Trouble Getting HIE Data in the ED? Maybe You Need a FHIR HIEdrant | Healthcare Informatics Magazine | Health IT | Information Technology Skip to content Skip to navigation

Trouble Getting HIE Data in the ED? Maybe You Need a FHIR HIEdrant

November 14, 2017
by David Raths
| Reprints
FHIR HIEdrant won first-place prize in Pitch IT contest at last week’s AMIA Symposium

When new patients with chest pain show up in an emergency room, physicians are often so busy that they don’t take the time to go beyond what is in their own health system’s EHR to search the regional or state health information exchange for relevant tests. A new SMART on FHIR app being developed at the Regenstrief Institute in Indianapolis retrieves that HIE data and integrates it in the clinical workflow in the EHR.

The team that developed the cleverly named FHIR HIEdrant won the first-place prize of $12,500 at the Pitch IT contest, a shark tank-like event held at last week’s AMIA Symposium in Washington, D.C.

Matthias Kochmann, M.D., a pediatrician and clinical informatics fellow at the Regenstrief Institute, said that as the team studied this problem, they first asked ED physicians which data elements residing in the HIE would be valuable to them. They mentioned five things:

• Last EKG

• Last cardiac catheterization report

• Last stress echocardiography

• Last cardiology note

• Last discharge summary.

One of the problems with accessing the HIE is that involves a separate workflow outside of the usual process of reviewing patient data in their EHR, without assurance that any meaningful data will be found. “We did a time-motion study to see how often ED physicians access the HIE in order to retrieve documents,” Kochmann said. “The result was that most of the time, they just do not have time. It wasn’t convenient for them. They would go with the information in the chart or provided by the patient.” If there is nothing in the EHR chart, then they order the tests.

The key point, he said is that the tests are sometimes invasive, expensive and take time. “Ideally, no tests would need to be repeated. The patient saves time; the health system saves money; and the physician can spent more time with the patient.”

The developers sought to bring that HIE information directly into the ED physicians’ workflow. They built a FHIR server connected to the Indiana HIE. “Using Smart on FHIR technology, we created a tab within the clinician’s work flow. They use it to open up the FHIR HIEdrant and it then searches the HIE for the physician,” Kochmann explained.

Starting Dec. 4, the FHIR HIEdrant will begin a four-month evaluation project at Indiana University Health Methodist Hospital.

If that pilot goes well, there are three ways the developers plan to branch out. One is to make the chest-pain app better with more functionality, Kochmann said. Another pathway is to try to get the app placed in the Epic and Cerner app stores. The third is to find different use cases in other hospital departments. One could be a surgery dashboard for pre-surgical clearance. Another could be pain management, which could be used in the ER, family medicine, or neurology.

The FHIR HIEdrant team is looking forward to another competition sponsored by Medstartr at the end of November in New York.



Get the latest information on Health IT and attend other valuable sessions at this two-day Summit providing healthcare leaders with educational content, insightful debate and dialogue on the future of healthcare and technology.

Learn More



Congratulations to the group. I would like to note however that that can be accomplished today in an HIE with or without FHIR as the protocol. The most valuable piece of information was, what does the use want. From there, the 5 documents lists can be viewed by the user using either a "tab" as noted in the project; which can be done with a patient context link into an HIE viewer. And it can be in the workflow TODAY if the EMR is integrated for pulling records. A custom on-demand document that aggregates for cardiac care can be created as the default for those users TODAY. What we need in the HIE world more, is users with specific requests that will get them to use HIE more (such as the list of 5 most valuable documents for this group of users) than it is protocol changes. That said, I have nothing against FHIR and I'm hopeful that it will increase population query integrations with HIEs. But we shouldn't be looking to protocols to solve what is really just a lack of good expression of what is needed and an implementation of what can then meet that need. This can be done today with common tooling that in many cases is hopefully already deployed.