Uptake of Generative AI Integrated With Electronic Health Records in US Hospitals

Area of Science:

• Healthcare Informatics
• Artificial Intelligence in Medicine
• Health Information Technology

Background:

• Generative artificial intelligence (AI) is generating enthusiasm in healthcare, but systematic evidence on its implementation within healthcare organizations is lacking.
• Understanding the adoption patterns of generative AI integrated with electronic health records (EHRs) is crucial for strategic planning and resource allocation.

Purpose of the Study:

• To describe the adoption of generative AI integrated with EHRs by US nonfederal acute care hospitals.
• To examine the relationship between generative AI adoption and prior experience with predictive AI.
• To identify hospital characteristics associated with generative AI adoption.

Main Methods:

• A survey study was conducted using the 2024 American Hospital Association (AHA) Information Technology (IT) Supplement.
• Data were collected from individuals knowledgeable about health IT at 2174 nonfederal acute care US hospitals.
• Adoption was categorized as early, fast follower, or delayed, and analyzed in relation to predictive AI experience, EHR developer, and hospital characteristics.

Main Results:

• In 2024, 31.5% of hospitals were early adopters of generative AI, 24.7% were fast followers, and 43.7% were delayed adopters.
• Hospitals with experience using predictive AI were more likely to be early or fast followers (26.2 percentage points difference).
• Epic EHR users were more likely to be early or fast followers compared to other EHR users (21.9 percentage points difference).

Conclusions:

• More than half of US hospitals anticipate implementing generative AI by the end of 2025.
• Support is needed to facilitate the adoption of beneficial generative AI in healthcare.
• Developing and disseminating best practices for evaluating generative AI is essential for responsible implementation.