Generative AI in Dutch General Practice: A Pilot Study of Current Use and Future Perspectives.

Area of Science:

• Primary Care Medicine
• Health Informatics
• Artificial Intelligence in Healthcare

Background:

• Dutch general practice faces significant challenges due to workforce shortages and administrative burdens.
• Generative artificial intelligence (GenAI) is proposed as a solution for documentation, communication, and clinical reasoning support in primary care.
• The real-world application and evaluation of GenAI in primary care settings are not well understood.

Purpose of the Study:

• To explore current GenAI usage among Dutch general practitioners (GPs).
• To identify perceived benefits and barriers associated with GenAI implementation in general practice.
• To determine necessary conditions for the responsible adoption of GenAI in primary care.

Main Methods:

• An exploratory pilot study involving 13 Dutch GPs.
• Semi-structured interviews were conducted until thematic saturation was achieved.
• Reflexive thematic analysis was used to identify key themes, supplemented by manual sentiment scoring of coded excerpts.

Main Results:

• Five major themes emerged: GenAI as support without replacing clinical judgment, the necessity of EHR and workflow integration, significant barriers related to privacy/governance/liability, tensions with professional identity, and trust linked to transparency.
• GPs perceive GenAI as potentially valuable when integrated and governed properly.
• Clear conditions for safe and acceptable GenAI use were articulated by participants.

Conclusions:

• GenAI holds potential value for supporting Dutch general practice, provided it is well-integrated and governed.
• Addressing concerns regarding privacy, governance, liability, professional identity, and trust is crucial for successful implementation.
• Findings are context-specific and hypothesis-generating, requiring further research for generalizability.