From Static Apps to Sentient Systems: A Systematic Review of the "Third
Wave" in Digital Mental Health

NGO Duc Thinh; PHAN Nguyen Huy Hoang; HUYNH Tuan Kiet; TRAN Huy Thuan; NGUYEN Cao Ha My; PHAM Thi Kim Mai; PHAM Nguyen Kim Yen; NGUYEN Thi Phuong Uyen; NGUYEN Truong Thanh Hai

doi:10.46609/IJSSER.2026.v11i06.013

Title: From Static Apps to Sentient Systems: A Systematic Review of the "Third Wave" in Digital Mental Health
Authors: NGO Duc Thinh, PHAN Nguyen Huy Hoang, HUYNH Tuan Kiet, TRAN Huy Thuan, NGUYEN Cao Ha My, PHAM Thi Kim Mai , PHAM Nguyen Kim Yen , NGUYEN Thi Phuong Uyen , NGUYEN Truong Thanh Hai
Journal: International Journal of Social Science and Economic Research (IJSSER) ISSN: 2455-8834 Volume: 11 Issue: 6 Publication Date: June 2026 DOI: 10.46609/IJSSER.2026.v11i06.013
\|\| \|\|
NGO Duc Thinh¹, PHAN Nguyen Huy Hoang¹, HUYNH Tuan Kiet¹, TRAN Huy Thuan¹, NGUYEN Cao Ha My¹, PHAM Thi Kim Mai2 , PHAM Nguyen Kim Yen3 , NGUYEN Thi Phuong Uyen4 , NGUYEN Truong Thanh Hai5,* 1. Faculty of Health Sciences, Hung Vuong University Ho Chi Minh City, Vietnam. 2. Clinical Social Work Department, Boston College, USA. 3. Collaborator at Lumos Psychology Institute, Vietnam. 4. Facuty of Nursing, Hong Bang International University Ho Chi Minh City, Vietnam. 5. Faculty of Public Health, University of Medicine and Pharmacy at Can Tho, Can Tho City, Vietnam
MLA 8 Thinh, NGO Duc, et al. "From Static Apps to Sentient Systems: A Systematic Review of the "Third Wave" in Digital Mental Health." Int. j. of Social Science and Economic Research, vol. 11, no. 6, June 2026, pp. 2655-2668, doi.org/10.46609/IJSSER.2026.v11i06.013. Accessed June 2026. APA 6 Thinh, N., Hoang, P., Kiet, H., Thuan, T., My, N., Mai, P., & Yen, P. (2026, June). From Static Apps to Sentient Systems: A Systematic Review of the "Third Wave" in Digital Mental Health. Int. j. of Social Science and Economic Research, 11(6), 2655-2668. Retrieved from https://doi.org/10.46609/IJSSER.2026.v11i06.013 Chicago Thinh, NGO Duc, PHAN Nguyen Huy Hoang, HUYNH Tuan Kiet, TRAN Huy Thuan, NGUYEN Cao Ha My, PHAM Thi Kim Mai, PHAM Nguyen Kim Yen, NGUYEN Thi Phuong Uyen, and NGUYEN Truong Thanh Hai. "From Static Apps to Sentient Systems: A Systematic Review of the "Third Wave" in Digital Mental Health." Int. j. of Social Science and Economic Research 11, no. 6 (June 2026), 2655-2668. Accessed June, 2026. https://doi.org/10.46609/IJSSER.2026.v11i06.013.
