Houston Methodist Deploys Ambient AI Operating Room Platform

Eight-hospital Houston Methodist is using an ambient AI operating room platform from Apella to improve capacity planning and performance improvement. Roberta Schwartz, Ph.D., the health system’s executive vice president and chief innovation officer, recently spoke with Healthcare Innovation about the implementation. 

San Francisco-based Apella’s Horizon platform seeks to help hospitals improve case duration accuracy and increase surgical utilization. As Schwartz says, when OR schedules rely on manually entered data, it can lead to inaccurate scheduling, which can impact resources and patients. 

As Apella CEO David Schummers explained in a statement, “For years, hospitals have made high-stakes OR scheduling and capacity decisions using limited data from EHRs — systems designed for documentation, not operations. Horizon replaces those subjective, inconsistent inputs with ground-truth data captured directly from the OR. Apella’s ambient sensors and machine learning enables more accurate predictions, smarter use of OR time, and stronger trust between physicians and perioperative leaders.”

Healthcare Innovation: Before talking about your work with Apella, could you describe your role and portfolio there as executive vice president and chief innovation officer. How do you work with the CIO and CMIO? 

Schwartz: I hold two very independent roles. I am the chief executive over our academic medical center. So I run a 1,000-bed hospital with 8,000 employees. That's my executive vice president role. In addition,  I run the innovation activities for our eight-hospital system. So I have that dual role. The chief information officer didn't work for me until about six months ago. Now that position reports to me, as does the CMIO. So I can now integrate those two portfolios and look at them together.

HCI: I read that Apella has ambient sensing technology and uses data captured from the operating room. Is that ambient sensing technology different from scribes used for physician offices to record conversations with the patients for physician notes?

Schwartz: It is similar, but different. Apella video-records what is happening in the OR and is taking the time stamps from every action and turning it into information. Patient on the table, patient draped, etc. What we found early on is that there were times where our people inputting this data — patient came in the room, patient left the room — were anywhere from 3 to 10 minutes inaccurate. When you're trying to move cases along and be more efficient, and understand  the average time for a particular doctor to do a spine case, those minutes add up. You’re trying to make the use of this precious resource incredibly efficient. Therefore, having this accurate information is great, and that's what we get from this.

HCI: Do the operating rooms use a separate application besides the EHR for scheduling and managing blocks of time?

Schwartz: Not usually. There are other companies that are doing this kind of work based on the Epic data. There are also a few companies in the space, including Apella, doing work using this video type of data. 

HCI: So the goal is to do predictive case scheduling?

Schwartz: Yes. Also, now the doctors get texts that say, patient wheeled in. It is sort of the same type of message you get with an airplane, right? We're boarding group one. They get that now from the operating rooms on their phone. 

HCI: Was this a solution that Apella had in place at other hospitals or was this something you co-developed with them? 

Schwartz: We had seen a little bit of it, but there was a lot of work that was done here at Houston Methodist building it out. And now we learn from their other customers, too. We're also excited about working on things like whether we have the right supplies and how long the average cleaning time was.

HCI: Do you have positive feedback from the surgical teams themselves about the efficiencies this is bringing to their work?

Schwartz: Actually, the beginning was difficult. There were some initial concerns, but once people understood that we really were using it for the information, they got more comfortable with it. I think it's made us more efficient. It's made us better. We've been able to fit in more cases, and our doctors like the texts. We are able to schedule better. We're able to look at our blocks of time better. 

Now we're developing more sophisticated algorithms. Not everyone's using it at the level of sophistication that they can and we're going back to say, how can we use this even better? Because it unlocks enormous amounts of new data. At the same time as we're maximizing what we’re doing, Apella is coming out with new features, which is really nice.

HCI: Are there some other high-priority areas where you see inefficiencies or administrative burdens that you're targeting with AI solutions or are there too many to name?

Schwartz: I have 76 projects on a list. I mean, there’s almost no area where we’re not looking to improve quality and make it more efficient. For example, we just signed a three-year agreement with Aidoc to put in our radiology areas — and not only for patient management, but to be a trial sight on their new foundational model. We have prioritized the imaging area.