ClinBench: A Standardized Multi-Domain Framework for Evaluating Large Language Models in Clinical Information Extraction

Ismael Villanueva-Miranda, Zifan Gu, Donghan Yang, Kuroush Nezafati, Jingwei Huang, Peifeng Ruan, Xiaowei Zhan, Guanghua Xiao, Yang Xie

Keywords: Benchmarking, Large Language Models, LLMs, Clinical NLP, Information Extraction, Evaluation Framework, Clinical Notes, Electronic Health Records (EHRs), Standardization, Reproducibility, Data-Centric AI

TL;DR: ClinBench is an open-source, multi-model, multi-domain framework for rigorously benchmarking large language models on clinical information-extraction tasks.

Abstract: Large Language Models (LLMs) offer substantial promise for clinical natural language processing (NLP); however, a lack of standardized benchmarking methodologies limits their objective evaluation and practical translation. To address this gap, we introduce ClinBench, an open-source, multi-model, multi-domain benchmarking framework. ClinBench is designed for the rigorous evaluation of LLMs on important structured information extraction tasks (e.g., tumor staging, histologic diagnoses, atrial fibrillation, and social determinants of health) from unstructured clinical notes. The framework standardizes the evaluation pipeline by: (i) operating on consistently structured input datasets; (ii) employing dynamic, YAML-based prompting for uniform task definition; and (iii) enforcing output validation via JSON schemas, supporting robust comparison across diverse LLM architectures. We demonstrate ClinBench through a large-scale study of 11 prominent LLMs (e.g., GPT-4o series, LLaMA3 variants, Mixtral) across three clinical domains using configurations of public datasets (TCGA for lung cancer, MIMIC-IV-ECG for atrial fibrillation, and MIMIC notes for SDOH). Our results reveal significant performance-efficiency trade-offs. For example, when averaged across the four benchmarked clinical extraction tasks, GPT-3.5-turbo achieved a mean F1 score of 0.83 with a mean runtime of 16.8 minutes. In comparison, LLaMA3.1-70b obtained a similar mean F1 of 0.82 but required a substantially longer mean runtime of 42.7 minutes. GPT-4o-mini also presented a favorable balance with a mean F1 of 0.81 and a mean runtime of 13.4 minutes. ClinBench provides a unified, extensible framework and empirical insights for reproducible, fair LLM benchmarking in clinical NLP. By enabling transparent and standardized evaluation, this work advances data-centric AI research, informs model selection based on performance, cost, and clinical priorities, and supports the effective integration of LLMs into healthcare. The framework and evaluation code are publicly available at https://github.com/ismaelvillanuevamiranda/ClinBench/.

Dataset URL: https://github.com/ismaelvillanuevamiranda/ClinBench

Code URL: https://github.com/ismaelvillanuevamiranda/ClinBench

Primary Area: AL/ML Datasets & Benchmarks for health sciences (e.g. climate, health, life sciences, physics, social sciences)

Submission Number: 1547