How Automation Can Improve the EHR Experience

By Matt Phillion

When electronic health records (EHR) arrived on the scene, they were supposed to be a magic bullet for improving healthcare. Moving away from paper records was a huge step forward, but the ever-increasing requirements for clinical documentation—while purposeful—add an administrative burden that takes time away from patients.

Meanwhile, healthcare faces a crisis on multiple fronts when it comes to efficiency and staffing. According to a Kaufman Hall report, roughly half of hospitals ended 2022 with a negative margin, while more than 1.7 million people left healthcare jobs last year, according to the U.S. Bureau of Labor Statistics.

We’ve heard the story time and time again: Overburdened staff burn out, staffing shortages rise, and patient safety is put at risk—all while hospitals face falling revenue. So where does the EHR come into play here? How does it go from being a passive repository for patient information to a driver of action, and how can it be leveraged for a resource- and staffing-starved industry?

“When I think back to the 2009–2010 era, we were predominantly on paper. It was a natural evolution we had to go through to get everything digitized,” says Greg Miller, chief growth officer with Lumeon. “Fortunately, the federal government was behind the adoption of the EHR. But it’s a matter of personal frustration for me that there’s still so much focus on the EHR as being it—that we can stop with innovation because we have it.”

When he talks with healthcare CIOs, Miller says he hears just how much time they spend focusing on maintenance of the EHR. “It’s expensive and time-consuming,” he says. “This is the genesis of the thought that [the EHR is] necessary but insufficient. When you talk to CIOs and others in the industry, there’s a universal agreement that it was designed as a system of records, to store information, and to be predominantly for regulatory compliance, billing, and scheduling.” There are outlier functions, but they are limited, Miller says.

A system of record architecture is designed very differently from a system of action that can efficiently leverage stored data. “The end game [of the push for EHR adoption] was not the EHR itself, it was improving outcomes, safety, and quality,” says Miller. “We should be thinking about that when it comes to all the time clinicians spend on documentation.”

Luckily, “the industry has started to pick up the innovation steam in the past few years,” he says. “There’s been an evolution of technology and new capabilities that will really leverage the EHR for what it’s there for, but also layer on top of that additional beneficial capabilities.”

One example: visibility when it comes to diagnoses. “Every clinician knows what they know to do for the patient, but they don’t know what everyone else is doing for that patient,” says Miller. We see this in the form of duplicate tests and procedures, friction between caregivers, and siloed information.

“This leads to not only safety issues but massively increases the cost and frustration not just to the care coordinator, but also to the patient, who is in this constant state of uncertainty,” he says. “They don’t necessarily know where they’re going or what they’re doing next. Technology has evolved to help with this, and legislation has evolved to promote interoperability to help solve some of that problem.”

Time-saving options

If one of the goals of improving the EHR is addressing how much time is spent on documentation, no conversation can be complete without addressing the growing presence of AI in the field.

“There’s a lot of buzz out there about AI and generative AI, and I’m a huge proponent,” says Miller. “It’s going to take a while for it to work its way into everyday technology, and there are some scary parts about it for obvious reasons, but I’m encouraged. To me, it represents a quantum leap forward in the art of the possible.”

At the end of the day, everyone wants to improve patient care, Miller notes, and there are many technologies we’re comfortable with in our daily lives—in banking, hospitality, and travel—that healthcare has been slow to adopt.

“In the past, it was the common belief in healthcare that we need to be cautious and therefore slow about adoption of these technologies, and that’s an excuse for not embracing the future,” he says. “We’re comfortable with it in other parts of our lives, and we’re at an inflection point where there’s enough proof out there that people should feel more comfortable progressing.”

A common stat in healthcare is that it takes 17 years for a new clinical practice to reach adoption, but things are changing too fast for that to remain the case.

“I was just talking to a friend whose friend was diagnosed with a brain tumor. The first oncologist said it was inoperable, and this patient obviously wanted a second opinion,” says Miller. “They got that second opinion, and they were able to remove 100% of the tumor. Two clinicians, two oncologists, very different opinions. That variability in how care is delivered is something technology can help with, standardizing treatment and helping get new treatments into the pipeline faster by making it more common knowledge.”

Incentivizing change

Miller explains that the U.S. suffers from a lack of alignment in care. “In the U.K., for example, with socialized medicine, one of the great things is that the incentives are aligned from the government down to the individuals,” says Miller. “In the U.S., there’s generally a desire among clinicians to do the best they can, to adopt new ways of doing things, but new ways are not always aligned with the competitive perspective.”

A change in incentives means a shift to value-based care, a model the U.S. has been slow to adopt. “My hope is that this will accelerate,” says Miller.

The other challenge is getting clinicians themselves to adopt the new technology. Lumeon watches the “override rate” when its product is deployed in a system to see how often the clinicians work around the solution.

“We want to do things that take the burden off the frontline staff, both administrative and clinical. Things they shouldn’t have to do but are forced to because of the EHR,” Miller says. “We monitor the override rate to see how often clinicians override the automations Lumeon instigates.” Generally, there are moderate override numbers in the first month or so, but by the second or third month, the override rate drops precipitously. “This is because they begin to trust it,” says Miller. “We were able to make them comfortable with the automation.”

Once that trust is established, Lumeon also sees a large drop in the time the clinician spends documenting in the chart—from 11 minutes per patient to two minutes. “This is where the rubber meets the road: reducing what’s known as ‘pajama time.’ Working on documentation is where people get really excited because then the clinician can do what they were trained to do, not data entry,” says Miller.

The idea behind the technology is to unobtrusively automate repetitive tasks clinicians end up doing because there hasn’t previously been an option to do otherwise.

“We model our engine on the various pathway decision points for a given type of patient, leveraging real-time data to see things like medical histories or current medications,” says Miller. “Say there’s a required pre-surgery program where every single patient is required to have a 30-minute visit with a clinician, for patient education, reviewing lab tests, lots of administrative stuff. The technology enables patients to be fast-tracked through automation. That’s a lot of time saved, especially in a high-patient-volume environment.”

There are some tasks a computer is designed to do better than a human being, Miller notes, and putting computers to work on these tasks can remove variability, improve consistency, and eliminate waste, which can all impact the cost, quality, and patient experience.

“When you look at the way healthcare has usually solved problems, normally we throw money and people at it. But we don’t have more money, everyone’s struggling financially, and we don’t have more people, and the people we do have are already stressed,” says Miller. “Ways to leverage tech that reduces the burden, especially on clinicians, should be a high priority for every organization.”

There are two parts to automation, Miller explains. The first is automation itself, which addresses a singular task, like dispensing medication, placing refills, or performing robotic surgery. The second is orchestration: leveraging automation for multiple tasks in an overall process that weaves functions together.

“EHRs are necessary but not sufficient: they have the ability to automate individual tasks, but they’re not architected for orchestration,” Miller says. “There’s no excuse for not adopting modern technology. You can dramatically reduce the amount of time clinicians and staff spend on non-value-added tasks, and offer ease, convenience, and reduce friction through automation.”

Matt Phillion is a freelance writer covering healthcare, cybersecurity, and more. He can be reached at matthew.phillion@gmail.com.