Utilization of Generative AI-drafted Responses for Managing Patient-Provider Communication.

Area of Science:

• Health Informatics
• Artificial Intelligence in Healthcare
• Clinical Communication

Background:

• Generative AI (GenAI) offers potential benefits and challenges for patient communication.
• Integrating AI into clinical workflows requires careful consideration of user adoption and efficiency.

Purpose of the Study:

• To assess the utilization of AI-generated drafts for patient messages by healthcare professionals (HCPs).
• To analyze the impact of GenAI on message turnaround time and workflow efficiency.
• To identify HCP preferences for AI-generated draft content and style.

Main Methods:

• Retrospective observational study analyzing Electronic Health Record (EHR) audit logs.
• Data collected from 75 healthcare professionals at a large health system in New York City.
• Analysis included utilization rates, prompt refinement impact, text analysis of drafts, and message turnaround times.

Main Results:

• Overall GenAI utilization was low at 19.4%, but improved with prompt refinements (12% to 20%).
• Physicians showed higher utilization compared to other roles.
• AI drafts reduced message turnaround time by 6.76% but slightly increased InBasket actions.
• HCPs preferred concise, information-rich drafts, with role-specific preferences for length and empathy.

Conclusions:

• Targeted deployment strategies are needed for effective GenAI integration into clinical workflows.
• Optimizing AI draft generation based on user roles and preferences can enhance efficiency.
• Further research is required to overcome barriers to GenAI adoption in patient communication.