As the Chief Clinical Officer at Eleos, I’ve observed a troubling trend in the mental health technology sector: the race to innovate often outpaces the evidence backing these advancements. It’s disheartening to see many companies encroach upon what were once revered clinical domains, driven more by the pressure to grow their business than the commitment to uphold the sacred duty we, as mental health providers, owe to our clients.
At Eleos, we believe an unwavering dedication to science and empirical data is an essential element of any healthcare tech firm—not solely to ensure product quality, but also to maintain the trust of the clinicians and clients we ultimately serve. We must not build on a foundation of hollow promises. Instead, our solutions—and our success as a company—must always remain firmly anchored in thorough and rigorous clinical study.
That goes for the organizations purchasing and using new technology as well. To ensure operational efficiency never comes at the expense of client welfare, providers must insist that their tech partners uphold certain standards:
- Evidence-Backed Value Statements: Vendors can claim their solution does pretty much anything—but proving it is another story. A healthcare tech platform’s “proof points” must be underpinned by solid data and research—not merely good marketing. Just like we, as clinicians, must adhere to the tenets of evidence-based practice, healthcare technology platforms must engage in evidence-based product development.
- Peer-Reviewed Research Publications: It’s time for healthcare tech companies to commit to publishing their findings in esteemed, peer-reviewed scientific journals. This step not only subjects our work to rigorous scrutiny but also contributes to the broader healthcare knowledge pool (instead of just boosting profits).
- Clinician and Expert Involvement: Behavioral health providers and experts must be deeply involved in research and product development efforts. This ensures solutions are not just theoretically sound, but also practically applicable in real-world care environments.
- Continuous Improvement: Like science, technology products should never be static. They must evolve through cross-pollination as well as continual updates and enhancements—guided by ongoing research and client feedback.
By upholding these principles, we—as a health tech community—can ensure that every technological leap we take is not just an advancement in innovation, but also a stride toward preserving trust and efficacy within the healthcare community.
1. Evidence-Backed Value Statements
The business compasses for behavioral health orgs and tech startups can point in wildly different directions.
Startups often need to chase rapid growth to satisfy investors—while health organizations answer to their communities’ call for reliable and effective care. But behavioral health tech companies cannot afford to behave like the typical startup. They must do more than make bold claims; they need to back them up with solid scientific proof.
That’s precisely why, in our first three years as a company, Eleos established the following proof points by way of rigorous scientific study (more on this in the next section!):
- 2x Higher Client Engagement: Clients whose providers use Eleos attend 67% more sessions compared to treatment as usual.
- 3–4x Better Symptom Reduction: Clients whose providers use Eleos achieve three to four times greater reduction of depression and anxiety symptoms compared to treatment as usual.
- 42% Reduction in Progress Note Completion Time: When providers use Eleos-generated session summaries, they spend 42% less time on documentation.
- High-Quality AI-Generated Note Content: Mental health professionals rated Eleos Health’s EHR-embedded progress notes as highly coherent (85% of sampled summaries), clinically relevant (89.2%), and comprehensive (78.4%).
- Greater Review of Homework Assigned in the Previous Session Linked to Improved Outcomes: A positive correlation was found between therapists reviewing action recommendations from previous sessions and clients with mild depression achieving better treatment outcomes.
These findings—particularly those regarding care quality and outcomes—tie in nicely with the increased emphasis on value-based care (VBC) in behavioral health. VBC payment and delivery are designed to incentivize the provision of effective, efficient health care—with an emphasis on driving positive outcomes for clients.
Continuous monitoring and objective outcome measurement are integral components of VBC, and AI solutions hold great promise in supporting both. But more science is needed—another reason for behavioral health organizations to prioritize working with vendors who have a solid research track record.
Objective research is key to bridging the gap between the fast-changing mental health space and everyday practice within community-minded health organizations. The more tech companies commit to telling value stories rooted in real data, the better—for clients and clinicians alike.
2. Peer-Reviewed Research Publication
It’s not easy to plan and execute a real scientific study—let alone publish the findings in a vetted scientific journal. That’s why many healthcare companies hire a Chief Clinical Officer like me.
