Perceptions of artificial intelligence in healthcare: findings from a qualitative survey study among actors in France
Open accessAbstract
Background: Artificial intelligence (AI), with its seemingly limitless power, holds the promise to truly revolutionize patient healthcare. However, the discourse carried out in public does not always correlate with the actual impact. Thus, we aimed to obtain both an overview of how French health professionals perceive the arrival of AI in daily practice and the perception of the other actors involved in AI to have an overall understanding of this issue. Methods: Forty French stakeholders with diverse backgrounds were interviewed in Paris between October 2017 and June 2018 and their contributions analyzed using the grounded theory method (GTM). Results: The interviews showed that the various actors involved all see AI as a myth to be debunked. However, their views differed. French healthcare professionals, who are strategically placed in the adoption of AI tools, were focused on providing the best and safest care for their patients. Contrary to popular belief, they are not always seeing the use of these tools in their practice. For healthcare industrial partners, AI is a true breakthrough but legal difficulties to access individual health data could hamper its development. Institutional players are aware that they will have to play a significant role concerning the regulation of the use of these tools. From an external point of view, individuals without a conflict of interest have significant concerns about the sustainability of the balance between health, social justice, and freedom. Health researchers specialized in AI have a more pragmatic point of view and hope for a better transition from research to practice. Conclusion: Although some hyperbole has taken over the discourse on AI in healthcare, diverse opinions and points of view have emerged among French stakeholders. The development of AI tools in healthcare will be satisfactory for everyone only by initiating a collaborative effort between all those involved. It is thus time to also consider the opinion of patients and, together, address the remaining questions, such as that of responsibility.
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?
Clinicians expressed the need to understand both the scientific and clinical bases of the recommendations provided by the AI to confidently validate and apply the decision.
“Clinicians expressed the need to understand both the scientific and clinical bases of the recommendations provided by the AI to confidently validate and apply the decision [19,20,36-38].”
From Barriers to and facilitators of artificial intelligence adoption in health care: scoping review (2024) · cites this paper as [37]
The Results (page unavailable) say physicians were open to training to understand how AI tools work, and the Discussion says they wanted proof of usefulness in clinical practice. The paper does not specifically report that clinicians needed to understand the scientific and clinical bases of individual AI recommendations in order to validate and apply a decision. (LLM, conf=0.84)
Experimental — accuracy not yet measuredThe dilemma around who owns the data (government, institution, or patient) is a barrier to adoption.
“In terms of data ownership, the dilemma around who owns the data, whether it would be the government, institution (eg, hospital), or patient, is a barrier to adoption, as it leaves questions around how data would be integrated or accessible for future AI advancement [37,50].”
From Barriers to and facilitators of artificial intelligence adoption in health care: scoping review (2024) · cites this paper as [37]
The Results (page unavailable) describe data ownership as a difficult and problematic question, and the Conclusion says ethical issues can obstruct AI adoption. The paper does not frame ownership as a choice among government, institutions, and patients, or say that this uncertainty leaves data integration or future access unresolved. (LLM, conf=0.85)
Experimental — accuracy not yet measuredLack of regulation and policies from the government, including uncertainty around legal direction or law, is a barrier to the application of AI technologies.
“A lack of regulation and policies from the government, including uncertainty around legal direction or law, was presented as a barrier to the application of AI technologies [37,45,57,60,77,78].”
From Barriers to and facilitators of artificial intelligence adoption in health care: scoping review (2024) · cites this paper as [37]
The Results report that industry interviewees perceived a legal loophole and unclear legal documents, while others complained that ministries were insufficiently proactive. This supports uncertainty about legal direction as a perceived obstacle, but not a general lack of government regulation or policies: the same Results describe existing regulatory action and complaints of excessive regulation. (LLM, conf=0.84)
Experimental — accuracy not yet measuredLack of accountability in the decision-making process is a challenge.
“The lack of accountability in the decision-making process is a challenge; therefore, framing this in the governance model could be a way to address adoption issues related to accountability [37,50,73].”
From Barriers to and facilitators of artificial intelligence adoption in health care: scoping review (2024) · cites this paper as [37]
Previous studies exploring the implementation of these models have focused on surveying clinician perspectives on future rollout.
“Previous studies exploring the implementation of these models have focused on surveying clinician perspectives on future rollout [42,49-51].”
From Integrating a Machine Learning System Into Clinical Workflows: Qualitative Study (2020) · cites this paper as [51]
Dissatisfaction was reported if recommendations were too similar, inappropriate, or not useful.
“This also included dissatisfaction if recommendations were too similar, inappropriate, or not useful [35,37,38,40,56].”
From Barriers to and facilitators of artificial intelligence adoption in health care: scoping review (2024) · cites this paper as [37]
The Results discuss physicians’ concerns about the utility of AI tools, and the Discussion mentions limited usefulness of point-of-care tests. Neither passage reports dissatisfaction with recommendations because they were too similar, inappropriate, or not useful. Searches for those terms and paraphrases found no passage addressing that claim. (LLM, conf=0.73)
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.
The citation lists haven’t been fetched yet
References 47 · Cited by 294
Research works printed but not linked: 1
- [40]Commissioner O of the. Statement from FDA Commissioner Scott Gottlieb, M.D. on steps toward a new, tailored review framework for artificial intelligence-based medical devices. FDA. 2019. https://news-events/press-announcements/statement-fda-commissioner-scott-gottlieb-md-steps-toward-new-tailored-review-framework-artificial. Accessed 5 May 2019.
- Authors
- Laï, M.-C., Brian, M., & Mamzer, M.-F.
- Affiliations
- Inserm (FR)
- Venue
- Journal of Translational Medicine
- Published
- 2020-01-09
- Volume
- 18
- Issue
- 1
- Pages
- 14-14
- ISSN
- 1479-5876 (ISSN-L)
- PMC
- PMC6953249
- OpenAlex
- W3000239480
- Cited by
- 294
- References
- 47
- Citations by year
- 2026: 35 · 2025: 54 · 2024: 51 · 2023: 61 · 2022: 51 · 2021: 32 · 2020: 10
- FWCI
- 11.30
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Locations
- Journal of Translational Medicine (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- HAL (submittedVersion, other-oa)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
- Funders
- CARPEM program (INCA-DGOS-INSERM_12561)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Safety Research › Ethics and Social Impacts of AI
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence Applications
- Keywords
- artificial intelligence in healthcare, stakeholder perspectives, healthcare workers, artificial intelligence adoption, health data access, AI regulation, research-to-practice transition, patient perspectives, AI ethics, social justice
- MeSH terms
- Artificial Intelligence, Delivery of Health Care, France, Humans, Public Opinion, Surveys and Questionnaires
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
