Investigating how ambient sensing can improve OR efficiency

Addressing OR efficiency with ambient sensor technology

Schwartz emphasized that OR efficiency is a challenge not just for Houston Methodist but also for health systems and providers everywhere.

"Everyone wants [ORs] to be more efficient," she explained. "Doctors and surgeons want to move their day along. They want to come in on time, leave early if they can get all their patients done, have turnover time and have a very pleasant day at the hospital."

She further noted that OR minutes are some of the most expensive on a health system's campus, creating an incentive for hospital leadership to pursue efforts that help fit more cases into the OR docket and reduce overtime.

"I don't think our problems are any different than any other hospital in America -- it's a generic set of problems," Schwartz remarked, indicating that OR performance improvement projects focused on reducing turnover or optimizing block time are not new ideas. "The question is, is there a better way to [drive] improvements on efficiency in the operating room?"

This is where emerging technologies like AI have a chance to shine.

Schwartz highlighted that when clinicians step into the OR, much of the information recorded in EHRs is entered manually, which can create hurdles to efficiency.

"The information we're pulling from becomes very human-facing -- wheels in, drapes up, all of those things," she said. "We've learned that humans are inaccurate, and, in our estimation, they are inaccurate about 60% of the time."

To overcome this, Houston Methodist undertook a pilot with Apella, a health technology company specializing in ambient sensor technology and AI for ORs. During the pilot, the health system introduced ambient sensors to collect video data in the OR, which is used to inform efficiency improvements.

The AI analyzes historical information and video data from procedures to make predictions about future OR schedules, staffing needs and surgery duration. This reduces the reliance on human judgment, allowing clinicians to focus on care while improving error rates and decision-making.

Schwartz compared the scenario to planning for a flight's departure, where having accurate, up-to-date information is crucial. She explained that if the flight is scheduled to leave at 8:30, but actually leaves at 8:20 or 8:50, passengers are likely to be unhappy.

She indicated that running the OR is similar.

"Having that accurate information allows you to make decisions… the accuracy of that information is allowing for improvements in the way we do our business," she said.

She emphasized that the tool is useful for augmenting the expertise and judgment of charge nurses, who typically make decisions about staffing needs and other resource allocations. Over time, these nurses learn how long their surgeons typically take, how much staff they will need to handle caseloads and how a patient's health status might impact time spent in the OR.

However, nurses are human, and clinical judgment can be negatively affected by many factors, such as job or personal stressors. Schwartz noted that the AI can make precise, accurate predictions about staffing needs for clinical decision support.

"[The technology] basically says, 'Here's the computer prediction.' The human can override it but can also use that information to determine how much staff they will need to keep at the end of the day," she said. "So, they can start asking for extra volunteers early, or they can start releasing people early and trust that if they release those people on time, they're still going to have enough [staff]."

DeDominico likened the AI-enabled workflow to ordering an item online and tracking its journey to the destination. A customer orders the product, and once it ships, the retailer provides real-time tracking insights that the customer can use to inform their decision-making, such as whether or not they might need to reschedule the delivery or be home to receive the package.

The OR efficiency process, driven by ambient sensing and AI, is similar in that its predictive modeling allows stakeholders to look at the timing of a patient case, the specialty equipment required, the staffing levels and additional considerations ahead of time.

Access to this information allows nurses, surgeons and other staff to keep track of various resources and communicate more effectively about how long preparation and surgery will take.

"The difference between video audio information [and human-entered EHR data] is that there's no debating what time the drape went up -- the computer sees it," Schwartz stated. "There's no longer a question of any judgment in that, which is something we've never had in our ORs. So, you're talking about your most expensive asset, the one that frustrates your surgeons -- who you're trying to attract to come to your campuses -- the most."

But with AI, these frustrations can be alleviated by providing accurate, easily accessible information. However, getting clinicians on board with ambient sensing technology can be challenging.

