Exploring the Potential of Ambient AI for Inpatient Documentation: A Qualitative Study with Junior Doctors

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Clinical Documentation Improvement

Background:

• Clinical documentation is vital but burdensome for physicians, especially trainees.
• Ambient AI scribes offer potential solutions by automating note generation.
• Inpatient ward round application of AI scribes remains under-explored.

Purpose of the Study:

• To explore early-career doctors' perceptions of ambient AI scribes in simulated inpatient orthopaedic ward rounds.
• To understand anticipated individual, team, and system-level work required for AI implementation.
• To assess perceived benefits, risks, and practical challenges of using AI scribes in this setting.

Main Methods:

• Semi-structured interviews with ten postgraduate year one doctors after simulated ward rounds.
• Use of Normalisation Process Theory (NPT) to guide the interview topic guide.
• Reflexive thematic analysis with a hybrid inductive-deductive approach for data analysis.

Main Results:

• Seven themes emerged across NPT constructs, highlighting understanding of AI's potential to reduce burden and improve records (Coherence).
• Participants recognized the need for individual engagement and AI output review (Cognitive Participation).
• Practical implementation challenges included workflow adaptation, consent, infrastructure, and governance (Collective Action & Reflexive Monitoring).

Conclusions:

• Ambient AI scribes are perceived as valuable for inpatient documentation, potentially enhancing efficiency and communication.
• Successful implementation requires significant individual, team, and organizational adaptation.
• Addressing workflow, consent, privacy, and infrastructure are critical for real-world adoption.