Assessment of information quality on insulin injection for type 2 diabetes patients based on DeepSeek

Read full text
Authors: RAN Shan, Yang Ying
Cited by: 0

Abstract

ObjectiveTo evaluate the quality of DeepSeek's responses to questions related to insulin injection in patients with type 2 diabetes.MethodsBased on expert consensus and clinical practice for type 2 diabetes patients, 10 frequently asked questions regarding insulin injection were selected and input into DeepSeek to obtain answers. The responses were scored based on the consensus report jointly issued by the American Diabetes Association(ADA) and the European Association for the Study of Diabetes(EASD),across three dimensions:effectiveness(1⁃4 points),safety(1⁃4 points),and practicality(1⁃4 points).ResultsThe average scores for DeepSeek's responses were 3.6 for effectiveness,4.0 for safety,and 4.0 for practicality.Incomplete information was identified in 40% of the responses,but no safety risks were found.ConclusionsDeepSeek can serve as an auxiliary tool for insulin injection education;however,its responses should be supplemented with guidance from healthcare professionals to ensure information completeness.

Fields of Study & Subjects

Artificial Intelligence in Healthcare and Education Diabetes Management and Education Diabetes Treatment and Management

Publication details

Journal / source
DOAJ (DOAJ: Directory of Open Access Journals)
Publication date
1 January 2026
Language
ZH