EUVIMEDEUVIMED entdecken
Crossref

Leveraging Large Language Models to Understand Reasons for SGLT2 Nontreatment Among Patients With Heart Failure

Elena Vasti, Minh Nguyen, Emily Alsentzer, Emily Chu, Xichong Liu, Madie Chakoumakos, Christian Haudenschild, Benjamin Wang, Jimmy Zheng, Neil M Kalwani, Ankeet S Bhatt, Andrew P Ambrosy und 4 weitere

Journal of Cardiac Failure · 2026-11 · 32 (11) · 1902-1911

DOI: 10.1016/j.cardfail.2026.08.014 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Cardiac Failure, 2026-11. Verfasst von Elena Vasti und 15 weiteren Autor:innen. Beteiligte Einrichtung: Division of Cardiovascular Medicine, Stanford Health Care, Stanford, California. Electronic address: ecvasti@stanford.edu. Hauptthemen laut MEDLINE: Heart Failure, Sodium-Glucose Transporter 2 Inhibitors. Die Arbeit nennt 30 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

BACKGROUND: Many patients with heart failure (HF) and a guideline indication for sodium-glucose cotransporter-2 inhibitors (SGLT2is) do not receive a prescription, but the reasons behind this are unclear. This study optimizes an large language model framework to identify reasons for the lack of SGLT2i prescriptions. METHODS AND RESULTS: We extracted clinical notes for patients >18 years old with a first HF diagnosis identified by International Classification of Diseases, 10th edition, code seen in cardiology clinics between August 10, 2021, and March 1, 2025. Five annotators labeled reasons for SGLT2i nonprescription (the test set); the remaining unlabeled notes served as the inference set. Using both zero-shot (prompts do not include examples) and few-shot prompting techniques, we compared 2 large language models from OpenAI on their ability to classify the documented reason for nonprescription into 1 of 6 themes and 1 of 14 subthemes. We identified 599 HF patients (with 1123 notes), with an average age of 71 years and predominantly having HF with preserved ejection fraction. We used 313 labeled and 705 unlabeled notes as the test and inference sets, respectively, for a total set of 1018 notes. GPT-o1 outperformed GPT-4o and was 84.7% accurate (95% confidence interval 80.2%-88.5%) in identifying theme reasons for SGLT2i nonprescription. The most frequent theme was physician-attributed factors (n = 292 notes [28.7%]), among the total set. Physician deferral to a future visit (n = 157 notes [15.4%]) and the physician perceiving the patient to be at low risk or unlikely to benefit (n = 75 [7.4%]) were the most prevalent subthemes. CONCLUSIONS: We optimized a large language model to accurately discriminate reasons for not prescribing SGLT2is from clinical notes. When a reason for SGLT2i nonprescription was documented, delayed initiation of SGLT2i and physician perception of patients as being at low risk were the most common reasons, offering valuable insight to improve guideline-indicated SGLT2i prescription.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Heart FailureSodium-Glucose Transporter 2 InhibitorsHumansAgedFemaleMaleLarge Language ModelsMiddle AgedUndertreatment

Stichwörter der Autor:innen

Heart failure, sodium glucose cotransporter-2 inhibitors, prescription patternsartificial intelligence

Substanzen

Sodium-Glucose Transporter 2 Inhibitors

Autor:innen und Institutionen

  • Division of Cardiovascular Medicine, Stanford Health Care, Stanford, California. Electronic address: ecvasti@stanford.edu.
  • Division of Cardiovascular Medicine, Stanford Health Care, Stanford, California; Department of Medicine, Veterans Affairs Palo Alto Health Care System, Palo Alto, California.
  • Department of Biomedical Data Science, Stanford Health Care, Stanford, California; Department of Computer Science, Stanford Health Care, Stanford, California.
  • Division of Cardiovascular Medicine, Stanford Health Care, Stanford, California.
  • Department of Pharmacy, Stanford Health Care, Stanford, California.
  • Division of Cardiovascular Medicine, Stanford Health Care, Stanford, California; Kaiser Permanente San Francisco Medical Center, San Francisco, California; Division of Research, Kaiser Permanente Northern California, Pl…

Zitieren

APA

Vasti, E., Nguyen, M., Alsentzer, E., Chu, E., Liu, X., Chakoumakos, M., Haudenschild, C., Wang, B., Zheng, J., Kalwani, N. M., Bhatt, A. S., Ambrosy, A. P., Jain, S. S., Rodriguez, F., Heidenreich, P., & Sandhu, A. T. (2026). Leveraging Large Language Models to Understand Reasons for SGLT2 Nontreatment Among Patients With Heart Failure. Journal of Cardiac Failure, 32(11), 1902-1911. https://doi.org/10.1016/j.cardfail.2026.08.014

Vancouver

Vasti E, Nguyen M, Alsentzer E, Chu E, Liu X, Chakoumakos M, et al. Leveraging Large Language Models to Understand Reasons for SGLT2 Nontreatment Among Patients With Heart Failure. Journal of Cardiac Failure. 2026;32(11):1902-1911. doi:10.1016/j.cardfail.2026.08.014

BibTeX
@article{Vasti2026,
  author = {Vasti, Elena and Nguyen, Minh and Alsentzer, Emily and Chu, Emily and Liu, Xichong and Chakoumakos, Madie and Haudenschild, Christian and Wang, Benjamin and Zheng, Jimmy and Kalwani, Neil M and Bhatt, Ankeet S and Ambrosy, Andrew P and Jain, Sneha S and Rodriguez, Fatima and Heidenreich, Paul and Sandhu, Alexander T},
  title = {Leveraging Large Language Models to Understand Reasons for SGLT2 Nontreatment Among Patients With Heart Failure},
  journal = {Journal of Cardiac Failure},
  year = {2026},
  volume = {32},
  number = {11},
  pages = {1902--1911},
  doi = {10.1016/j.cardfail.2026.08.014}
}
Kostenlos starten →