ENHANCING HEALTH EQUITY: CHATGPT 4.0 VS 3.5'S ROLE IN SIMPLIFYING LIVING KIDNEY DONATION INFORMATION

Back

E-Poster

https://storage.unitedwebnetwork.com/files/1099/7dd5b088e88f86baee58489fcf9e5b7d.pdf

Abstract Title

First Name *

Oscar

Last Name *

Garcia Valencia

Co-author 1

Iasmina Craici Craici.Iasmina@mayo.edu Mayo Clinic Division of Nephrology and Hypertension Rochester

Co-author 2

Jing Miao Miao.Jing@mayo.edu Mayo Clinic Division of Nephrology and Hypertension Rochester

Co-author 3

Carrie Jadlowiec Jadlowiec.Caroline@mayo.edu Mayo Clinic Division of Transplant Surgery Phoenix

Co-author 4

Shennen Mao Mao.Shennen@mayo.edu Mayo Clinic Division of Transplant Surgery Jacksonville

Co-author 5

Michael Mao Mao.Michael@mayo.edu Mayo Clinic Division of Nephrology and Hypertension Jacksonville

Co-author 6

Napat Leeaphorn Leeaphorn.Napat@mayo.edu Mayo Clinic Division of Nephrology and Hypertension Jacksonville

Co-author 7

Pooja Budhiraja Budhiraja.Pooja@mayo.edu Mayo Clinic Division of Nephrology and Hypertension Phoenix

Co-author 8

Charat Thongprayoon Thongprayoon.Charat@mayo.edu Mayo Clinic Division of Nephrology and Hypertension Rochester

Co-author 9

Wisit Cheungpasitporn Cheungpasitporn.Wisit@mayo.edu Mayo Clinic Division of Nephrology and Hypertension Rochester

Co-author 10

Co-author 11

Co-author 12

Co-author 13

Co-author 14

Co-author 15

Introduction

Addressing disparities in living kidney donation calls for making information comprehensible across literacy levels. With the average American adult reading at an 8th-grade level, simplifying complex medical information is imperative. This study evaluated the utility of ChatGPT, an advanced AI language model, in modifying text about living kidney donation to make it accessible at an 8th-grade reading level or below.

Methods

We selected 27 questions and answers from Donate Life America, a widely-used resource on living kidney donation. These texts were modified using two versions of ChatGPT to enhance readability for those with literacy levels at or below the 8th grade. The Flesch-Kincaid formula was employed to gauge the readability of both the original and modified texts. A paired t-test assessed the statistical significance of any changes in readability grade levels. A statistical comparison between Chat GPT 3.5 and Chat GPT 4.0 was conducted examining the mean difference (MD) with standard deviation (SD).

Results

The original FAQs had an average reading level of 9.6±1.9. After modification, Chat GPT 3.5 and Chat GPT 4.0 achieved average readability levels of 7.72±1.85 and 4.30±1.71, respectively, with a p-value <0.001, signifying a substantial improvement. Chat GPT 3.5 rendered 59.26% of the answers readable below the 8th-grade level, whereas Chat GPT 4.0 achieved this in 96.30% of cases. The grade level of the modified answers by Chat GPT 3.5 and Chat GPT 4.0 ranged from 3.4 to 11.3 and 1 to 8.1, respectively with MD of 3.43 (SD = 1.62, p-value <0.001).

Conclusions

Both Chat GPT 3.5 and Chat GPT 4.0 effectively reduced the readability grade levels of complex medical information. However, Chat GPT 4.0 was demonstrably more effective, indicating its promising potential for enhancing the accessibility of medical information. This performance suggests that ChatGPT could play a contributory role in advancing diversity and equity in living kidney donation. Although our findings confirm the ability of ChatGPT to significantly bring down readability levels, there exists scope for further refinement.

E-Poster Format Requirements

PDF file
Layout: Portrait (vertical orientation)
One page only (Dim A4: 210 x 297mm or PPT)
E-Poster can be prepared in PowerPoint (one (1) PowerPoint slide) but must be saved and submitted as PDF file.
File Size: Maximum file size is 2 Megabytes (2 MB)
No hyperlinks, animated images, animations, and slide transitions
Language: English
Include your abstract number
E-posters can include QR codes, tables and photos