Smile! Including patient photos with medical imaging may reduce doctor error

A study jointly conducted by Emory University and University of Arizona found that including patient photographs with medical imaging can reduce wrong-patient error while maintaining speed and accuracy.

Misidentification errors in radiology may be more common than expected, due to the uncertain nature of radiologic interpretation. The answer isn’t always normal or abnormal, cancer or no cancer. Rather, radiologists weigh the evidence and deliver opinions on a diagnosis.

According to a Pennsylvania study, 30 percent of serious errors in radiology were errorswhere medical images are attributed to the wrong patient. One method to decrease the incidence of wrong-patient error is to include photos of the patient’s face with his radiographs. Radiologists, however, found the photos distracting and thought they required more time per reading, according to an observational study.

Interestingly, the average interpretation time of 60 seconds per radiograph did not change with the introduction of the photo, so researchers from Atlanta's Emory University and the University of Arizona employed eye-tracking technology to investigate why there were no significant differences in interpretation time.

Ten radiologists viewed chest radiographs with and without patient photographs while eye position data was collected. A pre-and post-study survey collected basic demographic information as well as questions about how radiologists acquire information.

Researchers found that overall time viewing the cases did not differ with the photograph included or not, and the surveyed preference for including the photographs was 6.1 in 10. On average, readers scanned to the photographs about four times per search, and men spent significantly more time looking at the images. They noted that radiologists seem to take about 60 seconds per reading no matter the task or setting.

“This suggests that whether they are conscious of it nor not, the radiologists are adjusting their perceptual and cognitive strategies so accuracy and speed are maintained,” wrote Krupinski et al.

While there are structural limitations with the study, and there is the question of hospitals actually incorporating the photo acquisition into their workflow, the use of photographs as a patient identifier reduced wrong-patient error without a substantial increase in time per interpretation. With 98,000 deaths per year due to medical error, a low-investment improvement like this may be an easy decision.

Read the full study at Science Direct.com.