Elamur Research⌕
☰
home / papers / 2026 / SOAPTriage: SOAP-Guided Multi-View Clinical Text Modeling Framework for Automated ESI Prediction

SOAPTriage: SOAP-Guided Multi-View Clinical Text Modeling Framework for Automated ESI Prediction

Paper · 2026 · Enming Wang, Jianlei Wang, Xueping Peng et al.

Summary

Emergency departments assign Emergency Severity Index (ESI) levels, from 1 (critical) to 5 (non-urgent), mostly by reading triage notes by hand. This is slow and becomes inconsistent under heavy patient load. The authors identify two obstacles to automating it: real triage text is hard to share because of privacy rules, and existing LLM approaches rarely build in an explicit clinical reasoning structure.

SOAPTriage addresses both. Its Clinical Note Augmentation (CNA) module fills templates from structured MIMIC-IV ED records and uses retrieved real-world documentation to guide an LLM in writing natural-language triage notes. The SOAP-Guided Encoding (SGE) module prompts a frozen LLM to read each note from four angles: Subjective, Objective, Assessment and Plan. It takes intermediate-layer embeddings for each view. The SOAP-Aware Aggregation and Inference (SAAI) module then weights these views adaptively and predicts ESI in a coarse-to-fine, two-stage way.

The framework beats prompting-based, multi-agent and encoder-based baselines. This suggests that grounding models in established clinical reasoning schemes makes automated triage support more reliable.

Key points

  • The new dataset contains 15,393 synthetic, ESI-labelled triage notes derived from MIMIC-IV ED visits. Code and data are released on GitHub (xiaoyaoiii/SOAPTriage).
  • Evaluation uses Total Discordance, Under-Triage, Over-Triage, and clinically significant variants of the last two (S-Under, S-Over), where lower is better.
  • Ablations with a Qwen3-8B backbone show that single-stage direct prediction does worse and varies more than hierarchical inference. Using last-layer embeddings instead of mid-depth ones also reduces stability.
  • Backbones in the experiments range from 0.6B to 14B parameters, and all data is English only.
  • The authors present the system as decision support that needs clinical oversight, not as an autonomous triage tool.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Harbin Institute of TechnologyHarbin · China158 works · 28,841 citations🎓University of Technology SydneySydney · Australia18 works · 12,833 citations🎓Qilu University of TechnologyJinan · China4 works · 3 citations🎓Shandong Academy of SciencesJinan · China2 works · 1 citation

Also on the byline: National Supercomputing Center in Jinan.

 Fields, topics and concepts

📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksPsychology · 1,552 worksBiochemistry, Genetics and Molecular Biology · 234 worksMolecular Biology · 231 worksSocial Psychology · 306 works

Cite

Cite as Enming Wang, Jianlei Wang, Xueping Peng, Hongjiao Guan, Yinglong Wang, Sibo Wei, Jianbin Guo, Ruifeng Xu and Wenpeng Lu. 2026. SOAPTriage: SOAP-Guided Multi-View Clinical Text Modeling Framework for Automated ESI Prediction. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.1608

@inproceedings{wang2026soaptriage,
  title = {{SOAPTriage: SOAP-Guided Multi-View Clinical Text Modeling Framework for Automated ESI Prediction}},
  author = {Enming Wang, Jianlei Wang, Xueping Peng, Hongjiao Guan, Yinglong Wang, Sibo Wei, Jianbin Guo, Ruifeng Xu and Wenpeng Lu},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.1608},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1608}
}

