Worklist sharing adds balance to radiologist reading load
The average diagnostic image reading volume of a radiologist working in the United States, according to a disparate variety of sources, is between 17,000 to 20,000 or more studies annually—depending on the procedure mix. Outliers exist on both ends of the spectrum, and chest x-rays are easier to clear than a mammogram or neurological MRI study; however, it is clear that radiology groups are confronting an increasing number of exams each year.

Worklists are a tool used to shuttle each day’s exam volume among members of a practice. A radiologist will come in, open his or her worklist, and spend the next several hours reading the queued studies. Although this system works to clear the cases, it may not be the most efficient method to manage workflow, according to a scientific paper presentation at the 93rd scientific session and annual meeting of the Radiological Society of North America (RSNA) last month in Chicago.

“Despite the proliferation of PACS, worklists in many radiology departments are not widely shared among radiologists,” said Mark Halsted, MD, associate professor of radiology and chief of research for the Radiology Informatics Research Core at Cincinnati Children’s Hospital Medical Center.

Halsted presented the results of research conducted at the institution with the utilization of RadStream (sold commercially by Amicas); radiology worklist management software that runs in parallel with a single vendor combined RIS/PACS/speech recognition system to redistribute the departmental caseload across a radiology practice. The application was designed and developed at Cincinnati Children’s Hospital Medical Center by Halstead and his team.

“Queuing theory suggests that more broad sharing of radiology case worklists for triage and case distribution purposes can reduce overall departmental report turn-around times,” Halsted said. “Our study investigated whether combining radiology worklists in a large academic radiology practice reduced overall departmental report turn-around times.”

Citing the example of workflow in a bank, Halsted noted that it is more efficient when customers are queued in a single line and move to the next available teller than when they form multiple lines in front of each teller on duty.

Halsted said that his team used industrial operations management methodology to measure the degree to which increased sharing of worklists by multiple radiologists reduced overall report turn-around times.

He reported that despite the fact that the total departmental caseload volume increased following implementation of the new workload distribution software while the workforce remained stable, report turn-around times decreased for all case types.

“Median case reading time, defined as the time between case availability on PACS and time of case dictation, decreased 18 percent for the emergency department, 3 percent for inpatients, and 29 percent for outpatient radiology cases,” Halsted said. “Median sign-off time, defined as the time between transcription and final staff signature on the report, decreased 29 percent for the emergency department, 97 percent for inpatients, and remained zero minutes for outpatient reports.”

Halsted said that by using the RadStream tool, his facility was able to combine worklists and facilitate decentralized, dynamically balanced workloads across a group of radiologists and significantly shorten report turn-around times.