Narrowing the Needs-Provision Gap for Young People's Mental Health: Developing and Piloting an AI-Guided Behavioural Activation Intervention

Background

In the UK, one in three young people are impacted by poor mental health. However, due to limited resources and personnel within the National Health Service (NHS), NHS Talking Therapies for young people are largely characterised by low availability, poor accessibility, and waiting periods exceeding six months.

To help relieve the burden placed on the NHS, Artificial Intelligence (AI) chatbots - like OpenAI's ChatGPT - have been proposed as a means to deliver personalised, low-cost psychotherapy at scale.

However, despite widespread recognition of their potential, there is a significant lack of research examining the effectiveness of AI-guided psychotherapeutic interventions. Notably, few interventions have been co-produced with end-users, even fewer have targeted young people, and none have been tested within UK-based populations.

Aims and Objectives

To help resolve these research gaps, the overarching aim of this study will be to co-produce an AI-guided psychotherapeutic intervention for UK-based young people aged 16-25. This will be achieved through the following objectives: (A) Identify the features of that will make an AI-guided intervention maximally feasible, acceptable, and effective.

Using a combination of existing data from surveys, interviews, and clinical observations, along with newly collected primary data, this research will explore the factors that are crucial for successful implementation of AI-guided interventions. To achieve this, a blend of exploratory qualitative, quantitative, and mixed-methods approaches will be leveraged.

(B) Draw upon user input to inform the development of an AI-guided intervention.

Adhering to the Medical Research Council's framework for intervention development, the project will involve an iterative process of usability testing, participant interviews, and responsive intervention refinements. This approach is designed to ensure that the intervention is precisely tailored to the specific needs and preferences of the target population.

(C) Test the impact of the AI-guided intervention within a limited pilot study.

A two-arm, parallel-group randomized control pilot trial will be conducted to compare the feasibility, acceptability, and preliminary effectiveness of the AI-guided intervention with a wait-list control. External Partners

To maximize the real-world impact and reach of this project, we have partnered with Healthwatch, an independent consumer champion for UK-based healthcare services. Through this partnership, Healthwatch has agreed to provide us with exclusive access to their data repository, participant recruitment support, training opportunities, and channels for research dissemination.

Overseas Institutional Visits

This project will also involve an overseas institutional visit to a leading AI research hub. The visit will encompass attendance at academic conferences, participation in clinical innovation competitions, and advanced training in AI ethics and applications. It will also provide valuable opportunities for research collaboration, both on the present project and other related initiatives at the host institution.

ESRC Priorities

This project's economic and societal impact lies within its potential to bridge the gap between the needs of NHS patients and current service provisions. Aligning closely with SWDTP and ESRC priorities, this project directly contributes to key areas such as "improving health, wellbeing, and social care" and "improving public services".