CoChat

Conversational Agents in Health Care: Scoping Review and Conceptual Analysis

Open access
Tudor Car, Lorainne, Dhinagaran, Dhakshenya Ardhithy, Kyaw, Bhone Myint, Kowatsch, Tobias, Joty, Shafiq, Theng, Yin-Leng, Atun, Rifat · 2020 · Journal of Medical Internet Research

Published version of a preprint

The DOI matches its Crossref record; 141 of the 141 references it cites were found in a registry or an index; 39 of the 147 sources its claims cite were checked for support.

Abstract

Background: Conversational agents, also known as chatbots, are computer programs designed to simulate human text or verbal conversations. They are increasingly used in a range of fields, including health care. By enabling better accessibility, personalization, and efficiency, conversational agents have the potential to improve patient care. Objective: This study aimed to review the current applications, gaps, and challenges in the literature on conversational agents in health care and provide recommendations for their future research, design, and application. Methods: We performed a scoping review. A broad literature search was performed in MEDLINE (Medical Literature Analysis and Retrieval System Online; Ovid), EMBASE (Excerpta Medica database; Ovid), PubMed, Scopus, and Cochrane Central with the search terms “conversational agents,” “conversational AI,” “chatbots,” and associated synonyms. We also searched the gray literature using sources such as the OCLC (Online Computer Library Center) WorldCat database and ResearchGate in April 2019. Reference lists of relevant articles were checked for further articles. Screening and data extraction were performed in parallel by 2 reviewers. The included evidence was analyzed narratively by employing the principles of thematic analysis. Results: The literature search yielded 47 study reports (45 articles and 2 ongoing clinical trials) that matched the inclusion criteria. The identified conversational agents were largely delivered via smartphone apps (n=23) and used free text only as the main input (n=19) and output (n=30) modality. Case studies describing chatbot development (n=18) were the most prevalent, and only 11 randomized controlled trials were identified. The 3 most commonly reported conversational agent applications in the literature were treatment and monitoring, health care service support, and patient education. Conclusions: The literature on conversational agents in health care is largely descriptive and aimed at treatment and monitoring and health service support. It mostly reports on text-based, artificial intelligence–driven, and smartphone app–delivered conversational agents. There is an urgent need for a robust evaluation of diverse health care conversational agents’ formats, focusing on their acceptability, safety, and effectiveness.

Overview

CoChat Summary

TL;DR

This scoping review mapped 47 reports (45 studies and 2 ongoing trials) on conversational agents in health care, finding mostly text-based, AI-driven, smartphone app-delivered agents. The literature is largely descriptive and concentrated in treatment/monitoring, health service support, and patient education, with an urgent need for robust evaluation of acceptability, safety, and effectiveness.

Key findings

  • The search yielded 47 reports; agents were largely delivered via smartphone apps (n=23), used free text only as the main input (n=19) and output (n=30) modality, and were most often studied through case studies (n=18), with only 11 randomized controlled trials.
  • The 3 most commonly reported applications were treatment and monitoring, health care service support, and patient education; 40 studies were from high-income countries and 6 from low- and middle-income countries, with no studies from a low-income country.
  • Evaluation data were generally positive: Babylon triage/diagnosis agents were reported as more accurate, faster, and safer than doctors and nurses; Ghosh et al and Danda et al reported diagnostic precision of 82% and 86%; Tess reduced anxiety and depression by 18% and 13%; a breast cancer chatbot had 94% overall satisfaction; 26 studies commented on acceptability.
  • Publication counts increased from 3 in 2015 to 5 in 2016, 10 in 2017, and 23 in 2018, but the authors conclude the literature remains in its infancy and lacks robust evaluations.

Methods

  • Scoping review following Arksey and O’Malley guidance updated by Peters et al; searched April 2019 in MEDLINE, EMBASE, PubMed, Scopus, and Cochrane Central, plus gray literature sources including OCLC WorldCat, ResearchGate, Google Scholar, OpenGrey, and the first 10 pages of Google, using 63 search terms and reference-list checking.
  • Included primary research studies that evaluated and reported findings on a conversational agent implemented for a health care–specific purpose; excluded proposals, opinion pieces, poorly reported chatbot assessments, and embodied/relational/animated conversational agents.
  • Two reviewers independently screened and extracted data; findings were narratively synthesized and thematically analyzed, with personality codes adapted from de Haan et al.

Limitations

  • Stated: the novelty of the field and terminology may have omitted unpublished studies from niche conferences; theme classification may not be all-inclusive and agents with multiple themes were mostly classified by dominant application; excluding poorly reported data may have missed some agents; the personality framework is exploratory and adapted for this review.
  • Stated/evident: no measurements of cost, efficiency, or productivity were identified, and there was a lack of data on adverse effects; evidence from middle-income countries was scarce and no studies came from a low-income country.
  • Evident: the predominance of small case studies and only 11 randomized controlled trials limits robust evaluation; most agents were app-based, which the text notes is often associated with high dropout rates and low utilization.

