MedAlign: A Clinician-Generated Dataset for Instruction Following with Electronic Medical Records
Open accessAbstract
The ability of large language models (LLMs) to follow natural language instructions with human-level fluency suggests many opportunities in healthcare to reduce administrative burden and improve quality of care. However, evaluating LLMs on realistic text generation tasks for healthcare remains challenging. Existing question answering datasets for electronic health record (EHR) data fail to capture the complexity of information needs and documentation burdens experienced by clinicians. To address these challenges, we introduce MedAlign, a benchmark dataset of 983 natural language instructions for EHR data. MedAlign is curated by 15 clinicians (7 specialities), includes clinician-written reference responses for 303 instructions, and provides 276 longitudinal EHRs for grounding instruction-response pairs. We used MedAlign to evaluate 6 general domain LLMs, having clinicians rank the accuracy and quality of each LLM response. We found high error rates, ranging from 35% (GPT-4) to 68% (MPT-7B-Instruct), and 8.3% drop in accuracy moving from 32k to 2k context lengths for GPT-4. Finally, we report correlations between clinician rankings and automated natural language generation metrics as a way to rank LLMs without human review. We make MedAlign available under a research data use agreement to enable LLM evaluations on tasks aligned with clinician needs and preferences.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Previous health-related benchmarks graded short responses, extracted fixed entities from responses, or used weakly correlated natural language generation metrics.
“Previous health-related benchmarks graded short responses, extracted fixed entities from responses, or used weakly correlated natural language generation metrics (Jain et al., 2021; Fleming et al., 2024; McDuff et al., 2025).”
From HealthBench: Evaluating Large Language Models Towards Improved Human Health (2025) · cites this paper as Fleming et al., 2024
The Background and Related Work section says earlier EHR question-answering datasets focused on concise, factoid-style answers, supporting the short-response part. MedAlign also evaluates NLG metrics, but its Automated Evaluation section describes their correlations with clinician rankings as moderate and calls the strongest metric a reasonable, if coarse, measure. Our copy does not establish that previous benchmarks extracted fixed entities from responses. (LLM, conf=0.84)
Experimental — accuracy not yet measuredLarge language models have misinterpreted Electronic Health Record (EHR) data.
“misinterpreted Electronic Health Record (EHR) data6”
From Towards evaluating and building versatile large language models for medicine (2025) · cites this paper as [6]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
Wu, Y., Schuster, M., Chen, Z., Le, Q. V., Norouzi, M., Macherey, W., Krikun, M., Cao, Y., Gao, Q., Macherey, K., et al. (2016) · arXiv · doi:10.48550/arxiv.1609.08144 · Cited by 5,319 · see citing works
Kung, T. H., Cheatham, M., Medenilla, A., Sillos, C., De Leon, L., Elepaño, C., Madriaga, M., Aggabao, R., Diaz-Candido, G., Maningo, J., et al. (2023) · PLOS Digital Health · doi:10.1371/journal.pdig.0000198 · Cited by 4,115 · see citing works
Large language models encode clinical knowledge
Singhal, K., Azizi, S., Tu, T., Mahdavi, S. S., Wei, J., Chung, H. W., Scales, N., Tanwani, A., Cole-Lewis, H., Pfohl, S., et al. (2023) · Nature · doi:10.1038/s41586-023-06291-2 · Cited by 4,083 · see citing works
Large language models in medicine
Thirunavukarasu, A. J., Ting, D. S. J., Elangovan, K., Gutierrez, L., Tan, T. F., & Ting, D. S. W. (2023) · Nature Medicine · doi:10.1038/s41591-023-02448-8 · Cited by 3,944 · see citing works
