Using AI to Address Nursing’s Biggest Pain Points

McManus said San Diego-based Sharp HealthCare wants to “empower the nursing workforce with mobile-based technology so that they can document more at ease on the fly.”

“The more that we can reduce some of the information worker burden that is attached to the modern nursing discipline is a key step toward coming back to top of licensure and why someone wanted to become a nurse in the first place,” he noted.

Rocchio, part of a 50-hospital network spread across four states and including one completely virtual hospital, pointed out that nursing hasn’t seen a lot of innovation since Ida Jean Orlando-Pelletier’s Deliberative Nursing Process Theory was introduce in 1958.

“We’ve added the technology, we’ve added the electronic health record, we’ve added disparate technologies into the workflows of nurses, but what we haven’t done is [understand] how we could literally integrate all of this to clean up the work environment and workflows,” she said.

Nurses spend the most time in the EHR, she added, “and we’re left at the bedside to have to go through all those amounts of information to get to what we need to make good decisions.”

That includes patient handoffs. Rocchio said Mercy targeted that pain point with an AI tool, aiming to improve an often confusing process that accounts for roughly 80% of documentation errors. ED nurses often have to pull together disparate data, such as physician orders and EHR information, then pick up the phone and relay that information to the next member of the patient’s care team.

“We didn’t anticipate that unless the doctor signs their note in the emergency department, the items that we were pulling from that field weren’t always coming across correctly,” Rocchio said.

Aside from AI tools to improve interactions with the EHR and give nurses the patient data they need at the bedside, nursing leaders said they’d like to see the technology applied to the complex process of scheduling nurse shifts. A key factor in improving nurse workflows is giving them shifts that fit their work-life routine and aren’t overwhelming. That includes factoring in vacations and time off requests and breaks.

“There’s a lot of opportunity there to support improvements,” McManus pointed out.

McManus, who’s a member of the non-profit Coalition for Health AI (CHAI), which aims to establish standards for the safe and ethical use of AI in healthcare, said AI needs to be carefully planned and tested be health system leadership.

“You really need to practice micro to macro,” he said. “You have to do pilot-based work. You have to do integrated feedback loops with your care teams. You have to kind of grow these disciplines.”

“If you cannot ground in fact-based work, you’re not going to be able to build any type of trust in those workflows,” he said.

And as with any innovative idea, nurses should be included in planning and governance. Nurses need to be included in the deign of any new process or tool that they’ll be using, and they need to understand how to spot problems and react quickly to correct them.

“I’m not sitting in a boardroom deciding what this looks like for our nurses,” Rocchio noted. “I have an entire nursing informatics team that’s on the ground, at the elbow, with our nurses, developing.”

“Having our senior nurse leadership engaged in the topic, in the oversight, in the strategy development … is paramount,” McManus added.

He added that AI literacy is taught throughout the health system.

“We’re developing these foundational skills so that when we introduce specific features, we’re supporting the readiness to use, the readiness to understand and the expectation of transparency and feedback loops,” he said.

Addressing a common concern in healthcare that AI might replace jobs, both McManus and Rocchio pointed out that AI, to them, stands for augmented intelligence, and that the technology is a tool to be used by and for care providers, not instead of them.

“AI will never replace critical judgment,” Rocchio noted.