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

Integrity signals

Registry

Crossref · 10.2196/17158

Cites

141 of 141 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 2 PRISMA flow chart.
Figure 2 PRISMA flow chart.CC BY · PubMed Central

Verification

Claims the paper makes

24 Valid · 2 Not valid · 61 Not checked · 70 Own results and interpretations

66 more claims · 14 with their cited sources checked

Create a free account to see all 157 claims and to verify the 108 references still open.

See all claims — free account
  • It may state two separate claims as one. 61
  • Probably not a substantial claim. 38
  • The quote only partly states this claim. 1

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?

5 more claims about this paper

Create a free account to see every claim other papers make about this one.

See all claims — free account

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 141 references

Research works printed but not linked: 42

  1. 1992, Profiles Healthc Mark, 40
  2. Friedman, E, 1981, Hospitals, 55, 105
  3. de HaanHSnijderJvan NimwegenCBeunRChatbot Personality and Customer SatisfactionInfo Support Research20182020-07-18https://research.infosupport.com/wp-content/uploads/Chatbot-Personality-and-Customer-Satisfaction-Bachelor-Thesis-Information-Sciences-Hayco-de-Haan.pdf
  4. BallatiFCornoFde RussisL'Hey Siri, do you understand me?': Virtual Assistants and Dysarthria7th International Workshop on the Reliability of Intelligent Environments2018WoRIE'18June 25-28, 2018Rome, Italy
  5. Stasinaki, A, 2018, Swiss Medical Weekly, 148, 10s
  6. Heldt, K, 2018, Obesity Facts, 214
  7. L'Allemand, D, 2018, Obesity Reviews, 102
  8. RazzakiSBakerAPerovYMiddletonKBaxterJMullarkeyDSangarDTaliercioMButtMMajeedADoRosarioAMahoneyMJohriSA comparative study of artificial intelligence and human doctors for the purpose of triage and diagnosisCornell University2020-07-22https://arxiv.org/abs/1806.10698
  9. Griol, D, 2015, Artificial Computation in Biology and Medicine: International Work-Conference on the Interplay Between Natural andArtificial Computation, IWINAC 2015, Elche, Spain, June 1-5, 2015, Proceedings, Part I
  10. DandaPSrivastavaBMLShrivastavaMVaidya: A Spoken Dialog System for Health DomainProceedings of the 13th International Conference on Natural Language Processing2015International Conference on Natural Language Processing2016Varanasi, India
  11. Ghosh, S, 2018, Stud Health Technol Inform, 252, 51
  12. Ni, L, 2017, Knowledge and Systems Sciences
  13. GhandehariounAMcDuffDCzerwinskiMRowanKEMMA: An Emotionally Intelligent Personal Assistant for Improving WellbeingResearchgate2020-07-22https://www.researchgate.net/publication/330035444_EMMA_An_Emotionally_Intelligent_Personal_Assistant_for_Improving_Wellbeing
  14. CoieraEPaper Review: the Babylon ChatbotWordpress2020-07-22https://coiera.com/2018/06/29/paper-review-the-babylon-chatbot/
  15. Chatbots Meet eHealth: Automatizing HealthcareUniversity of Naples2020-07-22http://ceur-ws.org/Vol-1982/paper6.pdf
  16. 2016, Persuasive Strategies To Improve Driving Behaviour Of Elderly Drivers By A Feedback Approach
  17. Cooper, A, 2018, Stud Health Technol Inform, 252, 63
  18. Dharwadkar, R, 2018, Int J Comp Trends Technol, 60, -
  19. FadhilADiaconuMGabrielliSVillafioritaACoachAI: A conversational UI assisted e-coaching platformCornell Univeristy20172020-07-22https://arxiv.org/abs/1904.11961
  20. Beyond Patient Monitoring: Conversational Agents Role in Telemedicine & Healthcare Support For Home-Living Elderly IndividualsCornell University2020-07-22https://arxiv.org/abs/1803.06000
  21. Can a Chatbot Determine My Diet?: Addressing Challenges of Chatbot Application for Meal RecommendationCornell Univeristy2020-07-22https://arxiv.org/abs/1802.09100
  22. A Conversational Interface to Improve Medication Adherence: Towards AI Support in Patient's TreatmentCornell University2020-07-22https://arxiv.org/abs/1803.09844
  23. CARDIAC: An intelligent conversational assistant for chronic heart failure patient heath monitoringResearchgate2020-07-22https://www.researchgate.net/publication/228374471_CARDIAC_An_intelligent_conversational_assistant_for_chronic_heart_failure_patient_heath_monitoring
  24. Conversational System to Assist the User when Accessing Web Sources in the Medical DomainResearchGate2020-07-22https://www.researchgate.net/publication/261031810_Conversational_System_to_Assist_the_User_when_Accessing_Web_Sources_in_the_Medical_Domain
  25. Cooper, A, 2018, Stud Health Technol Inform, 252, 63
  26. Oh, KJ, 2017, Stud Health Technol Inform, 245, 1235
  27. A novel approach for medical assistance using trained chatbotResearchGate2020-07-22https://www.researchgate.net/publication/318474956_A_novel_approach_for_medical_assistance_using_trained_chatbot
  28. Chatbot Dimensions that Matter: Lessons from the TrenchesWeb Engineering2020-07-22https://link.springer.com/chapter/10.1007/978-3-319-91662-0_9 · DOI
  29. Automated Medical ChatbotSSRN2020-07-22https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3090881
  30. SleepBot: encouraging sleep hygiene using an intelligent chatbotResearchGate2020-07-22https://www.researchgate.net/publication/331428159_SleepBot_encouraging_sleep_hygiene_using_an_intelligent_chatbot
  31. HomeNL: Homecare Assistance in Natural Language. An Intelligent Conversational Agent for Hypertensive Patients ManagementCentre pour la Communication Scientifique Directe2020-07-22https://hal.archives-ouvertes.fr/inria-00519752/
  32. Chatbots and Conversational Interfaces: Three Domains of UseCEUR Workshop Proceedings2020-07-22http://ceur-ws.org/Vol-2101/paper8.pdf
  33. Towards Fully Automated Psychotherapy for Adults - BAS - Behavioral Activation Scheduling Via Web and Mobile PhoneSemantic Scholar2020-07-22https://www.semanticscholar.org/paper/Towards-Fully-Automated-Psychotherapy-for-Adults-Griffioen-Both-Cuijpers/43b3c78317a273c7ec46c1696370bf68e42b30fb
  34. Rizzo, AA, 2011, Stud Health Technol Inform, 163, 503
  35. Dr. Vdoc: A Medical Chatbot that Acts as a Virtual DoctorResearch & Reviews: Journal of Medical Science and Technology2020-07-22http://medicaljournals.stmjournals.in/index.php/RRJoMST/article/view/30
  36. Coach Me: A Platform For Promoting Healthy LifestyleResearchGate2020-07-22https://www.researchgate.net/publication/307573372_Coach_Me_A_Platform_For_Promoting_Healthy_Lifestyle
  37. Azevedo, RF, 2018, AMIA Annu Symp Proc, 2018, 185
  38. Denecke, K, 2019, Stud Health Technol Inform, 259, 77
  39. Designing for Health ChatbotsCornell University2020-07-22https://arxiv.org/abs/1902.09022
  40. Mindbot: A Social-Based Medical Virtual AssistantaIEEEXplore2020-07-22https://ieeexplore.ieee.org/document/7776377?reload=true
  41. Intelligent chatbot for analysis and diagnosis of the psychiatric disordersResearchGate2020-07-22https://www.researchgate.net/publication/329416982_INTELLIGENT_CHATBOT_FOR_ANALYSIS_AND_DIAGNOSIS_OF_THE_PSYCHIATRIC_DISORDERS
  42. Chatbots and the new world of HCIInteractions2020-07-22https://interactions.acm.org/archive/view/july-august-2017/chatbots-and-the-new-world-of-hci