Llama 2: Open Foundation and Fine-Tuned Chat Models
Touvron, H., Martin, L., Stone, K., Albert, P., Almahairi, A., Babaei, Y., Bashlykov, N., Batra, S., Bhargava, P., Bhosale, S., et al. (2023) · arXiv · doi:10.48550/arxiv.2307.09288 · Cited by 2,604 · see citing works
A Survey of Large Language Models
Zhao, W. X., Zhou, K., Li, J., Tang, T., Wang, X., Hou, Y., Min, Y., Zhang, B., Zhang, J., Dong, Z., et al. (2023) · arXiv · doi:10.48550/arxiv.2303.18223 · Cited by 2,090
Scaling Instruction-Finetuned Language Models
Chung, H. W., Hou, L., Longpre, S., Zoph, B., Tay, Y., Fedus, W., Li, Y., Wang, X., Dehghani, M., Brahma, S., et al. (2022) · arXiv · doi:10.48550/arxiv.2210.11416 · Cited by 1,441 · see citing works
COMET: A Neural Framework for MT Evaluation
Rei, R., Stewart, C., Farinha, A. C., & Lavie, A. (2020) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2020.emnlp-main.213 · Cited by 767 · see citing works
Capabilities of GPT-4 on Medical Challenge Problems
Nori, H., King, N., McKinney, S. M., Carignan, D., & Horvitz, E. (2023) · arXiv · doi:10.48550/arxiv.2303.13375 · Cited by 509 · see citing works
Creation and Adoption of Large Language Models in Medicine
Shah, N. H., Entwistle, D., & Pfeffer, M. A. (2023) · JAMA · doi:10.1001/jama.2023.14217 · Cited by 435 · see citing works
Self-citationLarge language models propagate race-based medicine
Omiye, J. A., Lester, J. C., Spichak, S., Rotemberg, V., & Daneshjou, R. (2023) · npj Digital Medicine · doi:10.1038/s41746-023-00939-z · Cited by 417 · see citing works
Melnick, E. R., Dyrbye, L. N., Sinsky, C. A., Trockel, M., West, C. P., Nedelec, L., Tutty, M. A., & Shanafelt, T. (2020) · Mayo Clinic Proceedings · doi:10.1016/j.mayocp.2019.09.024 · Cited by 413 · see citing works
LinkBERT: Pretraining Language Models with Document Links
Yasunaga, M., Leskovec, J., & Liang, P. (2022) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.18653/v1/2022.acl-long.551 · Cited by 374 · see citing works
Self-citationVoss, E. A., Makadia, R., Matcho, A., Ma, Q., Knoll, C., Schuemie, M. J., DeFalco, F., Londhe, A. A., Zhu, V. J., & Ryan, P. B. (2015) · Journal of the American Medical Informatics Association · doi:10.1093/jamia/ocu023 · Cited by 343 · see citing works
2018 n2c2 shared task on adverse drug events and medication extraction in electronic health records
Henry, S., Buchan, K., Filannino, M., Stubbs, A., & Uzuner, Ö. (2019) · Journal of the American Medical Informatics Association · doi:10.1093/jamia/ocz166 · Cited by 307 · see citing works
emrQA: A Large Corpus for Question Answering on Electronic Medical Records
Pampari, A., Raghavan, P., Liang, J., & Peng, J. (2018) · Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/d18-1258 · Cited by 191 · see citing works
Medical Documentation Burden Among US Office-Based Physicians in 2019: A National Study
Gaffney, A., Woolhandler, S., Cai, C., Bor, D., Himmelstein, J., McCormick, D., & Himmelstein, D. U. (2022) · JAMA Internal Medicine · doi:10.1001/jamainternmed.2022.0372 · Cited by 118 · see citing works
Towards a Unified Multi-Dimensional Evaluator for Text Generation
Zhong, M., Liu, Y., Yin, D., Mao, Y., Jiao, Y., Liu, P., Zhu, C., Ji, H., & Han, J. (2022) · Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2022.emnlp-main.131 · Cited by 117 · see citing works
DERA: Enhancing Large Language Model Completions with Dialog-Enabled Resolving AgentsPreprint
Nair, V., Schumacher, E., Tso, G., et al. (2023) · Clinical Natural Language Processing Workshop · doi:10.48550/arxiv.2303.17071 · Cited by 75 · see citing works
Carrell, D., Malin, B., Aberdeen, J. S., Bayer, S., Clark, C., Wellner, B., & Hirschman, L. (2012) · Journal of the American Medical Informatics Association · doi:10.1136/amiajnl-2012-001034 · Cited by 70 · see citing works
Electronic Health Records and Physician Burnout: A Scoping Review
