Barriers and Facilitators to the Use of Large Language Model-Based Conversational Agents in Mental Healthcare: A Systematic Review
Open accessAbstract
(1) Background/Objectives: Over one billion individuals globally live with mental health conditions, yet the treatment gap exceeds 75% in low- and middle-income countries. Large language model (LLM)-based conversational agents have emerged as a potentially scalable solution, though the evidence base remains nascent and largely pre-clinical. This review synthesises barriers and facilitators to their implementation in mental healthcare using the Consolidated Framework for Implementation Research (CFIR). (2) Methods: Eight databases were searched from January 2022 to January 2026. Study selection was managed using Covidence. Two reviewers independently screened, extracted, and appraised studies using the Mixed Methods Appraisal Tool. Directed content analysis guided by CFIR was used for synthesis. (3) Results: Twenty-seven studies (three RCTs, nine mixed methods, eight qualitative, four cross-sectional, three observational) comprising >22,000 participants across 12 countries met inclusion criteria. Five barrier domains (27 sub-themes) and four facilitator domains (22 sub-themes) were identified. Inadequate crisis detection (reported in 21/27 studies) and 24/7 availability (reported in 26/27 studies) are the most frequently reported barriers and facilitators, respectively. These figures represent study-level reporting frequencies, not population-level prevalence estimates. CFIR mapping revealed universal coverage for Knowledge and Beliefs (100%) and Patient Needs and Resources (96%) but critical gaps in the Process domain (Evaluating: 7%; Champions: 11%). (4) Conclusions: LLM-based conversational agents demonstrate substantial promise but present critical safety deficiencies. A tiered implementation framework, independent safety certification, and equity-sensitive design are recommended.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The present study provided preliminary theoretical support for applying the TAM to the acceptance of AR-enabled mental health interventions among undergraduate students with SAD.
“Therefore, the present study provided preliminary theoretical support for applying the TAM to the acceptance of AR-enabled mental health interventions among undergraduate students with SAD [92].”
From Advancing Undergraduate Student Mental Healthcare of Social Anxiety Disorder: Evaluating the Acceptance of AR-Assisted Cognitive Behavioral Therapy Through TAM-Based Constructs. (2026) · cites this paper as [92]
The Study Characteristics and Conclusions sections describe a review of LLM-based conversational agents in mental healthcare, not a study of AR-enabled interventions among undergraduate students with SAD. Searches for TAM, augmented reality, and that student population found no passage in our copy addressing the claimed theoretical support; the phrase “present study” appears to refer to the citing authors’ own study. (LLM, conf=0.76)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
The PRISMA 2020 statement: An updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., et al. (2021) · BMJ · doi:10.1136/bmj.n71 · Cited by 108,283 · see citing works
Three Approaches to Qualitative Content Analysis
Hsieh, H.-F., & Shannon, S. E. (2005) · Qualitative Health Research · doi:10.1177/1049732305276687 · Cited by 44,601 · cited 1× · see citing works
“Barriers and facilitators were coded using directed content analysis [25] guided by the CFIR 2022 update [19], with reference to the original 2009 taxonomy [18] where construct labels have been retained for consistency with prior DMHI literature.”
Damschroder, L. J., Aron, D. C., Keith, R. E., Kirsh, S. R., Alexander, J. A., & Lowery, J. C. (2009) · Implementation Science · doi:10.1186/1748-5908-4-50 · Cited by 14,496 · cited 3× · see citing works
“For the purposes of this review, implementation determinants are defined as factors that influence whether and how LLM-based CAs are adopted into routine mental healthcare practice, consistent with the CFIR definition [18,19].”
