve care quality by aiding treatment plan adherence and intervention fidelity. It could also identify the best approaches to specific phenotypes of autism and related comorbidities.
“When a BCBA has [so many] students, it’s really important that AI could point them in the right direction, find the concerns they should be looking out for,” Mordechai Meisels, CEO and founder of human services provider Chorus Software Solutions and founder and chief clinical officer of Encore Support Services, told ABN. “It could help you find the RBTs or the children that need some help, find the ones that are doing great and push them all toward mastery.”
Clinical AI tools could also expand access to diagnostics and identify the interventions patients need based on their specific combination of needs. Standardized assessments could be done with AI, and other inputs and coupled to any number of other large sources of data to assess needs and customize treatment plans, Marsh said.
“BCBAs may see [a need] and they may go to a repertoire of five or six things that they’ve done in the past versus an AI that would have hundreds of different things that could be considered,” Marsh said. “That would make treatment plans even more nuanced that what they would originally have in mind and may have better application to and adoption by the client.”
Getting things right on the front end of care impacts the end result and the immediate experience of patients and families. This is especially true when trying to integrate several care types in one setting, according to San Diego-based Cortica’s CEO Neil Hattangadi.
His company operates clinics that house several specialties relevant to autism therapy and care for those with other neurodivergences. Cortica’s homegrown tech system, called Axon, uses AI to help with scheduling complexities as well as “phenotype matching between clinicians and patients.” The latter tool helps with clinician load balancing, ensuring an even distribution of patients of varied clinical difficulty.
“We’re sitting on this really large data set that spans metabolic and genetic data, imaging data, long-term behavioral outcomes, medical changes,” Hattangadi said. “To be able to use that data and based on the initial testing of this child, those lab results and this genetic variant, we think they fit into this phenotype of autism, and therefore, we be most responsive to these kinds of therapies — that’s huge.”
Those kinds of insights are on the horizon but not yet a reality today, he added. The realization of those kinds of clinical insights would eliminate trial and error, especially when it comes to medication management in the context of other therapies.