Muhiyaddin, R., ElFadl, A. H., Mohamed, E., Shah, Z., Alam, T., Abd-Alrazaq, A., & Househ, M. (2022) · Studies in health technology and informatics · doi:10.3233/shti210962 · Cited by 67 · see citing works
How Physicians Spend Their Work Time: an Ecological Momentary Assessment
Toscano, F., O’Donnell, E. M., Broderick, J. E., May, M., Tucker, P., Unruh, M. A., Messina, G., & Casalino, L. P. (2020) · Journal of General Internal Medicine · doi:10.1007/s11606-020-06087-4 · Cited by 57 · see citing works
Extending Context Window of Large Language Models via Positional Interpolation
Chen, S., Wong, S., Chen, L., & Tian, Y. (2023) · arXiv · doi:10.48550/arxiv.2306.15595 · Cited by 50 · see citing works
Structured Chart Review: Assessment of a Structured Chart Review Methodology
Siems, A. L., Banks, R., Holubkov, R., Meert, K. L., Bauerfeld, C. P., Beyda, D. H., Berg, R. A., Bulut, Y., Burd, R. S., Carcillo, J. A., et al. (2020) · Hospital Pediatrics · doi:10.1542/hpeds.2019-0225 · Cited by 36 · see citing works
Clinical Reading Comprehension: A Thorough Analysis of the emrQA Dataset
Yue, X., Jimenez Gutierrez, B., & Sun, H. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.410 · Cited by 33 · see citing works
Annotating Electronic Medical Records for Question AnsweringPreprint
Raghavan, P., Patwardhan, S., Liang, J. J., et al. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1805.06816 · Cited by 12 · see citing works
Learning to Ask Like a Physician
Lehman, E., Lialin, V., Legaspi, K. E., Sy, A. J., Pile, P. T., Alberto, N. R., Ragasa, R. R., Puyat, C. V., Taliño, M. K., Alberto, I. R., et al. (2022) · Proceedings of the 4th Clinical Natural Language Processing Workshop · doi:10.18653/v1/2022.clinicalnlp-1.8 · Cited by 10 · see citing works
NLG Evaluation Metrics Beyond Correlation Analysis: An Empirical Metric Preference Checklist
Ni’mah, I., Fang, M., Menkovski, V., & Pechenizkiy, M. (2023) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.18653/v1/2023.acl-long.69 · Cited by 10 · see citing works
Flagged first, then by citations
Research works printed but not linked: 18
- [1]Banerjee S and Lavie A. Meteor: An automatic metric for mt evaluation with improved correlation with human judgments. In Proceedings of the acl workshop on intrinsic and extrinsic evaluation measures for machine translation and/or summarization, pages 65– 72, 2005.
- [8]Dao T, Fu D, Ermon S, Rudra A, and Ré C. Flashattention: Fast and memory-efficient exact attention with io-awareness. Advances in Neural Information Processing Systems, 35: 16344– 16359, 2022.
- [9]Fan J. Annotating and characterizing clinical sentences with explicit why-qa cues. In Proceedings of the 2nd Clinical Natural Language Processing Workshop, pages 101– 106, 2019.
- [10]Freitag M, Rei R, Mathur N, Lo C.-k., Stewart C, Avramidis E, Kocmi T, Foster G, Lavie A, and Martins AF. Results of wmt22 metrics shared task: Stop using bleu–neural metrics are better and more robust. In Proceedings of the Seventh Conference on Machine Translation (WMT), pages 46– 68, 2022.
- [12]Hendrycks D, Burns C, Basart S, Zou A, Mazeika M, Song D, and Steinhardt J. Measuring massive multitask language understanding. In International Conference on Learning Representations, 2020.
- [14]Jin D, Pan E, Oufattole N, Weng W-H, Fang H, and Szolovits P. What disease does this patient have? a large-scale open domain question answering dataset from medical exams. Applied Sciences, 11 (14): 6421, 2021.
- [18]Lin C-Y. Rouge: A package for automatic evaluation of summaries. In Text summarization branches out, pages 74– 81, 2004.
- [25]Oliveira L. E. S. e., Schneider ETR, Gumiel YB, Luz M. A. P. d., Paraiso EC, and Moro C. Experiments on portuguese clinical question answering. In Brazilian Conference on Intelligent Systems, pages 133– 145. Springer, 2021.
