There's an increasing movement within radiology to provide around-the-clock coverage, particularly in academic settings where residents are increasingly working overnight shifts, and some institutions are even providing 24 x 7 attending coverage for some studies, said lead author Dr. Thomas Loehfelm, PhD, now at Stanford University.
"Even private practices are starting to hire in-house night-shift radiologists, rather than using external nighthawks for preliminary reads and finally reading the overnight studies the following day," he told AuntMinnie.com. "Clinicians, particularly in the emergency room, are starting to expect these final reads."
The research team was also interested in quantitative metrics to gauge resident performance, and Emory's overnight call system provided the best opportunity to assess this, Loehfelm said. At Emory, residents provide preliminary interpretations on overnight CT, ultrasound, and MR; other residents and faculty read studies the next day and indicate whether they agree with the overnight resident's preliminary interpretations.
"This is one of the few times in residency where residents work mostly independently, and where we gather objective metrics such as the agree/disagree rate," Loehfelm said.
To identify specific factors that affected resident performance as measured by the discrepancy rate, the team examined nearly 18,500 studies that had been interpreted preliminarily by residents. Overall, the residents performed well on their "night float" duty, with a major discrepancy rate of only 1% to 2%, comparable to discrepancy rates found when attendings peer-review each other, Loehfelm said.
The researchers also discovered that less-experienced residents had a higher major discrepancy rate on busier nights, but the more-experienced residents performed well regardless of how busy the shift was, "suggesting that their additional training helped them effectively manage the higher workload of busy nights."
In a surprising finding, the number of consecutive shifts did not affect performance.
"This was unexpected and counterintuitive to the [Accreditation Council for Graduate Medical Education] rule that residents cannot work more than six nights in a row," Loehfelm said. "Our trend was [actually] toward improved performance with more consecutive nights worked, perhaps because the resident was getting better adjusted to the night-shift hours."
What else did they find? Check out their poster on Monday afternoon to learn more.
