UChicago Medicine Rolling Out Smart Hospital Platform System-Wide
Key Highlights
- UChicago Medicine is implementing Artisight’s platform across over 1,800 rooms to automate workflows and enhance patient safety.
- Initial use cases include virtual nursing, fall risk assessment, and real-time OR visibility, with plans to expand these capabilities enterprise-wide.
- The platform leverages computer vision, voice recognition, and RTLS to provide continuous, intelligent awareness in care settings.
- Automation aims to reduce clinicians’ administrative burden, allowing more focus on patient interactions and complex tasks.
- The hospital is reimagining workflows in a new cancer center and other clinical spaces, emphasizing iterative development and co-creation with clinicians.
Yeman Collier, CIO at UChicago Medicine, recently spoke with Healthcare Innovation about leading a system-wide rollout of Artisight’s smart hospital platform across more than 1,800 rooms. He described three initial use cases, including for virtual nursing workflows. Joining the conversation was Andrew Gostine, M.D., M.B.A, a practicing anesthesiologist and CEO of Artisight.
Artisight’s platform uses computer vision, voice recognition, and real-time location services (RTLS) to bring what it calls “continuous, intelligent awareness” to care settings. From a single platform, UChicago Medicine care teams can now deliver virtual nursing, virtual sitting, and bedside TV engagement while also leveraging RTLS-driven workflow automation, infusion monitoring, and ambulatory support.
UChicago Medicine’s first use case is in the operating room. The second is for virtual nursing workflows, and the third is for fall risk assessment and prevention.
“The list of capabilities is virtually endless,” Collier said. “It only lacks prioritization when you look at the potential value proposition for getting clinicians back to the bedside, interacting more compassionately with patients and families, and in some ways reducing the number of menial, trivial, repeatable, automatable tasks so that humans can actually focus on the more complex tasks.”
"There's a national shortage of nurses and any number of different specialists," he continued. “We are being asked to do more with what we have. Creating that capacity by leveraging automation and AI is our intention in terms of implementation at scale. It's not enough to do this with just a few activities or tasks. If you consider the number of tasks that a nurse does during a 12-hour shift and multiply that by thousands of nurses, that’s what we're after.”
UChicago Medicine already had some aspects of virtual nursing, but not the full complement of capabilities that Collier envisions for the future state.
“We really have a remarkable opportunity. We've got an almost 600,000-square-foot cancer center that's coming out of the ground, a billion-dollar structure. “The clinical workflows that are being re-imagined for that space, we expect to extend across other clinical settings, across the enterprise, but that's really the place that will be the proving ground for virtual nursing and developing and refining that workflow. “Oftentimes the intuition is to take what the nurses have been doing and try to plug that into a reimagined workflow, and often that doesn't work well. So we've got interdisciplinary groups that are working to redefine how we can leverage virtual nursing capability in the Artisight system.
In the OR and PACU, Artisight says its Smart OR capabilities provide real-time visibility and automated documentation. Every application runs on the same shared sensor fabric, the company says, seeking to eliminate the fragmented point solutions that create additional burden for both clinicians and IT teams.
The Artisight system connects to the Epic EHR (and other EHRs as well), said Gostine.
“The whole goal is to improve the quality and timeliness of data. You can almost think about any problem we have in healthcare, and if you could capture accurate data in real time, you're going to make a step in the right direction,” Gostine said. “For instance, we know Epic has a great epic deterioration index, but if you don't plug in the 31 variables for two hours, you just increase that patient's mortality rate by about 12%. We studied this, and the average time between nursing assessment and nursing data entry was 127 minutes. So there are big delays in getting data into the EMR. What we're really shooting for is to make the medical record read-only. I shouldn't have to sit down, log in, navigate to a patient chart, figure out which user interface and which field I need. That should happen automatically for me. By replicating human senses, you can get very accurate data in real time, and then you can re-imagine care.”
Following up on Gustine’s vision, Collier said he was visiting with a number of clinicians and technologists a few months ago and asked them the question: Can you imagine an exam room without a keyboard? “They looked at me like it was nuts,” he said. “But fast forward to where we are today, and it's really a clinical conversation, where you have patient, family, and care team enter the room. The care team member says we’re going to talk about the day’s schedule. We're going to draw some blood for labs, and you're going to have breakfast, and we’re going to get you down to infusion. That’s all documented now in the note, and, oh, by the way, it's able to pull in the recent labs and results and medications and allergies and imaging consults and all those things. Now you've got clinicians who are back to being clinicians, not documenting. They're simply having the conversation with the patient, family and care team.”
Another use case Collier described was helping with fall prevention. One of UChicago’s virtual capabilities is a “tele-hub,” where there are individuals staring at a screen with 12 different camera feeds from rooms of patients who are at high risk of a fall. If a patient attempts to stand, that observer has to look at the room, look at the unit, get in touch with the appropriate member of the care team on the floor, who has to get in that room to aid the patient to get back into the bed before they stand and perhaps fall. But what if we could automate that? What if I didn't have to have a nurse observer watching the patient all the time? What if we could automate both the observation and the notification? It is an area of very high risk. So that's one I think that we can fairly quickly deploy and show value from day one."
Collier said UChicago Medicine didn’t do a pilot before working with Artisight on an enterprise scale. “We have a fairly well developed innovation intake process, whether it's an opportunity that comes to us from the outside, from a board member, or from a vendor,” he said. “What I like is when we have a compelling use case that we can apply technology to help solve. In this particular case, that's exactly what it was. Here we were designing a greenfield space, and we're re-imagining what care could look like in a smart future state. One of the leaders in this space is Artisight and they happen to be right down the road. So that's how we landed here. But we did go through the due diligence to understand what our highest-value use cases would be and what we want to go after first.”
Gostine said the key to success is to develop new use cases in partnership with health systems. “Artisight was designed to be developed upon. It is a platform that has some capabilities to give you early wins, like Yeman talked about — where we can do falls reduction or remote nursing, or OR coordination. But once the hardware is installed, it allows a health system to iterate on it,” he said.
“The benefit of having all of this compute in the camera system and in the screens is that if a health system comes to us and says, we're getting great results, and next we want to co-develop something for this particular problem, we can do that co-development in an active clinical environment, and then deploy that to hardware that has available compute capacity, so that we don't have to spin up more compute in the cloud, which means we can deploy it without incremental cost structure at our site, which means I don't have to charge the customer,” Gostine said. “Once we get that initial deployment and we're generating the ROI from the first few use cases, we can then add more and more capabilities that are developed and learned from the customers without having to increase any fees. That's what becomes transformative.”
About the Author

David Raths
David Raths is a Contributing Senior Editor for Healthcare Innovation, focusing on clinical informatics, learning health systems and value-based care transformation. He has been interviewing health system CIOs and CMIOs since 2006.
Follow him on Twitter @DavidRaths