References

58 references, 10 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Jafar Bazyar, Mehrdad Farrokhi, Amir Salari, and Hamid Reza Khankeh. 2020. The principles of triage in emergencies and disasters: A systematic review. Prehospital and Disaster Medicine, 35(3):305–313. find
  2. Robet S Crausman. 1998. The ethics SOAP note. Chest, 113(2):1–1. find
  3. Haixing Dai, Zhengliang Liu, Wenxiong Liao, Xiaoke Huang, Yihan Cao, Zihao Wu, Lin Zhao, Shaochen Xu, Fang Zeng, Wei Liu, Ninghao Liu, Sheng Li, Dajiang Zhu, Hongmin Cai, Lichao Sun, Quanzheng Li, Dinggang Shen, Tianming Liu, and Xiang Li. 2025. AugGPT: Leveraging ChatGPT for text data augmentation.…
  4. Adebayo Da’Costa, Jennifer Teke, Joseph E Origbo, Ayokunle Osonuga, Eghosasere Egbon, and David B Olawade. 2025. AI-driven triage in emergency departments: A review of benefits, challenges, and future directions. International Journal of Medical Informatics, 197(1):1–7. find
  5. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 4171–4186.
  6. Rabie Adel El Arab and Omayma Abdulaziz Al Moosa. 2025. The role of AI in emergency department triage: An integrative systematic review. Intensive and Critical Care Nursing, 89(1):1–10. find
  7. Joseph L Fleiss. 1971. Measuring nominal scale agreement among many raters. Psychological Bulletin, 76(5):378–382. find
  8. Ari B Friedman, M Kit Delgado, and Gary E Weissman. 2024. Artificial intelligence for emergency care Triage-Much promise, but still much to learn. JAMA Network Open, 7(5):e248857–e248857. find
  9. Biyang Guo, Yeyun Gong, Yelong Shen, Songqiao Han, Hailiang Huang, Nan Duan, and Weizhu Chen. 2022. Genius: Sketch-based language model pre-training via extreme and selective masking for text generation and augmentation. arXiv preprint arXiv:2211.10330, pages 1–21. — [arXiv:2211.10330](https://arxiv.org/abs/2211.10330) find
  10. Judith A Hall. 2011. Clinicians’ accuracy in perceiving patients: Its relevance for clinical practice and a narrative review of methods and correlates. Patient Education and Counseling, 84(3):319–324. find
  11. Jeremiah S Hinson, Diego A Martinez, Stephanie Cabral, Kevin George, Madeleine Whalen, Bhakti Hansoti, and Scott Levin. 2019. Triage performance in emergency medicine: A systematic review. Annals of Emergency Medicine, 74(1):140–152. find
  12. Oleksandr Ivanov, Lisa Wolf, Deena Brecher, Erica Lewis, Kevin Masek, Kyla Montgomery, Yurii An-drieiev, Moss McLaughlin, Stephen Liu, Robert Dunne, and 1 others. 2021. Improving ED emergency severity index acuity assignment using machine learning and clinical natural language processing. Journal of… find
  13. Lee Jacobs. 2009. Interview with Lawrence Weed, MD-the father of the problem-oriented medical record looks ahead. The Permanente Journal, 13(3):1–6. find
  14. Sarah Jaroudi and J Drew Payne. 2019. Remembering Lawrence Weed: A pioneer of the SOAP note. Academic Medicine, 94(1):1–1. find
  15. A Johnson, L Bulgarelli, T Pollard, S Horng, LA Celi, and R Mark. 2023a. MIMIC-IV-ED (version 2.2). find
  16. A Johnson, L Bulgarelli, T Pollard, S Horng, LA Celi, and R Mark. 2023b. MIMIC-IV-Note: Deidentified free-text clinical notes (version 2.2). find
  17. Alistair Johnson, Lucas Bulgarelli, Tom Pollard, Steven Horng, Leo Anthony Celi, and Roger Mark. 2020. MIMIC-IV. PhysioNet, pages 49–55. find
