Interruptions during mammogram reading increase inaccuracies, radiologist burnout
When a physician is interrupted while reading a mammogram, it has adverse effects on their overall performance, as evidenced by higher recall rates, decreased accuracy, and reduced sense of job satisfaction, according to a study published in Journal of the American College of Radiology. [1]
Researchers from Duke University Medical Center recruited nine radiologists who specialize in breast cancer to interpret 150 mammogram screenings, previously conducted and accurately analyzed between 2008 and 2015. The 150 cases consisted of 125 mammograms that were benign and 25 with a diagnosis of subtle breast cancers.
In order to properly evaluate the impact of interruptions on their ability to make a diagnosis quickly and accurately, each reading made by the radiologists was placed into one of two categories: interrupted and not interrupted. Additionally, after each reading, the radiologists completed a survey assessing mental and physical effort, frustration levels, and performance satisfaction.
Measuring clinical performance revealed that recall rates were significantly higher (p=0.04) during interrupted reading sessions (35.4%), compared to uninterrupted sessions (31.4%). Accuracy also was significantly worse (p=0.049) in interrupted reading sessions (69.5%) compared to uninterrupted sessions (73.6%).
While the time it took to read each mammogram was barely affected by interruptions, radiologists reported feeling busier and experiencing elevated levels of physical fatigue during interrupted reading sessions.
“Although differences in reading time did not reach statistical significance, there was a trend toward shorter reading times during interrupted reading. The effect of interruptions on reader experience was also mostly negative with significant deleterious impacts on five of seven survey questions related to physician well-being,” lead author Sora Yoon, MD, and her colleagues wrote. “These findings raise important concerns about the tangible and intangible impact of interruptions on breast radiologists.”
It is important to note that while the results show interruptions did contribute to feelings often associated with physician burnout, the nine radiologists reported no feelings of insecurity over their ability to make an accurate diagnosis. Further research is necessary to better understand the impact distraction has on job satisfaction and physician well-being, especially given the nuances in the survey results.