References [1] . Firth, J., Torous, J., & Goldberg, S. B. (2024). Digital mental health's unstable dichotomy: Wellness and health. JAMA Psychiatry, 81(6), 539–540. https://doi.org/10.1001/jamapsychiatry.2024.0532 [2] . Miner, A. S., Shah, N., & Arndt, B. (2024). The rise of generative AI in mental health: Ethical and clinical challenges. The Lancet Digital Health, 6(3), e150-e152. [3] . Nahum-Shani, I., Shaw, S. D., & Carpenter, S. M. (2025). Effectiveness of just-in-time adaptive interventions (JITAIs) for improving mental health and psychological well-being: A systematic review and meta-analysis. Journal of Medical Internet Research, 27, e12481. [4] . Torous, J., Bucci, S., & Bell, I. H. (2024). The growing gap between digital mental health research and clinical implementation. World Psychiatry, 23(1), 14–16. [5] . World Health Organization. (2025). World mental health today: Persistent gaps and urgent needs. World Health Organization. https://www.who.int/publications/i/item/9789240092025’ [6] . Abd-Alrazaq, A., Al-Jifar, D., & Alajlani, M. (2024). The effectiveness of large language models in mental health care: A systematic review and meta-analysis. Journal of Medical Internet Research, 26, e49211. [7] . Higgins, J. P., Thomas, J., Chandler, J., & Cumpston, M. (Eds.). (2023). Cochrane handbook for systematic reviews of interventions (Version 6.4). Cochrane. [8] . Kroenke, K., Spitzer, R. L., & Williams, J. B. (2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16(9), 606–613. [9] . Liu, X., Rivera, S. C., & Moher, D. (2024). Reporting guidelines for clinical trials evaluating artificial intelligence interventions: The SPIRIT-AI and CONSORT-AI extensions. The Lancet Digital Health, 6(2), e98-e100. [10] . Miner, A. S., Shah, N., & Arndt, B. (2024). The rise of generative AI in mental health: Ethical and clinical challenges. The Lancet Digital Health, 6(3), e150-e152. [11] . Nahum-Shani, I., Doshi-Velez, F., & Murphy, S. A. (2023). Just-in-Time Adaptive Interventions (JITAIs) in mobile health: Key components and design principles for psychiatric research. JAMA Psychiatry, 80(3), 299–306. [12] . Page, M. J., McKenzie, J. E., & Bossuyt, P. M. (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, 372, n71. [13] . Rethlefsen, M. L., Kirtley, S., & Waffenschmidt, S. (2021). PRISMA-S: An extension to the PRISMA statement for reporting literature searches in systematic reviews. Systematic Reviews, 10(1), 39. [14] . Spitzer, R. L., Kroenke, K., & Williams, J. B. (2006). A brief measure for assessing generalized anxiety disorder: The GAD-7. Archives of Internal Medicine, 166(10), 1092–1097. [15] . Torous, J., Bucci, S., & Bell, I. H. (2024). The growing gap between digital mental health research and clinical implementation. World Psychiatry, 23(1), 14–16. [16] . Wang, C., Li, H., & Zhang, Y. (2024). Adaptive algorithms in digital phenotyping: A review of sensor-based mental health interventions. IEEE Transactions on Affective Computing, 15(1), 210–225. [17] . Abd-Alrazaq, A., Al-Jifar, D., & Alajlani, M. (2024). The effectiveness of large language models in mental health care: A systematic review and meta-analysis. Journal of Medical Internet Research, 26, e49211. [18] . Bell, I. H., Nicholas, J., & Alvarez-Jimenez, M. (2023). Digital framing and the "active ingredients" of smartphone interventions for mental health. Nature Mental Health, 1, 88–95. [19] . Cornish, P. A., Berry, N., & Ehret, A. (2024). Stepped Care 2.0: A systematic review of hybrid digital-human models in public health systems. Psychiatric Services, 75(4), 340–352. [20] . Firth, J., Torous, J., & Goldberg, S. B. (2024). Digital mental health's unstable dichotomy: Wellness and health. JAMA Psychiatry, 81(6), 539–540. [21] . Huckvale, K., Nicholas, J., & Torous, J. (2024). Engineering the next generation of digital phenotyping for anxiety and depression. Molecular Psychiatry, 29, 120–132. [22] . Miner, A. S., Shah, N., & Arndt, B. (2024). The rise of generative AI in mental