  18. Joshua W Joseph, Evan L Leventhal, Anne V Grosses-treuer, Matthew L Wong, Loren J Joseph, Larry A Nathanson, Michael W Donnino, Noémie Elhadad, and Leon D Sanchez. 2020. Deep-learning approaches to identify critically Ill patients at emergency department triage using limited information. Journal of… find
  19. Eleni Karlafti, Athanasios Anagnostis, Theodora Simou, Angeliki Sevasti Kollatou, Daniel Paramythiotis, Georgia Kaiafa, Triantafyllos Didaggelos, Christos Savvopoulos, and Varvara Fyntanidou. 2023. Support systems of clinical decisions in the triage of the emergency department using artificial intel… find
  20. Takeshi Kojima, Shixiang Shane Gu, Machel Reid, Yutaka Matsuo, and Yusuke Iwasawa. 2022. Large language models are zero-shot reasoners. In Proceedings of the 36th International Conference on Neural Information Processing Systems, pages 22199– 22213.
  21. J. Richard Landis and Gary G. Koch. 1977. The measurement of observer agreement for categorical data. Biometrics, 33(1):159–174. find
  22. Saikam Law, Brian Oldfield, Wah Yang, and Global Obesity Collaborative. 2024. ChatGPT/GPT-4 (large language models): Opportunities and challenges of perspective in bariatric healthcare professionals. Obesity Reviews, 25(7):1–6. find
  23. Jinhyuk Lee, Wonjin Yoon, Sungdong Kim, Donghyeon Kim, Sunkyu Kim, Chan Ho So, and Jaewoo Kang. 2020. BioBERT: A pre-trained biomedical language representation model for biomedical text mining. Bioinformatics, 36(4):1234–1240.
  24. Nicholas Lee, Thanakul Wattanawong, Sehoon Kim, Karttikeya Mangalam, Sheng Shen, Gopala Anumanchipalli, Michael Mahoney, Kurt Keutzer, and Amir Gholami. 2024. LLM2LLM: Boosting LLMs with novel iterative data enhancement. In Proceedings of the 62nd Annual Meeting of the Association for Computational… find
  25. David M Levine, Rudraksh Tuwani, Benjamin Kompa, Amita Varma, Samuel G Finlayson, Ateev Mehrotra, and Andrew Beam. 2024. The diagnostic and triage accuracy of the GPT-3 artificial intelligence model: An observational study. The Lancet Digital Health, 6(8):e555–e561. find
  26. Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, and 1 others. 2020. Retrieval-augmented generation for knowledge-intensive NLP tasks. In Proceedings of the 34th International Conference on Neur…
  27. Rensis Likert. 1932. A technique for the measurement of attitudes. Archives of Psychology, 22(33):5–55. find
  28. Michelle P. Lin, Dhruv Sharma, Arjun Venkatesh, Stephen K. Epstein, Alexander Janke, Nicholas Genes, Abhi Mehrotra, James Augustine, Bill Malcolm, Pawan Goyal, and Richard T. Griffey. 2024. The clinical emergency data registry: Structure, use, and limitations for research. Annals of Emergency Medici… find
  29. Danni Liu and Jan Niehues. 2025. Middle-layer representation alignment for cross-lingual transfer in fine-tuned LLMs. arXiv preprint arXiv:2502.14830.
  30. Junling Liu, Peilin Zhou, Yining Hua, Dading Chong, Zhongyu Tian, Andrew Liu, Helin Wang, Chenyu You, Zhenhua Guo, Lei Zhu, and 1 others. 2023. Benchmarking large language models on CMExam-a comprehensive Chinese medical exam dataset. In Proceedings of the 37th International Conference on Neural Inf… find
  31. Meng Lu, Brandon Ho, Dennis Ren, and Xuan Wang. 2024. TriageAgent: Towards better multi-agents collaborations for large language model-based clinical triage. In Proceedings of the 29th Conference on Empirical Methods in Natural Language Processing, pages 5747–5764. find
  32. Andrea Juliana Sanabria, Hector Pardo-Hernandez, Mónica Ballesteros, Carlos Canelo-Aybar, Emma McFarlane, Ena Niño de Guzman, Katrina Penman, Margarita Posso, Marta Roqué i Figuls, Anna Selva, Robin W. M. Vernooij, Pablo Alonso-Coello, Laura Martínez García, Arnav Agarwal, Sophie Blanchard, Laura Br… find