Santomauro, D. F., Mantilla Herrera, A. M., Shadid, J., Zheng, P., Ashbaugh, C., Pigott, D. M., Abbafati, C., Adolph, C., Amlag, J. O., Aravkin, A. Y., et al. (2021) · The Lancet · doi:10.1016/s0140-6736(21)02143-7 · Cited by 5,444 · see citing works
Synthesis without meta-analysis (SWiM) in systematic reviews: reporting guideline
Campbell, M., McKenzie, J. E., Sowden, A., Katikireddi, S. V., Brennan, S. E., Ellis, S., Hartmann-Boyce, J., Ryan, R., Shepperd, S., Thomas, J., et al. (2020) · BMJ · doi:10.1136/bmj.l6890 · Cited by 4,628 · cited 1× · see citing works
“Given study heterogeneity, narrative synthesis was conducted following Synthesis Without Meta-analysis (SWiM) guidelines [24].”
The updated consolidated framework for implementation research based on user feedback
Damschroder, L. J., Reardon, C. M., Widerquist, M. A. O., & Lowery, J. (2022) · Implementation Science · doi:10.1186/s13012-022-01245-0 · Cited by 4,452 · see citing works
The Lancet Commission on global mental health and sustainable development
Patel, V., Saxena, S., Lund, C., Thornicroft, G., Baingana, F., Bolton, P., Chisholm, D., Collins, P. Y., Cooper, J. L., Eaton, J., et al. (2018) · The Lancet · doi:10.1016/s0140-6736(18)31612-x · Cited by 3,692 · cited 1× · see citing works
“This disparity is driven by severe workforce shortages (a global median of 13 mental health workers per 100,000 population), prohibitive costs, geographical inaccessibility, and pervasive stigma [2,5].”
Fitzpatrick, K. K., Darcy, A., & Vierhile, M. (2017) · JMIR Mental Health · doi:10.2196/mental.7785 · Cited by 2,730 · see citing works
Llama 2: Open Foundation and Fine-Tuned Chat Models
Touvron, H., Martin, L., Stone, K., Albert, P., Almahairi, A., Babaei, Y., Bashlykov, N., Batra, S., Bhargava, P., Bhosale, S., et al. (2023) · arXiv · doi:10.48550/arxiv.2307.09288 · Cited by 2,604 · see citing works
Undertreatment of people with major depressive disorder in 21 countries
Thornicroft, G., Chatterji, S., Evans-Lacko, S., Gruber, M., Sampson, N., Aguilar-Gaxiola, S., Al-Hamzawi, A., Alonso, J., Andrade, L., Borges, G., et al. (2017) · British Journal of Psychiatry · doi:10.1192/bjp.bp.116.188078 · Cited by 1,121 · cited 1× · see citing works
“Despite this burden, the treatment gap remains vast: over 75% of individuals with severe mental disorders in low-and middle-income countries (LMICs) receive no treatment, and even in high-income settings the gap exceeds 50% [3,4].”
Inkster, B., Sarda, S., & Subramanian, V. (2018) · doi:10.2196/preprints.12106 · Cited by 1,063 · cited 1× · see citing works
“First-generation chatbots such as Woebot and Wysa, built on rule-based decision trees, demonstrated modest efficacy in delivering structured cognitive behavioural therapy (CBT) exercises but were constrained by inflexible dialogue [7,8].”
Loneliness and suicide mitigation for students using GPT3-enabled chatbots
Maples, B., Cerit, M., Vishwanath, A., & Pea, R. (2024) · npj Mental Health Research · doi:10.1038/s44184-023-00047-6 · Cited by 250 · see citing works
To chat or bot to chat: Ethical issues with using chatbots in mental health
Coghlan, S., Leins, K., Sheldrick, S., Cheong, M., Gooding, P., & D’Alfonso, S. (2023) · Digital Health · doi:10.1177/20552076231183542 · Cited by 247 · cited 2× · see citing works
“Coghlan et al. identified 15 categories of ethical violations committed by LLM counsellors, from inappropriate crisis navigation to reinforcement of negative self-beliefs [17].”
