CAN ARTIFICIAL INTELLIGENCE (CHATGPT) REPLACE HEALTHCARE PROVIDERS' WORKLOAD IN MANAGING TYPE 2 DIABETES?
Keywords:
CHATGPT, TYPE 2 DIABETES, ARTIFICIAL INTELLIGENCEAbstract
INTRODUCTION
ChatGPT is a large language model developed by OpenAI. It provides human-like responses to questions and solves cognitive problems by reinforcing learning via data from online human responses. Diabetes is a chronic illness and information found online can be misleading. This study aims to explore potentials of ChatGPT in providing accurate and patient-friendly diabetes knowledge.
METHODOLOGY
Open-ended questions pertaining to diabetes pathophysiology, complications, treatment and cure were input into ChatGPT via chat.openai.com platform. Responses were recorded and performance of ChatGPT was graded by two endocrinologists to assess the accuracy of the information.
RESULT
Overall, artificial intelligence performed well in delivering facts regarding general diabetes knowledge. AI explained in detail the pathophysiology of diabetes, emphasizing the role of insulin in hyperglycaemia. Complications of diabetes were laid out systematically and divided according to the organs involved, though explanations were oversimplified. The approach to treatment of diabetes was clear, where AI explained lifestyle modification followed by differentiating types of medications, including insulin and oral anti-diabetic agents. Errors were detected, such as AI mislabelling SGLT2 inhibitors as injectable medications. Monitoring and follow-up were included. Responses to inquiries regarding diabetes cure were interesting, as AI emphasized the chronic nature of diabetes. It also explored pancreatic transplantation and immunotherapy in type 1 diabetes.
CONCLUSION
Artificial intelligence has come a long way to provide humanoid responses to questions. It provides accurate information on most diabetes topics with nuances of diabetes educators. There was no inclusion of misinformation rampantly found in social media and web-based platforms. In conclusion, AI can assist healthcare providers in diabetic counselling and education.
Downloads
References
*
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2023 Goh Kian Guan, Mohd Syazwan Mohd Amin, Raja Nurazni Raja Azwan
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Journal of the ASEAN Federation of Endocrine Societies is licensed under a Creative Commons Attribution-NonCommercial 4.0 International. (full license at this link: http://creativecommons.org/licenses/by-nc/3.0/legalcode).
To obtain permission to translate/reproduce or download articles or use images FOR COMMERCIAL REUSE/BUSINESS PURPOSES from the Journal of the ASEAN Federation of Endocrine Societies, kindly fill in the Permission Request for Use of Copyrighted Material and return as PDF file to jafes@asia.com or jafes.editor@gmail.com.
A written agreement shall be emailed to the requester should permission be granted.