NIH funds trial giving AI-powered digital tools to clinicians treating teens for anorexia


Wed, 11/30/2022

author

Brendan M. Lynch

Photo illustration of young woman looking at phone.

LAWRENCE — Called a “pandemic within a pandemic,” the prevalence of anorexia nervosa (AN) among young people has spiked during COVID-19. According to research, in-patient interventions are effective for up to 75% of adolescents with AN. However, about 30% of these teens relapse following recovery in the “post-acute treatment window” as they return to their day-to-day lives.

“We know anorexia nervosa is one of the deadliest mental health conditions a person can have, and we also know that interventions tend to work best in youth,” said Kelsie Forbush, professor of clinical child psychology and director of the Center for the Advancement of Research on Eating Disorders (CARE) and the Center for Overcoming Problem Eating (COPE) at KU.

A screen capture from a suite of new digital mHealth tools, dubbed Smart Treatment for Anorexia Recovery, or STAR“In many cases an individual needs intensive focused treatment at a higher level of care for weight restoration,” Forbush said. “Upon restoring weight, they discharge to lower levels of care. But, what often happens is that individuals or their families lack resources in their community to get highly effective outpatient interventions — and they relapse. It ends up being a revolving door in which a teen goes to a high level of care, discharges back to their community, does not get evidence-based intervention and eventually needs to return again to higher levels of care.”

To improve outcomes and improve access to care for adolescents in this window of vulnerability to relapse, Forbush is leading a three-year clinical trial by researchers at the University of Kansas and Children’s Mercy Kansas City to assess a suite of new digital mHealth tools, dubbed Smart Treatment for Anorexia Recovery, or STAR.

The STAR intervention, developed by Forbush and her collaborators, will assist outpatient community therapists by providing them with support tools to assess and treat clients ages 13-21 during this post-acute treatment window. The randomized controlled trial will enroll 129 participants with AN and AN-related conditions (such as atypical anorexia nervosa, where AN traits are present but a body-mass index measure might be too high for a classic AN diagnosis). A pilot trial of usability and acceptability, already underway, is showing positive results.

Their work is supported by a new $700,000 award from the National Institutes of Health. Centers involved in this network include McCallum Place – Kansas City and St. Louis, InSight, Children’s Mercy Eating Disorders Center and the Eating Recovery Center in Denver.

A screen capture from a suite of new digital mHealth tools, dubbed Smart Treatment for Anorexia Recovery, or STAR“The majority of teens coming home from residential care are returning to communities without eating disorder specialists,” said co-primary investigator Sara Gould, director of the Eating Disorders Center at Children’s Mercy Kansas City. “This project has the potential to hugely increase access to proven interventions and to track teens’ response to the intervention. I have been honored to partner with Dr. Forbush and her team to develop the STAR app and am excited to see this project fully launch and benefit so many.”

The research collaborators at KU and CMKC developed STAR to improve outcomes for AN patients and aid clinicians who often juggle massive caseloads. Patients answer questions on an app that uses computerized adaptive testing to shorten the length of the test for users. Based on patient’s answers, they will get personalized evidence-based lessons, videos and interactive activities that are tailored to the problems that they experienced that week.

On the clinician’s side, the KU researcher said STAR uses statistics called machine learning to predict each patient’s likelihood of recovery each week, support decision-making in treatment, and alert a clinician when user-entered data points to a lack of progress in recovery. When a patient visits a clinician face-to-face, their time is better focused on that patient’s self-reported hurdles to recovery.

“The app will give feedback to the clinician each week,” Forbush said. “The clinician goes to an online dashboard and can chart how their patient is doing. That feedback alone has been shown in other types of mental health conditions to really improve treatment outcomes.

“Participants are entering information that's available to the clinician to guide a course of treatment that's going to be most effective. It's hard to know from just seeing somebody how at-risk they are and how they’re responding to treatment. A lot of times, clients come in and want us to feel good about how they're doing. They don't want to disappoint us. Here, when we get that report each week on how they're really doing, it can give clinicians deeper insight.”

Recovery success for patients tracked via the STAR tools will be compared in the randomized controlled trial against patients treated with Present-Focused Anorexia Nervosa Coping Treatment (PACT), a therapeutic approach that provides support for coping with life stressors.

A supplementary $230,000 NIH grant to the main STAR award will allow the collaborators to further develop the mHealth intervention to have maximum impact in diverse populations.

“The supplement was designed to help include diversity, equity, inclusion, access and belonging principles into the treatment from the get-go,” said Forbush. “Because usually treatments are more targeted toward young white females. But we know, in terms of who has an eating disorder, that eating disorders don’t discriminate. So, building an inclusive intervention to start with we thought could help improve the impact and reach of the intervention. And we want to provide impactful treatment to as many teens as possible to help them recover — why wait to be inclusive?”

Other KU co-investigators for the STAR trial include Alesha Doan, associate professor in the School of Public Affairs & Administration and the Department of Women, Gender & Sexuality Studies, who is spearheading the qualitative interviews for the pilot portion of the study; and Christopher Cushing, associate professor of applied behavioral science, who is advising on the use of digital interventions in youth. Kara Christensen, former KU postdoctoral scholar and now assistant professor at the University of Nevada Las Vegas, has been instrumental in content development for the intervention while serving as a consultant. Other KU personnel involved in the research are Yiyang Chen, a postdoctoral researcher and statistician at the CARE lab, and Trevor Swanson, former collaborator and statistician.

For more information about participating in the full randomized controlled trial, which will start enrolling in a few months, parents/guardians of a teen with anorexia nervosa or related conditions or therapists who are treating teens with anorexia nervosa or related conditions can email star@ku.edu.

Top image (photo illustration): To improve outcomes and improve access to care for adolescents vulnerable to relapse for anorexia nervosa, Kelsie Forbush is leading a three-year clinical trial by researchers at the University of Kansas and Children’s Mercy Kansas City to assess a suite of new digital mHealth tools, dubbed Smart Treatment for Anorexia Recovery, or STAR. Credit: iStock.

Right images: Screen captures from a suite of new digital mHealth tools, dubbed Smart Treatment for Anorexia Recovery, or STAR, designed to improve outcomes and improve access to care for adolescents in a time of vulnerability following in-patient treatment for anorexia nervosa. Credit: Kelsie Forbush.

Wed, 11/30/2022

author

Brendan M. Lynch

Media Contacts

Brendan M. Lynch

KU News Service

785-864-8855