What clinical decision support can learn from a helicopter cockpit

'Both pilots and physicians are dealing with peoples' lives, with mission-critical activities'
By Mike Miliard
10:24 AM
Share
Army Air Corps Apache helicopter

Before he co-founded clinical decision support developer medCPU, Eyal Ephrat, MD, was a practicing physician. Before that, he was a helicopter pilot. Want to guess which workplace had the more reliable information system?

This past May, Metairie, La.-based East Jefferson General Hospital partnered with medCPU for a new clinical decision support  system to help its clinicians make the most accurate decisions at the point of care.

As it worked to prevent adverse patient events such as deep venous thromboembolism, sepsis and stroke, EJGH deployed medCPU technology to help alert physicians and nurses better adhere to clinical protocols and its own treatment guidelines.

Sidney Raymond, MD, the hospital's chief medical information officer, sung the praises of the medCPU system, which aims to bring accurate, real-time data as needed to the patient's bedside.

"The system cross references a patient's history, physical findings, and laboratory and radiological data at the same time his or her doctor is assessing and prescribing a course of treatment," said Raymond, in a press statement. "The centerpiece of medCPU’s technology is the ability to read medical data (including free text notes or dictations) and cross-reference that within our medical database instantaneously.

"Once it identifies a departure from hospital guidelines and protocols, an alert message pops up on the computer screen at the point-of-care informing the clinical staff of such departure and highlighting the hospital’s expected approach in this case," he added.

The development of medCPU's CDS technology finds its roots in Ephrat's clinical experience, of course – he's an obstetrician – but it was also influenced by his time serving as an Apache helicopter pilot.

Simply, put he found the comparison between the decision support tools in the cockpit to those in clinical settings to be rather lacking.

"For me as a pilot, sitting in a cockpit, whether it was a helicopter or another aircraft, you are completely, 100 percent dependent on the accuracy of the computer system in order to have a safe flight and accomplish your mission," says Ephrat.

"If I'm comparing that to my time as a physician, practicing in hospitals in very intense environments, and seeing how very little you get reliable assistance from the IT systems, there's a tremendous, overwhelming difference between the two," he says.

"Think about the fact that both pilots and physicians are dealing with peoples' lives, with mission-critical activities, seeing how different it is from the experience perspective of the pilot getting very reliable assistance, compared to what I as a physician gets from the computer systems, you see how frightening the difference is," he adds.

The goal in launching medCPU was to reduce that gap, says Ephrat: "trying to reach a situation where the physician and the nurse at the point of care will have the same experience, in terms of credibility and reliability, that a pilot in a cockpit has."

Why has effective EHR-based CDS traditionally been so difficult – both for vendors to develop and for clinicians to use? "That's the billion dollar question," he says.

"The answer, at least from my perspective, is pretty simple, actually," says Ephrat, who founded medCPU with his colleague, company president, Sonia Ben-Yehuda. "And it all starts and ends with data. … The only thing that makes the barrier between the experience in the aviation world and with healthcare is the availability of data."

In the cockpit, he says, "100 percent of the information – everything about the flight – is available in the computer, so you become very accurate in doing the analysis. It becomes a critical assistance tool."

In the hospital setting, however, things are a bit different.

"The prosaic reason why it's not there in healthcare is that the way we communicate the clinical picture is mainly through narrative discussion," says Ephrat: "Follow-up notes. Visual summaries. Dictations. My experience is that somewhere between 50 and 70 percent of the data in healthcare resides in non-retrievable, non-usable information embedded in free text communication."

For the most part, that's a good thing: "free text communication is by far the easiest and most accurate way clinicians can communicate a clinical picture," he says. "But because the natural way of communicating is through natural language it's increasingly difficult to put it into the structured entries that health IT is expecting."

It's simply not possible to accurately describe a patient's condition "through checked boxes and pull-down menus," says Ephrat. But that's what's required in this data-driven age of quality reporting and value-based care.

"If 50-70 percent of the data is not entered into the right place on Epic or Cerner, the computer is bound to fail," he says. "If the computer doesn't see the complete picture you end up with alert fatigue, redundancy of prompts, and eventually something that really doesn't work like the experience I had as a pilot in the cockpit."

This is hardly the first time aviation has made its influence felt in a healthcare setting, of course.

In his 2009 book The Checklist Manifesto, surgeon and New Yorker staff writer Atul Gawande, MD, shows how pilots' exacting pre-flight list-making can be applied in hospitals for safer outcomes and saved lives.

And in a presentation at HIMSS13 in New Orleans two years ago, Robert Szczerba, corporate director of global healthcare initiatives at Lockheed Martin, made the case that the aerospace industry has much to teach healthcare.

The aviation giant's research on modeling and simulation, systems interoperability and data analytics have been applied in healthcare setting through initiatives such as Lockheed's  ICE STORM initiative. (It stands for Integrated Clinical Environment; Systems, Training, Operations, Research, Methods.)

No pilot would ever fly a plane without first practicing for hours on a simulator, for example. "Why can't you have simulation-based training in healthcare?" Szczerba asked. "The technology is out there. The capabilities are out there. It's in your hands."