NASHVILLE, Tennessee—The team at Humana's CenterWell unit has its eye on future growth, and that includes digging into potential use cases for artificial intelligence.
CenterWell President Sanjay Shetty, M.D., spoke at the ViVE Conference on Monday, and he said that the team has about 20 use cases for the tech beyond the "basic ideation" phase, focused in several different areas. One is focused around growth and general efficiency, he said.
Plus, the team is looking at ways to improve clinical outcomes, particularly by making clinicians' work easier. For example, CenterWell is deploying a scaled pilot around ambient AI, he said, and, so far, it's seeing physicians self-report lower levels of cognitive strain and burnout.
Last week, there were about 90,000 visit minutes recording through the pilot, Shetty said.
"That one really has been so interesting to watch as we think about, what are we going to do to make our clinical workforce love medicine again?" he said.
In addition, CenterWell's pharmacy teams are looking at ways AI could be used to engage with patients around their prescriptions, such as doing outreach around refills or better optimizing shipping for their home prescription delivery, Shetty said.
There are also AI pilots at its home health division as well as in its customer call centers, he said. In the latter case, AI tech can offer a nudge to the call center agent to ensure they're taking the best next step for the patient. If they don't, they can then circle back to that agent and learn more about the encounter to correct for the future.
For home health, there are potential use cases across multiple areas, including in the referral and intake process to documentation and billing, Shetty said.
"We've got pilots, sort of spanning that whole continuum of how we deliver care in the home health space," he said.
In terms of determining which of those pilots are generating a return on investment, Shetty said the calculations can be complex. For one, it means linking a lot of different individuals, from the data team to clinicians to financial experts.
Shetty said Humana also seeks to test these AI programs with the same level of rigor that it would for any clinical intervention, which means building a control group and examining the effects from top to bottom.
As an example, the team rolled out a new wound care tool for its home health teams, and, while it did closely examine both the financial and clinical implications of that program, it went beyond that by mapping out the compliance implications as well as the clinical teams' satisfaction with the tech, he noted.
Having an idea of how a program like this could work at scale is critical to accessing efficacy, he said.
"If we stopped at financial or stopped at clinical, we would not have gotten that whole picture for what is going to happen with that technology when we scale it across 350 branches," Shetty said.
"Every one of these ideas starts out as being something that makes perfect sense on paper, but you have to put it to the test to prove that it's real," he said.