Influential“It happened to be the perfect thing”: experiences of generative AI chatbots for mental health
Siddals, S., Torous, J., & Coxon, A. (2024) · npj Mental Health Research · doi:10.1038/s44184-024-00097-4 · Cited by 241 · see citing works
Assessing the Effectiveness of ChatGPT in Delivering Mental Health Support: A Qualitative Study
Alanezi, F. (2024) · Journal of Multidisciplinary Healthcare · doi:10.2147/jmdh.s447368 · Cited by 139 · see citing works
MindfulDiary: Harnessing Large Language Model to Support Psychiatric Patients' Journaling
Kim, T., Bae, S., Kim, H. A., Lee, S. H., Hong, H., Yang, C.-M., & Kim, Y. (2024) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3642937 · Cited by 123 · see citing works
Sharma, A., Rushton, K., Lin, I. W., Nguyen, T., & Althoff, T. (2024) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3642761 · Cited by 119 · see citing works
Hua, Y., Siddals, S., Ma, Z., Galatzer‐Levy, I., Xia, W., Hau, C., Na, H., Flathers, M., Linardon, J., Ayubcha, C., et al. (2025) · World Psychiatry · doi:10.1002/wps.21352 · Cited by 117 · cited 5× · see citing works
“A systematic review of 160 mental health chatbot studies confirmed that LLM-based systems surged to 45% of all new studies in 2024, overtaking rule-based approaches [11].”
InfluentialBlease, C., Worthen, A., & Torous, J. (2024) · Psychiatry Research · doi:10.1016/j.psychres.2024.115724 · Cited by 113 · see citing works
Performance of mental health chatbot agents in detecting and managing suicidal ideation
Pichowicz, W., Kotas, M., & Piotrowski, P. (2025) · Scientific Reports · doi:10.1038/s41598-025-17242-4 · Cited by 85 · cited 3× · see citing works
“Pichowicz et al. [13] corroborated these findings with standardised testing.”
Large Language Models as Mental Health Resources: Patterns of Use in the United States
Rousmaniere, T., Zhang, Y., Li, X., & Shah, S. (2025) · doi:10.31234/osf.io/q8m7g_v1 · Cited by 73 · see citing works
Scholich, T., Barr, M., Wiltsey Stirman, S., & Raj, S. (2025) · JMIR Mental Health · doi:10.2196/69709 · Cited by 72 · see citing works
Eisner, E., Faulkner, S., Allan, S., Ball, H., Basilio, D. D., Nicholas, J., Priyam, A., Wilson, P. M., Zhang, X., & Bucci, S. (2025) · JMIR Mental Health · doi:10.2196/65246 · Cited by 65 · cited 5× · see citing works
“Prior CFIR applications to digital mental health interventions (Bucci et al. [20]; Stiles-Shields et al. [21]) addressed rule-based systems and specific clinical populations without addressing LLM-specific challenges, including hallucination, crisis detection failure, and emotional dependency.”
InfluentialLee, H. S., Wright, C., Ferranto, J., Buttimer, J., Palmer, C. E., Welchman, A., Mazor, K. M., Fisher, K. A., Smelson, D., O’Connor, L., et al. (2025) · Frontiers in Psychiatry · doi:10.3389/fpsyt.2024.1505024 · Cited by 55 · see citing works
Hipgrave, L., Goldie, J., Dennis, S. J., & Coleman, A. (2025) · Frontiers in Digital Health · doi:10.3389/fdgth.2025.1606291 · Cited by 54 · see citing works
An Examination of Generative AI Response to Suicide Inquires: Content Analysis
Campbell, L. O., Babb, K., Lambie, G. W., & Hayes, B. (2025) · JMIR Mental Health · doi:10.2196/73623 · Cited by 48 · cited 2× · see citing works
“Pichowicz et al. [13] [14] confirmed that AI suicide responses improved over time but critical gaps persisted at higher risk levels.”
