Bad call: Telephone interruptions linked to higher radiology diagnostic discrepancies

Calls to a reading room within one hour of a resident generating a preliminary report increased the odds of a significant error by 12 percent, according to a study published online Sept. 30 in Academic Radiology.

"To the best of our knowledge, our research is the first to link distractions to errors in radiologic image interpretation," wrote lead author Brad J. Balint, of Indiana University Health Methodist Hospital in Indianapolis, and colleagues.

The study focused on interpretations from a pediatric radiology reading room over the span of 13 months. Resident radiologists were responsible for the generation of preliminary interpretations during a 14-hour call shift. A faculty radiologist was not present but was available for consultation during this period. All residents were in their third or fourth year of radiology residency training.

Over the study period, 15,504 resident preliminary reports were generated, while 14,950 calls were logged during these shifts. The number of calls in the hour preceding each discrepancy were recorded.

Of the 338 studied call shifts, 41 shifts logged 51 major discrepancies. Of these, 39 had documented error details and resident preliminary report time stamps.

The most frequent discrepancies were extremity fractures, chest infection and pneumothorax.

While there was no significant difference in calls per shift between discrepancy and non-discrepancy shifts, there was a significant difference in the number of calls in the hour preceding each discrepancy. There were an average of 4.23 calls in the hour preceding each documented discrepancy versus an average of 3.24 calls in the one hour period preceding an hour without a discrepancy.

More than half of the total discrepancies (53.8 percent) occurred during the first half of the call shift (7 p.m. to midnight) and an additional 10 discrepancies (25.6 percent) occurred during the last few hours of the shift.

The average telephone call length was 114.9 seconds, and more than half of all calls (54 percent) were less than one minute in length.

In the hour preceding the generation of a discrepant resident preliminary report, a single additional phone call above the average baseline increased the odds of a major discrepancy by 12 percent.

While authors noted that the radiology resident's primary responsibilities were generating imaging examination reports, they were also responsible for tasks like answering telephone calls, returning pages and in person-consultations.

"These functions are often performed simultaneously with the resident forced to toggle between tasks, frequently before the completion of the previous task," wrote Balint and colleagues. "These workflow disruptions are generally not scheduled, forcing the radiologist to disengage from the current task without completing it and 'forgetting' not only the specific task they were once doing but also the focused mind-set that existed before the interruption."

Despite the fact that residents will have other responsibilities besides simply churning out reports, the authors suggested that efforts should be made to minimize distraction in the reading room. "Ultimately, a compromise must be made that balances radiologist accessibility by consulting health care practitioners and at the same time minimizes unnecessary interruptions."