Adoption of Large Language Model AI Tools in Everyday Tasks: Multisite Cross-Sectional Qualitative Study of Chinese Hospital Administrators
View abstract on PubMed
Summary
This summary is machine-generated.Hospital administrators show low adoption of large language model (LLM) AI tools due to mistrust and skill gaps, despite their potential for administrative efficiency. Enhanced training and institutional support are crucial for broader integration and use.
Area Of Science
- Healthcare Administration
- Artificial Intelligence in Medicine
- Technology Adoption
Background
- Large language model (LLM) artificial intelligence (AI) tools offer significant potential to improve efficiency in healthcare administration tasks.
- However, the adoption of these advanced tools by hospital administrators is not well understood, especially at the individual level.
Purpose Of The Study
- To investigate the factors influencing the adoption and utilization of LLM AI tools among hospital administrators in China.
- To identify key enablers, barriers, and practical applications of these tools in daily administrative workflows.
Main Methods
- A multicenter, descriptive qualitative study was conducted.
- Semistructured interviews were performed with 31 hospital administrators across three tertiary hospitals.
- Thematic analysis using the Colaizzi method was employed to analyze participant experiences.
Main Results
- LLM AI tool adoption was generally low, with variations across sites.
- Technological familiarity and positive early experiences correlated with increased usage.
- Mistrust in accuracy, limited prompting skills, and inadequate training were identified as significant barriers.
Conclusions
- Facilitators for adoption include technological familiarity, positive experiences, and innovation openness.
- Barriers such as knowledge gaps and mistrust hinder widespread use.
- Enhanced usability and integration require structured training, institutional support, and trust-building strategies.