Real-world evaluation of an AI mental health app

Many people experience mental health challenges that can have significant negative impacts on their health and well-being. The prevalence of these challenges means that the demand for mental health services in the UK is greater than the capacity to provide support, and many people face weeks on the waitlist before accessing treatment. 

Our lab co-led a mixed-methods randomised control trial investigating whether a digital health app, Wysa, could support people during this waiting period to manage their mental health and improve their resilience. Wysa uses an artificial intelligence-based chatbot to provide self-management support for patients. The research project examined its impact on patients’ symptoms of anxiety and depression during the referral process for standard UK mental health services and explored users’ engagement and experiences with the app. The project was one of a handful funded by a £36 million boost from central government for AI technologies to transform NHS care, in the second wave of the NHS AI Lab’s AI in Health and Care Award, alongside funding from the National Institute for Health Research (NIHR). 

Relevant papers: 

• Artificial intelligence and machine learning in mobile apps for mental health: A scoping review | PLOS Digital Health 

Evaluation of a mental health app recommender system

Mental health challenges are highly prevalent in the UK, and the demand for mental health services far outstrips the supply. People face wait times of up to 6 months to receive treatment despite initiatives such as NHS Talking Therapies, which aimed to reduce this time to less than 6 weeks for 75% of patients. Support while on the waiting list is negligible, making it necessary for patients to turn to mental health apps for self-management support. There are tens of thousands of apps available, however, and not many of them have strong evidence of effectiveness. Patients can also find it difficult to find an app that will meet their individual needs. 

Syndi, an integrated digital health platform, aims to address this need by providing highly personalised app recommendations to patients looking for self-management support for their mental health. In collaboration with Syndi Ltd., we won an Innovate UK Biomedical Catalyst award to fund the development and evaluation of their recommender system for application in the NHS. Our lab is currently investigating Syndi’s acceptability, usability, and usefulness at generating mental health app recommendations and improving mental health and well-being.

Co-producing a mobile app for monitoring and managing bipolar disorder

There is an increasing burden of mental health issues on the healthcare system, which is straining the capacity of mental health services and professionals, especially in the wake of the COVID-19 pandemic. Early onset disorders, like bipolar disorder, can have a significant and increasing impact on patients’ overall health and well-being as they age (for example, by increasing risk of dementia and other conditions). By helping patients learn to manage their condition from a young age, early interventions have the potential to reduce the impact of mood episodes over their lifetime. Digital tools, such as mobile apps, can be easily and privately accessed by people who may otherwise be unable or unwilling to seek mental health support services. 

In collaboration with Dr Aditya Sharma’s (Clinical Senior Lecturer at Newcastle University and Honorary Consultant in Child and Adolescent Psychiatry) team, our lab won funding from the NIHR Newcastle Biomedical Research Centre to build on previous co-production work to design a prototype for a digital health intervention that can meet the evolving needs of patients with mood disorders throughout their lives. A series of workshops with young people, their family and carers, and healthcare professionals will be conducted to identify needs - and how these might have changed since Dr Sharma’s initial work in 2017 - and to co-produce improvements to the app to incorporate technological advancements and new suggestions. We will evaluate how the patient’s carers (a likely diverse population including older adults) interact with this technological approach to mental health monitoring and explore the implications for re-using this app for similar interventions in different populations (e.g. monitoring the mental health of older adults). The refined app that will be designed by the end of this project can then be developed and evaluated for acceptability, usability, and real-world impact in future evaluations.