Minimizing central line-associated bloodstream infections (CLABSIs) has been a focus of hospitals since the first central line was placed. These hospital-acquired blood infections bring a 10% to 30% mortality rate, and with it incalculable sorrow. CLABSIs also cost U.S. hospitals about $46,000 per patient to treat, adding as much as $2.3 billion to the U.S. health care system’s annual tab.

Decades of work to minimize these infections cut their prevalence roughly in half. Now, Boulder-based Amplifire has shown it can cut it roughly in half again – in just 26 minutes, says Bob Burgin, the company’s CEO.

That’s the average amount of time it takes a health care provider to go through the training-technology company’s CLABSI module. And while the UCHealth CARE Innovation Center partner’s success with CLABSI and other areas of medicine have yet to become universal, Amplifire is making a big difference on multiple fronts at UCHealth and 25 other U.S. health care systems.

UCHealth started working with Amplifire in 2018. The initial focus was on training physicians, nurses, and staff on the Epic electronic health record that contains every UCHealth patient’s medical information. Soon, though, it was clear that Amplifire’s approach would work for clinical learning as well, says JoAnn DelMonte, UCHealth’s vice president of Professional Development & Practice.

“We have 9,300-plus nurses at UCHealth,” DelMonte said. “To target all those learners and keep them engaged in learning is challenging.”

It’s not that nurses, doctors, and other health care providers aren’t good learners. It’s that they’re extremely busy with high-stakes patient care. Plus, hospital-based health care involves biology, tricky combinations of illnesses and injuries, and distributed decision-making happening 24 hours a day. Burgin calls it “the most complex work environment on Earth.”

Amplifire’s learning system manages to keep about 2 million people a year engaged by melding a cognitive-science-derived, artificial-intelligence-enhanced learning approach with the expertise of UCHealth and other experts. The approach is based on decades of academic research on how we retain information. Put simply, we learn best when information pairs with a dose of emotion.

Into the brain: how learning happens

As Burgin explains it, someone learning raw information engages with it using their prefrontal cortex. Add an emotional context to that raw information and the amygdala gets involved. When both the prefrontal cortex and the amygdala wrangle the same bit of information, the hippocampus lights up.

That combination, and the hippocampus’s involvement in particular, has been shown to supercharge learning.

“Traditional learning has less than 5% retention in 12 months, and about 30% retention in 90 days,” Burgin said. “We can load information into the human brain with somewhere between a 50% and 80% reduction in time, and with 84% retention in 12 months.”

Amplifire didn’t come up with these cognitive connections – but several of those who did are now on the company’s science advisory-board, Burgin says. Beginning in 2011, the startup took it from theoretical construct to learning tool.

The tool gets the hippocampus engaged by asking a factual question while at the same time asking the learner how confident she is in her answer. The nature of the system’s factual questions is also strategic. It emphasizes what Burgin calls “known mistakes” – what health care specialists working with Amplifire have zeroed in on as the most common and costly ones. The system also takes into account what previous learners have gotten wrong – and with 3.6 billion learning interactions and counting, Amplifire has a solid grip on that. It also hammers at something called “confidently held misinformation.”

Overconfidence can kill

For example, in the CLABSI training, about half of those taking it nationally get a typical answer right and are confident in their answer. Another 30% get the answer wrong and admit to being unsure about their answer. The final 20% get the answer wrong but are convinced they’re right. That’s confidently held misinformation, and it turns out that a disproportionate share of medical errors happens because of it.

Medical errors are, statistically speaking, rare. But in aggregate, the numbers are staggering. One study estimated about 250,000 annual U.S. deaths from medical errors – about six times the number who die in auto accidents, and that’s probably an undercount, because the figure only included deaths among the roughly 33 million U.S. inpatients admitted each year.

In addition to CLABSIs, Amplifire and its health care partners have developed clinical training for sepsis, pressure injuries, patient falls, c. diff infections, and catheter-associated urinary tract infections (CAUTIs), among others. UCHealth codeveloped a framework for Epic electronic health record training now used by 18 hospital systems. It also developed training related to dealing with nurse safety and workplace violence, which Burgin described as “extremely important right now” at hospitals around the country using Amplifire.

DelMonte, who oversaw that codevelopment, offered an example of its content.

“If a visitor is resistant to put on a mask, what are your resources and how do you help to de-escalate that situation?” DelMonte said.

UCHealth also codeveloped with Amplifire a program to train patient care assistants who are helping nurses on UCHealth acute-care inpatient units. Since the program went live in 2021, UCHealth has trained more than 300 patient care assistants, with Amplifire’s training prepping new hires prior to hands-on training at UCHealth hospitals, DelMonte says. UCHealth and Amplifire are now working on training for incoming infectious disease and palliative care staff, she added.

Burgin described UCHealth as “the number-one health system for co-development that we work with in the country,” in large part because the CARE Innovation Center enables quick decision-making and the ability to test and then roll out innovations at scale.

The respect goes both ways, DelMonte says.

“They don’t claim to have subject matter expertise in the health care space, but they have subject matter expertise in the science of learning,” she said. “That creates a good partnership that’s enabling great learning.”