Do severe injuries distract radiologists from finding minor issues in the same image?

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

When radiologists miss one image abnormality due to the presence of another, it is known as the satisfaction-of-search (SOS) effect. Do significant abnormalities lead to a greater SOS effect?

A recent study published by the Journal of the American College of Radiology showed that severe injuries do not lead to an increase in missed abnormalities. The authors said this result was not what they had anticipated going into the study.

“We had expected, as do many radiologists, that the most serious injuries are responsible for the greatest portion of SOS errors,” wrote Kevin M. Schartz, PhD, University of Iowa Roy J. and Lucille A. Carver College of Medicine, and colleagues. “Our present results, together with those of a prior study, fail to support this hypothesis. The clinical significance of a detected abnormality does not seem to determine whether it will suppress the detection of another abnormality.”

Schartz et al. tested their hypothesis by finding 35 cervical spine CT cases that showed severe spine injuries. Each case was paired with a similar case that showed no injuries, but then 16 of those similar cases had a minor fracture added via image modification software. The authors then asked 20 different residents and fellows to read the cases, noting each time one of the minor fractures was missed due to the presence of a severe spine injury.

Overall, the team found no significant evidence of any sort of SOS effect on the detection of the simulated minor fractures.

“There was a nonsignificant finding of poorer detection in the presence of cases with severe injuries,” the authors wrote. “However, the magnitude of the effect was no greater than has been observed for less severe distracting injuries.”

The team also analyzed the readers’ median inspection times. Cases that included software-added fractures took the readers 18 seconds longer than those that did not, and cases with “major native injuries” took 42 seconds longer than those that did not.

“Reading time was directly related to how much injury was present,” the authors wrote.