Dive Brief:
- Ochsner Health is launching a pilot program this month that will use generative artificial intelligence to draft “simple” messages to patients.
- About a hundred clinicians across the New Orleans-based health system will participate in the first phase of the program, where AI will prepare responses to patient questions unrelated to diagnoses or clinical judgments. The messages will be reviewed and edited by providers before being sent to patients, according to a news release.
- Ochsner is part of an early adopter group of Microsoft’s Azure OpenAI Service, which integrates with the Epic electronic health record. The health system will test the messaging feature over three phases through this fall, and Ochsner will collect patient feedback to improve the system.
Dive Insight:
Several large technology companies have developed healthcare applications for generative AI, which can be used to autonomously generate content like text or images.
Microsoft and Epic, the nation’s largest health records vendor, expanded their partnership this spring to integrate the Azure OpenAI Service and begin implementing capabilities like message responses, interactive data analysis and natural language query tools.
Last month, the partners said they would accelerate their work on generative AI, working to “rapidly deploy dozens” of AI technologies in healthcare.
Other tech giants have their own healthcare offerings. Amazon has announced its own clinical documentation service and Oracle, now a competitor to Epic after its Cerner acquisition, unveiled generative AI tools that can be integrated into EHRs.
Many of the AI technologies aim to cut down on providers’ administrative tasks and health records upkeep, like responding to patient messages or taking notes on visits. Clinicians report spending hours per day on EHR-related tasks, potentially fueling burnout and cutting into time spent with patients.
Last year, more than four million medical advice requests were sent to physicians through Ochsner’s patient portal, according to the health system.
“The AI will generate a draft for the clinician to review and send. It’s meant to help clinicians respond more quickly to patient messages, so patients can get answers to their questions sooner,” Ochsner Chief Application Officer Amy Trainor said in a statement. “And it will reduce time our clinicians are spending on computers so that they can spend more time doing what they do best — direct patient care.”
Each phase of the pilot will incorporate different physicians across the system, according to an Ochsner spokesperson. The program will start with primary care and internal medicine clinicians, as they receive the most messages.
Some experts have raised concerns about the rapid deployment of generative AI in healthcare. Accuracy is a big concern, as some models have been caught making up incorrect information. Others question who would be held liable for errors, and whether AI could perpetuate biases and deepen existing healthcare inequities.
