Doctors and nurses at Novato Community Hospital are using artificial intelligence to bolster training for emergency care of infants.
The hospital, part of the Sutter Health system, has acquired a robotic replica of a newborn girl. It can simulate a number of medical conditions and scenarios in real time that emergency room staff might rarely experience — and therefore only be able to practice infrequently.
“I don’t know anyone who doesn’t get more emotionally revved up when it’s a pediatric patient. It naturally happens,” said Zachary Schoenberger, an emergency room nurse at the hospital. “Everything feels more intense, everyone’s more ramped up. The reality of working with a smaller patient is, everything is more delicate and smaller so doing the procedures is more difficult.”
“So having this doll to refine our skills on just makes it so much easier and more successful for us when we have to do this with true patients,” he said.
The robot, which the hospital dubbed “Luna,” not only moves, cries and babbles like a real baby, but it has a pulse, simulated breathing and a readable temperature. Its skin color can even change if it is simulating insufficient oxygen or jaundice.
“On the sim, you’ll actually see changes in the shape of the chest,” Schoenberger said. “You’ll start to see the skin on her chest start to retract as she’s breathing in, or maybe you’ll see her belly become more involved in the motion of the breathing.”
Using a connected tablet computer, the staff can pick from 14 pre-loaded scenarios or customize different ones. Schoenberger said this can be used to replicate any medical issues that might be more prevalent, or to recreate real emergencies staff have experienced, to find ways to improve treatment.
Luna can also respond to various treatments or instruments used by medical staff, Schoenberger said. For instance, employees can check for blood oxygen levels, use a stethoscope to listen to the baby’s lungs and intestines, insert an intravenous line or place a tracheal tube down its throat.
The robot can also show whether the treatment was performed correctly, such as having only one lung filled with oxygen if a tracheal tube is inserted too deep, Schoenberger said.
The hospital, which says it is the first in Marin in get a robotic training baby, received state designation as a pediatric receiving center in 2021. To receive this designation, a hospital must demonstrate to the local emergency medical services agency that it has appropriate staff, training, processes and equipment to treat children as well as the ability to determine if a child needs more specialized care. MarinHealth Medical Center and Kaiser Permanente’s San Rafael Medical Center received the designation prior to Novato.
Matthew Mullis, director of critical care, emergency and trauma services at MarinHealth Medical Center, said about 15 to 20% of its daily emergency department cases are pediatric patients.
“We have to be prepared. The equipment is different, the training is different, the medication,” Mullis said. “Any designation in the county, even at a basic level, shows a commitment of that hospital to see pediatric patients.”
Dustin Ballard, chief medical officer of Marin County Emergency Medical Services, said having the designations at all three Marin hospitals is notable because only a handful of hospitals have met these criteria. A 2019 study published in the medical journal Pediatrics found that mortality risk was nearly four times higher among children treated at emergency rooms with the lowest pediatric readiness scores.
“There is evidence that preparedness makes a big difference because sick kids are fortunately a rare thing for most hospitals,” Ballard said. “While that’s fortunate, oftentimes when it does happen the preparation is not as good as it should be.”
Schoenberger said the Novato hospital also plans to offer further training sessions with Novato Fire Protection District’s EMS providers to improve care while an infant is transported to the hospital in an ambulance.
“Having this doll allows us to get in that real-world practice in the most true-to-life situations we can build so our team is prepared as we possibly can be,” Schoenberger said.