September 16, 2025

Investigating patients’ perceptions of ChatGPT as a health information resource: A qualitative study

Key Points

Patients reported improved health literacy and efficiency when using ChatGPT for health information, enhancing engagement with their health.
Among 28 outpatient participants in interviews, 92.9% noted increased health literacy, indicating a significant facilitator in using ChatGPT as a resource.
The qualitative design involved outpatients using ChatGPT for two weeks, followed by interviews to assess perceptions, barriers, and facilitators.
The findings suggest benefits and limitations of ChatGPT as a health resource, calling for further studies to assess its broader impact on diverse populations.

Abstract

BackgroundThe rapid adoption of artificial intelligence-powered tools like ChatGPT has introduced new avenues for patients to access health information independently. Understanding how patients perceive and engage with such tools is essential to evaluating their trustworthiness, usability, and potential impact on health decision-making.AimThe purpose of this study is to investigate the facilitators and barriers of using ChatGPT as a health information resource for patients' health management.MethodsA qualitative research design was adopted in this study. The participants included outpatients at a public hospital. Participants interacted with ChatGPT (version 3.5) for at least 15 min daily over 2 weeks to explore health-related topics before participating in semi-structured interviews. A total of 28 outpatients participated in the interviews.ResultsThe findings from this study have indicated both positive and negative aspects of ChatGPT as a health information resource. Among the 28 participants, the most frequently reported facilitators included improved health literacy (reported by 26 participants, 92.9%), effectiveness and efficiency (24 participants, 85.7%), cost-effectiveness (23 participants, 82.1%), accessibility (17 participants, 60.7%), empowerment (13 participants, 46.4%), and anonymity (11 participants, 39.3%). Reported barriers included lack of personalized information (15 participants, 53.6%), limited reliability (9 participants, 32.1%), restricted diagnostic capability (6 participants, 21.4%), lack of human interaction (14 participants, 50%), privacy concerns (4 participants, 14.3%), legal and ethical issues (9 participants, 32.1%), and lack of emotional support (3 participants, 10.7%).ConclusionAlthough ChatGPT has significant benefits of being used as a health information resource, to arrive at specific conclusions, there is a need to extend these kinds of studies across the regions to assess the impact of ChatGPT on different populations for promoting health literacy.

Mark Helpful

Bookmark

Relay