Hospital acquired pressure injuries are a global challenge, placing a significant burden on health care systems despite being preventable. Prevention can be simple, but treatment is complex and resource intensive.
Each year, preventable pressure injuries cost hospitals millions. At Mount Sinai Health System, an AI-driven solution is transforming care by accurately identifying patients at risk and preventing pressure injuries before they start.
Follow along as wound care nurses refine workflows, leverage artificial intelligence, and enhance patient outcomes. This innovative approach is making Mount Sinai more efficient—reducing costs and saving lives. The future of wound care is here.
Pressure injury is a global issue. It's taxing our health care system when something is preventable, it is really important to prevent it from happening because prevention is easy but treatment taking care of it is really really hard. Hospital aqui pressure injury costs millions every year to the hospital. If they occur in the hospital, then it's the hospital's fault for letting it happen and it costs an extra 12 to $45,000 per injury. When they happen for the hospital to cover the care cost, it's not reimbursed by Medicare. But Medicaid, it's a big burden for patients and hospitals. Pressure injuries in a perfect world should be a never event. Unfortunately, due to patients, clinical conditions, comorbidities, they do happen and you are far more at risk for developing other complications as of it, it's just a huge setback, detriment to the patients. The traditional methods have shown that it doesn't really work as it was missing more than half of the patients as the nurses were more reacting to the wounds than preventing the wounds. We need other clinical data points in identifying these at risk patients. Vicky came to us with the idea of having a predictive engine to be able to target the patients who are most at risk of developing press injury before it happens and take preventative measures. We developed the model for many months, shadowing the wound care nurses in the hospital first to understand their workflow, then came up with a model design that was the most appropriate for this workflow in the hospital. After multiple rounds of optimization, we had a final great model that was working and validated by everyone, data science side but also clinical side, it will identify the patients admitted with no pressure injuries, but at risk immediately, the nurses doesn't have to be there as soon as the patient arrives to the unit. Even if the nurse did not see the patient yet, the artificial intelligence identifies patients at risk. There was an increase of 9% of patients discharged without any pressure compared to before the pilot started. It's 50% more efficient at identifying the patients at risk. We found out that patients who started in the pilot units and moved to non plot units develop pressure injuries. The model will learn from the nurses who then tells the model this is right. This is not right and it is iteratively trained over time optimized every time to get better and better, more accurate. The tool really does help to empower nurses. It makes them more aware. I think it's simple, easy to understand. It helps me stay so organized and I think that the patients really have benefit from it. We are in the progress of scaling to the whole hospital at Mount Sinai Hospital. And the goal is to scale it up to the whole health system. We can roll it out to the Mount Sinai health system. The whole system will have lesser pressure injury rates and then bring it out to the public to other facilities and globally, hopefully decrease pressure injuries. Globally, it's really about getting the best information you can to make the clinical decision appropriately for the right patients. I love to help people. I love to heal. I am excited when there's a wound and the patient goes home with no wound or healed wound. I am so so happy.
