AI has the potential to revolutionise mental healthcare, with studies published in respected journals, such as The Lancet, identifying it as the prime solution to a range of challenges, including delayed, inaccurate and inefficient care delivery.
With around one in six adults thought to suffer from a common mental disorder, this technology provides a timely solution to a growing problem. NHS Talking Therapy Services provide highly effective support to people struggling with anxiety and depression. However, with capacity challenges exacerbated by Covid-19, there is now a waiting list for mental health services of between 1.6m and 1.8m according to organisations like NHS Providers, the Royal College of Psychiatrists and Mind.
Studies show that long waits are associated with a variety of negative outcomes for patients and, with 68.6% of NHS Talk Therapy psychological wellbeing practitioners (PWPs) suffering burnout even before the pandemic, the backlog will also be impacting on NHS Talking Therapies staff.
Depression and anxiety are costly to individuals, the NHS and wider society. People with depression and anxiety have a lower life expectancy and are more at risk of a wide range of physical conditions, including heart disease, stroke, lung disease and epilepsy. Mental health costs the NHS £14.3bn per year, while the cost to society is reported to be at least £117.9bn a year – supporting the need for urgent solutions that can make services more efficient and deliver improved patient outcomes.
There is no doubt in our minds that AI is the primary solution to these challenges – we agree with the Lancet report that came to the same conclusion. However, if this technology is to be used to support the NHS in its mission to improve health and wellbeing, it must be exercised responsibly and safely. That means ensuring that it meets the rigorous standards expected by regulators and outlined by leaders including the NHS AI Lab, NHS England’s Transformation Directorate and the Medicines and Healthcare Products Regulatory Agency (MHRA). Only then can we expect clinicians and patients to place their confidence and trust in this formidable technology.
Limbic Access, our multi-award-winning e-triage and assessment mental health chatbot, recently became the first and only AI mental health chatbot in the world to be regulated as a UKCA Class II Medical Device status. This recognises its clinical effectiveness, safety and risk management. It also means that our evidence-based decision support tool, which is already being used by more than 25% of NHS Talking Therapy services, is able to be safely incorporated into the psychological therapy pathway to support patient self-referral.
The awarding of this certification came after a rigorous audit into Limbic Access’s evidence and processes. From an examination of more than 60,000 NHS patient referrals, UKCA auditors verified that, when compared to standard referral methods such as telephone calls and online forms, our conversational AI mental health chatbot helped to deliver:
- 53% improvement in recovery rates
- 45% fewer changes in treatment, due to increased triage accuracy
- 5% reduction in assessment time – a saving of 12.7 minutes per referral
- 18% reduction in treatment dropouts
- 13% reduction in wait time for assessment
- 5% shorter wait time for treatment
Built around a friendly interface and an AI algorithm that has been trained on data gathered from real-world NHS patient interactions, our mental health chatbot assists patients at the beginning of their referral journey, as they seek support from the NHS Talking Therapy services for mental health issues like anxiety or depression. Engaging them with natural, conversation, Limbic asks the key questions that would typically be asked by a healthcare professional or via long, static form that can be challenging for service users with common mental disorders.
Limbic Access is then able to screen the patient for service eligibility, capture key demographic information and identify common mental health disorders with a 93% accuracy. All of this information supports the clinician conducting their clinical assessment with high quality data that enables them to focus on the most pertinent presenting issues – saving significant time, admin and paperwork.
Evidence shows that we’re improving patient outcomes and providing a better experience for both patients and clinicians, which is really important to us. At its heart, psychological therapy is intrinsically human and we think it should stay that way, after all, NHS Talking Therapy services have already been proven to be effective. Limbic Access is designed to augment the clinical journey rather than fully automate it.
We think that twin focus on safety and impact should be at the heart of any AI developments in healthcare. AI has amazing potential, of that there can be no doubt. The health service must prioritise those technologies that can prove positive outcomes, improved experiences and also that they meet rigorous standards of safety and risk management. Only then will clinicians feel confident in utilising these technologies to deliver even better quality mental healthcare to even more people.
About the author
Dr Ross Harper is co-founder and CEO of Limbic. He studied neuroscience at Cambridge University, mathematical modelling at UCL, and then married these two fields together during his PhD in computational neuroscience at the world-renowned Gatsby Unit.
