LLM Hallucinations in Clinical Documentation: Scoping Review with ☸️SAIMSARA.

Name: SAIMSARA Evidence Object digital::LLM_HALLUCINATIONS_SS
Creator: SAIMSARA
License: https://saimsara.com/license/

SAIMSARA

doi:10.62487/saimsarab885dc08

LLM Hallucinations in Clinical Documentation: Scoping Review with ☸️SAIMSARA.

Issue 3, Volume 1, 2026

Editorial note

• Last update: 2026-05-18 08:59:51

What is this paper about

LLM documentation is not simply “safe” or “unsafe”: this evidence map shows where hallucination risk becomes measurable, where it falls below human benchmarks, and which safeguards actually matter—RAG, verification loops, fine-tuning, and human review. Built from 46 references and 32 original studies, it turns scattered clinical AI documentation evidence into a practical safety map for deciding whether, where, and how LLMs should enter real clinical workflows.

Human-verified editorial review Verified by World ID proof-of-human. This editorial layer was submitted from a SAIMSARA account verified as a unique human.

Evidence preview · Did you know?

Realistic clinical documentation scene with physician reviewing an AI-generated patient note.

Good-looking notes can still hallucinate

Did you know? In 97 real outpatient encounters, ambient LLM notes had more hallucinations than physician “gold” notes: 31% vs 20%.

The notes may appear organized and useful while still adding unsupported clinical content.

Realistic hospital AI dashboard showing variable hallucination risk across clinical documentation tasks.

The risk range is surprisingly wide

Did you know? Reported hallucination rates ranged from 0.73–2.0% in constrained workflows to 31–35% in higher-risk tasks.

Clinical LLM safety depends heavily on task design, source grounding, and workflow constraints.

Realistic hospital governance scene with clinicians reviewing AI documentation safeguards.

Safeguards may change the outcome

Did you know? RAG, Generator-Verifier-Judge loops, and fine-tuning reduced major hallucinations to human-comparable or lower levels in several workflows.

The key question is not whether LLMs write notes, but which control layers make them clinically acceptable.

Swipe sideways on mobile · full evidence map opens after unlock

Abstract:

Keywords:

Review Stats

Final search date and database lock: 2026-05-14 22:11:23 CEST

Plan: Pro (expanded craft tokens; source: Semantic Scholar)

Source: Semantic Scholar

Total Abstracts/Papers: 93

Downloaded Abstracts/Papers: 93

Included original and non-original Abstracts/Papers (all): 53

Included original Abstracts/Papers (Vote counting by direction of effect): 32

Reference Index (links used in paper): 46

Total participants/sample observations (topic deduplicated ΣN): 32465

Get access to the full paper

Unlock the full evidence map

The full evidence review, including the Introduction, Methods, Results, Discussion, Conclusion, figures, and complete reference index, opens after purchase or sign-in. The Evidence Object JSON is a separate machine-readable evidence product: a concentrated synthesis of results, topic-level evidence, and discussion across original and non-original studies. It can be directly input into your LLM, agent, or RAG workflow.

Reference Index (46)

