AI-Driven Clinical Interview for Depression and Anxiety

(SEQUOIA-1 Trial)

Age: 18+

Sex: Any

Trial Phase: Academic

Sponsor: Deliberate Solutions Inc.

Must not be taking: Benzodiazepines, Antipsychotics, Stimulants, others

Disqualifiers: Cognitive impairment, Schizophrenia, Neurodegenerative, others

No Placebo Group All trial participants will receive the active study treatment (no placebo)

Compensation: Varies

Number of Visits: Unknown

120 Participants Needed

Trial Summary

What is the purpose of this trial?

The SEQUOIA-1 study evaluates the effectiveness of Artificial Intelligence (AI) in measuring depression and anxiety severity in adults. Investigators from Deliberate Solutions, Inc. and Baylor College of Medicine are conducting this study to determine whether AI can provide reliable clinical assessments of mood and anxiety disorders. In clinical trials for new depression and anxiety treatments, human clinicians typically conduct interviews to evaluate participants' symptoms. These assessments are critical but may vary based on the clinician's experience or interview style, potentially affecting the reliability of research findings. To address this challenge, the study team developed an AI-based Clinical Outcome Assessment tool, called AI-COA®, which analyzes video interviews to measure symptoms of depression and anxiety consistently and objectively. AI-COA® has been accepted by the U.S. Food and Drug Administration (FDA) into the Innovative Science and Technology Approaches for New Drugs (ISTAND) pilot program. The primary objectives of the SEQUOIA-1 study are to collect additional data to improve model accuracy and to evaluate model performance across diverse demographic groups. The study also pilots the use of an AI interviewer-an interactive digital agent-to conduct remote assessments. During the study, participants will complete questionnaires about their symptoms and perform brief tasks. Participants will also provide feedback regarding their experience interacting with the AI interviewer. All assessments will be securely video-recorded. Recorded videos will be analyzed by AI-COA® to determine depression and anxiety symptom severity. These results will be compared to assessments conducted by human clinicians. The development and validation of reliable, AI-driven assessment tools through this study aim to enhance the accuracy of mental health evaluations, potentially improving the testing and approval processes for new treatments targeting depression and anxiety.

Do I need to stop taking my current medications to join the trial?

The trial requires that you stop taking certain medications, such as fast-acting treatments for depression or anxiety, benzodiazepines, antipsychotics, mood stabilizers, stimulants, and epilepsy medications, for specific periods before joining. If you're on any of these, you may need to stop them for 2 to 6 weeks, depending on the medication.

What data supports the effectiveness of the treatment AI-COA® for depression and anxiety?

Research on AI-driven tools like Wysa shows that users can form a strong therapeutic bond with AI, similar to traditional therapy, which is important for effective treatment. Additionally, AI-based systems like Aifred have been found useful by physicians in selecting treatments for depression, suggesting potential benefits for AI-COA® in managing depression and anxiety. ^{1 2 3 4 5}

Is the AI-Driven Clinical Interview for Depression and Anxiety safe for humans?

The research does not provide specific safety data for the AI-Driven Clinical Interview for Depression and Anxiety, but it does mention that subjects felt comfortable and less embarrassed giving information to a computer compared to a clinician. ^{4 5 6 7 8}

How is the AI-COA® treatment different from other treatments for depression and anxiety?

AI-COA® is unique because it uses artificial intelligence to conduct clinical interviews, potentially offering a more objective and precise assessment of depression and anxiety symptoms compared to traditional self-reported measures, which can be biased. This approach may enhance early detection and monitoring of these conditions by analyzing language patterns and other digital biomarkers. ^{1 3 9 10 11}

Eligibility Criteria

The SEQUOIA-1 study is for adults with depression or anxiety who are interested in contributing to the development of an AI tool that assesses mental health. Participants should be willing to complete questionnaires, perform tasks, and interact with an AI interviewer on video.

Inclusion Criteria

HAM-D 17 score greater than 10

English fluency

Access to a laptop or other computer with a well-functioning microphone and webcam, and a stable Internet connection

Exclusion Criteria

Vulnerable or protected populations (e.g. prisoners)

Acute intoxication at the time of the assessments

Impairment that would prevent participants from completing an online survey and/or engaging in clinician assessment interviews (e.g., visual impairment, motor impairment, hearing impairment)

Timeline

Screening

Participants are screened for eligibility to participate in the trial

2-4 weeks

AI Interview and Assessment

Participants undergo AI-based video interviews and complete questionnaires about their symptoms. Feedback is collected regarding their experience with the AI interviewer.

2 days

Remote assessments

Follow-up

Participants are monitored for concordance of AI and human assessments of depression and anxiety symptoms.

