Healthcare IT leaders have told Becker's that artificial intelligence will only work if it is completely integrated within the EHR; Cerner and Epic have looked to deepen their AI reach.
In 2020, Epic CEO Judy Faulkner had looked toward AI as a tool that could be used on the clinical side to alert providers about patient deterioration.
"I think where AI really works well is when there are so many things to think about that the human brain doesn't work as well as a machine," Ms. Faulkner said in a 2020 podcast. "There are many inputs – for example, the patient is deteriorating or has sepsis – that AI can alert healthcare givers to hours before they would have been able to recognize it on their own. And that's going to be saving a lot of lives – it already is."
As the hype for generative AI increased earlier this year, Epic landed a partnership with Microsoft to use its GPT-4 offerings to reduce administrative burden.
Epic's main rival, Cerner, has also looked to harness the power of AI as an administrative tool.
In April, Cerner Enviza launched a two-year partnership with John Snow Labs to use AI as a way to extract clinical notes from the EHR.
Leadership at Cerner has expressed caution on AI and urged developers to train AI on equitable datasets.
"It's a real contradiction — the results [of AI and ML] are better than humans, but it's not suitable for everyone," Cerner President and CEO David Feinberg, MD, told Becker's.
"We have to adapt to this technology capability, but there needs to be a very clear label on when not to use it. It's similar to a nutrition label we find on food products: warn me in bold letters about problems that will arise if I apply this to a certain population."
So far, both Epic and Cerner have looked to make AI-powered voice assistants available to their customers. In May, Epic connected with Suki, a company that uses AI to generate clinical notes by ambiently listening to patient and clinician conversations. Cerner also synced with Suki in August.