Here are four strategies to ensure your PACS is ready for an interoperable, digital healthcare system.
The University of Utah Health Science Center's picture-archiving and communication system currently generates about 15 to 20 terabytes of data annually. That is the rough equivalent to the number of cataloged books stored at the Library of Congress—if you estimate one megabyte of text per book. "We generate more data in a day than the rest of the hospital does in a year," says Richard Wiggins III, MD, director of imaging informatics.
PACS is often referred to as a space hog that monopolizes a hospital or health system's storage capacity, and as more organizations start digitizing mammography and implementing cardiology PACS, those storage needs are only going to increase. A chest X-ray may be about 10 megabytes in size, but one digital mammography image can be 40 to 50 megabytes, Wiggins explains. "Most larger academic sites are looking at about a 10% increase in the data that they generate every year."
Finding solutions to store the volume of data generated by PACS is an ongoing concern for chief information officers. So is ensuring that their facilities have the infrastructure in place to support a digital, interoperable healthcare system. It's a good thing IT executives are more PACS savvy today than they were just five years ago.
The majority of deals right now are PACS migrations, not installations, says Wiggins. Nearly 76% of the CIOs and technology executives surveyed for the 2009 HealthLeaders Media Industry Survey already had PACS in place at their facilities, and the 15.6% of those who didn't planned to implement it within two years.
"People are not coming into PACS without any idea of what it is and how it works," Wiggins says. "People are coming in with an idea of what they want it to do and accomplish."
Here's a look at how two institutions are ensuring that their PACS data is quickly accessible to physicians, radiologists, and clinicians from any location.
UUHSC, which implemented its first PACS in 1997, knew it wanted its next generation PACS to be fast, stable, and enterprisewide to get images to referring physicians and clinicians easily and quickly, says Wiggins.
The four-hospital system in Salt Lake City selected iSite PACS from Andover, MA-based Philips Healthcare because of its strong background as an enterprise solution. It began the implementation in January 2009 and finished migrating all of its data to the iSite system in September.
Prior to its new system, UUHSC's PACS was based on a distributed architecture. Studies were sent to an archive where they were sorted based on modality, region of the body—brain, knee, shoulder—and ordering physician. From there they would be stored on a specific computer. "I could go to that computer and know that it had a bunch of MRIs of the brain that needed to be read," says Wiggins. The good thing was if the network went down clinicians could still work because all of the images and studies were stored right there on the computer—but that was also the problem.
Under a centralized model, the study is still sent to an archive. The difference is clinicians can query the archive for a list of all brain studies that need to be read from any location. The system generates a list of studies that can be selected and viewed, Wiggins says. "In general, the industry is moving from a distributed to a centralized architecture because the networks are fast enough and smart enough."
But a centralized system also means organizations are completely dependent on the network. That's why it is essential to have a fast network with a lot of redundancy, says Wiggins. "So if somebody is doing construction a half mile away and cuts through your network, they don't leave you completely dead in the water."