An AI-powered therapy app that currently requires a clinician to guide it will be redesigned to let patients work through it on their own, using avatars and automated triage to replace human oversight. NHS mental health services are overwhelmed, with children and young people waiting months for treatment. Existing digital tools lack personalisation and often fail to keep patients engaged. This project aims to solve that by building a self-guided version of an immersive digital therapy platform already used in six NHS Trusts. The current version cuts treatment from 8–16 sessions to 2–4, and an independent health economic report found it saves £21,000 per 100 patients while delivering five additional QALYs. But it still needs a clinician to run it, limiting how many people can access it. If the feasibility study succeeds, the self-guided version could let far more patients receive effective therapy without adding to clinician workloads. That would shorten waiting lists, improve access for children and young people, and give the NHS a scalable digital tool that reduces workforce strain. The project will also test whether AI triage can accurately replace clinician judgement in patient referrals, which could streamline intake across mental health services.
View original technical description
Objective: This project will investigate the feasibility of an AI-powered, self-guided digital therapy intervention for mental health treatment, reducing reliance on clinician involvement while maintaining clinical effectiveness. Background: NHS mental health services face critical workforce shortages and increasing demand, resulting in long wait times and limited access to therapy. Children and young people are especially affected, often waiting months to access services, increasing risk and reducing treatment effectiveness. While digital interventions such as online CBT modules exist, they lack personalisation and often show low engagement and efficacy. For example, a systematic review of digital CBT interventions (Carlbring et al., 2018) highlighted lower effectiveness in unguided formats compared to clinician-supported delivery. Our company has developed a clinically validated, immersive digital therapy platform, currently deployed across six NHS Trusts, that enhances CBT and graded exposure techniques using interactive, scenario-based environments. The intervention drastically reduces treatment times as it only takes 2-4 sessions as opposed to 8-16 for conventional CBT, in fact an independent health economic report by CHEATA found that NHS departments that used our intervention could save £21,000 per 100 patients whilst producing positive patient outcomes and delivering an additional 5 QUALYs. However, its current format requires a clinician to guide the intervention, limiting scalability and availability. Innovation: We will develop and prototype a self-guided version of our intervention by integrating: • AI-driven avatars using custom language models for natural, engaging therapeutic interactions. • Mobile app functionality, incorporating guided relaxation, coping strategies, emotion regulation, and gamification. • AI-powered triage, automating patient referrals based on referral data and user-reported symptoms to reduce clinician burden. Methodology: The project will follow a 12-month feasibility study to assess user engagement, clinical impact, and NHS workflow integration. We will: 1. Conduct stakeholder workshops to co-design self-guided functionality with clinicians and patients. 2. Develop and refine AI models through iterative testing for conversational accuracy, empathy, and risk identification. 3. Deploy a pilot study with NHS partners (Sheffield Children’s, GM CAMHS, and South West London CYP services) to assess: - Engagement metrics (completion rates, time-on-task) - Treatment readiness improvements (using standardised tools such as WSAS, GAD-7) - Clinical safety and acceptability - AI triage accuracy (measured against clinician judgement) Expected Impact: This project will generate key insights into AI-enhanced digital mental health interventions and assess their potential for NHS adoption. It supports current NHS priorities to expand access through scalable digital tools and reduce reliance on over-stretched clinical teams. If successful, it will improve access to mental health support, reduce waiting times, and alleviate clinician workload, providing a foundation for future large-scale trials and commercial rollout within NHS pathways.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know