How EHR Clinical Workflow-Driven Design Enhances Health IT Usability

After experiencing EHR clinical workflow challenges as an internist and interventional cardiologist, Bill Hayes, MD, entered the health IT industry to bring a clinician's voice to the table.

Hayes, who is now chief medical officer at health IT vendor CPSI, noted that the shift from paper-based health records to EHRs about a decade ago affected clinical workflows drastically.

"Some of the early systems I got exposed to were frustrating, to be honest," he told EHRIntelligence in an interview. "They changed workflow sometimes from a 10- to 12-hour day to a 14- to 15-hour day. That was difficult for a lot of us."

After talking to his wife one day after work about the challenges he experienced using the EHR system, she suggested he try to do something about it.

One thing led to another, and Hayes found himself drafting a letter to all the major EHR developers in the country asking if they would be willing to receive clinician input on their products.

"There was a very positive response," Hayes said. "I started doing some consulting work in that field and then realized I probably can't do cardiology and consulting really well full-time. Eventually, one of the EHR companies offered me a full-time position, and I took it."

Hayes said that the problem with many of the early EHR platforms was that the vendors created them as transactional systems instead of workflow-driven systems. He noted that vendors originally designed EHRs to measure encounters, financials, and document E&M codes, not to present longitudinal health data for care delivery and clinical decision support.

For example, Hayes explained that the system would represent information for a pregnant woman's monthly check-ups as nine separate events when it's really one pregnancy that's continuous over nine months.

"That's one event that needs to be seen from beginning to end for the quality of the care and patient outcomes," Hayes said. "The systems were all essentially designed in a way in which the applications dictated what the workflow was.”

He noted that to this day, EHRs often have a "huge navigator bar" of applications that serve individual purposes, such as order entry, dictation entry, and social history.

"Electronic health records changed healthcare for the better in some ways, and yet I still think there's tremendous potential to improve it even much more from a computerization standpoint," Hayes said.

However, Hayes said that the industry is working towards creating clinical workflows that are supported by applications rather than application-defined workflows.

"When I talk about whether a system is driving your workflow because it's just made of siloed applications, or whether it's a workflow supported by the applications, it is because I am a huge believer in interoperability," he explained.

Hayes said that as the digital health transformation progresses, connections to interoperability networks like CommonWell and HL7 interfaces will help advance EHR data exchange to improve patient outcomes across the care continuum.

However, he emphasized that as the industry looks toward widespread interoperability, it will be essential to ensure that the sheer amount of clinical data does not overwhelm providers at the point of care.

Take, for example, a patient from Florida who goes to Colorado and gets into a terrible skiing accident that requires emergency surgery, Hayes said.

Suppose the patient is unconscious and no one is there to speak for them. In that case, their safety could be at risk, even if the hospital in Colorado has access to the patients' complete EHR through an interoperability network like CommonWell, Hayes said.

"When we think about how much data is in a single patient's chart in their lifetime, it's an amazing amount of data," he explained. "To care for a patient in an optimal way, especially as we go to value-based care, the physician needs to know high-value information once they get to the computer."

For instance, say the patient has malignant hyperthermia, an adverse reaction to anesthesia administration that can result in death.

"That is a very high-value piece of information," Hayes said. "It could save the patient's life for the anesthesiologist in Vail to know that. It doesn't matter who the EHR vendor was, just the fact that the patient record was patient-centric, profile level, longitudinal, and not transactional. The value of interoperability is quintessential because it can save lives."

Hayes said that computers have the power to sort through EHR data to bring forth relative information at the point of care based on what the provider is treating.

He said that the industry is starting to move towards this kind of functionality by taking clinical concepts and making "clinical connections" relevant to those concepts. Hayes noted that this kind of EHR optimization has the potential to revolutionize patient care delivery.

For example, if a patient has renal insufficiency, the EHR should present certain relevant information in the provider's workflow immediately, Hayes said.

"Clinical end users don't want all the data," he concluded. "It would take a day or days to go through all the information in a patient's chart in their lifetime, but a computer that's given certain relative high-value relationships to what's being cared for has the ability to bring that information to the end-user