Type of publication:
Journal article
Author(s):
Adindu K.N.; Akubue N.; Jude N.O.; Onakoya A.; Chukwunonye C.; Odion O.; *Okengwu C.G.; Uchechukwu N.; Osita-Obasi P.Z.; Ezike A.; Bello I.; Olenloa E.; Eruteya O.O.; Oyewole S.A.
Citation:
SSRN. (no pagination), 2025. Date of Publication: 20 May 2025. [preprint]
Abstract:
Background: Adolescence constitutes a critical developmental stage marked by the onset of mental health difficulties, yet timely access to effective mental health care remains a significant challenge for many adolescents in the United Kingdom (UK). Artificial intelligence (AI)-enabled digital therapies present innovative opportunities to address these gaps. Objective(s): This systematic review critically assesses current evidence on AI-driven digital interventions for adolescent mental health within the UK, highlighting their potential opportunities, barriers to implementation, and pertinent ethical considerations. Method(s): Employing a mixed-methods design, a systematic literature review adhering to PRISMA guidelines was combined with thematic analysis of semi-structured interviews. Comprehensive database searches (MEDLINE, PsycINFO, Web of Science; 2013-2023) targeted studies involving UK adolescents (ages 11-19) using AI-based mental health technologies. Included studies underwent rigorous quality appraisal (Cochrane RoB 2.0, ROBINS-I, CASP). Additional insights were gathered through stakeholder interviews (clinicians, AI developers, adolescent users). Result(s): Twenty-seven studies met inclusion criteria, investigating interventions such as AI chatbots, predictive analytics, mobile apps, and virtual environments targeting anxiety and depression. Key opportunities identified include enhanced accessibility for underserved populations, personalization through adaptive algorithms, proactive early-risk detection, scalability, cost-efficiency, and improved engagement via interactive interfaces. Significant implementation barriers encompassed technical infrastructure limitations, data security concerns, insufficient longitudinal efficacy data, socioeconomic disparities, and clinician scepticism. Ethical challenges emphasized informed consent, algorithm transparency, potential biases, unclear accountability, and clinician deskilling risks. Conclusion(s): AI-driven digital interventions offer substantial promise for augmenting adolescent mental health services in the UK. However, realizing their full potential necessitates addressing infrastructural, ethical, and evidentiary challenges through robust governance frameworks and continued rigorous research.
DOI: 10.2139/ssrn.5253224
Link to full-text [open access - no password required]
