Expanding Hispanic/Latinx Participation in Digital Health Research Through Large Language Models

Advances in digital health exclude Hispanic/Latinx populations with limited English language proficiency, despite their high cardiovascular risk and smartphone usage. Large language models (LLMs) offer promising English-to-Spanish translational solutions, with comparable accuracy to professional medical translators. In this work, we present data and propose a hybrid workflow that combines automated LLM translation with professional review to reduce costs and improve Hispanic/Latinx inclusion in cardiovascular digital health research.