Khorasani details 3 building blocks to integrate EMR data into rad workflow

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
 - blocks

The adoption of EMRs brings a major performance opportunity for radiology, according to Ramin Khorasani, MD, MPH, vice chair of informatics at Brigham & Women’s Hospital in Boston and author of a column in the December issue of the Journal of American College of Radiology. However, the process of cobbling EMR data in to radiology workflow is riddled with obstacles.

EMR data, such as physician notes, operative notes and pathologic results, offer a rich source of relevant information that can help radiologists deliver high quality patient care, according to Khorasani. Existing strategies to integrate these data into radiology workflow are problematic.

Radiology requisitions, the conventional process for sharing clinical information, will not suffice for information beyond the basics required to justify the imaging exam. Thus, new processes are needed. Khorasani detailed the pros and cons of three options.

The first option, sending relevant data to the RIS or PACS, would provide radiologists with straightforward access to clinical data within existing radiology applications. The hitch is that most RIS and PACS are not configured to display EMR data such as discharge summaries and problem lists. This model also requires replicating and maintaining data in disparate databases. “As such, the strategy of sending relevant EMR data to your RIS or PACS is highly unlikely to be practical, scalable or ultimately successful,” wrote Khorasani.

Implementing and adopting EMR applications in radiology might provide a viable option. However, this approach presents workflow issues, including logging into multiple systems and navigating the EMR. Khorasani suggested IT integration activities to improve EMR utility in radiology. These are:

  • Single sign-on to enable end users to login to the PACS, RIS and EMR with the same username and password;
  • Launching the EMR in context at the PACS workstation, which eliminates duplicate data entry to sync patients in the PACS and EMR; and
  • Creation of a radiologist-centric view, such as a single screen with key data for imagers in the EMR.

“Without these three integration activities, the adoption of EMR applications for radiologists' use becomes inefficient and unscalable given current constraints in workflow,” wrote Khorasani.

The third option leverages system-oriented architecture (SOA), a framework that enables users to consume data from reusable components. For example, discharge summaries for a given patient might be packaged as a reusable component and made available to applications that can call for it using an established IT protocol. The model allows PACS, rather than the EMR, to define which data are displayed and how they are displayed to the end user.

The challenge with this approach is that it likely requires substantial modifications to existing IT implementations. “Although SOA shows great promise, it may not be practical for many practices,” wrote Khorasani. He recommended integrating the EMR into radiology as the most viable approach for the short term.