Human judgment: The magic ingredient for making AI work across aging services disciplines

https://www.mcknights.com/news/human-judgment-the-magic-ingredient-for-making-ai-work-across-aging-services-disciplines/

From tackling a scabies outbreak in a nursing home to scheduling shifts and helping write plans of correction that satisfy regulators, a range of aging services providers are eagerly already putting artificial intelligence to work.

“Machines can inform us. They can help us be more efficient. They certainly can help us be more effective,” said Lori Strubbe, CEO of Focused Post Acute Care Partners, a chain of mostly rural Texas nursing homes.

But, she said, AI can’t replace human rapport or value judgments, the wisdom that comes from lived experience, or discernment in a crisis.

“AI doesn’t bring the empathy necessary to have the conversation with a scared family. It doesn’t notify us when a resident is failing and how we have to hold their hand,” she added. “Only human discernment can bring that clarity when the stakes are highest.”

Strubbe said she shares that message with her staff members regularly, noting that some of them are afraid that AI may take their jobs. But, she and other panelists emphasized, AI truly should be used to make caregivers more efficient with everyday tasks and give them more time to focus on the actual job of delivering care and services.

Strubbe said she and her team are regularly using AI to help write plans of correction after facility inspections and for other repetitive rules-based tasks, especially those that require writing. The chain recently started using an AI-based tool to speed the admission process across its footprint, and it is now using it to review staff member phone calls and suggest customer service improvements.

Work flow improvements

The ability to process data quickly is the key component that nursing management expert Amy Stewart said she has come to love.

The chief nursing officer for the American Association of Post-Acute Care Nurses, Stewart echoed Strubbe’s belief that AI helps nurses with rote tasks and data interpretation, but not with critical judgments about care.

Stewart praised the ability for AI to process survey and nursing home quality improvement information and determine where efforts should be focused, rather than on the falls and pressure ulcer categories that everyone falls back on.

When working with care teams, she said, she also praises its ability to interpret notes and data in the background and offer early predictive warnings in between formal cognition or depression assessments.

“If I knew I could stop someone from having this big depressive episode by intervening early, or slowing cognition issues by giving them activities that are more stimulating, that’s going to be huge as we move forward,” Stewart said.

That ability to process data without hands-on involvement is a win-win across settings, including in home health, she said.

“It will lead to higher satisfaction for families and nurses,” Stewart noted.

Integration and usefulness — for resident/patient/client and healthcare professional alike — has to be the focus of any intervention tool, pointed out Patrick Tarnowski, the chief commercial officer for OneStep. His company has developed a product that can assess someone with a phone-based app using short videos of the person’s gait.

He added that any AI strategy has to have clinical, economic and operating values, “meaning it fits into, complements and enables operating strategy.” Otherwise, Tarnowski added, any well-intentioned app or tool simply may be another burden for caregivers or butt up against limits imposed by setting or user capability.

But intuitive tools that can work across the senior living and care continuum are a “real win,” he said.

Increasing therapy, clinical uses

Renee Kinder, executive vice president of clinical services for Broad River Rehab, said she first appreciated AI when she saw it being used for in-house medical record reviews to ensure compliance for clients.

Now, it is beginning to help therapists write treatment plans.

The therapy sector, she noted, has worked with the American Medical Association to create multiple billing codes that allow therapists to use the technology to varying degrees to look at resident/patient data and develop clinical plans and intended outcomes.

“How fortunate are we to be in this time?” Kinder wondered aloud, calling AI evolution the follow-up to telehealth and remote monitoring advances. “But when we’re given these allowances, we have to use them in a responsible way.”

Like others, she insisted that human involvement and intelligence remains critical in understanding, questioning and refining whatever recommendations AI cranks out.

Seeing is not believing

Geriatrician and Post-Acute and Long-Term Care Medical Association Board President Sabine von Preyss-Friedman, MD, picked up on that concern, acknowledging that AI still “completely hallucinates. It makes things up.”

But, said the chief medical director at Avalon Healthcare and Caldera Care, awareness of that reality and a willingness to fact-check and dig deeper into sources still can get healthcare professionals the answers they need, more quickly.

von Preyss-Friedman said she has used AI for diagnostic assistance in the case of someone who was on a ventilator and had a very complex health crisis, with the tool leading her to the right answer. She also recently used Open Evidence, an AI-powered platform providing peer-reviewed medical information, to learn more about scabies during an outbreak at one of the facilities she oversees.

She said that, with the tool, she was able to quickly come up to speed on treatment and containment strategies, then took what she found, dropped in a second AI-powered tool and customized a treatment protocol for her staff members to use.

“I’ve always done that kind of thing, but it used to take me hours,” von Preyss-Friedman said.

Next, she said, she’s looking forward to more reliable electronic health records system query tools. They allow providers to quickly scour hundreds or thousands of pages in someone’s chart to find details. In the case of someone with diabetes, for instance, she might use an AI search function to learn what medication a hospital used to treat him two years ago or when he had his last vision exam.

All of those tools are adding up to clinical teams that are more empowered, better prepared and more responsive to the people in their care — if providers continue to bring their “A” game when it comes to interpretation and fact-checking.

And one other positive noted by Stewart: AI is “buzzy” enough that the right investments could net senior living and care providers new staff members, “because, finally, we have something the hospital doesn’t have.”

The session was moderated by McKnight’s Long-Term Care News Executive Editor James M. Berklan as part of the 2025 McKnight’s Tech Awards + Summit. The full webinar and other webinars that were part of the event are available to watch for free on demand. Additional coverage of the McKnight’s Tech Awards + Summit will be featured online, in print and on social media by McKnight’s Long-Term Care News, McKnight’s Senior Living and McKnight’s Home Care.

Read Full Message