Autonomous artificial intelligence increases real-world specialist clinic productivity in a cluster-randomized trial

Area of Science:

• Ophthalmology
• Artificial Intelligence in Healthcare
• Health Services Research

Background:

• Autonomous artificial intelligence (AI) shows potential for increasing healthcare productivity.
• Real-world evidence demonstrating AI's impact on clinical efficiency is limited.
• Diabetic eye exams are crucial for preventing vision loss in diabetic patients.

Purpose of the Study:

• To test the hypothesis that an AI for diabetic eye exams increases clinic productivity.
• To evaluate the impact of AI on the number of completed care encounters per hour per specialist physician.
• To generate testable hypotheses and study designs for AI implementation in clinical settings.

Main Methods:

• A preregistered cluster-randomized clinical trial was conducted.
• 105 clinic days were randomized to either AI intervention or standard care (control).
• Clinic productivity was measured as completed care encounters per hour per physician.

Main Results:

• AI use resulted in a 40% increase in clinic productivity compared to control (1.59 vs. 1.14 encounters/hour).
• The primary endpoint was met, demonstrating statistically significant improvement (p < 0.001).
• The secondary endpoint, measuring productivity across all patients, was also met.

Conclusions:

• Autonomous AI significantly enhances productivity in diabetic eye care settings.
• Increased healthcare system productivity through AI may improve patient access to care.
• AI implementation holds potential for reducing health disparities by optimizing resource allocation.