Journal article · 2024 · Open access
Detecting and Mitigating LLM Hallucinations in Clinical Decision Support: A Safety Framework for Healthcare AI Deployment
African Journal of Biomedical Research · Published 29 May 2024 · DOI 10.65327/59902c30
In brief
“Detecting and Mitigating LLM Hallucinations in Clinical Decision Support: A Safety Framework for Healthcare AI Deployment” is a 2024 journal article by Gangadhar Vasanthapuram, published in African Journal of Biomedical Research. OpenAlex counts 0 citations as of 19 November 2026. It is open access under a CC BY licence: the full text is free to read at African Journal of Biomedical Research.
- Citations
- 0
- References
- 2
- Authors
- 1
Abstract
Large language models (LLMs) are increasingly integrated into clinical workflows — medical question answering, differential diagnosis, discharge summary drafting, and patient communication — yet exhibit hallucinations: syntactically coherent but factually incorrect or clinically dangerous outputs. In healthcare, hallucinated drug dosages, fabricated contraindications, or misinterpreted laboratory values can directly harm patients. Existing LLM safety research focuses predominantly on general-domain factuality with minimal investigation of clinical hallucination patterns. This paper presents ClinicalGuard, a comprehensive framework for detecting, classifying, and mitigating hallucinations in clinical LLM outputs. We developed a six-category hallucination taxonomy — dosing errors, fabricated citations, contraindication inversions, lab value misinterpretations, temporal confabulations, and demographic attribute errors — from clinician evaluation of 14,200 LLM outputs across four medical specialties (Cohen's κ=0.87). A DeBERTa-v3 classifier fine-tuned on 28,400 annotated outputs achieved macro-F1 of 0.912. Four mitigation strategies — RAG, chain-of-thought prompting, semantic entropy, and human-in-the-loop escalation — combined to reduce hallucination rate from 18.7% to 3.2% (82.9% relative reduction). ClinicalGuard provides a reproducible, regulatory-aligned safety layer for responsible LLM deployment in healthcare.
Cite this paper
Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.
APA 7
Vasanthapuram, G. (2024). Detecting and Mitigating LLM Hallucinations in Clinical Decision Support: A Safety Framework for Healthcare AI Deployment. African Journal of Biomedical Research. https://doi.org/10.65327/59902c30
MLA 9
Vasanthapuram, Gangadhar. “Detecting and Mitigating LLM Hallucinations in Clinical Decision Support: A Safety Framework for Healthcare AI Deployment.” African Journal of Biomedical Research, 2024. https://doi.org/10.65327/59902c30.
Chicago
Vasanthapuram, Gangadhar. 2024. “Detecting and Mitigating LLM Hallucinations in Clinical Decision Support: A Safety Framework for Healthcare AI Deployment.” African Journal of Biomedical Research. https://doi.org/10.65327/59902c30.
Harvard
Vasanthapuram, G. (2024) ‘Detecting and Mitigating LLM Hallucinations in Clinical Decision Support: A Safety Framework for Healthcare AI Deployment’, African Journal of Biomedical Research. doi:10.65327/59902c30.
IEEE
G. Vasanthapuram, “Detecting and Mitigating LLM Hallucinations in Clinical Decision Support: A Safety Framework for Healthcare AI Deployment,” African Journal of Biomedical Research, 2024, doi: 10.65327/59902c30.
BibTeX
@article{vasanthapuram2024detecting,
title = {{Detecting and Mitigating LLM Hallucinations in Clinical Decision Support: A Safety Framework for Healthcare AI Deployment}},
author = {Vasanthapuram, Gangadhar},
journal = {African Journal of Biomedical Research},
year = {2024},
doi = {10.65327/59902c30},
url = {https://doi.org/10.65327/59902c30}
}
Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)
Topic and keywords
Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- large language models
- hallucination detection
- clinical decision support
- hallucination taxonomy
- hallucination mitigation
- drug dosage errors
- human-in-the-loop
- retrieval-augmented generation
Work with this paper
Most-cited papers in Medicine
- Yaşlı Kadınlarda Üreme SağlığıDemet Kokanalı, Yaprak Engin Üstün · Jinekoloji-Obstetrik ve Neonatoloji Tıp Dergisi · 2019 · 169,630 citations
- Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 CountriesHyuna Sung, Adalberto M. Filho, Mathieu Laversanne and 5 more · CA A Cancer Journal for Clinicians · 2021 · 119,885 citations
- “Mini-mental state”Marshal F. Folstein, Susan E. Folstein, Paul R. McHugh · Journal of Psychiatric Research · 1975 · 90,424 citations
- Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countriesFreddie Bray, Jacques Ferlay, Isabelle Soerjomataram and 3 more · CA A Cancer Journal for Clinicians · 2018 · 87,396 citations
- Hallmarks of Cancer: The Next GenerationDouglas Hanahan, Robert A. Weinberg · Cell · 2011 · 69,758 citations
Questions about this paper
How do I cite this paper in APA style?
Vasanthapuram, G. (2024). Detecting and Mitigating LLM Hallucinations in Clinical Decision Support: A Safety Framework for Healthcare AI Deployment. African Journal of Biomedical Research. https://doi.org/10.65327/59902c30
How many times has this paper been cited?
0 times, according to OpenAlex on 19 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.
Is this paper free to read?
Yes. It is open access at African Journal of Biomedical Research: sign in and open the PDF from this page.
Who wrote this paper?
Gangadhar Vasanthapuram.
What is the DOI of this paper?
10.65327/59902c30, permanently at https://doi.org/10.65327/59902c30. Sign in and open it from this page.
Bibliographic data from OpenAlex (CC0), last checked 19 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.