  33. Rocío Sánchez-Salmerón, José L Gómez-Urquiza, Luis Albendín-García, María Correa-Rodríguez, María Begoña Martos-Cabrera, Almudena Velando-Soriano, and Nora Suleiman-Martos. 2022. Machine learning methods applied to triage in emergency services: A systematic review. International Emergency Nursing, 6… find
  34. Martin Scherer, Dagmar Lühmann, Agata Kazek, Heike Hansen, and Ingmar Schäfer. 2017. Patients attending emergency departments: A cross-sectional study of subjectively perceived treatment urgency and motivation for attending. Deutsches Ärzteblatt International, 114(39):1–9. find
  35. Qingyang Shen, Xiaozhi Zhang, Haomin Ren, Quan Guo, and Zhang Yi. 2025. Knowledge-embedded large language models for emergency triage. Knowledge-Based Systems, 318(1):1–13. find
  36. Wenpeng Lu, Kangjun Liu, Jianlei Wang, Xueping Peng, Tao Shen, Fa Zhu, Weiyu Zhang, Jiabing Zhu, Tao Xin, and Athanasios V. Vasilakos. 2025. Advancing Chinese conversation-based patient guidance with a benchmark and knowledge-evolvable assistant. IEEE Journal of Biomedical and Health Informatics, 1(… find
  37. Oscar Skean, Md Rifat Arefin, Dan Zhao, Niket Nikul Patel, Jalal Naghiyev, Yann LeCun, and Ravid Shwartz-Ziv. 2025. Layer by layer: Uncovering hidden representations in language models. In Proceedings of the 42nd International Conference on Machine Learning, pages 1–22. find
  38. Guozheng Ma, Zhen Wang, Zhecheng Yuan, Xueqian Wang, Bo Yuan, and Dacheng Tao. 2025. A comprehensive survey of data augmentation in visual reinforcement learning. International Journal of Computer Vision, 133(10):7368–7405. find
  39. Chenxu Wang, Fei Wang, Shuhan Li, Qing-wen Ren, Xiaomei Tan, Yaoyu Fu, Di Liu, Guangwu Qian, Yu Cao, Rong Yin, and 1 others. 2025. Patient triage and guidance in emergency departments using large language models: Multimetric study. Journal of Medical Internet Research, 27(1):1–16. find
  40. Xin Men, Mingyu Xu, Qingyu Zhang, Qianhao Yuan, Bingning Wang, Hongyu Lin, Yaojie Lu, Xianpei Han, and Weipeng Chen. 2025. ShortGPT: Layers in large language models are more redundant than you expect. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics, pages 2…
  41. Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc V Le, Ed H. Chi, Sharan Narang, Aakanksha Chowdhery, and Denny Zhou. 2023. Self-Consistency improves chain of thought reasoning in language models. In Proceedings of the 19th International Conference on Learning Representations, pages 1–24.
  42. Sophie Rentschler, Martin Riedl, Christian Stab, and Martin Rückert. 2022. Data augmentation for intent classification of German conversational agents in the finance domain. In Proceedings of the 18th Conference on Natural Language Processing, pages 1–7. find
  43. Christopher B Reznich, Dianne P Wagner, and Mary M Noel. 2010. A repurposed tool: The programme evaluation SOAP note. Medical Education, 44(3):298– 305. find
  44. Zaitian Wang, Pengfei Wang, Kunpeng Liu, Pengyang Wang, Yanjie Fu, Chang-Tien Lu, Charu C. Aggarwal, Jian Pei, and Yuanchun Zhou. 2026. A comprehensive survey on data augmentation. IEEE Transactions on Knowledge and Data Engineering, 38(1):47– 66. find
  45. Zhenhailong Wang, Shaoguang Mao, Wenshan Wu, Tao Ge, Furu Wei, and Heng Ji. 2024. Unleashing the emergent cognitive synergy in large language models: A task-solving agent through multi-persona self-collaboration. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguist…
