How Children’s Hospitals Approach Change in Patient Care

In Ohio, Akron Children’s handled nearly 1.3 million patient encounters in 2021. Similar to other hospitals worldwide, it used a Pediatric Early Warning Score system to reveal signs that seemingly stable patients might need more immediate intervention. However, the system required a great amount of coordination and manual calculations from nurses. The hospital wanted a more comprehensive and automated solution to better monitor patients, something that could use AI to analyze millions of real-time data points such as vitals and lab results from the electronic health record (EHR) system to better predict adverse events.

“Nothing happens overnight,” Erin Pitchure, quality initiative specialist at Akron Children’s, says of building the new analytical system. “We worked together to add rules and indications of conditions so that predictions would be more accurate and trigger the appropriate warnings.”

LEARN MORE: A Texas-based children’s hospital transforms its recovery strategy.

The system is anything but static, since nursing staffers offer continuous feedback. “Someone is always asking us to add something. By adding more rules to the system, we’re able to paint a clearer picture of the patient,” says Ericka Moore, clinical informatics specialist at Akron Children’s. (Moore and Pitchure have also held nursing positions in the past.)

After implementing the new system, Akron Children’s saw a 39 percent decrease in medical response team calls and a nearly 18 percent decrease in unplanned transfers to the intensive care unit. In addition, the team noted improved situational awareness and a reduced cognitive workload for clinicians.

Looking ahead, the team hopes to expand the system to the emergency department. More work needs to be done, but with its success so far, Moore and Pitchure say that the implementation will be incredibly useful.

“The emergency department will do the head-to-toe assessment, generating a score that will help provide guidance for the providers to think about next steps. It will help get the patient to the right level of care from the beginning,” Pitchure says.

Easing Fears with Enhanced Patient Experiences

Creating a calming environment for patients and their families to begin what can be a nerve-wracking care journey is a goal for many children’s hospitals. In Southern California, when Loma Linda University Children’s Hospital needed a new facility to comply with updated seismic codes, leaders saw the opportunity to create a high-tech, welcoming environment.

The result is Loma Land, a 60-foot-long digital interactive experience in the hospital lobby. Visitors interact with digital wall screens to choose an animated forest creature to personalize. Then, as they begin their walk down the hallway, their customized friend accompanies them in a captivating walk. Motion sensors have the animals follow their “owners” and even respond to their waves.

DIVE DEEPER: How modern HVAC systems support smart hospital transformations.

“Digital technologies are getting to the point where they can help us see or imagine things that weren’t possible before,” says Loma Linda University Health CIO Mark Zirkelbach. “We can imagine being in a dark spot. Maybe you’re scared or away from home. You can have a moment to smile or relieve stress, which is important to the healing process.”

The stories on the wall and the scenarios can all be changed over time. Zirkelbach says that similar technologies could also be used for in-room education.

“Imagine sitting in a room with a doctor and talking about your child’s X-ray,” he says. “With a 3D digital model, we could help parents and patients visually connect the dots and offer them a much greater understanding of their situations. To visualize that is very powerful.”

Safer Surgeries Through AI

Even with some reservations, parents have so far been generally receptive to the use of AI in certain pediatric acute cases, a Chicago-based children’s hospital found in a 2022 survey. That has been encouraging given the growing clinical interest in AI.

At Boston Children’s Hospital, Dr. Kathy Jenkins is the Kobren Family Chair for Safety and Quality and has spent her career as a pediatric cardiologist. She is particularly interested in how AI can be used in a pediatric cardiac environment to improve safety.

DISCOVER: How can healthcare balance the reward and risk of AI?

Jenkins is currently conducting more research on the use of data in improving the quality of difficult procedures, such as those that take place in the cardiac catheterization lab.

“We took advantage of the fact that we already have a large amount of data,” Jenkins says. “These procedures are technically challenging and invasive. We’re putting catheters into hearts, taking measurements and doing interventions. It’s typically more diverse and complex than an adult procedure.”

Jenkins and her colleagues and collaborators from MITRE brought together data from national registries and the health system’s EHR to evaluate how well AI could predict the likelihood of an adverse event in a new case. They were able to embed the application of those models into real-life operations and come up with probability values on whether there would be a serious complication.

When the model predicts a very high risk of complication, the next step is to decide what to do about it. This is where the human element comes in. “The team makes a plan: We will have enhanced communication and preparation for this case. But within the framework, the team has the flexibility to divert from the recommended plan based on their own thoughts and experience,” she adds.

Jenkins has found that all members of the team — anesthesiologists, nurses and technicians — have a better understanding of procedures and patient needs thanks to the AI’s output.

READ MORE: Learn three keys to success with a generative AI platform.

Jenkins says that she can envision many more uses for AI that improve quality and safety.

“We could use AI for risk-based scheduling,” she says. “Instead of scheduling out of convenience, we could use AI to schedule based on staff availability and time needed for each individual procedure.”

Jenkins also hopes to use AI to continuously learn about and help clinicians improve procedures.

“In a broader sense, the entire goal is that the information we’re collecting all the time teaches the AI, so we’re not doing the same thing tomorrow as we’re doing today,” she says.