Talk to the bot: AI assistant marks breakthrough for UK mental health

Talk to the bot: AI assistant marks breakthrough for UK mental health

Source Node: 1919559
<!--

-->

An artificial intelligence (AI) driven assessment tool for diagnosing mental health disorders has become the first mental health chatbot to secure a Class IIa UKCA (UK Conformity Assessed) medical device certification.

Using machine learning, Limbic Access is designed to support patient self-referral through digital conversations that are incorporated into the psychological therapy pathway. The chatbot can classify common mental health disorders treated by NHS Talking Therapies (IAPTs) with an accuracy of 93%.

The certification comes as NHS Improving Access to Psychological Therapies (IAPT) services are experiencing significant capacity challenges in the face of record demand. In 2021-22, 1.24 million referrals accessed IAPT services, compared to 1.02 million in 2020-21, an increase of 21.5%.

With services overstretched, the UK government is under pressure to boost investment. On January 23, prime minister Rishi Sunak announced a £150 million investment to support mental health services and ease pressure on healthcare staff. The funds will be used for 150 new projects centered on supporting the provision of mental health crisis response and urgent mental health care.

However, the investment has been labeled inadequate. Commenting on the announcement, Paul Spencer, Head of Health, Policy & Campaigns at mental health charity Mind called for further investment to address the growing crisis.

“Mental health services have faced decades of underfunding in comparison to physical health, and mental health received none of the Covid-19 recovery funding announced in 2021, despite huge backlogs in mental health care and the well-known impact of the pandemic on the nation’s mental health,” said Spencer. “We’re seeing the human cost of that playing out now as people wait days in our A&Es for mental health beds and people experiencing complex mental health problems struggle without help around us in our communities.”

Following its breakthrough certification, Limbic Access is aiming to deepen ties with the NHS to augment the psychotherapy process. The London, UK-based health tech developer has already trialed the AI therapy assistant with more than 130,000 NHS patients. UKCA auditors, SGS, reviewed clinical evidence generated from more than 60,000 referrals and found that when compared to standard referral methods such as telephone calls and online forms, the chatbot showed a 53% improvement in recovery rates and 45% fewer changes in treatment, due to increased triage accuracy.

In an interview with Medical Device Network, Dr. Ross Harper, co-founder, and CEO of Limbic discusses how AI tech can impact mental health assessment. The interview has been edited for clarity and length.

CL: What was your vision when developing Limbic Access for mental health care?

Dr. Ross Harper: Our mission from the beginning has always been that we want to use technology to make the highest quality mental health care available to everybody everywhere, regardless of their socio-economic background.

AI is the perfect technology to make this world a reality so we began building something that can continuously learn about a patient’s mental state – whether through clinical screening questions, free text conversational data or a variety of other data streams. We combine this with algorithms in the background which deliver the probability of possible mental health issues that the patient might be experiencing.

One crucial piece of the puzzle we had to solve was how are we going to make sure this is used safely in a clinical setting? The answer could have been around substituting humans with AI but that wasn’t appropriate in the case of our technology. Instead, we focused on how we can use AI alongside trained mental health professionals to augment the therapy process and empower them to support more people.

The system algorithms help manage the self-referral by doing an initial analysis and identifying symptoms of mental illness and this can then be shared with the clinician in a way that allows them to really get to the root of the patient’s issue faster and identify the most appropriate treatment pathway from there on. But the final step in using machine learning models is ensuring they are regulated appropriately to meet the highest clinical and technical standards. That’s why our Class IIa medical device certification is important – it independently verifies that we can do, what we say, and that we can be used as an integral tool in psychotherapy.

CL: Do you see this certification as significant for the evolution of the digital mental health space and AI therapy?

Dr. Ross Harper: Yes, it’s a landmark as this is first of its kind. It’s also exciting as this is one of the first examples of AI working alongside clinicians in psychotherapy. The amalgamation of AI and human therapist is the future and the only way to safely deliver these sorts of technologies in mental health care. I think it’s important that we haven’t just tried to digitize existing treatment exercises into an app and deliver that up as a tool. What we really wanted was to develop probabilistic models and use AI to provide a decision support tool, which can be hugely valuable and fundamentally support mental health services, which are very much overburdened globally.

CL: How do you see your relationship with the NHS evolving?

Dr. Ross Harper: We have a lot of admiration for the NHS and it’s the market we have chosen to develop in. We believe in the organization and are proud of Limbic as a company founded in in the UK. We hope to continue providing this sort of technology to the NHS and listen to their problems and challenges to see whether our expertise in AI can support them because we believe very much in their mission and what they stand for.

CL: Are you looking to expand into additional markets outside of the UK?

Dr. Ross Harper: We’ve got our eyes open, and we have explored other markets. But I think for a company like ours, limited focus is critical because you can’t do everything to begin with. For the time being we are focused on how we can augment NHS psychotherapy, support clinicians and patients and services within this care system. It goes without saying that a lot of the benefits that I have described that we’re able to support the NHS with, they do translate to other markets delivering psychological therapies so we have explored that purely from a gathering information point to understand how our solutions may fit within those markets. But right now, it’s the NHS for us.

CL: The digital mental health space is expanding with new technologies. How will Limbic differentiate from other companies?

Dr. Ross Harper: It’s tough to comment on other companies but I can talk with authority about Limbic. We’re the only technological solution that is regulated to this level to support triage and assessment in adult mental health care. We are also the most widely used tool of its kind in the NHS so, while others are quite rightly looking at entering the space – currently Limbic is the leader. We’ve also got the evidence that shows our decision support can support both patients and clinicians because that’s a key part of this puzzle rather than being a straight business to consumer product. Other technologies are also gearing towards wellbeing, and we aren’t positioning ourselves as a wellbeing tool for corporate or direct consumers. Nor is this an area we plan to push. We consider ourselves to be a clinical tool – we were born in mental health care and plan to stay in this space.

CL: What are some of the key milestones for Limbic in 2023?

Dr. Ross Harper: The key things are continuing to develop our triage and assessment support tool. We also have a product called Limbic Care, that we plan to add some exciting new functionalities to in 2023. We’ve also got new machine learning models coming out of our lab, which will be used to create service efficiencies and improve patient experience and clinical outcomes. All of these are currently in the pipeline so it’s going to be an exciting year.

<!-- GPT AdSlot 3 for Ad unit 'Verdict/Verdict_In_Article' ### Size: [[670,220]] --
googletag.cmd.push(function() { googletag.display('div-gpt-ad-8581390-1'); });
!-- End AdSlot 3 -->

Time Stamp:

More from Medical Device Network