Using data and technology to improve autism care

About 1 in 100 children around the world are diagnosed with autism spectrum disorder (ASD) each year. And while there isn’t one standard treatment for autism, the U.S. health care system has become much better at diagnosing and managing autism, helping many individuals reduce symptoms and improve their day-to-day lives.

“Applied behavior analysis (ABA) is currently among the most well-researched and effective interventions that focus on the unique experiences and challenges of individuals with autism,” said Darren Sush, a doctor of psychology and board certified behavior analyst who is the head of Autism and Psychology within the Behavioral Health division of Evernorth Health Services. “Despite the success of ABA intervention, challenges exist because there are no industry-wide, standardized performance measures to assess quality of care and health outcomes among ABA providers.”

That’s precisely why Evernorth partnered last year with the Behavioral Health Center of Excellence (BHCOE), an international accrediting body focused on measuring the effectiveness of ABA services delivered to people diagnosed with autism. The goal of the partnership is to develop metrics and key performance indicators around effective ABA therapy to be able to better assess quality ABA care and direct patients to the best therapy possible. As a next step, Evernorth is partnering with data and analytics provider SpectrumAI to focus on individualization and effectiveness of services by bringing innovative measurement capture to the point of care.

To better understand the autism care landscape, ABA, and the partnerships with BHCOE and SpectrumAI, we sat down with Evernorth’s Sush, who walked us through the outcomes of the collaborations, as well as what’s in store for the future.

Talk to me about some of the biggest challenges we face today in autism care. What does the ABA landscape look like?

The biggest challenges we see are not enough ABAs for the demand and a lack of ability to define success and standardized measures.

ABA therapists are predicted to experience a more than 20% job growth rate by 2029, according to the Bureau of Labor Statistics, so there are more professionals joining this growing area of therapy, which is great news for accessing care. However, supply is still not meeting demand in that there was a 1,098% increase in demand for ABA therapists between 2010 and 2018. Increasing the number of ABA providers is the first challenge that we need to overcome as an industry in order to expand access to care.

The other challenge is that there is limited transparency for patients and their caregivers as it relates to the quality of care they are receiving from their ABA providers. Add to that the fact that there’s also a lack of standardized measurement in the value of the care people receive, which makes it difficult to identify the differences in service delivery have the most impact on patient health outcomes. Our partnerships with BHCOE and SpectrumAI both center on improving the quality of care that people receive. 

How are Evernorth and the BHCOE working to improve quality of care?

As more people enter the ABA field, the importance in looking at the quality of those services and patient outcomes grows. There isn’t an industry standard in how we specifically define quality ABA intervention, which makes it hard to compare providers. This is a challenge for the families who are dealing with an autism diagnosis, as it’s difficult to choose a provider and to understand if the care your child is receiving is the right care for them.

The BHCOE has relationships with many large provider groups, and through their accreditation work, they have insight into the best practices, standard protocols, and other guidance to ensure that the outcomes of ABA services are reliable and effective. We’ve worked together over the last year to develop quality metrics and key performance indicators for ABA agencies and organizations. Much of this work relies on the data that BHCOE has available from their years of experience working with agencies directly, as well as Evernorth claims data and provider impact data.

We are developing metrics around ethics, integrity/professionalism, clinical documentation, service delivery, and more. These metrics and performance indicators help us to understand what makes a quality ABA provider, so that we can guide members and their families to the best possible care.

Okay, so the work with BHCOE is really around defining “quality care.” What is the relationship with SpectrumAI all about?

SpectrumAI is focused on the collection and analysis of data and information during the ABA treatment process. SpectrumAI offers an ABA data collection and documentation technology that ABA providers can use to harvest insights to improve care. As part of our work with them, Evernorth is looking to use that aggregated data from ABA providers to identify patterns and trends, so that we can make more informed decisions on treatment programs.

The characteristics associated with the diagnosis of autism may present differently across individuals, so the provision of service needs to be individualized depending on each person’s needs. This data has the potential to inform which treatment strategies work better within different populations or tend to have better prognosis with particular individuals. We are hoping that our partnership with SpectrumAI can assist ABA service providers in identifying uniformity in autism treatment across similar individuals based on the strategies and interventions that work for them. It will help us to become more effective and more efficient across populations of individuals who may be more similar to each other and with those who may have specific strengths and needs. I also believe it will help us reduce overtreatment. Working in this more pointed and focused manner could reduce the overall duration of services, which can help cut costs for health plans.

The goal is to add to what we already did in instituting a measurement framework with BHCOE, and make autism care more personalized and thereby more effective.  

So the collection of data by the ABA provider is key, right? Do you worry about human error?

The collection of data is very important in quality provision of ABA services. What’s really interesting about SpectrumAI is that they have developed a video recording system, which ABA staff members can use during an ABA session (with the patient or their caregiver’s consent). How it works is the ABA session is recorded, and SpectrumAI analyzes the recording and pulls key details into the documentation automatically. This gives the therapist who is working with the patient more time to focus on the therapy, as opposed to having to pause and collect data. The video recording will essentially work toward collecting data for the provider and then inputting that data to then be analyzed. That will also increase the accuracy and consistency in how the data is inputted.

How will these videos help improve autism care?

Typically, a board-certified behavior analyst (BCBA) develops the treatment program, oversees the staff, and makes sure the staff is trained to implement the strategies and interventions with integrity. But that supervisor isn't necessarily implementing the services. Having that video allows the supervisor to review and train more efficiently and effectively to make sure that treatment is being delivered with integrity and to look at differences and data across staff members and analyze why that might be. This allows for continuous improvement in care.

What’s the future of autism care, and how will data and AI help shape that future?

The future of autism care will be rooted in a value-based care model that is driven by unique measurement capture. Evernorth is taking a role to unite the ecosystem because we cannot do it in silos.

I see a lot of benefits in leveraging technology like AI and data to identify quality services and maintain consistency in care. Establishing who the quality providers are will help us direct people to the best possible care and improve health outcomes while lowering costs.

The opportunities with AI, specifically in guiding treatment decisions, will make autism care more efficient and more effective because the technology allows you to see patterns and connections that wouldn't otherwise be available.

Data collection is the key here, but it takes time and is susceptible to human error. By using AI, we can remove that burden from providers and leave them more time to serve the individual patient.