Evaluating Generative AI Psychotherapy Chatbots Used by Youth: Cross-Sectional Study
Sobowale, K., Humphrey, D. K., & Zhao, S. Y. (2025) · JMIR Mental Health · doi:10.2196/79838 · Cited by 40 · cited 1× · see citing works
“This framework aligns conceptually with Hua et al. [11], California SB 243 [15], and the EU AI Act’s high-risk classification [38,60,61], but has not been empirically validated and should be tested through implementation trials before adoption in clinical guidelines.”
“It’s Not Only Attention We Need”: Systematic Review of Large Language Models in Mental Health Care
Bucher, A. M., Egger, S., Vashkite, I., Wu, W., & Schwabe, G. (2025) · JMIR Mental Health · doi:10.2196/78410 · Cited by 37 · see citing works
Hasei, J., Hanzawa, M., Nagano, A., Maeda, N., Yoshida, S., Endo, M., Yokoyama, N., Ochi, M., Ishida, H., Katayama, H., et al. (2025) · Frontiers in Digital Health · doi:10.3389/fdgth.2025.1543543 · Cited by 34 · see citing works
Governing AI in Mental Health: 50-State Legislative Review
Shumate, J. N., Rozenblit, E., Flathers, M., Larrauri, C. A., Hau, C., Xia, W., Torous, E. N., & Torous, J. (2025) · JMIR Mental Health · doi:10.2196/80739 · Cited by 33 · cited 2× · see citing works
“California’s SB 243 mandates crisis prevention protocols, transparency disclosures, and annual reporting for companion chatbots effective January 2026 [15], and 11 US states had enacted 20 relevant laws by May 2025 [16].”
Boit, S., & Patil, R. (2025) · JMIR Mental Health · doi:10.2196/75078 · Cited by 24 · see citing works
Ma, Y., Zeng, Y., Liu, T., Sun, R., Xiao, M., & Wang, J. (2024) · Frontiers in Public Health · doi:10.3389/fpubh.2024.1475867 · Cited by 22 · see citing works
Schäfer, L. M., Köhler, S., & Krause, T. (2025) · Frontiers in Digital Health · doi:10.3389/fdgth.2025.1576135 · Cited by 22 · see citing works
AI-Enhanced Virtual Reality Self-Talk for Psychological Counseling: Formative Qualitative Study
Zisquit, ., Shoa, A., Oliva, R., Perry, S., Spanlang, B., Klomek, A. B., Slater, M., & Friedman, D. A. (2025) · JMIR Formative Research · doi:10.2196/67782 · Cited by 18 · see citing works
Zhao, Y., Qian, W., Chen, Y.-R., Wu, D., Luo, Y., Gao, C. F., Wu, K., & Liu, Z. (2025) · Applied Psychology Health and Well-Being · doi:10.1111/aphw.70067 · Cited by 17 · see citing works
Strudwick, G., Kassam, I., Torous, J. B., & Patenaude, S. (2025) · JMIR Mental Health · doi:10.2196/78791 · Cited by 14 · see citing works
Li, Z., Zhu, Z., Gui, X., & Luo, Y. (2025) · International Journal of Human-Computer Studies · doi:10.1016/j.ijhcs.2025.103555 · Cited by 13 · see citing works
Vecchione, B., & Singh, R. (2025) · Big Data & Society · doi:10.1177/20539517251383884 · Cited by 10 · see citing works
Zhang, X., Zayed, A. M., Hamrin, J. R., Özgür, A. G., & Kuoppamäki, S. (2025) · Journal of Medical Internet Research · doi:10.2196/78829 · Cited by 7 · see citing works
Tan, K. S., Cervin, M., Leman, P. J., Nielsen, K., Kumar, P. V., & Medvedev, O. N. (2025) · Journal of Psychology and AI · doi:10.1080/29974100.2025.2545258 · Cited by 7 · cited 1× · see citing works
“Tan et al. [53] found that LLMs struggled with integrative clinical reasoning across seven models.”
