PACS showdown proves a winner -- sort of

May 16, 2006

The first-ever PACS showdown had many highlights, not least the sheer technical accomplishments needed to pull it off. What it won’t have is the public announcement of an official winner.

The first-ever PACS showdown had many highlights, not least the sheer technical accomplishments needed to pull it off. What it won't have is the public announcement of an official winner.

The showdown at the annual meeting of the Society of Computed Body Tomography and Magnetic Resonance in April involved six vendors: Agfa Healthcare, Amicas, Fuji, GE Healthcare, McKesson Provider Technologies, and Philips Medical Systems. Some vendors expressed apprehension about publicizing the outcome, while others welcomed it.

"The society asked us if we wanted the results released, and we said yes," said Barry Gutwillig, executive director of marketing and business development for Amicas. "We wouldn't be here if we didn't feel we could rank. I think it's a testament to the vendors who are here and maybe more so to the ones who are not."

The vendors with more sensitive concerns, however, won this debate. To ensure a healthy participation at the next showdown, the SCBT/MR will not publish how the audience ranked each vendor.

But if a dozen attendees were asked to rate PACS, each vendor's name would rank at the top at least once. Workflow function, like beauty, is in the eye of the beholder.

In one test, for example, the range of "excellent" rankings for each vendor went from 13% to 53%. This variable pattern remained consistent with each demonstration.

While no clear winner emerged for Dr. Richard Hallett, a cardiovascular radiologist at Riverview Hospital in Indiana, he selected - when pushed - McKesson, Agfa, and Fuji as his first-, second-, and third-place favorites, respectively.

"I realize this showdown is a small fraction of everything PACS entails, but just by the demo cases and radiologist interaction, I got a good feel for how user-friendly and robust the interfaces are," Hallett said.

Dr. Dennis Foley, chief of digital imaging at the Medical College of Wisconsin, said that Philips, McKesson, and Fuji performed reasonably well in handling large data sets and doing relatively routine daily work. None of these companies, however, had well-integrated 3D solutions in their packages, he said. His nod went to GE for its hanging protocols, exam comparison, access to prior reports, and recovery from interruption.

An electronic survey of the audience of about 150 revealed that two-thirds had a PACS. Slightly more than half of them owned one system, while the rest owned two or three. Surprisingly, 64% of these were dissatisfied with their PACS, and 32% will be shopping for a new system within the next year. Two-thirds of attendees read CT studies on PACS, while 22% use 3D workstations. Another 42% are shopping for an advanced visualization workstation to interpret CT.

The first test put to the PACS was handling multimodality studies. The second scenario focused on workflow and included an interruption. The final demonstration asked the radiologists to maneuver through a large trauma CT angiography data set.

The audience rated key aspects of PACS in these scenarios:

  • ease of navigation and use of standard tools

  • speed, especially through large data sets

  • navigation among hanging protocols

  • ability to re-sort prior studies when interrupted

  • integration with 3D and advanced imaging

"How PACS integrate advanced imaging is a hot and very rapidly changing process," said comoderator Dr. Eliot Siegel, chief of radiology and nuclear medicine at the VA Maryland Health Care System. "It's fascinating that the PACS vendors are embedding multiplanar capabilities into their products in addition to integrating with 3D and advanced workstations."

Who eventually will do advanced imaging is not yet clear, Siegel told Diagnostic Imaging, calling it an interesting tug of war. It could be part of a routine PACS, or it could be a very tight integration. Or the advanced visualization vendors could begin driving some of the workflow. The trend for now seems to be using workstations to interpret the majority of CT studies.

More than a few attendees, while happy with their PACS, were surprised to find their system was capable of performing at a much higher level. Attendees seemed to cherry-pick certain functions in the different scenarios. One liked the way the mouse moved for GE in a certain exercise. Someone commented on the Philips' color scheme for a particular window. Another was partial to Amicas' seemingly infinite number of allowable open windows.

The demonstrators' personalities also influenced opinions, as an attendee described one as very efficient and another as too humorous.

As radiology increasingly depends on PACS, however, the focus of the competition should move away from counting clicks and toward assessing workflow efficiency, Gutwillig said. The intervariability between vendors is getting small. The question is now about the C in PACS, the communication. It's not the people or the pictures. It's about how workflow is distributed.

"We didn't talk about how quickly you can go through and interpret a case. We talked about how those tools translate into productivity and how productivity across your entire department or practice comes into play as well. It can't be done in this kind of scenario, but it's a very interesting question to ask," he said.

For more information from the Diagnostic Imaging archives:

Radiologists face repetitive stress injuries

TRIP evolves to keep up with image overload

Computer simulation predicts impact of equipment on PACS

Workstations progressing, but still lack intuitive function