Navigating AI in clinical practice: A practical guide for physicians

Area of Science:

• Clinical Informatics
• Medical Ethics
• Health AI

Background:

• Artificial intelligence (AI) is increasingly integrated into clinical settings.
• Physicians, particularly in smaller practices, often lack adequate guidance for safe AI tool utilization.
• Clinicians bear full accountability for AI-assisted care outcomes, as medical boards regulate physicians, not AI developers.

Purpose of the Study:

• To outline the risks associated with AI in clinical practice, including opaque algorithms, hallucinations, omissions, and performance inequities.
• To clarify physicians' professional responsibilities concerning AI, referencing the Federation of State Medical Boards' 2024 guidance.
• To provide a practical toolkit for evaluating AI tools and ensuring responsible physician engagement.

Main Methods:

• Review of current research on AI in healthcare.
• Analysis of the Federation of State Medical Boards' 2024 guidance on AI.
• Development of a practical toolkit for physician evaluation and implementation of AI tools.

Main Results:

• Identified key risks: opaque algorithms, AI hallucinations, data omissions, and biased model performance.
• Clarified physician responsibilities: competence, documentation, informed consent, privacy, and bias mitigation.
• Provided actionable steps for AI tool evaluation, verification, performance monitoring, and patient transparency.

Conclusions:

• Responsible physician engagement with AI is crucial for harnessing its benefits.
• Physician accountability remains paramount in AI-assisted care.
• Prioritizing patient safety, professional integrity, and equitable care is essential when implementing AI tools.