Pittsburgh-based Highmark Health is employing generative artificial intelligence from Google to create marketing materials and document encounters on both its provider and payer sides.
Highmark Health, the parent company of payer Highmark and Pittsburgh-based Allegheny Health Network, is using the technology to create images for early marketing storyboards via text commands and to generate audio for marketing content, said Richard Clarke, PhD, the health system's chief analytics officer. The system also plans to tap generative AI to reduce documentation burden by summarizing Allegheny patient visits and Highmark customer service calls.
"We fundamentally believe this is something that can transform the way that we do business internally, and frankly, the way we interface with our patients and our members," Dr. Clarke told Becker's. "Right now, (generative AI) is a little hyped. You can't go two minutes without hearing someone ask about it or talk about it. But frankly, we're utilizing that excitement to enable good positive dialogue around how we can transform processes, because people are so interested."
The work expands on Highmark Health's Living Health program, a data analytics platform started in partnership with Google in 2020.
The health system is also using generative AI to help its software developers complete code working internally and with Google to reduce any inherent bias from the models.
In addition, Highmark Health is adopting Google's Vertex AI Search to find clinical insights in unstructured data such as patient notes and scanned documents and testing its Med-PaLM 2 healthcare large language model.
"There couldn't be a more interesting time to be in like data and analytics and healthcare," Dr. Clarke said.
He said the technology will also help with Highmark Health's shift to value-based care. Providers might one day interact with a value-based reimbursement program by typing questions like "What are my largest opportunities for improvement?" or "Which are the patients that I really need to focus on?" and getting data-backed responses instantaneously.
"One of our goals is to bring actual insights to providers inside of that value-based care structure in the easiest way possible, to just try to reduce the barrier for action adoption as low as possible. And I think gen AI is a great way to help us achieve that," he said.