- [28]Papineni K, Roukos S, Ward T, and Zhu W-J. Bleu: a method for automatic evaluation of machine translation. In Proceedings of the 40th annual meeting of the Association for Computational Linguistics, pages 311– 318, 2002.
- [29]Popović M. chrf++: words helping character n-grams. In Proceedings of the second conference on machine translation, pages 612– 618, 2017.
- [35]Soni S, Gudala M, Pajouhi A, and Roberts K. Radqa: A question answering dataset to improve comprehension of radiology reports. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 6250– 6259, 2022.
- [38]Topsakal O and Akinci TC. Creating large language model applications utilizing langchain: A primer on developing llm apps fast. In Proceedings of the International Conference on Applied Engineering and Natural Sciences, Konya, Turkey, pages 10– 12, 2023.
- [41]Trotman A, Puurula A, and Burgess B. Improvements to bm25 and language models examined. In Proceedings of the 19th Australasian Document Computing Symposium, pages 58– 65, 2014.
- [43]Wei J, Bosma M, Zhao V, Guu K, Yu AW, Lester B, Du N, Dai AM, and Le QV. Finetuned language models are zero-shot learners. In International Conference on Learning Representations, 2022. URL https://openreview.net/forum?id=gEZrGCozdqR.
- [44]Wei Z, Xu X, Wang C, Liu Z, Xin P, and Zhang W. An index construction and similarity retrieval method based on sentence-bert. In 2022 7th International Conference on Image, Vision and Computing (ICIVC), pages 934– 938. IEEE, 2022.
- [45]Wu C, Zhang X, Zhang Y, Wang Y, and Xie W. Pmc-llama: Further finetuning llama on medical papers. arXiv preprint arXiv:2304.14454, 2023. · arXiv
- [49]Yue X, Zhang XF, Yao Z, Lin S, and Sun H. Cliniqg4qa: Generating diverse questions for domain adaptation of clinical question answering. In 2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pages 580– 587. IEEE, 2021.
- [50]Zhang T, Koshre V, Wu F, Weinberger K, and Artzi Y. Bertscore: Evaluating text generation with bert. In International Conference on Learning Representations, 2020. URL https://openreview.net/forum?id=SkeHuCVFDr.
Cited by
Sorted by citations
Adapted large language models can outperform medical experts in clinical text summarization
Van Veen, D., Van Uden, C., Blankemeier, L., Delbrouck, J.-B., Aali, A., Bluethgen, C., Pareek, A., Polacin, M., Reis, E. P., Seehofnerová, A., et al. (2024) · Nature Medicine · doi:10.1038/s41591-024-02855-5 · Cited by 891 · see citing works
Self-citationTowards evaluating and building versatile large language models for medicine
Wu, C., Qiu, P., Liu, J., Gu, H., Li, N., Zhang, Y., Wang, Y., & Xie, W. (2025) · npj Digital Medicine · doi:10.1038/s41746-024-01390-4 · Cited by 52 · see citing works
Holistic evaluation of large language models for medical tasks with MedHELM
Bedi, S., Cui, H., Fuentes, M., Unell, A., Wornow, M., Banda, J. M., Kotecha, N., Keyes, T., Mai, Y., Oez, M., et al. (2026) · Nature Medicine · doi:10.1038/s41591-025-04151-2 · Cited by 59 · see citing works
Self-citationZambrano Chaves, J. M., Huang, S.-C., Xu, Y., Xu, H., Usuyama, N., Zhang, S., Wang, F., Xie, Y., Khademi, M., Yang, Z., et al. (2025) · Nature Communications · doi:10.1038/s41467-025-58344-x · Cited by 45 · see citing works
Self-citationHealthBench: Evaluating Large Language Models Towards Improved Human Health
Arora, R. K., Wei, J., Hicks, R. S., Bowman, P., Quiñonero-Candela, J., Tsimpourlas, F., Sharman, M., Shah, M., Vallone, A., Beutel, A., et al. (2025) · arXiv · doi:10.48550/arxiv.2505.08775 · Cited by 28 · see citing works
