How clinicians are helping shape EHR design

Designing together

Clinician feedback of EHR design is no longer a late-stage development afterthought. Now, it's fundamental, experts say. Not only must the software support workflows for some of the most complex jobs in any industry, it has to work across wide variations between facilities, clinical areas and numerous clinical roles within those areas. 

"We're not trying to design for clinicians. We're trying to design with clinicians, " says Dana Close, director of UX design at Epic. "All technology expectations are rising — not just functionality, but beauty, simplicity and joy to use." 

To inform everything from user interface choices to future roadmaps, vendors are embedding clinicians directly into product development cycles. At Epic, this includes internal clinicians, specialty-specific steering committees, usability labs during large user conferences and hospital site visits for designers and developers.

"It’s a requirement that everyone goes out into hospitals to see their work in action," said Close. "We see how clinicians are using the software in conjunction with the physical space and human interaction. There’s no substitute for that empathy."

Close reported that 1,700 user tests were performed at a recent useability lab — and that number rises every year. These task-based tests validate the useability of new features, such as its AI-powered chart summary. Clinical users then give feedback that drives pre-release adjustments. 

At Meditech, designers lean on a dedicated physician advisory council for design input and strategic direction. "Even before a project is a twinkle in an eye, they're already influencing the roadmap — reporting challenges, things they're missing and hopes for AI to accomplish specific tasks," says Jen Lindquist, manager of development at the healthcare tech provider. "It’s by physicians, for physicians." 

These clinician collaborators range from frontline physicians and nurses to clinical IT specialists and administrative leaders. The design team also uses early adopter programs, feedback channels through clinical informatics and real-time iterative design sessions with clinicians to create solutions that go beyond simple feature delivery. 

Blending customizations with best practices

But, with so many providers practicing across vastly different specialties, matching functionality to workflows is an exceptional challenge in EHR design. "We have hundreds of thousands of users that all want different things, " reports Close. 

To solve this problem, developers at Epic lean on standardization informed by physician steering boards in fields like cardiology, primary care, pediatrics and more, while also offering customizable features. 

"Our foundation system incorporates standards that you need to have consistent across instances. To optimize localized software, we test it in the local environment and go on immersion trips, " says Dr. Jackie Gerhart, — Epic’s vice president of clinical informatics. 

"We have two delivery models," says Caitlin Heatwole, lead software designer at Meditech. "There’s room for customization and local build." 

On the other hand, smaller or rural organizations with limited IT staff often choose a pre-configured software-as-a-service model. 

Making an impact with applicable AI

Seventy percent of clinicians feel that AI tools are the most helpful route to reduce their mental burden. Health tech providers prioritize embedding AI in EHR workflows where it fits intuitively, with a focus on easing cognitive load and helping clinicians navigate health data.

"We went to school to practice medicine, not to curate charts," Gerhart said. "AI tools like chart summaries are helping me get the information I need faster, so I can make better decisions."

Ambient listening is frequently cited as a high-impact use case. Representatives from both Epic and Meditech reported positive feedback from clinicians using it to reduce documentation time.

"The feedback we’re hearing is that providers can return to patient care focus. The cognitive burden goes away and the background tool is helping with the day-to-day items,” reports Heatwole. 

Still, not all AI solutions are ready for clinical use. Leaders emphasized the importance of validating functionality. "If someone has to review and re-enter what the AI generated every single time, we’ve failed," said Lindquist. "Even if we miss the mark the first time with an enhancement, we circle back and try again until we deliver a complete solution."

Elizabeth Stricker comes from a nursing and healthcare leadership background, and covers health technology and leadership trends for B2B audiences.