CDrugRed: A Chinese Drug Recommendation Dataset for Discharge Medications in Metabolic Diseases

2025-10-23

AI Summary

The scarcity of non-English real-world electronic health record (EHR) data—particularly Chinese—hinders the development of multilingual intelligent medication recommendation systems. Method: To address this, we introduce CMED-DR, the first publicly available Chinese EHR dataset tailored for discharge medication recommendation in metabolic diseases. It comprises 5,894 de-identified records from 3,190 patients, encompassing structured clinical information including demographics, medical history, diagnoses, and treatment details. Contribution/Results: CMED-DR fills a critical gap in non-English discharge medication recommendation resources. Leveraging it, we conduct supervised fine-tuning of large language models, achieving benchmark F1 and Jaccard scores of 0.5648 and 0.4477, respectively—demonstrating both task difficulty and dataset validity. This work establishes a foundational resource and methodological paradigm for interpretable, accurate cross-lingual clinical decision support research.

Technology Category

• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Natural Language Processing: Machine Translation, Multilinguality, Cross-Lingual NLP

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Intelligent drug recommendation based on Electronic Health Records (EHRs) is critical for improving for improving the quality and efficiency of clinical decision-making. By leveraging large-scale patient data, drug recommendation systems can assist physicians in selecting the most appropriate medications according to a patient's medical history, diagnoses, laboratory results, and comorbidities. However, the advancement of such systems is significantly hampered by the scarcity of publicly available, real-world EHR datasets, particularly in languages other than English. In this work, we present CDrugRed, a first publicly available Chinese drug recommendation dataset focused on discharge medications for metabolic diseases. The dataset includes 5,894 de-identified records from 3,190 patients, containing comprehensive information such as patient demographics, medical history, clinical course, and discharge diagnoses. We assess the utility of CDrugRed by benchmarking several state-of-the-art large language models (LLMs) on the discharge medication recommendation task. Experimental results show that while supervised fine-tuning improves model performance, there remains substantial room for improvement, with the best model achieving the F1 score of 0.5648 and Jaccard score of 0.4477. This result highlights the complexity of the clinical drug recommendation task and establishes CDrugRed as a challenging and valuable resource for developing more robust and accurate drug recommendation systems. The dataset is publicly available to the research community under the data usage agreements at https://github.com/DUTIR-BioNLP/CDrugRed.

Problem

• Addressing scarcity of public Chinese EHR datasets for drug recommendation
• Developing discharge medication recommendation for metabolic diseases in Chinese
• Evaluating LLMs' performance on clinical drug recommendation tasks

Innovation

• Created first Chinese drug recommendation dataset
• Benchmarked large language models on medication tasks
• Used supervised fine-tuning to improve model performance