High-volume shifts associated with increased discrepancy rates among radiology residents

High-volume shifts can cause discrepancy rates to increase for radiology residents, according to a study published in Current Problems in Diagnostic Radiology.

“Regardless of training level, errors do occur,” wrote lead author Benjamin Wildman-Tobriner, MD, with the department of radiology at Duke University Hospital in Durham, North Carolina, and colleagues. “There are both radiologist-specific factors and systems-based processes that contribute to errors. Study volume represents one such factor.”

With so many academic medical groups relying on residents to make preliminary reads after hours, the authors wanted to examine the impact of high-volume shifts on resident discrepancies. They retrospectively studied data from more than 500 weekend call shifts between 2011 and 2015 at a single institution, focusing on abdomen-pelvis (AP) CT exams and chest-abdomen-pelvis (CAP) CT exams.

Overall, Wildman-Tobriner et al. found that the average number of CT reads during each weekend shift increased each year; it was more than 58 in 2011 and more than 75 in 2015. More than 4,500 AP or CAP CT exams were interpreted during the study period, with an overall discrepancy rate of 4.8 percent. This included 80 category 1 (“incidental") discrepancies, 134 category 2 (“significant and time dependent”) discrepancies and 11 category 3 (“critical/urgent”) discrepancies.

To dive deeper into these numbers, the authors developed the concept of “peristudy volume,” which they defined as “the number of CTs read within 60 minutes (±30 minutes) of a given CT.” This helps determine how busy the residents were at a given time.

Peristudy volume was found to have “a significant effect on the likelihood of having a category 2 or category 3 discrepancy,” the authors wrote. In addition, the rate of category 2 and category 3 discrepancies increased as the overall volumes increased. From 2013 to 2015, for instance, when volumes increased at the institution, the discrepancy nearly doubled.

“Our residents had increased odds of making an error both when reading during overall high-volume shifts and when interpreting during high-volume moments in time,” the authors wrote. “Most radiologists are familiar with pressures related to turnaround time, the feeling of reading a study ‘too fast,’ and the intuition that something may have been missed. Indeed, our data showed a nearly threefold higher odds of a miss when reading during a higher volume shift.”

Wildman-Tobriner and colleagues noted that these statistics have resulted in changes for the institution’s call schedule and structure. A radiology coordinator was also implemented during peak hours of certain shifts. “Having a coordinator to help manage telephone calls and questions from providers and technologists can be extremely helpful for busy radiologists,” the authors wrote.