Cited by

See all 553

Metadata

Authors
Tudor Car, L., Dhinagaran, D. A., Kyaw, B. M., Kowatsch, T., Joty, S., Theng, Y.-L., & Atun, R.
Affiliations
  • Nanyang Technological University (SG)
  • University of St.Gallen (CH)
  • Harvard University (US)
Venue
Journal of Medical Internet Research
Published
2020-08-07
Volume
22
Issue
8
Pages
e17158
ISSN
1438-8871 (ISSN-L)
OpenAlex
W3048100966
Cited by
553
References
139
Citations by year
2026: 109 · 2025: 102 · 2024: 93 · 2023: 106 · 2022: 87 · 2021: 53 · 2020: 3
FWCI
4.73
Type
journal article
Language
en
Open access
Licence
other-oa
Locations
  • Journal of Medical Internet Research (publishedVersion, other-oa)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Alexandria (UniSG) (University of St.Gallen) (acceptedVersion)
  • Alexandria (UniSG) (University of St.Gallen) (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • DR-NTU (Nanyang Technological University) (submittedVersion, cc-by)
  • Spiral (Imperial College London) (submittedVersion, other-oa)
  • Repository for Publications and Research Data (ETH Zurich) (submittedVersion, cc-by)
  • ETH Zürich Research Collection (cc-by)
Funders
  • National Research Foundation
  • Nanyang Technological University
Topics
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
  • Social Sciences › Psychology › Applied Psychology › Digital Mental Health Interventions
  • Social Sciences › Business, Management and Accounting › Marketing › AI in Service Interactions
Keywords
chatbots, conversational AI, health care applications, patient education, mobile application, text-based interaction, chatbot evaluation
MeSH terms
Communication, Delivery of Health Care, Humans, Software

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 2026-10-04, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-04 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (qwen3.8-omni-flash and 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.