Artificial intelligence interventions in the mental healthcare of adolescents
Bibliographic record
Abstract
Background: Adolescence is a critical phase in a person's life since it might affect behaviors and conditions that impact physical and mental health and contribute to adulthood illnesses. Primary care physicians (PCPs) are increasingly acknowledged for their critical role in detecting and managing adolescents' mental health problems. PCPs however face several challenges when providing mental healthcare to this population. Artificial Intelligence (AI) interventions may provide some solutions if they are adapted for primary care use. Given the rise and importance of mental health problems among adolescents, it is critical to identify such interventions and assess PCPs potential interest in using them. Objective: Two studies were conducted to respond to these objectives. The first sought to identify AI interventions tested and/or implemented in adolescents’ mental healthcare. The second explored perceived challenges of primary care physicians in providing adolescents’ mental healthcare, along with their perceived needs for AI interventions that would be helpful in dealing with adolescents’ mental health issues.Methods: In the first study a systematic scoping review was conducted to identify AI interventions tested and/or implemented in adolescents’ mental healthcare. Using the Levac et al. framework, we searched five electronic databases (MEDLINE, Embase, Web of Science Core Collection, Compendex, INSPEC) from inception date until February 2020. Two independent reviewers identified articles based on title and abstract, and full text. Inclusion criteria were patients aged 10 to 19 receiving mental healthcare from healthcare professionals (HCPs) and any HCP that provides for this demographic and listing of AI interventions that were tested and/or implemented. Outcomes were any related to patients, HCPs, or the healthcare system. Setting: any healthcare setting.The second study was a qualitative descriptive, based on focus group (FG) discussion with a sample of PCPs in the Montreal, Canada. Through purposeful sampling, we recruited four PCPs with specific interest in adolescent mental healthcare and AI interventions. FG discussions were conducted and audio-visually recorded through Zoom software and lasted 1 ¼ hours. The discussion was transcribed verbatim, followed by thematic analysis using A-priori and inductive coding.Results: Scoping review: 30 papers were retained for analysis from 1044 retrieved. AI interventions were most commonly reported for Autism Spectrum Disorder (n=3), Unspecified Outcomes of Psychological Stress/Pressure Level (n=3), Substance Use Disorder (n=2) and Dysfunctional Behavior (n=2). The application of AI within the continuum of mental healthcare for adolescents was used for the mediation of diagnostic processes (n=23), monitoring and evaluation (n=8), treatment (n=5), and prognosis (n=2). Focus Group study: PCPs saw AI interventions as potentially cost-effective, able to handle large amounts of data, and relatively credible. They envisioned AI to assist in collecting patients' data, suggesting a diagnosis, and establishing a treatment plan. However, they were concerned about these interventions' performances and outcomes and feared losing clinical competency. Participants highlighted systematic challenges PCPs face while giving care to adolescents, including parental involvement and psychosocial influences. PCPs desired interventions that were user-friendly. Conclusion: To implement AI appropriately, greater participation and critical understanding of patients', physicians', and data scientists' opinions on AI use in clinical processes is required, resulting in a feedback loop of co-designing future AI initiatives. It is predominantly believed that the promise of AI in adolescents' mental health was considerable. These first stages and analyses provide the foundation for future work examining the practical usability, application, and effectiveness of these interventions in adolescents' mental healthcare
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.014 | 0.054 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.003 | 0.003 |
| Bibliometrics | 0.006 | 0.005 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".