Epic is tapping into Microsoft's AI expertise to roll out more tools to help clinicians save time and access critical data at the point of care. These "copilot" solutions aim to help with medical note summarization, offer coding suggestions and provide access to real-world evidence for patient care.
Epic and Microsoft announced the latest generative AI tools for healthcare in conjunction with the EHR giant's annual conference for healthcare and health IT leaders, called its Users Group Meeting, held at its campus in Verona, Wisconsin.
Back in April, Epic announced it was working with Microsoft to integrate large language model tools and AI into its electronic health record software. The health IT company also is collaborating with Nuance Communications, Microsoft's speech recognition subsidiary, to integrate its Dragon Ambient eXperience (DAX) Express into its EHR software. The integration of DAX Express into Epic workflows will act as a copilot for Dragon Medical users to help cut down on clinicians' administrative workloads, the companies said.
The tech giant and Epic want to bring "AI to healthcare at scale, integrating conversational, ambient and generative AI technologies across the Epic EHR ecosystem," the two companies said in a blog post.
"We are working together to rapidly deploy dozens of copilot solutions that securely unlock the potential value that the Microsoft Cloud and our AI technologies enable as health systems strive to overcome the urgent staffing, financial and clinical access challenges they face today," Eric Boyd, corporate vice president, AI platform at Microsoft, wrote in the blog post.
During the event, Epic CEO Judy Faulkner told the health IT audience that the company is trying to make work easier for clinicians harnessing the power of AI.
"We're planning to reduce the MyChart in-basket workload," Faulkner said, referring to Epic's EHR platform. "AI will write a draft of the response to the patient, the physician can review it and send it on as is or can edit it and send it on."
She added, "We're going to make charting easier. The AI will listen to the discussion and create a note in seconds. Then the doctor will review it. The AI can summarize patient information from the chart."
The new AI-powered features in the Epic EHR include note summarization to support faster documentation through suggested text and rapid review with in-context summaries and tools to reduce manual, labor-intensive processes. One solution provides medical coding staff with suggestions based on clinical documentation in the EHR to improve accuracy and streamline the entire coding and billing processes, according to the blog post.
Built on top of Microsoft's Azure OpenAI Service, Epic is now delivering generative AI exploration for an initial set of users via SlicerDicer to fill gaps in clinical evidence using real-world data and to study rare diseases, the company said.
By 2025, the U.S. Department of Health and Human Services predicts that there will be a nationwide shortage of 90,000 physicians. Additionally, 40% to 60% of clinicians report they are experiencing burnout. Epic contends that generative AI tools can help address some of healthcare's most urgent needs, from workforce burnout to staffing shortages.
According to McKinsey & Company, nearly a quarter of U.S. national health expenditure goes toward administrative costs, which could be reduced through technology.
During the event, Sumit Rana, Epic's senior vice president, referred to generative AI as a "paradigm-shifting innovation" that will require collaborations to deploy effectively in healthcare. He called out Epic's partnership with Microsoft combining "advanced engineering and cutting-edge research."
"It's going to take a village. It's going to take people building AI models, people building cloud infrastructure, building chips and building power plants," he told the audience.
Microsoft CEO Satya Nadella also joined Rana onstage to dive into the tech giant's work with Epic. He described generative AI as the result of "two breakthroughs" that happened simultaneously—natural language capabilities and a reasoning engine to gain insights from data.
"If [we] take these two things, a natural user interface as an ultimate dream machine, a reasoning engine, and then you put it together with a design aesthetic or a design pattern of a copilot, because the human is always in the loop. And, you design around the human, so this generated AI capability where anything you do starts with a draft. That, I think, is the best way to conceptualize why I fundamentally believe it's a paradigm-shifting technology, at least at the same level as graphical user interface or relational databases or even on the web," Nadella said.
The real power of the technology will be in the hands of the practitioners to help streamline workflows, gain productivity and cut down on administrative burdens, Nadella said.
On the topic of responsible AI and ethics, Rana pointed out that the generative AI tools are run in a HIPAA-compliant environment so patient data stay protected. "There is an expectation that a clinician will review everything that comes back for accuracy, so nothing is done automatically by the AI," he noted.
There are also thorny cybersecurity issues with generative AI.
"I think that's a real challenge because after all generative AI will get used by the cyber criminals and state actors," Nadella said. "The first thing that we have to then in order to defend ourselves better is to use this technology faster and better than any threats." He pointed to Microsoft's "security copilot" tool, an AI-powered security analysis tool that enables analysts to respond to threats quickly, process signals at machine speed and assess risk exposure in minutes.