Wearing a face mask may increase dictation errors in radiology reports

Wearing a face covering may increase physicians’ rate of dictation errors in unedited radiology reports, according to a new analysis published Friday.

Many healthcare employers have mandated the use of masks amid the COVID-19 pandemic. But this can present problems when radiologists read their recommendations into common speech-recognition software, experts detailed in the Journal of Digital Imaging.

Wanting to explore this challenge, University of North Carolina researchers conducted a prospective study, charging six randomized rads with reading with and without a face covering. They believe masks may impact the specialty’s work and suggest further exploration on this topic.

“We conclude that wearing a mask while dictating results in at least a near-significant increase in the rate of dictation errors in unedited radiology reports created with speech recognition, a difference which may be accentuated in some groups of radiologists,” corresponding author Benjamin Mervak, MD, with the UNC School of Medicine’s Department of Radiology, and colleagues wrote Sept. 24. “Notably, however, most errors are minor single incorrect words and are unlikely to result in a medically relevant misunderstanding.”

To reach their conclusions, the team tasked rads with dictating word-for-word content from 40 model report two times, including once with a mask. This resulted in 80 dictations for each individual and 480 total across the study group. Mervak et al. then measured these new dictations up against the model reports and calculated the frequency of mistakes.

Unmasked radiologists logged roughly 21.7 errors per 1,000 words, the study found, compared to 27.1 in the masked group. However, 1 of the 6 participants was a rad resident with an accent who recorded a higher error rate. Removing that individual from the study dropped the masked error rate to 20.1 per 1,000 versus 16.9 without a face covering (a 19% difference). Incorrect-word errors were somewhat higher among masked group of five attendings, while other types of dictation errors did not occur at a significantly higher clip. Meanwhile, error rates appeared higher for MRI and radiography compared to CT.

The majority of mistakes were deemed “clinically consequential,” often resulting from a single replaced word.

“Additional errors resulting from mask-wearing would therefore also be expected to be minor,” the authors advised. “On one hand, minor errors are a nuisance in that they can affect the perceived quality of radiology reports, including from a medicolegal standpoint, or put undue pressure on the reader when assessing the results of a radiologic study…On the other hand, minor errors do not generally impact patient care, making them of far lesser clinical importance than moderate or severe errors.”

Around the web

The Department of Defense is gifting Case Western researchers a grant to study the use of AI in determining whether patients require surgery. 

Research publications by radiologists increased by 167% during the second year of the mentorship program, doctors at West Virginia University reported.

The American Medical Association highlighted Medicare spending for all specialties in a new report. 

Trimed Popup
Trimed Popup