Radiology IT system hiccups that result in downtimes are generally mild and only modestly impact radiologist workflows, according to a study published online Nov. 4 in the Journal of the American College of Radiology.

Even as radiologists work in increasingly complex, multi-site medical systems, they are still being called upon to produce high quality, accurate work in a timely manner. To meet these challenges, wrote the study’s authors, including Murray Becker, MD, PhD, with the Department of Radiology at Rutgers Robert Wood Johnson Medical School in New Jersey, enterprises often create complex IT informatics infrastructures that can aggregate cases from multiple sites and distribute them to geographically dispersed radiologists.

For this study, Becker and colleagues investigated a live system used by a hybrid academic-private practice that aggregates 900,000 studies per year from six hospitals, 10 outpatient imaging centers and a number of off-hours sites. The researchers compiled all unplanned downtimes of the IT system over a 12-month period using five downtime metrics: duration, etiology, failure type, extent and severity.

In the year-long time period, there were 117 unplanned downtimes. The median outage length was 3.5 hours (34 percent were less than an hour and a half and 30 percent were greater than 12 hours).

Software malfunctions caused 87 percent of the outages, while 13 percent were attributed to hardware malfunctions.

As far as extent of failures, all sites experienced downtimes, but they were always localized to a subset of sites and no complete system-wide failures occurred.

In terms of severity, 47 percent of the outages had minimal impact on radiologists’ workflow and 3 percent severely impacted workflow.

“Downtimes have a relatively modest impact on radiologist workflow for two reasons:

1. Usually, they are caused by quickly resolved software errors; and
2. The architecture of the system incorporates inherent redundancy, because data are duplicated in both the hospital IT and the Super PACS or common dictation systems,” the authors wrote.

Additionally, they found, other IT system redundancies like server virtualization and redundant network connections helped bolster system-wide uptime.

Becker and team concluded that the study showed that it’s possible to measure and characterize downtimes in radiology IT systems.

“The very nature of a complex, multisite system with disparate PACS and radiology IT systems clearly increases the number of data transfers by adding transfers between the hospital system and Super PACS or common dictation systems, and interruptions of these transfers most commonly causes downtimes, but the impacts on radiologist workflow are relatively modest and manageable, and in and of themselves, are not a barrier to delivering radiology 24 hours a day, every day of the year,” the authors wrote.