2 days

Treatment Details

Interventions

• AI-COA®

Trial Overview This trial tests an AI-driven clinical interview tool called AI-COA® designed to measure depression and anxiety severity more consistently than human clinicians. It involves participants completing assessments which will be compared against traditional clinician evaluations.

Participant Groups

Group I: B2: A1: AI interview second and measures set 2 Experimental Treatment 1 Intervention

Group II: B1: AI interview second, and measures set 1 Experimental Treatment 1 Intervention

Group III: A2: AI interview first, and measures set 2 Experimental Treatment 1 Intervention

A2: AI interview first, and measures set 2

Group IV: A1: AI interview first, and measures set 1 Experimental Treatment 1 Intervention

A1: AI interview first, and measures set 1

Who Is Running the Clinical Trial?

Deliberate Solutions Inc.

Lead Sponsor

Baylor College of Medicine

Collaborator

Trials

1,044

Recruited

6,031,000+

Findings from Research

In a study involving 1,205 users of the Wysa app for mental health, users reported a strong therapeutic alliance with the AI conversational agent, with scores comparable to traditional therapy methods like CBT.

Over a brief period of 3 days, users' perceptions of their bond with Wysa improved, indicating that digital mental health services can effectively foster a therapeutic relationship similar to that found in face-to-face therapy.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Evaluating the Therapeutic Alliance With a Free-Text CBT Conversational Agent (Wysa): A Mixed-Methods Study. Beatty, C., Malik, T., Meheli, S., et al.[2022]

Project AFFECT aims to create an AI-based software, LyssnCBT, to automatically assess the quality of cognitive behavioral therapy (CBT) sessions, which could enhance training and supervision for therapists and improve treatment outcomes for clients.

In a planned Phase II trial involving 1,875 clients, the effectiveness of LyssnCBT will be evaluated to see if it can improve therapist skills, client outcomes, and reduce client drop-out rates, marking a significant step towards scalable quality assurance in psychotherapy.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Enhancing the quality of cognitive behavioral therapy in community mental health through artificial intelligence generated fidelity feedback (Project AFFECT): a study protocol. Creed, TA., Salama, L., Slevin, R., et al.[2023]

A study involving 308 participants with subclinical depression demonstrated that a conversational agent can effectively collect thought record data, making it feasible for patients to engage in cognitive therapy techniques.

Participants preferred discussing interpersonal situations over achievement-related ones, highlighting the importance of social contexts in their thought processes, but providing richer feedback did not increase motivation for recording thoughts.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Using a conversational agent for thought recording as a cognitive therapy task: Feasibility, content, and feedback. Burger, F., Neerincx, MA., Brinkman, WP.[2023]

References

1. Switzerland pubmed.ncbi.nlm.nih.gov

Evaluating the Therapeutic Alliance With a Free-Text CBT Conversational Agent (Wysa): A Mixed-Methods Study. [2022]

2. England pubmed.ncbi.nlm.nih.gov

Understanding the relationship between patient language and outcomes in internet-enabled cognitive behavioural therapy: A deep learning approach to automatic coding of session transcripts. [2021]

3. United States pubmed.ncbi.nlm.nih.gov

Analysis of Automated Clinical Depression Diagnosis in a Chinese Corpus. [2023]

4. United States pubmed.ncbi.nlm.nih.gov

Out with AI, in with the psychiatrist: a preference for human-derived clinical decision support in depression care. [2023]

5. Ireland pubmed.ncbi.nlm.nih.gov

Evaluating the perceived utility of an artificial intelligence-powered clinical decision support system for depression treatment using a simulation center. [2022]

6. England pubmed.ncbi.nlm.nih.gov

Computerized and clinician assessment of depression and anxiety: respondent evaluation and satisfaction. [2006]

7. England pubmed.ncbi.nlm.nih.gov

Enhancing the quality of cognitive behavioral therapy in community mental health through artificial intelligence generated fidelity feedback (Project AFFECT): a study protocol. [2023]

8. Canada pubmed.ncbi.nlm.nih.gov

Repeated Digitized Assessment of Risk and Symptom Profiles During Inpatient Treatment of Affective Disorder: Observational Study. [2020]

9. Switzerland pubmed.ncbi.nlm.nih.gov

A Study of Novel Exploratory Tools, Digital Technologies, and Central Nervous System Biomarkers to Characterize Unipolar Depression. [2021]

10. United States pubmed.ncbi.nlm.nih.gov

Depression and anxiety have distinct and overlapping language patterns: Results from a clinical interview. [2023]

11. Switzerland pubmed.ncbi.nlm.nih.gov

Using a conversational agent for thought recording as a cognitive therapy task: Feasibility, content, and feedback. [2023]