Patient Photo Reduces Errors Between X-ray Image Sets

Pam Harrison

April 24, 2013

Adding a photograph of a patient to an x-ray image increases the likelihood that radiologists will detect labeling errors without requiring any additional interpretation time, new research shows.

"Sometimes an x-ray gets placed on another patient's folder, and if that patient is similar in age and sex, you can be easily fooled," Srini Tridandapani, MD, from the Emory University School of Medicine in Atlanta, Georgia, told Medscape Medical News.

"With a photo embedded with each x-ray at the point of care, our eyes can readily see the difference between patients. This helps radiologists detect any mislabeling that may have occurred between 2 sets of images," he explained.

These findings were presented at the American Roentgen Ray Society 2013 Annual Meeting in Washington, DC.

Dr. Tridandapani and colleagues obtained simultaneous portable x-rays and photos from 34 patients. From these, "we generated 88 pairs of chest x-rays — 1 recent and 1 previous — from which we compiled lists of 20 pairs for reader review," they report. The researchers then purposely inserted 2 to 4 mismatched pairs of x-rays.

Ten radiologist readers, blinded to the presence of mismatches, interpreted 20 x-ray pairs that were not accompanied by any patient photo. The same readers were subsequently given a second set of 20 x-ray pairs; these also contained mismatches, but were accompanied by a photo of the patient taken at the time of imaging. Readers were not told why the photos had been attached to the x-rays.

Wrong Patient

When no photos accompanied the x-rays, the mismatch detection rate was only 12.5. However, when patient photos accompanied the x-rays, the mismatch detection rate was 64% (P = .0003).

Interpretation times were faster when a photo accompanied the x-ray than when it didn't (26.5 vs 33.9 minutes), although the difference was not significant.

"We're not sure why this happened, but it could be because the photograph provided clinical clues that assisted radiologists in making the diagnosis," Dr. Tridandapani speculated.

In a second study involving 5 radiologists, readers were told to use the photos in their interpretation of the x-rays, although they were not told about any mismatch errors.

In that study, 94% of mismatch errors were detected, Dr. Tridandapani reported.

In these studies, an additional person was initially required to take the patient photos immediately after the x-ray.

However, the researchers have now developed a prototype in which a camera can be attached to a portable x-ray machine, so a photo can be taken without creating additional work for the technician. The researchers hope to implement their prototype in imaging facilities at Emory in the near future.

It is estimated that about 1 in 10,000 exams have wrong-patient errors.

When there is no patient mismatch, "technology like this could help a lot by informing us which lines and tubes are external and which ones are internal. It would also help identify positioning of the patient and sources of artifact," said Nabile Safdar, MD, from the University of Maryland Medical Center in Baltimore, who was asked by Medscape Medical News to comment on the findings.

Overall, Dr. Safdar noted, the simple attachment of photos to x-rays will improve efficiency and reduce alert fatigue on the part of intensivists. "There is so much rich information available in a visual light image," he noted. "If achievable, I don't see why you wouldn't take advantage of this."

Dr. Tridandapani and Dr. Safdar have disclosed no relevant financial relationships.

American Roentgen Ray Society (ARRS) 2013 Annual Meeting: Abstract 17. Presented April 15, 2013.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....