Stepping up to engage in peer review shows a company’s true intentions: to meaningfully participate in the healthcare field, rather than just profit from it. Allowing other scientists to “check your math” introduces a higher level of accountability and integrity. This opens our findings and methodologies to criticism, further validating their credibility.
As noted in this Frontiers in Digital Health article by Jennifer Huberty, PhD, et al., “Science should not be an afterthought. Science should be deeply integrated into companies’ infrastructure (e.g., internal or external science teams, roadmaps, communication pathways) and operate as a key component of business strategy. Incorporating science into business strategy ensures that decisions are rooted in evidence, which fosters innovation and drives companies towards sustainable growth and success.”
Publishing research in peer-reviewed journals also adds to the knowledge base of mental health professionals around the world. When we publish, we prove our commitment to elevating behavioral health as a whole.
Here’s a snapshot of Eleos Health’s public body of research work, including both published studies and papers currently under peer review. Note that our Director of Data Science, Samuel Jefroykin, played a pivotal role in many of these papers.
- Published: Randomized Controlled Trial (RCT): Effects of an Artificial Intelligence Platform for Behavioral Interventions on Depression and Anxiety Symptoms. Research conducted in collaboration with The Ozark Center, The Freeman Health Center, Missouri.
- Published: Algorithm Development and Validation: Machine Learning Model to Predict Assignment of Therapy Homework in Behavioral Treatments. Research conducted with Josef Ruzek, PhD, from the VA, and the late C. Barr Taylor, MD, Stanford University.
- Published: Review Paper: Pragmatic AI-augmentation in mental healthcare: Key technologies, potential benefits, and real-world challenges and solutions for frontline clinicians. Written in collaboration with Katherine C. Kellogg, PhD, MIT Sloan School of Management.
- Published: Perspective: Using technology to enhance experiential learning through simulated role plays: A teaching note. Authored and published by Hanni Flaherty, PhD, Yeshiva University School of Social Work.
- Published: Retrospective Study: A Comparison Between Clinical Guidelines and Real-World Treatment Data in Examining the Use of Session Summaries. Written in collaboration with Simon Rego, PsyD, Chief of Psychology and Director of Psychology Training at Albert Einstein College of Medicine and Montefiore Medical Center.
- Published: Case Report: Leveraging the power of nondisruptive technologies to optimize mental health treatment. Written in collaboration with Steve Hollon, PhD, Vanderbilt University.
- Published: Perspective: Using AI-Supported Supervision in a University Telemental Health Training Clinic. Coauthored with Donna Sheperis, PhD, Palo Alto University.
- Accepted for Publication: The Impact of Action Recommendations and Review in Community-Based Therapy on Depression and Anxiety Outcomes: A Machine Learning Approach. Led by Amit Spinard, MSc, and co-authored with Drs. Taylor and Ruzek.
- Accepted for Publication: Training the Psychologist of the Future in the Use of Digital Therapeutics. Written with Drs. Taylor and Ruzek, Ellen Fitzsimmons-Craft, Washington University at St. Louis, and Eduardo Bunge, Palo Alto University.
- Submitted for Publication: Revolutionizing Mental Healthcare Services through AI-Augmentation: A New Model. Written with Steve Hollon, PhD, Vanderbilt University.
- Submitted for Publication: Use of Natural Language Processing to Identify Climate Distress in Therapy. Led by Lidor Bahar, MSc, and co-authored with Dr. Rego.
- Submitted for Publication: Artificial Intelligence to Capture Social Determinants of Health Discussions in Behavioral Health Sessions. Submitted for Publication – New Methodology: The Sound of Mental Health: Audio Features as Indicators of Depression and Anxiety Symptoms in Behavioral Treatment. Led by Daniel Sand, PhD, and written with Björn Schuller, PhD, Imperial College London.
- Preprint available: Integrating Artificial Intelligence in a Telemental Health Clinic. Conducted with Victoria Guinn, MS, Roshni Joseph, MS, and Drs. Sheperis and Bunge.
