Kim-Anh-Nhi (Nhi) Nyugen (left) and Maria “Vickee” Sevillano
In 2022, Maria “Vickee” Sevillano, BSN, RN, CWCN, COCN, Wound Care Specialist at The Mount Sinai Hospital, attended a virtual informational discussion on artificial intelligence (AI) that clicked like a light bulb in her mind. “Even before attending that lecture I had been wondering if we could create a machine-learning application for pressure injury prevention. I was aware that big strides had been made using AI to interpret radiologic images. But could we use AI to identify patients at risk for pressure injuries? Most likely.” So Vickee reached out to Robbie Freeman, DNP, RN, Chief Digital Transformation Officer at the Mount Sinai Health System, and asked the question.
Vickee quickly became deeply involved in exploring how AI could be used in wound care as a member of a workgroup with Kim-Anh-Nhi (Nhi) Nyugen, MSc, Senior Clinical Data Scientist, Icahn School of Medicine at Mount Sinai. “Nhi needed to know our workflow, from placing the consult to chart review and completing a consults. Shadowing the nurses at The Mount Sinai Hospital to understand how wound care prevention, assessment, and treatment currently existed, we identified more than 300 clinical data points that would indicate a profile of patients at risk,” Vickee says. “From there, we created a model called the Pressure Injury Prevention Artificial Intelligence (PIPAI) Tool that would align best with our workflow. I validated the model every day for four months on two pilot units, and after some finetuning, we scaled to five additional units. I did unit-to-unit in-service with the staff to increase tool utilization. We deployed the tool to additional units, and currently the PIPAI tool is in use in 15 units.”
The initial results from the pilot units were overwhelmingly positive. There was a nine percent increase in patients discharged without pressure injuries, compared with before the pilot started. And the model was 50 percent more accurate in identifying patients at risk compared to the current risk assessment tool, the Braden Score.
“Pressure injuries are a global issue, and a heavy financial burden for hospitals, in addition to contributing to complications for patients,” Nhi says. “Traditional methods miss more than half of patients who will develop pressure injuries. “We need to create a tool to be more proactive in prevention rather than reactive after a pressure injury happens. The tool runs independently and continuously, thus at-risk patients are identified as soon as they arrive in the unit, even before the nurses see the patients.”
The model will be rolled out to Mount Sinai Morningside later in 2025 and likely to the rest of the Health System over time.
Vickee’s face lights up with a big smile when she speaks about the impact of this AI tool. “I am so excited and happy when a patient goes home with a healed pressure injury or no injury,” she says. “With this tool, maybe we can reduce pressure injuries around the world!”