Eprint first made available on ORBi (E-prints, working papers and research blog)
Large Language Models in Belgian Primary Care: Clinical Applications, Ethical Challenges, and Epistemological Reflections
Jamoulle, Marc; Angenot, Valentin; Ittoo, Ashwin
2025
 

Files


Full Text
LLM_CARE_20251005.pdf
Author postprint (377.16 kB)
Download

All documents in ORBi are protected by a user license.

Send to



Details



Keywords :
Large Language Models (LLMs); Artificial Intelligence in Medicine; Primary Care; Family Medicine; Clinical Documentation; Human Phenotype Ontology (HPO); Medical Ethics; Epistemology of Medicine; Health Information Systems; Belgium
Abstract :
[en] Background: Large language models (LLMs) are increasingly integrated into healthcare, with potential applications spanning clinical documentation, patient communication, de- cision support, and biomedical research. Their use, however, raises significant challenges in terms of ethics, epistemology, and health system organization. Objective: This narrative review critically examines the opportunities and limitations of LLMs in medicine, with a particular focus on their implications for Belgian primary care and family medicine. Methods: We conducted a structured narrative review of the recent literature (2023–2025) on LLMs in healthcare, complemented by clinical case illustrations and analysis of the Belgian information system. The review emphasizes practical applications, ethical issues, and epistemological reflections relevant to clinicians and policymakers. Results: LLMs can enhance efficiency in clinical documentation, support patient-centered communication across languages, and facilitate evidence retrieval and research through standardized terminologies such as the Human Phenotype Ontology. However, their inte- gration into Belgian healthcare is hindered by fragmented information systems and rigid data governance. Ethical concerns include reliability, bias, confidentiality, transparency, patient trust, and environmental impact. Epistemological biases in scientific production and the invisibilization of field knowledge further limit the reliability of LLM outputs. Im- portantly, LLMs cannot substitute for the therapeutic relationship, which remains central to care. Conclusion: LLMs represent promising auxiliary tools in Belgian and international medicine, capable of reducing administrative burden and supporting clinical practice and research. Their safe and effective integration requires ethical oversight, systemic reform of fragmented health information structures, and renewed attention to the physician– patient relationship. For Belgian primary care, these technologies should be adopted with caution, transparency, and critical vigilance to ensure they serve patients and reinforce, rather than undermine, the human dimension of medicine.
Disciplines :
General & internal medicine
Computer science
Author, co-author :
Jamoulle, Marc  ;  Université de Liège - ULiège > HEC Liège : UER > UER Opérations : Systèmes d'information de gestion
Angenot, Valentin;  ULiège - Université de Liège > Business Management System, HEC Liège
Ittoo, Ashwin ;  Université de Liège - ULiège > HEC Liège Research > HEC Liège Research: Business Analytics & Supply Chain Mgmt
Language :
English
Title :
Large Language Models in Belgian Primary Care: Clinical Applications, Ethical Challenges, and Epistemological Reflections
Publication date :
07 October 2025
Number of pages :
14 p
Available on ORBi :
since 05 October 2025

Statistics


Number of views
67 (4 by ULiège)
Number of downloads
70 (1 by ULiège)

Bibliography


Similar publications



Contact ORBi