AI Scribes in Health Care: Balancing Transformative Potential With Responsible Integration

Area of Science:

• Medical Informatics and Artificial Intelligence
• Clinical Documentation and Workflow Optimization
• Health Care Practitioner Well-being and Burnout

Background:

• Administrative burden in clinical documentation contributes to health care practitioner burnout.
• Ambient AI scribes, particularly those using large language models (LLMs), offer a potential solution for automating clinical note generation and reducing clinician workload.
• Existing AI transcription software utilizing automated speech recognition and natural language processing forms the foundation for these emerging technologies.

Discussion:

• Studies indicate ambient AI scribes can reduce burnout, cognitive load, and documentation time, enhancing patient-physician interaction by allowing clinicians to be more present.
• Concerns persist regarding the accuracy, consistency, language, and style of AI-generated notes, necessitating diligent clinician oversight.
• User experience varies by specialty and workflow, with ethical, legal, bias, and autonomy concerns also noted.

Key Insights:

• Ambient AI scribes demonstrate potential for significant time savings and improved clinician well-being.
• Enhanced patient-physician interaction is a consistently reported benefit.
• Critical challenges include ensuring accuracy, addressing algorithmic bias, and managing ethical and practical implementation concerns.

Outlook:

• Further research is needed to evaluate the adoption and impact of AI-driven documentation technologies across diverse health professions.
• Responsible deployment is crucial to ensure AI scribes support efficient, equitable, and patient-centered care.
• JMIR Medical Informatics is launching a new section to advance the science and understanding of ambient AI scribes and AI-driven documentation technologies.