Skip to main content
2025 ASCRS Annual Scientific Meeting Main banner
Final Program


Visit ASCRS U
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
Presentation Icons
CME
CME
Invitation/Ticketed Event
Invitation/Ticketed Event
Invited Lectureship
Invited Lectureship
Award Recipient
Award Recipient
Self-Assessment
Self-Assessment
Social
Social
Industry
Industry
Back

Session: ePoster: Colon and Colorectal Cancer 14 (Mon1015)

EP717 - Using ChatGPT-4 to Improve Online Organizational Health Literacy for Spanish-speaking Patients with Colorectal Cancer

Monday, May 12, 2025
10:53 AM – 10:58 AM PT

    ePoster Presenter(s)

  • Munir H. Buhaya, MD (he/him/his)

    Resident
    University of Texas Southwestern
    University of Texas Southwestern
    Dallas, TX, United States

    All of the relevant financial relationships listed below have been mitigated.
    This individual has no financial relationships with ineligible companies.

    • Author(s)

  • AA

    Andres Abreu, MD

    Resident
    UT Southwestern
    Dallas, Texas, United States

    Disclosure information not submitted.

  • AB

    Amanda Benavides, n/a

    Medical Student
    UT Southwestern
    Dallas, Texas, United States

    Disclosure information not submitted.

  • HL

    Haley Loan, n/a

    Medical Student
    UT Southwestern
    Dallas, Texas, United States

    Disclosure information not submitted.

  • MP

    Morgan F. Pettigrew, MD

    Resident
    UT Southwestern
    Dallas, Texas, United States

    Disclosure information not submitted.

  • SC

    Samy Castillo, MD

    Research Fellow
    UT Southwestern
    Dallas, Texas, United States

    Disclosure information not submitted.

  • JS

    Javier Salgado Pogacnik, MD

    Faculty
    UT Southwestern
    Dallas, Texas, United States

    Disclosure information not submitted.

  • Emina Huang, MD

    Professor
    Surgery
    University of Texas
    Dallas, Texas, United States

    All of the relevant financial relationships listed below have been mitigated.
    This individual has no financial relationships with ineligible companies.

    • Senior Author(s)

  • PP

    Patricio Marcelo Polanco, MD

    Faculty
    UT Southwestern
    Dallas, Texas, United States

    All of the relevant financial relationships listed below have been mitigated.
    This individual has no financial relationships with ineligible companies.

All of the relevant financial relationships listed below have been mitigated.
Munir H. Buhaya, MD: This individual has no financial relationships with ineligible companies.
Emina Huang, MD: This individual has no financial relationships with ineligible companies.
Patricio Marcelo Polanco, MD: This individual has no financial relationships with ineligible companies.
Image or Table (Do Not add text here):

Purpose/Background: Hispanics, representing 20% of the U.S. population, face unique colorectal cancer (CRC) disparities, including rising incidence and mortality in younger patients. Language barriers and low health literacy worsen these outcomes. Most patient-facing, cancer education materials from healthcare organizations exceed recommended readability levels (8th grade according to the National Institutes of Health), and Spanish-language content is scant. This study aims to use Chat GPT-4 to translate and improve readability of CRC online content for Spanish-speaking patients from the National Comprehensive Cancer Network (NCCN) institutions’ websites, addressing both language and literacy gaps.

Methods/Interventions: In March 2024, 34 NCCN institution websites were reviewed for CRC content in English. ChatGPT-4 was used to translate to Spanish and improve readability. Readability was analyzed using validated formulas for both English (Fry and SMOG) and Spanish languages (GPM and SOL). The readability scores of ChatGPT-4 output were compared to the original English source and to a Spanish translation generated with Google Translate. Student’s t test was performed, and statistical significance was defined as p < 0.05.

Results/Outcomes: All 34 NCCN institutions' websites provided English-language colorectal cancer (CRC) information. Only four (12%) had Spanish language content. The mean readability scores for the English language content were 14 and 13.4 using the Fry and SMOG formulas, respectively. When this content was translated to Spanish and rewritten using an ChatGPT-4, readability improved significantly, with scores of 10.4 and 10.9 according to GPM and SOL formulas, respectively. In comparison, the Google Translated text obtained a score of 14.2 (GPM) and 13.2 (SOL). A paired T-test showed significant readability improvements with the ChatGPT-4 output (p < 0.001) but not with Google Translate (p = 0.24).

Conclusion/Discussion: Although ChatGPT-4 did not achieve the NIH-recommended 8th-grade reading level, it significantly improved the readability of Spanish-translated CRC content. Healthy People 2030 emphasizes that medical organizations must provide health information that is easy to understand, yet many cancer centers fail to meet these standards, particularly for Spanish-speaking patients. We believe that AI technologies like ChatGPT4 can be leveraged to help reduce the organizational health literacy gap that exists in the United States for Spanish-speaking patients. The next step in this research is to assess the quality of the AI-generated text to ensure it maintains accuracy and clarity.