The Role of AI in Safety Monitoring in Redefining Patient Care
Narinder Singh highlights the transformative potential of AI in patient safety by creating safety zones that enable continuous monitoring. This technology alerts caregivers when patients move outside their safe areas, allowing healthcare professionals to focus their expertise where it’s needed most. By overseeing multiple patients at once, AI can identify those at risk for falls or pressure injuries, maximizing nursing effectiveness. As healthcare systems work to integrate AI into standard care practices, addressing implementation challenges in diverse environments will be essential for successful adoption.
Listen to Narinder Singh and Eric Yablonka discuss the impact of AI on patient safety and the future of healthcare technology, emphasizing the need for collaboration and innovation to enhance care delivery.
Video Transcript
Narinder Singh:
So this is one example where we look at AI that creates a safety zone around the patient and then it allows a human to change that, adjust that, move it around the chair to apply their human judgment around this. But the point is what it allows us to do is see when we’re not there and then the patient is moving inside the safety zone, no problem. As soon as they start to get outside of it, then it alerts a person who can decide is that a dangerous action or is that them just stretching their arms or legs? And I think that’s compelling, but where it gets transformative is the ability for the AI to be everywhere. It can watch every patient. In my mom’s case, it can be in every room at once and then simply surface to an expert on fall prevention or on nursing or on pressure injury, the people that they need to pay attention to so they can go through and take action of trying to keep someone in bed or decide that we need to engage in a different way.
So this ability for AI to be everywhere and then to say, I’m going to alert you if you are moving too much and risking for falls and safety or moving too little and potentially at risk for pressure injury, but now I as the person can intervene, that really is kind of getting us to the point of creating massive leverage without having to redefine workflows or disintermediate nurses because we still rely on them for the expertise. We just let them be everywhere at once. And so as I show this, I think back to the fact that many of these concepts were invented at Stanford when you were there. I’m interested in those early days and what you saw as the challenges and opportunities then and how you think about what we need to do to make this a part of standard of care versus just a researching initiative.
Eric Yablonka:
Well, we did a lot of really interesting projects at Stanford and partnership with the faculty, particularly on the research side, really unbelievable work in a large part. We’re seeing more of that come out of Stanford even today. I think when we opened up the new hospital, we put in some machine vision cameras that were related to a research project by one of our preeminent faculty. It was fascinating the ideas of both monitoring patients and also things like hand washing and other challenges that healthcare organizations have, proprietary cameras, research algorithms, all really, really good stuff. But it really wasn’t ready to roll out and we didn’t have the initiative to roll that out to the entire hospital.
Narinder Singh:
Yeah, I think the proprietary is such an important piece because it doesn’t matter how interesting it is, if we lift the burden for a hospital to put it in, we’re suddenly going to be left with something too heavy, right? It’s just something where our environments are diverse. Rural hospitals, urban hospitals, room configurations. I think there’s a lot of really interesting compelling parts of ai, but the last mile problem of the engineering aspect is critical to unleashing that onto actual clinical workflows. Let me talk about that IT architecture side a little bit more. It’s the place that you and I have had many good conversations. Every situation at every large entity inside of healthcare and outside has some version of this sprawl problem where they’ve got applications everywhere, generations of technology, and then someone comes along, whether it’s the ERP or the smart hospital and says one solution to rule them all.