Ethical advancements in healthcare: Ensuring accurate and culturally relevant translations of kidney transplant information for the Hispanic population
Oscar Garcia Valencia1, Charat Thongprayoon1, Caroline Jadlowiec2, Shennen Mao3, Napat Leeaphorn3, Pooja Budhiraja2, Iasmina Craici1, Maria Gonzalez Suarez1, Wisit Cheungpasitporn1.
1Mayo Clinic, Rochester, MN, United States; 2Mayo Clinic, Phoenix, AZ, United States; 3Mayo Clinic, Jacksonville, FL, United States
Introduction: Achieving health equity remains a critical challenge, especially in providing essential medical information for the Spanish-speaking Hispanic community, particularly in accessing information about kidney transplantation. The crucial role of providing accurate and culturally sensitive information in bridging the informational gap cannot be overstated. For this purpose, we assess the capabilities of Chat GPT 3.5 and 4.0 in providing Spanish translations of kidney transplant FAQs for the Hispanic community. The aim is to address the language barrier in medical information, ensuring that the Spanish-speaking Hispanic population receives accurate and culturally appropriate information about kidney transplantation, thereby addressing a critical aspect of healthcare accessibility and equity.
Methods: Our study involved translating 54 questions related to kidney transplantation, sourced from the Organ Procurement and Transplantation Network (OPTN), the National Health Service (NHS), and the National Kidney Foundation (NKF). These questions, categorized into 19 from OPTN, 15 from NHS, and 20 from NKF, were translated using Chat GPT versions 3.5 and 4.0. The translations were evaluated for accuracy and cultural sensitivity specifically tailored to the Spanish-speaking Hispanic community, using a rubric scale ranging from 1 to 5. The assessments were carried out by native Spanish-speaking nephrologists to further validate the translations' precision and cultural appropriateness.
Results: Our results showed that both versions of Chat GPT exhibited high performance in translating FAQs for the Hispanic community. Chat GPT 3.5 had the highest accuracy in the NHS subgroup (4.93) and perfect cultural sensitivity in the NKF subgroup (5.0). For GPT 4.0, there was a slight improvement in accuracy in the OPTN subgroup (4.95), and it achieved perfect cultural sensitivity in both OPTN and NHS subgroups. A minor decrease in cultural sensitivity was noted for the NKF subgroup in GPT 4.0. (Fig. 1)
Conclusion: The findings indicate that both Chat GPT 3.5 and 4.0 have the potential to proficiently translate kidney transplant information from English to Spanish. This ability holds significant importance in addressing the demand for accessible and culturally pertinent medical information, particularly concerning kidney transplantation among the Hispanic population. Such English-to-Spanish translations have the potential to enhance access to crucial kidney transplant information for Spanish-speaking Hispanics, thereby potentially advancing health equity within this domain.
[1] Kidney Transplant
[2] Minorities
[3] health disparity
[4] Artificial Intelligence
[5] Accessible information
[6] Cultural sensitivity
[7] Health Equity