Yang, R., Zeng, Q., You, K., Qiao, Y., Huang, L., Hsieh, C.-C., Rosand, B., Goldwasser, J., Dave, A., Keenan, T., et al. (2024) · Journal of Medical Internet Research · doi:10.2196/60601 · Cited by 16 · see citing works
Hallucination Detection-Guided Preference Optimization for Clinical Summarization
Seethakantha, S. K., Thai, D. N., Gudi, V. P., Tiwari, S., Matar, R., Mitra, A., Zhao, W., McCallum, A., & Salloum, W. (2026) · arXiv · doi:10.48550/arxiv.2605.28910
Clinical Text Summarization: Adapting Large Language Models Can Outperform Human Experts
Veen, D. V., Uden, C. V., Blankemeier, L., Delbrouck, J.-B., Aali, A., Bluethgen, C., Pareek, A., Polacin, M., Reis, E. P., Seehofnerova, A., et al. (2023) · doi:10.21203/rs.3.rs-3483777/v1 · Cited by 892 · see citing works
Self-citation
- Authors
- Fleming, S. L., Lozano, A., Haberkorn, W. J., Jindal, J. A., Reis, E., Thapa, R., Blankemeier, L., Genkins, J. Z., Steinberg, E., Nayak, A., Patel, B., Chiang, C.-C., Callahan, A., Huo, Z., Gatidis, S., Adams, S., Fayanju, O., Shah, S. J., Savage, T., Goh, E., Chaudhari, A. S., Aghaeepour, N., Sharp, C., Pfeffer, M. A., Liang, P., Chen, J. H., Morse, K. E., Brunskill, E. P., Fries, J. A., & Shah, N. H.
- Affiliations
- Stanford Medicine (US)
- Stanford University (US)
- Hospital Israelita Albert Einstein (BR)
- Stanford Health Care (US)
- Vanderbilt University (US)
- Mayo Clinic (US)
- ORCID
- Fleming, Scott L.
- Lozano, Alejandro
- Jindal, Jenelle A.
- Reis, Eduardo
- Thapa, Rahul
- Blankemeier, Louis
- Genkins, Julian Z.
- Steinberg, Ethan
- Nayak, Ashwin
- Patel, Birju
- Chiang, Chia-Chun
- Callahan, Alison
- Huo, Zepeng
- Gatidis, Sergios
- Adams, Scott
- Fayanju, Oluseyi
- Shah, Shreya J.
- Savage, Thomas
- Goh, Ethan
- Chaudhari, Akshay S.
- Aghaeepour, Nima
- Sharp, Christopher
- Pfeffer, Michael A.
- Liang, Percy
- Chen, Jonathan H.
- Morse, Keith E.
- Brunskill, Emma P.
- Fries, Jason A.
- Shah, Nigam H.
- Venue
- Proceedings of the AAAI Conference on Artificial Intelligence
- Published
- 2024-03-24
- Volume
- 38
- Issue
- 20
- Pages
- 22021-22030
- ISSN
- 2159-5399 (ISSN-L)
- 2374-3468
- PMC
- PMC12826664
- OpenAlex
- W4393157213
- Cited by
- 66
- References
- 28
- Citations by year
- 2026: 11 · 2025: 29 · 2024: 23 · 2023: 3
- FWCI
- 16.87
- Type
- conference paper
- Language
- en
- Open access
- diamond
- doi.org
- Locations
- Proceedings of the AAAI Conference on Artificial Intelligence (publishedVersion)
- PubMed (publishedVersion)
- PubMed Central (publishedVersion)
- arXiv (Cornell University) (submittedVersion, cc-by)
- AAAI Publications (The Association for the Advancement of Artificial Intelligence (AAAI)) (submittedVersion)
- arXiv (Cornell University) (cc-by)
- Funders
- National Heart, Lung, and Blood Institute (75N92020C00008, 75N92020C00021, R01HL167974)
- National Institute of General Medical Sciences (R35GM138353)
- National Institute of Arthritis and Musculoskeletal and Skin Diseases (R01AR077604, R01AR079431)
- National Institute of Biomedical Imaging and Bioengineering (P41EB027060, R01EB002524)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning in Healthcare
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Keywords
- electronic health records, natural language instructions, large language models, LLM benchmark, clinical documentation burden, GPT-4
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