  46. Jason Wei and Kai Zou. 2019. EDA: Easy data augmentation techniques for boosting performance on text classification tasks. In Proceedings of the 24th Conference on Empirical Methods in Natural Language Processing, pages 6382–6388.
  47. Christopher YK Williams, Brenda Y Miao, Aaron E Kornblith, and Atul J Butte. 2024a. Evaluating the use of large language models to provide clinical recommendations in the emergency department. Nature Communications, 15(1):1–10. find
  48. Christopher YK Williams, Travis Zack, Brenda Y Miao, Madhumita Sushil, Michelle Wang, Aaron E Kornblith, and Atul J Butte. 2024b. Use of a large language model to assess clinical acuity of adults in the emergency department. JAMA Network Open, 7(5):e248895–e248895. find
  49. Richard C Wuerz, Debbie Travers, Nicki Gilboy, David R Eitel, Alex Rosenau, and Ramine Yazhari. 2001. Implementation and refinement of the emergency severity index. Academic Emergency Medicine, 8(2):170–176. 2023. Exchange-of-thought: Enhancing large language model capabilities through cross-model c… find
  50. Zaifu Zhan, Jun Wang, Shuang Zhou, Jiawen Deng, and Rui Zhang. 2025. MMRAG: Multi-mode retrieval-augmented generation with large language models for biomedical in-context learning. Journal of the American Medical Informatics Association, 32(10):1505– 1516. find
  51. Kaiyuan Zhang, Qian Liu, Luyang Zhang, Chaoqun Zheng, Shuaimin Li, Bing Xu, Muyun Yang, Xinxiao Qiao, and Wenpeng Lu. 2025a. MADAWSD: Multi-Agent debate framework for adversarial word sense disambiguation. In Proceedings of the 30th Empirical Methods in Natural Language Processing, pages 22294–22313… find
  52. Luyang Zhang, Shuaimin Li, Yishuo Li, Kunpeng Kang, Kaiyuan Zhang, Cong Wang, and Wenpeng Lu. 2025b. RoDEval: A robust word sense disambiguation evaluation framework for large language models. In Proceedings of the 30th Empirical Methods in Natural Language Processing, pages 17095–17126. find
  53. Lei Xu, Wenzhe Zhao, and Xin Huang. 2025. Diagnosis and triage performance of contemporary large language models on short clinical vignettes. Journal of Medical Systems, 49(1):1–12. find
  54. Siyi Xun, Yue Sun, Jingkun Chen, Zitong Yu, Tong Tong, Xiaohong Liu, Mingxiang Wu, and Tao Tan. 2025. MedIQA: A scalable foundation model for prompt-driven medical image quality assessment. In Proceedings of the 28th International Conference on Medical Image Computing and Computer-Assisted Intervent… find
  55. Diji Yang, Jinmeng Rao, Kezhen Chen, Xiaoyuan Guo, Yawen Zhang, Jie Yang, and Yi Zhang. 2024. IM-RAG: Multi-round retrieval-augmented generation through learning inner monologues. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval, page… find
  56. Li-Hung Yao, Ka-Chun Leung, Chu-Lin Tsai, Chien-Hua Huang, and Li-Chen Fu. 2021. A novel deep learning–based system for triage in the emergency department using electronic medical records: retrospective cohort study. Journal of Medical Internet Research, 23(12):1–15. find
  57. Liang Yao, Zhe Jin, Chengsheng Mao, Yin Zhang, and Yuan Luo. 2019. Traditional Chinese medicine clinical records classification with BERT and domain specific corpora. Journal of the American Medical Informatics Association, 26(12):1632–1636. find
  58. Zhangyue Yin, Qiushi Sun, Cheng Chang, Qipeng Guo, Junqi Dai, Xuan-Jing Huang, and Xipeng Qiu. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wang-2026-soaptriage-soap-guided-multi-view/