Student or educator? Sign up with your school email and get 6 months of Pro free.

The DOI matches its Crossref record; 102 of the 109 references it cites were found in a registry or an index; 232 of the 321 sources its claims cite were checked for support.

Examining the role of AI technology in online mental healthcare: opportunities, challenges, and implications, a mixed-methods review

Open access
Gutierrez, Gilmar, Stephenson, Callum, Eadie, Jazmin, Asadpour, Kimia, Alavi, Nazanin·2024·Frontiers in Psychiatry

Abstract

Introduction: Online mental healthcare has gained significant attention due to its effectiveness, accessibility, and scalability in the management of mental health symptoms. Despite these advantages over traditional in-person formats, including higher availability and accessibility, issues with low treatment adherence and high dropout rates persist. Artificial intelligence (AI) technologies could help address these issues, through powerful predictive models, language analysis, and intelligent dialogue with users, however the study of these applications remains underexplored. The following mixed methods review aimed to supplement this gap by synthesizing the available evidence on the applications of AI in online mental healthcare. Method: We searched the following databases: MEDLINE, CINAHL, PsycINFO, EMBASE, and Cochrane. This review included peer-reviewed randomized controlled trials, observational studies, non-randomized experimental studies, and case studies that were selected using the PRISMA guidelines. Data regarding pre and post-intervention outcomes and AI applications were extracted and analyzed. A mixed-methods approach encompassing meta-analysis and network meta-analysis was used to analyze pre and post-intervention outcomes, including main effects, depression, anxiety, and study dropouts. We applied the Cochrane risk of bias tool and the Grading of Recommendations Assessment, Development and Evaluation (GRADE) to assess the quality of the evidence. Results: Twenty-nine studies were included revealing a variety of AI applications including triage, psychotherapy delivery, treatment monitoring, therapy engagement support, identification of effective therapy features, and prediction of treatment response, dropout, and adherence. AI-delivered self-guided interventions demonstrated medium to large effects on managing mental health symptoms, with dropout rates comparable to non-AI interventions. The quality of the data was low to very low. Discussion: The review supported the use of AI in enhancing treatment response, adherence, and improvements in online mental healthcare. Nevertheless, given the low quality of the available evidence, this study highlighted the need for additional robust and high-powered studies in this emerging field. Systematic review registration: https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=443575, identifier CRD42023443575. Copyright © 2024 Gutierrez, Stephenson, Eadie, Asadpour and Alavi.

Overview

CoChat Summary

## TL;DR

This mixed-methods review found that AI-augmented online mental healthcare interventions were generally promising for reducing mental health symptoms and supporting engagement, but the evidence base was small and low quality. Across 29 studies, AI applications ranged from triage and psychotherapy delivery to predicting dropout and adherence, with self-guided AI interventions often performing well.

## Key findings

  • The review included 29 studies and identified AI uses in triage, psychotherapy delivery, treatment monitoring, therapy engagement support, identification of effective therapy features, and prediction of treatment response, dropout, and adherence.
  • For AI-augmented interventions, the meta-analysis found a significantly large effect size for overall mental health symptoms, a significant medium effect size for depression, and a significant small effect size for anxiety.
  • In the network meta-analysis, AI-self-guided ACT was ranked best for reducing overall mental health symptoms, AI-self-guided eCBT was ranked best for reducing depression, and AI-guided eCBT was ranked best for reducing anxiety; however, the anxiety comparisons were not statistically significant.
  • No included studies reported adverse events. For dropout, reported rates ranged from 0% to 61% for online or remote psychotherapy, 4% to 60% for psychoeducation, and 5% to 34% for no intervention; AI-self-guided eCBT had a 14% lesser risk of dropping out than psychoeducation in one significant comparison.
  • User satisfaction was generally high: several studies reported AI tools were perceived as helpful and encouraging by around 60% to 90% of users, though around 30% or less found them unhelpful or bothersome.

## Methods

  • Mixed-methods systematic review registered in PROSPERO (CRD42023443575) and conducted according to PRISMA and PRISMA-NMA guidance.
  • Searched MEDLINE, CINAHL, PsycINFO, EMBASE, and Cochrane for English-language studies published up to October 2023; included RCTs, observational studies, non-randomized experimental studies, and case studies.
  • Extracted pre/post outcomes and AI applications; used meta-analysis and network meta-analysis for symptom and dropout outcomes, plus narrative synthesis for machine-learning prediction studies; assessed quality with the Cochrane risk of bias tool and GRADE.

## Limitations

  • The paper states that the evidence was low to very low quality under GRADE, with many studies showing high or unclear risk of bias, especially allegiance bias (n=14) and selective reporting (n=8).
  • The authors note a major limitation was the scarcity of studies: many interventions/comparators were represented by one study only, which increased imprecision, indirectness, and bias; several NMA comparisons were therefore based on indirect evidence.
  • Evident from the text: the review relied largely on Western, English-language studies and mostly female samples, which the authors say may limit generalizability; also, adverse events may be underreported in this literature, so “none reported” does not necessarily mean none occurred.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3389/fpsyt.2024.1356773

