CoChat

Artificial Intelligence-Based Chatbots for Promoting Health Behavioral Changes: Systematic Review.

Open access
Aggarwal, Abhishek, Tam, Cheuk Chi, Wu, Dezhi, Li, Xiaoming, Qiao, Shan · 2023 · Journal of medical Internet research

Published version of a preprint

The DOI matches its Crossref record; 38 of the 38 references it cites were found in a registry or an index; 93 of the 113 sources its claims cite were checked for support.

Abstract

Artificial intelligence (AI)-based chatbots can offer personalized, engaging, and on-demand health promotion interventions. The aim of this systematic review was to evaluate the feasibility, efficacy, and intervention characteristics of AI chatbots for promoting health behavior change. A comprehensive search was conducted in 7 bibliographic databases (PubMed, IEEE Xplore, ACM Digital Library, PsycINFO, Web of Science, Embase, and JMIR publications) for empirical articles published from 1980 to 2022 that evaluated the feasibility or efficacy of AI chatbots for behavior change. The screening, extraction, and analysis of the identified articles were performed by following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. Of the 15 included studies, several demonstrated the high efficacy of AI chatbots in promoting healthy lifestyles (n=6, 40%), smoking cessation (n=4, 27%), treatment or medication adherence (n=2, 13%), and reduction in substance misuse (n=1, 7%). However, there were mixed results regarding feasibility, acceptability, and usability. Selected behavior change theories and expert consultation were used to develop the behavior change strategies of AI chatbots, including goal setting, monitoring, real-time reinforcement or feedback, and on-demand support. Real-time user-chatbot interaction data, such as user preferences and behavioral performance, were collected on the chatbot platform to identify ways of providing personalize…

Overview

CoChat Summary

TL;DR

This systematic review found that AI-based chatbots showed efficacy for several health behavior changes, especially healthy lifestyles, smoking cessation, and medication/treatment adherence, but evidence was mixed for feasibility, acceptability, and usability. The authors caution that most included studies had moderate to high risk of internal validity issues and limited reporting of AI methods, so stronger RCTs are needed.

Key findings

  • The review included 15 studies; AI chatbots were reported as efficacious for healthy lifestyles (6/15, 40%), smoking cessation (4/15, 27%), treatment or medication adherence (2/15, 13%), and reduction in substance misuse (1/15, 7%).
  • Feasibility evidence was limited: 11/15 (73%) studies reported feasibility, but safety was only reported in 1/15 (7%) study, and engagement often declined over time.
  • Acceptability was mixed: 7/15 (47%) studies reported acceptability/engagement, but the overall proportion of satisfied participants was less than 50%.
  • Usability was also mixed: 11/15 (73%) studies reported usability, with ease of use generally low to moderate and 1/15 (7%) studies reporting technical issues for 82.3% of participants.

Methods

  • Systematic review following PRISMA; searched 7 databases (PubMed, IEEE Xplore, ACM Digital Library, PsycINFO, Web of Science, Embase, JMIR publications) for empirical English-language articles from 1980 to 2022.
  • Included studies had to evaluate AI-based chatbots for health behavior change; from 1961 records screened, 15 studies met inclusion criteria.
  • The authors extracted study characteristics, intervention features, and outcomes; quality was assessed using the NIH quality assessment tool for controlled intervention studies and CONSORT-AI guidance.

Limitations

  • The paper states that no meta-analysis was conducted because of heterogeneity in design, outcomes, and measures.
  • The review states it did not cover a comprehensive set of behavioral outcomes because of strict inclusion criteria; the included studies focused on healthy lifestyle, smoking cessation, and treatment/medication adherence.
  • The text makes evident that many included studies had pre-post designs without control groups, small samples in some studies, and limited description of AI techniques; the paper explicitly notes moderate to high internal validity risk and limited generalizability.

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

Integrity signals

Registry

Crossref · 10.2196/40789

Cites

38 of 38 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Eligibility screening process. AI: artificial intelligence.
Figure 1 Eligibility screening process. AI: artificial intelligence.CC BY · PubMed Central

Verification

Claims the paper makes

51 Valid · 5 Not valid · 6 Not checked · 114 Own results and interpretations

90 more claims · 28 with their cited sources checked

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

See all claims — free account
  • It may state two separate claims as one. 66
  • Probably not a substantial claim. 27

Create a free account to see these claims.

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?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

2 of 2

Citations

Cites

Show all 38 references

Research works printed but not linked: 1

  1. Radziwill, NM, 2017, arXiv

Cited by

Items 401–500 of 657

Metadata

Authors
Aggarwal, A., Tam, C. C., Wu, D., Li, X., & Qiao, S.
Affiliations
University of South Carolina (US)
Corresponding author
Qiao, Shan
Venue
Journal of medical Internet research
Published
2023-02-24
Volume
25
Pages
e40789
ISSN
1438-8871 (ISSN-L)
OpenAlex
W4321748146
Cited by
683
References
36
Citations by year
2026: 239 · 2025: 263 · 2024: 135 · 2023: 46
FWCI
34.57
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • Journal of Medical Internet Research (publishedVersion, cc-by)
  • medRxiv (acceptedVersion)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • University of South Carolina (BDHSC-2021-14)
  • National Institutes of Health
  • National Institute of Allergy and Infectious Diseases (R01AI127203)
Topics
  • Health Sciences › Health Professions › General Health Professions › Mobile Health and mHealth Applications
  • Social Sciences › Psychology › Applied Psychology › Digital Mental Health Interventions
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
Keywords
AI chatbots, health behavior change, health promotion, smoking cessation, medication adherence, substance abuse, behaviour change theory, personalized interventions, feasibility and acceptability, chatbot usability, real-time feedback
MeSH terms
Artificial Intelligence, Delivery of Health Care, Health Promotion, Humans, Software, Health Behavior

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 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (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

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.