Clinical Informatics and Artificial Intelligence: What the Working Emergency Physician Needs to Know

Area of Science:

• Clinical Informatics
• Artificial Intelligence (AI)
• Large Language Models (LLMs)

Background:

• Clinical Informatics encompasses the electronic health record (EHR) and its impact on clinical care.
• Machine learning and AI have been introduced into clinical settings, with a recent surge in Large Language Models (LLMs) for Emergency Departments (EDs).

Purpose of the Study:

• To highlight the critical role of Clinical Informaticists in the safe integration of AI and LLMs into the Emergency Department.
• To address the gap in clinical reasoning capabilities of LLMs and the need for specialized expertise in Emergency Medicine.

Main Methods:

• Review of current trends in AI and LLM adoption in clinical settings.
• Analysis of the expertise required for safe implementation of AI in the ED.
• Discussion of the challenges and importance of training Emergency Physicians (EPs) in Clinical Informatics.

Main Results:

• LLMs show high performance on some clinical measures but lack proven clinical reasoning abilities, often providing confident but false information.
• There is a significant shortage of board-certified Clinical Informaticists, particularly among Emergency Physicians (EPs).

Conclusions:

• AI systems, including LLMs, will increasingly be used in the ED, necessitating a strong understanding of their benefits and limitations by EPs.
• EPs trained in Clinical Informatics are essential for the safe integration of these technologies.
• Increased exposure to Clinical Informatics in residency and fellowship programs is vital for Emergency Medicine.