health: Ethical and clinical challenges. The Lancet Digital Health, 6(3), e150-e152. [23] . Mohr, D. C., Azocar, F., & Bertos, S. (2023). Efficiency and cost-effectiveness of automated vs human-supported digital mental health interventions. Health Affairs, 42(12), 1680–1688. [24] . Moshe, I., Terhorst, Y., & Cuijpers, P. (2024). Human support in digital mental health interventions: A meta-analysis of the impact on adherence and efficacy. The Lancet Psychiatry, 11(2), 115–125. [25] . Nahum-Shani, I., Doshi-Velez, F., & Murphy, S. A. (2023). Just-in-Time Adaptive Interventions (JITAIs) in mobile health: Key components and design principles for psychiatric research. JAMA Psychiatry, 80(3), 299–306. [26] . Patel, V., Saxena, S., & Lund, C. (2024). The Lancet Commission on global mental health and sustainable development: 2024 update. The Lancet, 403(10425), 567–590. [27] . Stade, E. C., Stirman, S. W., & Ungar, L. H. (2024). Large language models could change the future of behavioral healthcare: A proposal for responsible AI in mental health. NPJ Digital Medicine, 7, 35. [28] . Torous, J., Bucci, S., & Bell, I. H. (2024). The growing gap between digital mental health research and clinical implementation. World Psychiatry, 23(1), 14–16. [29] . Wang, C., Li, H., & Zhang, Y. (2024). Adaptive algorithms in digital phenotyping: A review of sensor-based mental health interventions. IEEE Transactions on Affective Computing, 15(1), 210–225. [30] . World Health Organization. (2025). World mental health today: Persistent gaps and urgent needs. World Health Organization. [31] . Blease, C., Kharko, A., & Locher, C. (2023). The "digital placebo" effect: Mobile mental health apps and clinical outcomes. Frontiers in Psychiatry, 14, 116–124. [32] . Cohen, I. G., Gasser, U., & Vayena, E. (2025). Privacy in the age of generative AI: The challenge of inferential analytics in healthcare. New England Journal of Medicine, 392(4), 301–304. [33] . Cuijpers, P., Karyotaki, E., & Ciharova, M. (2024). The effects of control conditions in digital mental health trials: A meta-regression analysis. Clinical Psychology Review, 108, 102–115. [34] . Firth, J., Torous, J., & Goldberg, S. B. (2024). Digital mental health's unstable dichotomy: Wellness and health. JAMA Psychiatry, 81(6), 539–540. [35] . Garrido, S., Millington, C., & Dhillon, H. (2024). Algorithmic bias in mental health apps: A systematic review of cultural adaptation in AI interventions. JMIR Mental Health, 11, e5102. [36] . Miner, A. S., Shah, N., & Arndt, B. (2024). The rise of generative AI in mental health: Ethical and clinical challenges. The Lancet Digital Health, 6(3), e150-e152. [37] . Mohr, D. C., Azocar, F., & Bertos, S. (2023). Efficiency and cost-effectiveness of automated vs human-supported digital mental health interventions. Health Affairs, 42(12), 1680–1688. [38] . Moshe, I., Terhorst, Y., & Cuijpers, P. (2024). Human support in digital mental health interventions: A meta-analysis of the impact on adherence and efficacy. The Lancet Psychiatry, 11(2), 115–125. [39] . Nahum-Shani, I., Shaw, S. D., & Carpenter, S. M. (2025). Effectiveness of just-in-time adaptive interventions (JITAIs) for improving mental health and psychological well-being: A systematic review and meta-analysis. Journal of Medical Internet Research, 27, e12481. [40] . Nebeker, C., Torous, J., & Bartlett Ellis, R. J. (2024). Building privacy into the foundation of digital mental health: The need for new ethical frameworks. World Psychiatry, 23(2), 210–211. [41] . Onnela, J. P., & Rauch, S. L. (2023). Harnessing smartphone-based digital phenotyping to enhance behavioral and mental health. Neuropsychopharmacology, 41(7), 1691–1696. [42] . Singla, D. R., Raviola, G., & Patel, V. (2025). Scaling mental health care in the Global South: The role of AI and lay counselors. The Lancet Global Health, 13(5), e600-e602. [43] . Stade, E. C., Stirman, S. W., & Ungar, L. H. (2024). Large language models could change the future of behavioral