AI-Facilitated Cognitive Reappraisal via Socrates 2.0: Mixed Methods Feasibility Study.
Held, P., Pridgen, S. A., Szoke, D. R., Chen, Y., Akhtar, Z., & Amin, D. (2025) · JMIR Mental Health · doi:10.2196/80461 · Cited by 7 · see citing works
Stiles‐Shields, C., Gustafson, E. L., Lim, P. S., Bobadilla, G., Thorpe, D., Williams, F. S., Donenberg, G. R., Julion, W. A., & Karnik, N. S. (2024) · Journal of Pediatric Psychology · doi:10.1093/jpepsy/jsae058 · Cited by 7 · see citing works
Wang, Y., Wang, Y., Ye, X., et al. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2501.15599 · Cited by 6 · see citing works
Wang, L., Carrington, D., Carrington, D., et al. (2025) · PubMed · doi:10.48550/arxiv.2506.11376 · Cited by 2 · see citing works
World Health Organization (2025). World Mental Health Today, WHO.
(n.d.)
UnresolvedWorld Health Organization (2025). Mental Health Atlas 2024, WHO.
(n.d.)
UnresolvedWorld Health Organization (2023). Mental Health Gap Action Programme (mhGAP), WHO.
(n.d.)
UnresolvedLanguage Models Are Few-Shot Learners
Brown (2020) · Adv. Neural Inf. Process. Syst.
UnresolvedTouvron, H., Martin, L., Stone, K., Albert, P., Almahairi, A., Babaei, Y., Bashlykov, N., Batra, S., Bhargava, P., and Bhosale, S. (2023). Llama 2: Open Foundation and Fine-Tuned Chat Models. arXiv.
(n.d.)
UnresolvedCalifornia Legislature (2025). Senate Bill 243: Companion Chatbots, Chapter 677.
(n.d.)
UnresolvedHong, Q.N., Pluye, P., Fabregues, S., Bartlett, G., Boardman, F., Cargo, M., Dagenais, P., Gagnon, M.-P., Griffiths, F., and Nicolau, B. (2018). Mixed Methods Appraisal Tool (MMAT), Version 2018.
(n.d.)
UnresolvedAn empathic GPT-based chatbot to talk about mental disorders with Spanish teenagers
(2025) · Int. J. Hum.–Comput. Interact.
UnresolvedChatGPT as therapy: A qualitative and network-based thematic profiling of shared experiences, attitudes, and beliefs on Reddit
Collins (2025) · J. Psychiatr. Res.
UnresolvedWang, Y., Wang, Y., Xiao, Y., Escamilla, L., Augustine, B., Crace, K., Zhou, G., and Zhang, Y. (2025). Evaluating an LLM-powered chatbot for cognitive restructuring: Insights from mental health professionals. arXiv.
(n.d.)
UnresolvedExamining the efficacy of large language models for mitigating depression and anxiety among Chinese students: A randomized controlled trial
Ye (2025) · Comput. Inform. Nurs.
UnresolvedWang, J., Sheng, Y., He, Q., Zhou, W., Zhang, Y., Fung, H.H., Chen, A.T., and Liang, Y. (2025). LLM-Powered Conversational Agent Delivering Problem-Solving Therapy for Caregivers. arXiv.
(n.d.)
UnresolvedRandomized trial of a generative AI chatbot for mental health treatment
Heinz (2025) · NEJM AI
UnresolvedSafety of large language models in addressing depression
Heston (2023) · Cureus
UnresolvedUnderstanding the benefits and challenges of using large language model-based conversational agents for mental well-being support
Ma (2024) · AMIA Annual Symposium Proceedings
UnresolvedSpring Health (2026, February 15). VERA-MH: A Framework for Evaluating AI Chatbot Safety in Mental Health Contexts. Available online: https://springhealth.com/vera-mh.
(n.d.)