Gaining clinician buy-in

Schwartz noted that adding ambient sensing tools, including video, to areas where they have not been deployed before can be a "nerve-wracking" process for everyone involved.

Clinicians can experience anxiety about being recorded, but she explained that addressing some of those feelings and gaining buy-in requires stakeholders to emphasize the benefits of each newly integrated tool.

DeDominico indicated that highlighting the ambient sensors' accuracy and the AI's capability to streamline communication is a key aspect of encouraging clinician buy-in. With these tools, clinicians receive texts with relevant updates, such as when a patient is leaving their room to be prepared for the OR. These texts typically include a prediction of how long the clinician has before they need to start preparing themselves for surgery.

The AI is also useful for streamlining workflows that might impact ORs -- such as medical imaging.

"Collecting accurate and precise data makes a difference," DeDominico explained. "By viewing [these data,] departments like imaging services -- which cross over with surgical imaging -- can say, 'Hey, your MRI is going to open up earlier,' or 'MRI is going to be a little bit late because they have an OR patient that's using the MRI and then going back to the OR.' [Healthcare facilities] can add in patients and move them around to maximize their utilization of expensive equipment."

He further noted that text messages have increased satisfaction among clinicians by making information available and actionable.

"When they get that text message that says, 'Your patient's in the room,' they start their timer, and they know when to come into the room instead of circling. So, they can go discharge patients, write orders, see someone in their office, and go see a patient in another room. Then, they come back and know when that patient entered the room and their times because [the AI] can tell them."

Because of this, even clinicians who were initially more wary of using ambient sensors in the ORs have reported being satisfied with the improved information and communication the tools enable.

Schwartz indicated that transparency whenever clinicians asked questions about liability and consent in the context of the ambient sensors' use was also key to building the trust necessary for buy-in.

She stated that if a video ever needs to be pulled for review, there is a process that each review request must go through.

"If we ever want to pull a video, we keep them for 28 days," she said. "Then, three individuals will have to sign off -- the chief quality officer, the head of the department, and the head of the operating room committee. If those three individuals agree, we can then check the video."

She elaborated that in the majority of cases, the person requesting video review is the lead clinician on a patient's case or a leader from the quality department. Since deploying the ambient sensors, only a handful of videos have needed to be pulled.

These videos are handy for team training and clinician education, which can lead to efficiency improvements at the care team level.

Measuring success

When targeting OR efficiency, one of the key performance indicators that health systems track is OR block utilization.

Ambient sensing technology allows for the collection of highly accurate data about how much block time clinicians typically use for different procedures. These insights can identify underutilized slots and offer them to surgeons who might need additional block time.

This reduces the amount of unused OR blocks and can help optimize block scheduling depending on clinicians' needs and preferences.

"The number one thing that most surgeons care about is their time here. From the moment they walk in, whether they're 15 minutes late or 15 minutes early, their clock starts. They want to be efficient when they're here," DeDominico stated.

He explained that by optimizing block utilization, clinicians can reduce their turnaround time -- the time between one patient leaving and another coming in.

The use of ambient sensors can also contribute to successes in infection control. In the event of an infection, stakeholders can review the video data to help determine whether the cause is related to something in the OR or something that the patient did before the procedure.

This can lead to significant cost reductions, as treating an infection can add up to thousands of dollars. Addressing potential causes of infection is also critical for improving patient safety and outcomes.

Schwartz emphasized that driving these successes by deploying AI and ambient sensing technologies requires a comprehensive change management strategy that prioritizes engagement and communication across the enterprise.

"Implementing technology like this comes along with a huge change management effort. It's new for staff, it's new for doctors, it's new for administration. It's nerve-wracking," she noted. "So, there's lots and lots of questions, and it's really essential that people do a great job of communicating -- working with your surgeons, getting all of the questions on the table -- and not being afraid to be honest about what you can and can't do."

"The return on investment is very obvious. The hardest part is the change management," Schwartz concluded.

Shania Kennedy has been covering news related to health IT and analytics since 2022.