Download Full Text (329 KB)




This study addresses the critical need for accessible clinical screening in communities with a high incidence of Patients with Limited English Proficiency (PLEP). Recognizing the limitations of existing interpreter services and the scarcity of validated translations for standard clinical surveys like PHQ-9 and GAD-7, we developed a novel approach leveraging Large Language Models (LLMs). Our method utilizes GPT-4 to create bilingual versions of these surveys, which are then formatted into printable PDFs via a Python script and LuaLaTeX compiler. The resulting surveys, validated for translation accuracy and cultural competency, are made accessible through a Google repository. Preliminary results demonstrate that GPT-4 can consistently produce high-quality, culturally sensitive translations in various languages, including Spanish, Arabic, Nepali, and Somali. This innovative approach not only improves the accessibility of clinical screening tools but also enhances the efficiency of medical practice, especially in settings with diverse linguistic needs. Future directions include professional validation of the bilingual surveys and expansion of the repository to encompass a wider range of languages and forms. This study highlights the potential of LLMs in bridging language barriers in healthcare, offering a scalable solution to improve healthcare outcomes for PLEP.

Clinical Site

Community Health Centers of Burlington, Southend


Vermont, Language Accessibility, Translation, Limited English Proficiency, Artificial Intelligence, Screening

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.


Medical Education | Primary Care | Systems and Integrative Engineering

Bridging Language Barriers In Clinical Screening: Leveraging Large Language Models (LLMs) to Generate Bilingual Screening Surveys for Patients with Limited English Proficiency (PLEP)