AI-Supported Therapy for Depression and Anxiety Compared With Standard CBT

Evaluating the Efficacy of an AI Delivered, Neurosymbolic, Human Supervised Digital Intervention for Depression and Anxiety Versus Standard Cognitive Behavioural Therapy in Young Persons and Adults

This study is a pivotal, randomised, controlled, non-inferiority trial evaluating "Nook," an AI-delivered, neurosymbolic, clinician-supervised digital psychological intervention for depression and anxiety, compared with standard cognitive behavioural therapy (CBT). The trial will recruit 400 participants aged 16-64 years in the UK with moderate depression and/or anxiety symptoms. Participants will be randomised to receive either Nook or therapist-delivered CBT.

The primary objective is to determine whether Nook is non-inferior to CBT in reducing depression and anxiety symptoms, measured using the PHQ-9/PHQ-A and GAD-7 scales. Secondary outcomes include quality of life, functional impairment, sleep quality, treatment engagement, participant satisfaction, safety outcomes, and exploratory health economic measures.

The intervention incorporates clinician oversight and predefined escalation pathways for suicidality and clinical deterioration. Outcomes will be analysed using longitudinal mixed-effects models under an intention-to-treat framework.

Inclusion Criteria:

• Have symptoms of Generalized Anxiety Disorder (GAD) and/or symptoms of --•Major Depressive Disorder (MDD) as the primary reason for seeking treatment (formal diagnosis not required).(as determined by a Psychological Well-being Practitioner).
• Meet symptom-severity criteria on either validated screening measure:
• Depression: PHQ-9 (or PHQ-A for young persons) total score between 10 and 19, corresponding to moderate to moderately-severe symptoms, and/or
• Anxiety: GAD-7 total score between 8 and 21, corresponding to mild to severe symptoms.
• Aged 16-64 years
• If taking psychotropic medication for depression and/or anxiety, be on a stable regimen for at least 6 weeks prior to screening, with no initiation, discontinuation, or dose change during that period.
• Have reliable access to a compatible, internet-connected device and are able to use it for screening/eligibility, and the intervention and assessments (any potential costs to participants will be clearly noted in the PIS).
• Possess sufficient English language proficiency and cognitive capacity to engage with the digital therapeutic content and complete questionnaires.
• Provide informed consent.
• Willing to be randomised and to participate in a clinically-supervised CBT-based AI programme, including completion of scheduled outcome assessments.

Exclusion Criteria:

• Depression: PHQ-9 (or PHQ-A for young persons) total score >≥ 20
• Present with a primary or comorbid diagnosis (or history of) that is unsuitable for a digital CBT-based intervention (as judged by the Investigator), including:
• Post-traumatic stress disorder (PTSD) or complex trauma
• Psychotic disorder, bipolar disorder, and/or mania
• Complex or treatment-resistant obsessive-compulsive disorder (OCD)
• Personality disorder
• Eating disorder
• Substance or alcohol use disorder
• Exhibit high-risk clinical concerns, including:
• Current suicidal ideation with intent or plan (as indicated by PHQ-9 score and/or participant disclosure)
• Suicide attempt within the past 12 months
• Ongoing self-harming behaviours
• Requirement for urgent or crisis mental-health intervention (as indicated by •PHQ-9 score and/or participant disclosure)
• In participants aged 25 years and under, current treatment with an antidepressant medication initiated or dose-adjusted within the past 12 weeks (due to the recognised