Check out our Science page for more information on our past research efforts and new studies that are underway. This is a huge point of pride for us, and something you won’t find with most healthcare IT vendors. While some may conduct internal research and publish the results in a whitepaper, it’s important to remember that at the end of the day, a peer-reviewed journal article—unlike most whitepapers—is intended to advance health care as a whole by contributing to the body of scientifically vetted knowledge.
3. Clinician and Expert Involvement
Real-world clinical expertise is the heartbeat of effective healthcare technology. As a clinical professional myself, I know that no amount of customer research or market analysis is a substitute for the deep insights gained over years of clinical practice.
I’d go so far as to say that occasional consultations aren’t enough—it’s vital that clinical experts are deeply entrenched in product development from top to bottom, continually reviewing user activity and feedback, identifying potential pitfalls, and suggesting improvements. You simply can’t build a technology for behavioral health providers without consistent, thoughtful input from behavioral health providers.
At Eleos, our “secret sauce”—what really separates us from all the other AI solutions out there—is our diverse in-house team of behavioral health providers. Together, they boast over 180 years of experience in behavioral health practice.
On top of that, we’ve filled our clinical advisory group with experts across the healthcare spectrum, including:
- Dr. Denise Ernst: Motivational Interviewing Expert
- Dr. Anupam B. Jena: Professor of Health Care Policy, Harvard Medical School
- Dr. Dennis Morrison: Owner, Morrison Consulting
- Dr. Abigail Hirsch: Co-Founder & Chief Clinical Officer, Lin Health, and former Chief Clinical Officer, MyStrength
- Douglas Kim: Former revenue and commercialization leader at Kyruus, Cogito, Pegasystems; Fellow at MIT
- Prof. Steven Hollon: Gertrude Conaway Vanderbilt Professor of Psychology
- Dr. Simon Rego: Chief of Psychology, Montefiore Medical Center
- Dr. David Shulkin: The ninth United States Secretary of Veterans Affairs
Lastly, we regularly recruit external clinical professionals to test and validate the work we’re doing in-house—including those who participate in the research studies noted in the section above.
Integrating therapists of different backgrounds into our development process helps us create solutions that are both technologically advanced and deeply rooted in the realities of behavioral health practice. This ultimately makes our products more effective and usable—and guards against disruption to the therapeutic relationships our clinician users strive to protect at all costs.
4. Continuous Improvement
I firmly believe that technological innovation in behavioral health should benefit everyone involved—researchers, therapists, businesses, and most importantly, the people seeking our care. By moving research from the lab to the real world, we transform scientific breakthroughs into practical tools that do just that.
At Eleos, we’ve honed the art of translating complex research into practical, easy-to-use tools intended to move behavioral healthcare forward. Our recipe for success boils down to four ingredients:
- Cross-Functional Collaboration: Our development team is a melting pot of multidisciplinary expertise, with folks from product-dev, science, and clinical backgrounds collaborating closely and frequently. This synergy ensures each update or new feature is prioritized and created with a balanced perspective. For example, conducting research on climate anxiety in therapy necessitated teamwork across multiple departments within the company.
- Responsiveness to Client Needs: Drawing from our roots in behavioral health, we place a strong emphasis on listening. By having regular, candid conversations with both new and existing users, we can make sure we’re tailoring our technology to meet their most pressing needs. For example, our research on how social determinants of health are discussed and documented in the context of therapy was an essential response to the changing discourse on care equity in behavioral health.
- Proactive Adaptation: At Eleos, we take a proactive approach to incorporating new research findings into our technology. For example, we integrated the findings from our study on AI detection of therapy homework into the Eleos product almost immediately. Our agility is key to staying at the forefront of innovation.
- Goal Alignment: Throughout our product development process, we check to make sure our goals are in sync with our overarching mission of enhancing the care experience for behavioral health providers and their clients.
Science is the driving force behind progress—across all facets of human existence, but especially in health care.
In the health tech space, the challenge is not letting the rush to innovate distract us from grounding that innovation in evidence. The stakes—client health and well-being—are just too high.
And just as we must hold our tech vendors accountable to this high standard of scientific due diligence, so too must we hold ourselves accountable to working only with technology partners who can show—objectively and without a doubt—that science is baked into their DNA.