Evaluating large language model workflows in clinical decision support for triage and referral and diagnosis
Оценка рабочих процессов на основе больших языковых моделей для поддержки клинических решений при сортировке, направлении к специалистам и диагностике
2025-05-09
SCID: 54.1/bzazw9n6
Discuss with AI
MIMIC databaseclinical decision supportlarge language modelsretrieval-augmented generationtriage and referral
Figures from the paper
Abstract (AI)
Accurate medical decision-making is critical for both patients and clinicians. Patients often struggle to interpret their symptoms, determine their severity, and select the right specialist. Simultaneously, clinicians face challenges in integrating complex patient data to make timely, accurate diagnoses. Recent advances in large language models (LLMs) offer the potential to bridge this gap by supporting decision-making for both patients and healthcare providers. In this study, we benchmark multiple LLM versions and an LLM-based workflow incorporating retrieval-augmented generation (RAG) on a curated dataset of 2000 medical cases derived from the Medical Information Mart for Intensive Care database. Our findings show that these LLMs are capable of providing personalized insights into likely diagnoses, suggesting appropriate specialists, and assessing urgent care needs. These models may also support clinicians in refining diagnoses and decision-making, offering a promising approach to improving patient outcomes and streamlining healthcare delivery.
Key Findings
1
LLM-based decision support may help clinicians refine diagnoses and make more timely, accurate decisions from complex patient information.
2
The evaluated models provide personalized insights into likely diagnoses, recommend appropriate medical specialists, and assess patients’ urgent-care needs.
3
The findings indicate potential for LLM workflows to improve patient outcomes and streamline healthcare delivery, although the abstract does not report quantitative performance comparisons.
4
The study benchmarks multiple large language model versions and a retrieval-augmented generation workflow using 2,000 curated medical cases from the MIMIC database.
Research Object
large language model workflows for clinical decision support
Research Subject
their performance in triage, specialist referral, diagnosis, and assessment of urgent-care needs for patients and clinicians
Publication Details
Publication Date
2025-05-09
Journal
Publisher
ISSN
Cited by
164
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai4
Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks2020
MIMIC-IV, a freely accessible electronic health record dataset2023
Prompt engineering in consistency and reliability with the evidence-based guideline for LLMs2024
Diagnostic reasoning prompts reveal the potential for large language model interpretability in medicine2024