Evaluating Mental Health Apps & Digital Therapy

• Digital therapies, like internet-based CBT, demonstrates comparable effectiveness to face-to-face therapy for mild to moderate mental health conditions.
• Adherence to digital therapies is a challenge, with unguided iCBT showing a 54% adherence rate.
• Dropout rates for unguided iCBT were 29%, compared to 19% for waiting lists.
• The effectiveness of digital therapies is closely linked to patient engagement, personalization, and the availability of clinical support.

Evaluating Mental Health Apps

• Of the numerous mental health apps available, approximately 3% are evidence-based.
• Frameworks for evaluating mental health apps include clinical safety, data protection, technical assurance, interoperability, usability, and accessibility.
• Clinical safety involves ensuring the app is evidence-based and includes risk assessment features.
• Data protection requires secure storage and GDPR compliance.
• Technical assurance covers functionality and reliability.
• Interoperability ensures integration with health records.
• Usability and accessibility focus on user engagement and inclusivity.
• Clinicians should understand app features, discuss risks and benefits with young people, and ensure app use complements professional support.

NICE Early Value Assessment (EVA) – 2023

• The NICE EVA aims to expedite the adoption of digital mental health interventions for children and young people.
• Digital CBT shows promise in aiding mild/moderate anxiety. Reducing therapy time via digital tools.
• Cost-effectiveness is a key benefit, with digital therapies requiring significantly less therapist time than face-to-face therapy.
• Remote access is preferred by neurodivergent individuals, enhancing equality.
• An initial clinical assessment is required to determine suitability for digital interventions.
• A primary risk is low adherence due to varying engagement rates.
• Further research is needed on effectiveness in specific subgroups, such as neurodivergent youth. NICE will reassess these interventions within 3 years before full NHS adoption.

Role of AI & Data Science in Mental Health

• AI-driven monitoring can analyze various sources for early signs of relapse or treatment response.
• Predictive analytics can aid in detecting disease development and relapse risk.
• AI-enhanced therapeutic fidelity analyzes therapy session transcripts.
• Robotic Processing (RPA) reduces clinician workload by automating tasks.
• Virtual Reality (VR) aids treatment and training.
• Social media and peer support platforms are avenues for rapid knowledge-sharing.

CAMHS Digital Lab Origins

• High demand and long waiting times in CAMHS services prompted digital intervention.
• Complex referral and treatment paths, along with limited integration between NHS and third-sector services, posed significant challenges.
• Inefficient use of digital tools and gaps in preventive care and early intervention prompted changes.
• Clinicians lacking digital product development expertise hindered progress.
• NHS processes did not incentivize continuous improvement, which became impetus for change.
• Effective digital solutions weren’t sustained long-term due to organizational constraints.

CAMHS Digital Lab Objectives

• Expansion of access to evidence-based interventions for those outside secondary care became a core mission.
• Service delivery enhancement through digital innovations.

Strategic focuses

• Inclusive Design: Focus on co-design with young people for accessibility.
• Health Intelligence: Focus on using data-driven insights for mental health services.
• Digital Therapeutics & Assessment: Focus on improving treatment access.
• Data Science & Discovery: Focus on using AI-driven analytics for risk prediction.
• Commercial Development: Focus on partnering with industry to sustain innovation.
• Education & Training: Focus on upskilling clinicians in digital mental health.

Clinical & Population Analytics

• Emergency service use for self-harm incident data in schools provides key clinical insights.
• Routinely collected data is used to identify risk patterns.
• Machine learning can be used for ADHD risk prediction in infant school data.
• Tools such as CRIS/BI, linked external data sources, and digital survey tools are employed.

Digital Therapeutics & Assessment

• The myHealthE system digitizes symptom tracking for more efficient monitoring, providing better data for clinicians to tailor treatment.
• Digital tools are used to personalize care pathways.
• Racial and socioeconomic disparities in CAMHS services are specifically addressed.

Training & Outreach

• Inclusive design principles are used to ensure diverse engagement.
• Training is provided for clinicians and researchers to develop and evaluate digital products.
• Design Council’s Double Diamond Framework is used for co-designing innovations.

Data Science & Discovery

• Actigraphic motion tracking assesses ADHD in school/home settings.
• AI-driven automated detection of expressed emotion in caregivers’ speech.
• Machine learning analyzes mother-infant interactions.
• Speech & movement AI are used.
• Wearable sensors enable real-time behavioral tracking.

CAMHS Digital Lab Governance & Structure

• Project teams are formed as needed, characterized by organizational flexibility.
• Strategic oversight at Trust level involving Maudsley, KCL, and NHS.
• The technical team comprises experts in AI, informatics, UX design, and clinical psychiatry.
• Collaborations with academia and industry accelerate digital solutions.

Impact Summary

• CAMHS Digital Lab helps to provide faster access to interventions and more efficient tracking of mental health symptoms.
• Furthermore, it helps to provide AI-driven analytics for better risk assessment and enhanced clinician training in digital health.
• Reduced disparities in CAMHS service delivery are achieved.