[2] A Survey of LLM-based Agents in Medicine: How far are we from Baymax? — https://doi.org/10.48550/arxiv.2502.11211
[3] A framework to assess clinical safety and hallucination rates of LLMs for medical text summarisation — https://doi.org/10.1038/s41746-025-01670-7
[4] Challenges of Implementing LLMs in Clinical Practice: Perspectives — https://doi.org/10.3390/jcm14176169
[6] Effective prompt design for large language models in clinical practice — https://doi.org/10.1080/17843286.2026.2613903
[7] Improving Clinical Documentation with AI: A Comparative Study of Sporo AI Scribe and GPT-4o mini — https://doi.org/10.48550/arxiv.2410.15528
[8] Extracting International Classification of Diseases Codes from Clinical Documentation Using Large Language Models — https://doi.org/10.1055/a-2491-3872
[9] Reducing administrative burden in cardio-oncology: an LLM-powered approach for automated summarization of oncology and cardiotoxicity histories in dutch medical records — https://doi.org/10.1093/eurheartjsupp/suaf083.227
[10] LLM-Based Dictation Detection from Doctor-Patient Conversations — https://doi.org/10.1109/asru65441.2025.11434612
[11] Feasibility Study for Using Large Language Models to Identify Goals-of-Care Documentation at Scale in Patients With Advanced Cancer. — https://doi.org/10.1200/op-24-00992
[12] Large Language Models in Medicine: Clinical Applications, Technical Challenges, and Ethical Considerations — https://doi.org/10.4258/hir.2025.31.2.114
[13] From Notes to Billing: Large Language Models in Revolutionizing Medical Documentation and Healthcare Administration — https://doi.org/10.36347/sjams.2025.v13i08.005
[14] Integrating a Large Language Model to Streamline Nursing Handover Documentation Across Multiple Hospitals in Taiwan: Development and Implementation Study — https://doi.org/10.2196/81604
[15] MEDAI-LLM-SUMM: a reporting checklist for medical text summarization studies using large language models — https://doi.org/10.3389/fdgth.2026.1761601
[16] The Crisis of Ground Truth in Medical AI: Evaluating the Tools Used to Detect LLM Hallucinations — https://doi.org/10.22214/ijraset.2026.79309
[17] A web-based, LLM-powered AI symptom summarization tool (ASST) for monitoring of breast cancer treatment toxicity. — https://doi.org/10.1200/jco.2024.42.16_suppl.e13622
[18] KOSMOS: Ontology-Based Knowledge Graph Scaffolding for Medical Documentation Generation — https://doi.org/10.3390/info17040355
[19] Assessing the quality of AI-generated clinical notes: validated evaluation of a large language model ambient scribe — https://doi.org/10.3389/frai.2025.1691499
[20] CARESUM: A Verification-Driven Framework for Clinical Dialogue Summarization with an Intelligent Chatbot Interface — https://doi.org/10.1109/iconscept66142.2025.11436852
[21] Large Language Models for Clinical Narrative Processing: Methods, Applications, and Challenges — https://doi.org/10.3390/mps9030069
[22] Enhancing patient participation in emergency department through patient-friendly clinical notes generated by large language models — https://doi.org/10.1038/s41598-025-31113-y
[23] Evidence-based artificial intelligence: Implementing retrieval-augmented generation models to enhance clinical decision support in plastic surgery. — https://doi.org/10.1016/j.bjps.2025.03.053
[26] Towards Inpatient Discharge Summary Automation via Large Language Models: A Multidimensional Evaluation with a HIPAA-Compliant Instance of GPT-4o and Clinical Expert Assessment — https://doi.org/10.1101/2025.04.03.25325204
[27] Prompting for Healthcare Professionals: Enhancing Clinical Decision-Making with Artificial Intelligence — https://doi.org/10.36253/if-3198
[29] The Role of Generative Artificial Intelligence and Large Language Models in Atrial Fibrillation: Clinical Research and Decision Support. — https://doi.org/10.1097/crd.0000000000001042
[30] RAGClin: A Privacy-Preserving Retrieval-Augmented Generation Framework for Clinical Question — https://doi.org/10.1109/icerect65215.2025.11377112
[31] Physician- and Large Language Model-Generated Hospital Discharge Summaries. — https://doi.org/10.1001/jamainternmed.2025.0821
[32] AI Scribes in Health Care: Balancing Transformative Potential With Responsible Integration — https://doi.org/10.2196/80898
[33] Will AI Replace Physicians in the Near Future? AI Adoption Barriers in Medicine — https://doi.org/10.3390/diagnostics16030396
[35] Adoption and Use of LLMs at an Academic Medical Center — https://doi.org/10.48550/arxiv.2602.00074
[36] Large Language Models in Cardiovascular Prevention: A Narrative Review and Governance Framework — https://doi.org/10.3390/diagnostics16030390
[37] Developing Interpretable Large Language Models for High-Stakes Decision-Making in Healthcare: Insights from an XAI Review Perspective — https://doi.org/10.55248/gengpi.6.0525.1615
[38] Preserving Privacy, Increasing Accessibility, and Reducing Cost: An On-Device Artificial Intelligence Model for Medical Transcription and Note Generation — https://doi.org/10.1101/2025.07.01.25330679
[39] Accurate and efficient cardio-oncology outpatient clinic summarization with open-source large language models for non-english text — https://doi.org/10.1093/eurheartj/ehaf784.4159
[40] Medication safety and artificial intelligence: evidence, failure modes, and how to vet AI tools before they touch patient care — https://doi.org/10.59652/hrnhkr35
[41] Physician-Reported Safety Outcomes of AI-Generated Hospital Course Summaries — https://doi.org/10.1001/jamanetworkopen.2026.16556
[42] Large language models to identify advance care planning in patients with advanced cancer. — https://doi.org/10.1016/j.jpainsymman.2024.11.016
[44] What We Know About the Role of Large Language Models for Medical Synthetic Dataset Generation — https://doi.org/10.3390/ai6060109
[45] Privacy-Preserving Generation of Structured Lymphoma Progression Reports from Cross-sectional Imaging: A Comparative Analysis of Llama 3.3 and Llama 4 — https://doi.org/10.1007/s10278-025-01618-z
[46] UltraReporter for transforming spoken diagnostic cues into structured ultrasound reports with large language models — https://doi.org/10.1038/s41598-026-41439-w
[47] 0348 Utility of Large Language Models for Extracting Sleep Apnea Specific Symptoms and Major Adverse Cardiovascular Events from Electronic Health Records — https://doi.org/10.1093/sleep/zsag091.0348
[48] Quality assessment of large language model–generated prior authorization letters in nephrology — https://doi.org/10.3389/fdgth.2026.1767648
[49] Large Language Model Extraction of PD-L1 Biomarker Testing Details From Electronic Health Records — https://doi.org/10.1089/aipo.2024.0043
[50] Using large language models to assess adherence to ASCO patient-oncologist communication standards. — https://doi.org/10.1200/jco.2025.43.16_suppl.12105
[51] Can Local Vision-Language Models Improve Activity Recognition Over Vision Transformers? - Case Study On Newborn Resuscitation — https://doi.org/10.1109/icipw68931.2025.11385873
[52] Responsible AI for Personalized Patient Education and Engagement Across Medical Conditions: Leveraging Multi-Agent LLMs, Ambient Technology, and NotebookLM—A Case Study in Diabetes Education and Limb Preservation — https://doi.org/10.3390/japma116030030
[53] Performance validation of large language models for drug-drug interaction consultations: Potential pharmacist assistant in outpatient pharmaceutical services. — https://doi.org/10.1016/j.japh.2026.103077

LLM Hallucinations in Clinical Documentation: Scoping Review with ☸️SAIMSARA.

Digital Health

Issue 3, Volume 1, 2026

Good-looking notes can still hallucinate

The risk range is surprisingly wide

Safeguards may change the outcome

Get access to the full paper