Using Tech To Address The Relapse Problem

By Monica E. Oss, Chief Executive Officer, OPEN MINDS

Relapse is part of the mental health and addiction treatment and recovery process. For schizophrenia, relapse often occurs early in the disease course, with 81.9% of consumers experiencing a relapse within five years of a first episode of schizophrenia or schizoaffective disorder.

With depression, after treatment of the first episode, approximately half of all consumers will relapse, and this risk increases for every subsequent episode—70% and 90% after a second and third episode, respectively. For anorexia nervosa and bulimia, research indicates that more than a third of all patients relapse within the first few years of completing treatment. The highest risk for relapse from anorexia nervosa occurs in the first 18 months after treatment, with 35% falling back into eating disordered behaviors.

For alcohol use disorders, 50% of consumers relapse after their first treatment episode. And 72% to 88% of people relapsed between one and three years of quitting heroin. What these statistics confirm is that behavioral health conditions are lifelong chronic diseases and need to be treated as such.

The question for managers of behavioral health treatment programs is how to reduce the likelihood that consumers will relapse—and how new technologies may assist in that process. This was the focus of the session, The Future Is Now – Current Innovative Uses Of AI, at The 2022 OPEN MINDS Technology & Analytics Institute. The session featured Rachel Wood, Ph. D., Vice President, Learning Health Systems for Discovery Behavioral Health, and my colleagues, OPEN MINDS’ senior associates Carol Clayton and Paul Duck.

To address consumer needs for better long-term outcomes, DBH has created a virtual engagement program. The concept is to reach out to consumers using a virtual platform when they first contact the organization, throughout the treatment process, and post-treatment to track progress and intervene before any relapse. The program, which is built on the Videra platform, creates individual consumer profiles identifying the particular relapse risks for each consumer. After treatment, the AI-driven platform is used to monitor each consumer’s status with automated assessments using video and audio interactions. These periodic assessments measure word complexity and rate of speech. The assessments inquire about consumer sentiments (happiness, anger, calm, etc.) and they use standardized tools like the PHQ and GAD. Then using AI-driven natural language processing, facial analysis, and voice analysis, the platform can proactively identify consumers at risk for relapse.

This system has helped the DBH team rethink the discharge planning process. “The discharge plan, which is arguably the most important part of the treatment plan, is sometimes the most neglected,” said Ms. Wood. “And it’s not because we’re not focusing on it. It’s simply that we are trying to create a plan for a situation that is largely out of our control. We’re trying to put the onus of recovery and reporting on our patients as they transition from a predictable environment into an unpredictable environment. And then when relapse occurs, we just repeat the process.” Technology allows the treatment team to eliminate that void by creating an ongoing communication and mental health assessment mechanism.

And the platform has addressed the “black box problem”—what happens with the consumer between discharge and first follow up. “The consumer may have had an accident, lost a job or been divorced, maybe they couldn’t get their medications,” said Ms. Wood. “AI is a solution that gives us visibility into the black box so we can prevent relapse. We get visibility into how they are doing while getting great data. Because our platform is video based, it gives the clinician or staff the very information that you would be looking for if the patient were sitting across from you in your office. What would you be assessing? What would you be looking for or listening for? We know as clinicians there’s a difference between what someone says and how they say it.”

For the DBH treatment team, the platform provides alerts and notifications for high-risk behaviors and symptoms. And the data is also aggregated to provide real-time population outcome data. Launched in September 2022, this DBH initiative has already served over 4,000 consumers.

As Ms. Wood looks ahead, she sees continuing leverage of technology at DBH. Some of the technology under consideration include the use of predictive analytics, population health/value-based care management tools, clinical tools for consumer screening and triaging, and remote monitoring wearables.

Use of these tools is critical for allowing organizations like DBH to improve treatment and outcomes. Said Paul Duck, Senior Associate for OPEN MINDS: “In the coming years, AI will play an ever larger role in supporting mental health treatment. The two largest opportunities I see for AI is in improving the accuracy of diagnosis in treatment as well as utilizing AI in helping with workforce demand. As the market continues to change with new competition from retail, out-of-market mega competitors, payers acquiring provider service capacity and others, the need for digital solutions to gain efficiency will become more critical to grow competitive advantages and creating a solid sustainable future for organizations. Those organizations that are able to leverage technologies like AI are setup to enjoy much future success.”