Computer-led protocol selection could automate workflow, save radiologists time
As their duties are seemingly always piled sky-high, radiologists often wonder which aspects of their job they can trust to automated technology. Selecting protocols for standard CT and MRI exams could be one of them, a group of Detroit researchers reported this month in the Journal of the American College of Radiology.
“As radiology transitions from volume to value-based reimbursement, methods to automate or eliminate time-consuming tasks with low value contribution are becoming increasingly necessary,” first author James Tudor, MD, and colleagues wrote in JACR. “When changes in examinations happen during the process of protocol selection, they tend to follow predictable patterns, which makes protocol selection of most radiology orders for advanced imaging amenable to workflow automation.”
The majority of protocol work falls to radiologists, residents and radiologic technologists at the authors’ locations within the Henry Ford Health System, they wrote. Still, those hospitals perform around 1.5 million imaging studies each year—including 250,000 advanced CT, MRI and and nuclear medicine studies—and their in-house health system has been electronically selecting protocols since 2008.
“Since that time, our department has collected information regarding ordered studies and how the protocol was determined before completion,” Tudor et al. said.
The researchers retrospectively compiled a data set of more than 237,000 cross-sectional examinations from four hospitals and isolated the 12 most frequently requested exams, which represent nearly half of the institution’s advanced imaging studies. The team then compared protocol behaviors between technologists, radiologists and residents.
Between April 2015 and November 2016, 128,222 exams were ordered at the primary hospitals, while 146,091 exams were ordered at satellite locations. Initially, Tudor and co-authors wrote, it seemed as if radiologists altered the exam code more frequently, but the data explain those shifts are likely due to either consolidated multiple orders or clinicians ordering general exams with special, organ-specific protocols.
The authors said the results suggest protocol could be a workable target for reducing workload among radiologists—but there will always be specific moments when specialists are needed to protocol certain exams.
“Based on these data, we conclude that technologists rarely change examination protocols from the ordered study,” Tudor and colleagues wrote. “These results suggest that a radiologist or technologist selecting protocols of the most frequently ordered common CT and MRI examinations rarely results in value-added activity, regardless of who makes that protocol decision.”