UnresolvedShah, A.B., Green, F.M., Boiani, J.A., and Chung, E.T. (2026, February 15). Novel AI Laws Target Companion AI and Mental Health, National Law Review, Available online: https://natlawreview.com/article/novel-ai-laws-target-companion-ai-and-mental-health.
(n.d.)
UnresolvedFuture of Privacy Forum (2025). Understanding the New Wave of Chatbot Legislation: California SB 243 and Beyond, Future of Privacy Forum.
(n.d.)
UnresolvedEuropean Parliament (2024). Regulation (EU) 2024/1689 Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act). Off. J. Eur. Union, 1689.
(n.d.)
UnresolvedOpenAI (2025). GPT-5 Model Release Notes: Reducing Sycophancy and Encouraging Human Support, OpenAI.
(n.d.)
Unresolved
Flagged first, then by citations
20 of the 61 references couldn’t be matched to a known paper.
Cited by
Sorted by citations
Zhou, Z., Zhou, Y., Ouyang, B., Man, S. S., & Chan, A. H. S. (2026) · Healthcare · doi:10.3390/healthcare14131978
“Therefore, the present study provided preliminary theoretical support for applying the TAM to the acceptance of AR-enabled mental health interventions among undergraduate students with SAD [92].”
Yurkiv, H., Booth, R. G., Garnett, A., & McIntyre, A. (2026) · doi:10.2196/preprints.105389
Mental Health Innovation Is Scaling Faster Than Accountability
Zavaleta‐Monestel, E., Herrera-Jiménez, L. G., & González, R. M. (2026) · Cureus · doi:10.7759/cureus.113012
“These statutes are notable less for their jurisdiction than for their mechanism: they attach enforceable duties to function and foreseeable use rather than to medical claims, which is exactly the move that deployment-level governance of general-purpose AI requires [24].”
Triantafyllidis, A. K., Segkouli, S., Lithoxoidou, E. E., Alexiadis, A., Votis, K., Koutra, K., Kilintzis, V., Kondylakis, H., Arana‐Arri, E., Haug, S., et al. (2026) · Information · doi:10.3390/info17080770
Alsalman, Z. H., Almaqhawi, A., ALMULHIM, A., Alqadeep, A. F., Almutawah, G. M., Khalifa, F. A. A., Alghadeer, F. R., Aleid, H. A., Alkhalifah, W. M., Alabdullah, Z. A., et al. (2026) · Frontiers in Public Health · doi:10.3389/fpubh.2026.1926484
Xiao, Y., Su, P., Yujun, L., Zhou, P., & Jin, L. (2026) · Research Square · doi:10.21203/rs.3.rs-10741905/v1
- Authors
- Shankar, R., Lim, A., & Xu, Q.
- Affiliations
- Tan Tock Seng Hospital (SG)
- National University of Singapore (SG)
- University of Bristol (GB)
- Corresponding author
- Shankar, Ravi
- Venue
- Healthcare
- Publisher
- MDPI AG
- Published
- 2026-05-07
- Volume
- 14
- Issue
- 10
- Pages
- 1267
- ISSN
- 2227-9032 (ISSN-L)
- OpenAlex
- W7160523257
- Cited by
- 9
- References
- 61
- Citations by year
- 2026: 9
- FWCI
- 11.06
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Healthcare (publishedVersion, cc-by)
- PubMed (publishedVersion)
- PubMed Central (publishedVersion, cc-by)
- Explore Bristol Research (publishedVersion, cc-by)
- Bristol Research (University of Bristol) (submittedVersion, cc-by)
- Topics
- Social Sciences › Psychology › Applied Psychology › Digital Mental Health Interventions
- Health Sciences › Health Professions › General Health Professions › Health Policy Implementation Science
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Keywords
- large language models, chatbots, mental health services, implementation barriers, implementation facilitators, Consolidated Framework for Implementation Research, crisis detection, 24/7 availability, AI safety, low- and middle-income countries, treatment gap, implementation framework
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
