Crossover Evaluation of Two Ambient AI Scribe Tools in the Emergency Department

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Clinical Documentation

Background:

• Ambient AI scribes are emerging tools to alleviate documentation burden in healthcare.
• Emergency departments (EDs) face significant documentation demands, impacting clinician workflow and satisfaction.

Purpose of the Study:

• To compare Dragon Ambient eXperience (DAX) and Abridge, two ambient AI documentation tools, within an ED setting.
• To assess their impact on work burden, usability, documentation quality, satisfaction, and user preference.

Main Methods:

• A 6-week prospective crossover study involving 18 faculty in an ED.
• Participants used DAX and Abridge in alternating 3-week phases, with pre-, during, and post-surveys.
• Outcome measures included perceived burden, usability (SUS, UMUX-Lite), documentation quality (PDQI-9), and satisfaction.

Main Results:

• DAX demonstrated a significantly greater reduction in overall work burden compared to Abridge (p=0.025).
• Both tools exhibited high and comparable usability (SUS/UMUX-Lite scores).
• DAX showed a trend towards better documentation quality (PDQI-9, p=0.011) and higher satisfaction (p=0.015), though order effects influenced preferences.

Conclusions:

• Both ambient AI scribes are feasible and perceived to reduce documentation burden in the ED while maintaining note quality.
• Findings support the value of ambient AI in clinical settings and suggest the need for larger, multi-site studies.
• Further research with objective outcomes is warranted to fully evaluate these technologies.