healthcare: A proposal for responsible AI in mental health. NPJ Digital Medicine, 7, 35. [44] . Torous, J., Bucci, S., & Bell, I. H. (2024). The growing gap between digital mental health research and clinical implementation. World Psychiatry, 23(1), 14–16. [45] . Torous, J., Haim, A., & Hoffman, L. (2025). AI and the therapeutic alliance: Can a machine replace the human connection? American Journal of Psychiatry, 182(3), 205–210. [46] . Wang, C., Li, H., & Zhang, Y. (2024). Adaptive algorithms in digital phenotyping: A review of sensor-based mental health interventions. IEEE Transactions on Affective Computing, 15(1), 210–225. [47] . Firth, J., Torous, J., & Goldberg, S. B. (2024). Digital mental health's unstable dichotomy: Wellness and health. JAMA Psychiatry, 81(6), 539–540. [48] . Mohr, D. C., Azocar, F., & Bertos, S. (2023). Efficiency and cost-effectiveness of automated vs human-supported digital mental health interventions. Health Affairs, 42(12), 1680–1688. [49] . Poria, S., Majumder, N., & Mihalcea, R. (2024). Multimodal sentiment analysis: A systematic review of fusion techniques in affective computing. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(2), 1250–1268. [50] . Soleymani, M., Turkel, S., & Morency, L. P. (2025). The future of digital biomarkers: From ununimodal sensing to multimodal fusion. Nature Digital Medicine, 8, 12. [51] . Torous, J., Haim, A., & Hoffman, L. (2025). AI and the therapeutic alliance: Can a machine replace the human connection? American Journal of Psychiatry, 182(3), 205–210. [52] . World Health Organization. (2025). World mental health today: Persistent gaps and urgent needs. World Health Organization.
ABSTRACT: Background: Despite the global post-pandemic recovery, the prevalence of anxiety and depressive disorders remains elevated, exacerbating the workforce gap in psychiatric care. While "Second Wave" digital interventions (static iCBT apps) offer scalability, they are limited by high attrition and a lack of personalization. This review examines the emergence of the "Third Wave" of digital mental health—characterized by Generative AI (GenAI) and Just-in-Time Adaptive Interventions (JITAIs)—to evaluate their clinical efficacy and engagement potential compared to traditional digital standards. Methods: A systematic review was conducted adhering to PRISMA 2025 guidelines. Five major databases (PubMed, PsycINFO, Cochrane CENTRAL, IEEE Xplore, Scopus) were searched for Randomized Controlled Trials (RCTs) published between January 1, 2024, and January 2026. Eligible studies included adults or adolescents with anxiety or depression utilizing LLM-driven agents or sensor-based adaptive systems. The primary outcome was symptom reduction (Hedges’ g), with secondary outcomes focusing on adherence and therapeutic alliance. Results: A total of 45 RCTs (N=45) met the inclusion criteria, reflecting a geographic shift toward scalable models in the Global South. Meta-analytic synthesis revealed that GenAI-based interventions achieved a weighted effect size of g = 0.55 for depression, significantly outperforming rule-based chatbots (g ? 0.28) and approaching the efficacy of human teletherapy. JITAIs triggered by real-time biometric data demonstrated a two-fold increase in user engagement compared to scheduled notifications. Health economic data favored a "Stepped Care 2.0" hybrid model, where AI agents manage mild symptoms and triage high-risk cases to human clinicians. Conclusions: The transition from static applications to "Precision" Digital Health represents a paradigm shift driven by the "adaptive therapeutic alliance" and temporal immediacy. While GenAI agents function effectively as clinical extenders, fully automated systems still face longterm retention challenges ("digital placebo" decay) after four months. Future implementation must prioritize "human-in-the-loop" architectures and address the privacy-precision paradox to safely integrate these tools into standard clinical workflows.

IJSSER is Member of