November 13, 2026 · Journal of cardiac failure · DOI: 10.1016/j.cardfail.2026.10.009

Establishing a passive registry of patients with heart failure enabled by natural language processing and language models.

Listen to this summary

Disease registries are costly to build and maintain, and key heart failure data such as ejection fraction, NYHA class, and social determinants are often stored as unstructured text in EHRs. This study asks whether natural language processing and large language models can automatically build a passive heart failure registry from a Boston tertiary care system's EHRs, linked to Medicare, Medicaid, and vital statistics data. The authors identified 69,007 patients with HF and successfully extracted EF for over half, NYHA class for 23.4%, and social history for 97.7%, suggesting the approach can support future HF research at low cost.

Gerard T Portela, Nils Krüger, Sebastian Schneeweiss, Kerry Ngan, Bowen Gu, Joyce Lii, Jie Yang, Muthiah Vaduganathan, Rishi J Desai

This is one of 33,000+ journals available on OSLR. Try it free for 14 days.

Free 14-day trial. 33,000+ journals. Cancel anytime.

14-day free trial. No commitment.

“

"Oslr has become part of my weekly routine on my day off. The clinical relevance of the summaries is outstanding — I'd rate it 9/10. Being able to consume research hands-free is a huge advantage for busy physicians."

Dr. Jennifer Thompson

Dr. Jennifer Thompson

Portland, OR

Stay current without falling behind

33,000+ journals. 3-minute audio summaries. Free for 14 days.

Download on the App StoreGet it on Google Play