Performance of large language models in delivering accurate and comprehensible patient information on heart failure and cardiomyopathy - Takeaways - MDSpire

Performance of large language models in delivering accurate and comprehensible patient information on heart failure and cardiomyopathy

  • By

  • Christoph Reich

  • Jule Leverenz

  • Charlotte Brand

  • Lasse Niemeier

  • Isabel Branzei

  • Mustafa Yildirim

  • Farbod Sedaghat-Hamedani

  • Ali Amr

  • Norbert Frey

  • Benjamin Meder

  • June 9, 2026

  • 0 min

Share

  • 1

    This study evaluated six large language models (LLMs) for their accuracy and comprehensibility in providing patient information on heart failure and cardiomyopathy.

  • 2

    Gemini received the highest composite mean rating for readability and factual reliability among the evaluated LLMs, followed by Grok.

  • 3

    The evaluation involved 50 expert-curated questions and responses rated by twelve reviewers across nine domains, including appropriateness and empathy.

  • 4

    All LLMs demonstrated good accuracy in avoiding medical misinformation, though variability existed in readability and comprehensiveness.

  • 5

    The study highlights the need for rigorous evaluation of LLMs to ensure their reliability and accessibility for patient education in chronic disease management.

Original Source(s)

Performance of large language models in delivering accurate and comprehensible patient information on heart failure and cardiomyopathy

  • by Christoph Reich, Jule Leverenz, Charlotte Brand, Lasse Niemeier, Isabel Branzei, Mustafa Yildirim, Farbod Sedaghat-Hamedani, Ali Amr, Norbert Frey, Benjamin Meder

  • June 9, 2026

Related Content

Latent transition analysis of home-based fluid management during the vulnerable phase in patients with chronic heart failure: impact on symptom burden

“Double-hit” precipitates fulminant cardiac dysfunction in a child with homozygous CAP2 variant: a case report

ChatGPT response consistency to the 2025 ESC/EACTS guidelines for the management of valvular heart disease: A test–retest study using binary and multiple-choice questions