At Durham, N.C.-based Duke University Health System, Chief Nursing Executive Theresa McDonnell, DNP, RN, believes healthcare’s digital transformation must begin with one guiding principle: “Technology should care for the people who care for patients.”
As health systems navigate staffing shortages and rising demand, Dr. McDonnell sees AI and virtual reality as tools to ease — not replace — the human side of care.
“There’s a strong narrative right now that AI is replacing jobs,” she said. “In healthcare, we need to be very careful not to remove the human element. Technology can augment and support decision-making, but there still needs to be that human eye for high-level processing and accuracy.”
At Duke, that belief is shaping how AI and virtual reality are being introduced to the workforce. The health system has used VR in its training environment for several years to help newer clinicians — many of whom entered the field during or after the pandemic — build confidence, de-escalate tense situations and reduce the fear that can lead to burnout.
“When we talk about burnout, really, it’s around anxiety — fear of making a wrong decision, fear of making a mistake, anxiety around not feeling fully prepared for what’s to come,” Dr. McDonnell said. “As much as we can augment learning with VR, we can help build in that confidence and decrease that anxiety.”
Beyond training, Duke is turning to AI to lighten workloads and give time back to staff. The system is using computer vision and advanced analytics to help synthesize patient charts — cutting down on the time clinicians spend combing through records to get a current picture of a patient’s condition.
“Think about a clinician doing 45 minutes to an hour of deep research into a chart to try and get a current picture of the clinical condition of the patient they’re caring for,” she said. “If you can cut that down and leverage AI to synthesize an entire chart so that you’ve got a high-level view of everything that patient has experienced, that decreases a lot of burden on people, which helps with de-escalation, which then helps with people getting back time and decreasing anxiety.”
That regained time has already been measurable. In Duke’s ambulatory settings, clinicians who once spent hours at night finishing documentation — a practice known as “pajama time” — are now logging off before they leave work.
“Where people historically would have spent three or four hours in the evening eating into their family time, their own personal time ensuring that documentation is complete, we’re seeing that now come to zero, which is huge,” she said.
The benefits extend to patients, too. Dr. McDonnell said Duke has seen higher satisfaction scores from patients who said they feel their providers are more engaged.
A key reason these tools work, she added, is they are designed by the people who use them. Duke has created integrated design teams — made up of physicians, nurse practitioners, physician assistants, nurses, nurse scientists, informatics and data integration experts — to lead its technology initiatives.
“We made sure the people closest to the work were part of the design from the beginning,” she said. “When you invest up front, you mitigate the need for rework on the back end.”
Looking ahead, Duke plans to extend this human-centered design approach to robotics and other emerging technologies through partnerships with the university’s schools of nursing, medicine, engineering and business.
“We’re now starting to see where we can come together in an integrated fashion, again with nurses, engineers, technology professionals, to design what enhanced robotics looks like in the healthcare system,” Dr. McDonnell said. “So there’s no end to where we can go with all of this.”
Her advice for other health systems looking to implement similar innovations is simple: “Include your clinicians in the discussion, in the decision-making,” she said. “When you include the people who are most affected, the outcomes are improved.”
