Dartmouth researchers unveil remarkable findings from the first clinical trial of Therabot, an AI-powered therapy chatbot, showing significant symptom reductions in depression, anxiety and eating disorders — potentially revolutionizing mental health care.
In an era where access to mental health services is increasingly strained, a recent study by Dartmouth researchers offers new hope. The first clinical trial of Therabot, an AI-powered therapy chatbot, has demonstrated significant improvements in mental health symptoms among participants, shedding light on the potential for AI to complement traditional mental health care.
Significant Findings and Methodology
Published in the New England Journal of Medicine AI on March 27, the study involved 106 individuals across the United States diagnosed with major depressive disorder, generalized anxiety disorder, or an eating disorder.
The participants interacted with Therabot through a smartphone app, either responding to prompts or initiating conversations.
The results were compelling:
- Depression: Participants experienced an average reduction in symptoms by 51%, significantly improving mood and overall well-being.
- Anxiety: Those with generalized anxiety saw a 31% average reduction in symptoms, with many moving to less severe categories.
- Eating Disorders: Users at risk for eating disorders showed a 19% average reduction in concerns about body image and weight, outperforming a control group.
Potential Impact and Future Considerations
“The improvements in symptoms we observed were comparable to what is reported for traditional outpatient therapy, suggesting this AI-assisted approach may offer clinically meaningful benefits,” senior author Nicholas Jacobson, an associate professor of biomedical data science and psychiatry at Dartmouth’s Geisel School of Medicine, said in a news release.
Jacobson, who is the director of the treatment development and evaluation core at Dartmouth’s Center for Technology and Behavioral Health (CTBH), emphasized the growing demand for mental health services, noting that for every available provider in the United States, there are about 1,600 patients with depression or anxiety.
“There is no replacement for in-person care, but there are nowhere near enough providers to go around,” he added.
The aim, he explained, is for AI like Therabot to complement traditional therapy, potentially providing real-time support for those without regular access to mental health professionals.
Real-World Application and Safety Measures
Therabot, developed in Dartmouth’s AI and Mental Health Lab since 2019, has undergone rigorous review to ensure responses align with therapeutic best practices.
During the trial, participants were given four weeks of unlimited access to Therabot, with follow-ups at eight weeks. The researchers monitored conversations to ensure appropriateness and provided prompts for emergency services if high-risk content, such as suicidal ideation, was detected.
First author Michael Heinz, an assistant professor of psychiatry at CTBH, highlighted that while the findings are promising, caution is required.
“While these results are very promising, no generative AI agent is ready to operate fully autonomously in mental health where there is a very wide range of high-risk scenarios it might encounter,” Heinz, who also is an attending psychiatrist at Dartmouth Hitchcock Medical Center, said in the news release.
The team underscored the need for continuous clinician oversight during the use of such technologies.
Building Trust and Engagement
The study also revealed a surprising degree of therapeutic alliance between users and Therabot. The researchers were pleased to find that participants not only responded to the bot’s prompts but also initiated conversations, indicating a level of trust and comfort.
“We did not expect that people would almost treat the software like a friend. It says to me that they were actually forming relationships with Therabot,” Jacobson said.
He added that the non-judgmental nature of the AI might make users more comfortable.
Future Directions
The research team emphasizes that despite the promising results, extensive further research and development are needed to ensure the safe and effective use of AI in mental health care.
“This trial brought into focus that the study team has to be equipped to intervene — possibly right away — if a patient expresses an acute safety concern such as suicidal ideation, or if the software responds in a way that is not in line with best practices,” Heinz added. “Thankfully, we did not see this often with Therabot, but that is always a risk with generative AI, and our study team was ready.”
In a field where the need far outweighs the availability of care, the potential for AI to alleviate some of this burden cannot be overstated.
“We’re talking about potentially giving people the equivalent of the best treatment you can get in the care system over shorter periods of time,” added Jacobson.
As this pioneering research continues to unfold, it sets the stage for a future where AI and traditional therapy can work hand-in-hand to provide comprehensive mental health support for all.
Source: Dartmouth College