Cites

102 of 102 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 PRISMA guidelines – Study selection process.
Figure 1 PRISMA guidelines – Study selection process.CC BY · PubMed Central
Figure 2 Network plot of RCTs included in this review, comparing the identified interventions (“AI self-guided eCBT”: AI agent delivering eCBT with no therapist guidance; “AI guided eCBT”: AI agent delivering eCBT with asynchronous therapist guidance; “Remote CBT”: Standard CBT delivered over the phone; “AI self-guided ACT”: AI agent delivered ACT with no therapist guidance; “Combined”: Remote CBT and AI self-guided eCBT; “AI chatbot”: AI agent providing general conversational responses not associated with CBT or therapy; “Psychoeducation”; and “No intervention”). Circle sizes represent the aggregated sample size relative to each intervention, and the thickness of the lines represents the number of studies comparing the respective interventions. (A) Main effects of all the included RCTs (n=6), (B) Depression scores in RCTs specific to the study of interventions for depression symptoms (n=4), (C) Anxiety scores in RCTs specific to the study of interventions for anxiety symptoms (n=4), (D) RCTs reporting participant dropouts (n=6).
Figure 2 Network plot of RCTs included in this review, comparing the identified interventions (“AI self-guided eCBT”: AI agent delivering eCBT with no therapist guidance; “AI guided eCBT”: AI agent delivering eCBT with asynchronous therapist guidance; “Remote CBT”: Standard CBT delivered over the phone; “AI self-guided ACT”: AI agent delivered ACT with no therapist guidance; “Combined”: Remote CBT and AI self-guided eCBT; “AI chatbot”: AI agent providing general conversational responses not associated with CBT or therapy; “Psychoeducation”; and “No intervention”). Circle sizes represent the aggregated sample size relative to each intervention, and the thickness of the lines represents the number of studies comparing the respective interventions. (A) Main effects of all the included RCTs (n=6), (B) Depression scores in RCTs specific to the study of interventions for depression symptoms (n=4), (C) Anxiety scores in RCTs specific to the study of interventions for anxiety symptoms (n=4), (D) RCTs reporting participant dropouts (n=6).CC BY · PubMed Central
Figure 3 Forest plots of meta-analysis comparing pre- and post-intervention effects of five AI augmented interventions (Subgroup 1: AI self-guided eCBT, Subgroup 2: AI guided eCBT, Subgroup 3: AI modified eCBT, Subgroup 4: AI self-guided ACT, and Subgroup 5: Combined). (A) Main effects of all the included studies (n=10), (B) Interventions specific for the management of depression symptoms (n=7), (C) Interventions specific for the management of anxiety symptoms (n=4).
Figure 3 Forest plots of meta-analysis comparing pre- and post-intervention effects of five AI augmented interventions (Subgroup 1: AI self-guided eCBT, Subgroup 2: AI guided eCBT, Subgroup 3: AI modified eCBT, Subgroup 4: AI self-guided ACT, and Subgroup 5: Combined). (A) Main effects of all the included studies (n=10), (B) Interventions specific for the management of depression symptoms (n=7), (C) Interventions specific for the management of anxiety symptoms (n=4).CC BY · PubMed Central
Figure 4 Forest plots of network meta-analysis of RCTs. (A) Main effects of all the included RCTs (n=6), (B) Interventions specific for the management of depression symptoms (n=4), (C) Interventions specific for the management of anxiety symptoms (n=4), (D) RCTs reporting participant dropouts (n=6).
Figure 4 Forest plots of network meta-analysis of RCTs. (A) Main effects of all the included RCTs (n=6), (B) Interventions specific for the management of depression symptoms (n=4), (C) Interventions specific for the management of anxiety symptoms (n=4), (D) RCTs reporting participant dropouts (n=6).CC BY · PubMed Central

Verification

Claims the paper makes

82 Valid · 9 Not valid · 12 Not checked · 73 Own results and interpretations

84 more claims · 38 with their cited sources checked

Create a free account to see all 176 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 35
  • It may state two separate claims as one. 65

Create a free account to see these claims.

Citations

Cites

Show all 109 references

Research works printed but not linked: 4

  1. 2018, Mental health in the balance: ending the healthcare disparity in Canada
  2. 2020, Mental Illness and Addiction: Facts and Statistics
  3. Covidence Systematic Review Software2022
  4. Lipsey, 2001, Practical meta-analysis. Practical meta-analysis, ix, 247

Cited by

See all 71

Metadata

Authors
Gutierrez, G., Stephenson, C., Eadie, J., Asadpour, K., & Alavi, N.
Affiliations
Queen's University (CA)
Corresponding author
Alavi, Nazanin
Venue
Frontiers in Psychiatry
Published
2024-05-07
Volume
15
Pages
1356773-1356773
ISSN
1664-0640 (ISSN-L)
OpenAlex
W4396704075
Cited by
71
References
109
Citations by year
2026: 28 · 2025: 36 · 2024: 7
FWCI
4.30
Type
journal article
Language
en
Open access
Locations
  • Frontiers in Psychiatry (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (acceptedVersion)
Topics
  • Social Sciences › Psychology › Applied Psychology › Digital Mental Health Interventions
  • Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Telemedicine and Telehealth Implementation
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
Keywords
artificial intelligence, digital mental health interventions, medication adherence, treatment discontinuation, depression, anxiety, AI triage, treatment response prediction, self-guided intervention

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.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

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

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Examining the role of AI technology in online mental healthcare: opportunities, challenges, and implications, a mixed-methods review. https://app.cochat.ai/papers/10.3389/fpsyt.2024.1356773
The paper
Gutierrez, G., Stephenson, C., Eadie, J., Asadpour, K., & Alavi, N. (2024). Examining the role of AI technology in online mental healthcare: opportunities, challenges, and implications, a mixed-methods review. Frontiers in Psychiatry. https://doi.org/10.3389/fpsyt.2024.1356773

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.