New emergency rad-report system passes test

Radiologists and ED physicians at Brown University have developed a simple, five-category system for triaging imaged emergency patients based on their radiology reports, and the team’s test of the system has shown very good interobserver agreement.

Radiologist David Swenson, MD, emergency medicine physician David Portelli, MD, and colleagues describe their work in an article published online Oct. 8 in Emergency Radiology.

Called RADCAT, the system places imaging exams in categories from normal through life-threatening.

The team developed and tested the approach in the emergency room of Brown-affiliated Rhode Island Hospital/Hasbro Children’s Hospital in Providence.

In the pilot phase, four radiologists used the system real-time to categorize around 400 radiology reports in the ED.

From this set of reports, the researchers semi-randomly selected 58 to be de-identified, stripped of their original categorization and recategorized based on the narrative radiology report by six rads and six ED docs.

The team found agreement proved substantial among radiologists and ED physicians (κ = 0.73, p < 0.0001), as well as within each of the five categories (all κ > 0.60, p < 0.0001).

The lowest agreement levels occurred with RADCAT-3, which indicated “important but non-urgent,” while the highest agreement was with RADCAT-1, for normal imaging exams showing no abnormalities.

“Our RADCAT system is understandable between radiologists and ED physicians for categorizing a wide range of imaging studies, and warrants further assessment and validation,” the authors write. “Based upon these pilot results, we plan to adopt this RADCAT scheme and further assess its performance.”