Home

Cover Story

Table of Contents

E-Newsletter

Article Archive

Editorial Calendar

Datebook

Writers' Guidelines

Orgs/Links

Opinion Polls

Reprints

Forum

Search


Florida Hospital College

 

February 26, 2007

Looking to the Future of PACS
By J. K. Bucsko
Radiology Today
Vol. 8 No. 4 P. 14

While you probably won’t end up with your head hardwired for DICOM, the next generation of PACS will be designed to share information across a broader network than most systems were originally designed to do.

At nearly 30 years old, PACS can validly be considered “mature.” Its technology is well-understood, its clinical usefulness universally acknowledged, and ongoing research reveals new secondary benefits—from faster diagnosing and easier consultations to systemic cost savings. But in a computer-driven world, one of technology’s chief ironies is that you can usually measure a device’s value by how frequently it’s modified and how swiftly any particular “state of the art” is superseded.

In PACS, a relatively thin but steady flow of enhancements through the early years swelled to a flood of significant innovations beginning in the late 1990s, as digital developments from the corporate and consumer arenas found their way into medical applications. A slew of advanced imaging, storage, and communication tools are finally bringing enterprise PACS within reach for many institutions. While that’s not an entirely new idea, today the final outcome is often envisioned more as a series of interlinked, collaborative departmental systems, with less emphasis on radiology as the hub controlling access to imaging files.

Following the lead of cardiologists who want to take ownership of advanced modalities such as cardiac computed tomography angiography and real-time 4-D MRI, other specialties are slowly evolving their own requirements for PACS usage. The emergency department probably tops the list, but orthopedists, neurologists, oncologists, pathologists, and surgeons aren’t far behind. Using tailored workstations, multiple departments and specialty practices would be able to operate like a standalone, dedicated PACS, but, crucially, they would operate equally as members of the larger collective. The idea is to make sure that as new PACS come online, they don’t end up as radiology PACS all too often have in the past—an independent island, or silo, of vital but difficult-to-share information.

To get some perspective on this envisioned wave of change, Radiology Today interviewed two experts in the field.

Sharing Intelligence
Richard S. Bakalar, MD, is chief medical officer of IBM Healthcare and Life Sciences and coauthor of the company’s most recent white paper, “Healthcare 2015: Win-win or lose-lose?” IBM, in partnership with Bycast Inc., makes the Grid Medical Archive Solution (GMAS), an intelligent storage management architecture designed to run under an enterprise PACS or imaging system. Current GMAS users include Iowa Health System, Geisinger Health System of New York and Pennsylvania, University Health Care System Augusta, and the University of Pittsburgh Medical Center.

Says Bakalar, “In the last five years, with the development of department-centric PACS into enterprisewide PACS, we’ve seen that it’s not just the 10% of clinicians who are radiologists using the same stored images. It’s the entire community across the enterprise.” He characterizes today’s medical imaging user groups according to four categories: “The diagnosticians—the radiologists; referring specialists who have to access clinical and imaging information at a very high level and do the integration in their assessment and report; primary care physicians who generally focus on the reports and use reference quality images to interpret for and to educate patients; and the allied health professionals who typically used film in the past but have now become somewhat orphaned as film becomes more rare and difficult to access.”

Integrating all these users into a cohesive network requires optimizing the total workflow and for that, he says, improving the ability to share long-term archives is critical.

“We’ve learned that the extension of departmental PACS medical archives is not a direct, linear relationship,” says Bakalar. “Archives designed for enterprises, or even hospitals, are much different than departmental archives. One of the greatest challenges is the silos of information that have accumulated through the successful implementation of PACS, primarily in hospital imaging departments and now in enterprises.”

Extending Imaging Services
However, he adds, when planners and administrators look to integrate departmental databases, including PACS, it’s important that they don’t confuse the interoperability of PACS applications with the intelligent storage management functions. There’s much more to putting in PACS than just giving everyone free access to image files: “Trying to accomplish [imaging data distribution] in a hospitalwide or even an enterprisewide setting brings out a lot of governance changes and service-related activities that have to be addressed before the archives can be expanded to broaden the PACS services.

“In the more traditional approach, you have a system administrator at each location to coordinate resources,” Bakalar says. “The problem is that most other specialties outside radiology don’t have PACS administrators. They don’t have the technical resources to manage workflow and systems as radiology departments do. So what’s happened is that information technology officers try to manage PACS throughout the enterprise. And that’s put them in a position of managing not only the PACS applications [and] the workflow applications, but also the networks.”

IBM’s solution for integrating the use of diverse data streams is the GMAS, which uses a service-oriented architecture that provides a universal access layer of software intelligence. This distributed intelligence enables supporting PACS applications wherever the appropriate hardware is interfaced. Bakalar says this solution, modeled from the IT world, offers medical imaging access to features that traditional PACS approaches can’t support, including the ability to use existing resources anywhere in the network; monitor network use and change network rules dynamically from a single portal; and provide hot copies of files, including DICOM files, as needed, even if some sites are inoperable.

Because GMAS focuses on managing fixed content imaging data rather than the transactional data that typically comprises clinical records and information systems, the system supports multiple specialties equally. Imaging data from any source—spinning disk, tape, online, hard drive—is seen by the applications as one virtual hard drive, independent of local operating systems, and the innate network intelligence automatically performs storage management functions without disrupting applications.

Shifting Emphasis
IBM believes shifting the emphasis toward a distributed and interactive model opens up medical imaging in ways the healthcare industry hasn’t fully formulated yet.

“When we’re talking about taking electronic health record [EHR] information, PACS information, other registry information and integrating all of that vertically, on a service level as opposed to an application level, I think we’re going to see a lot of extended benefits related to workflow efficiencies and better decision making by both providers and patients,” says Bakalar. In particular, he says, everyone stands to reap major benefits from new capabilities for integrating healthcare information into “superwarehouses” or “data-marts” of clinical information: “The idea is not simply to replace paper with digital records but to establish a secure digital, interoperable environment. [That will enable us to] perform data mining and advanced analytics [by] accessing large volumes of clinical trial data as well as individual longitudinal clinical records.”

Such complexities at present are only within the means of the largest marquee institutions, such as the Mayo Clinic. But, says Bakalar, “we can build the personal EHR incrementally. We can improve processes, optimize workflow distribution, maximize our ability to aggregate information, enhance data mining and analytics. All it requires is an IT infrastructure that’s already available in retail, in entertainment, in financial industries. Even though it hasn’t caught up in the healthcare space yet, it will be part of the future.”

National Goals
“Imagine a collaborative, distributed network that lets different sites schedule appointments for one another, lets radiologists share the reading workload across office and home sites, and lets a central administrative office monitor or control multisite electronic billing and multisite reporting without taking away each site’s autonomy.” That’s the vision of Cyrus Hazari, president of Meta Fusion, Inc. His company makes a line of scalable, integrateable PACS and RIS products, including the turnkey Helium, MetaProxy mini-PACS, and MetaWorld online imaging archive service. Meta Fusion is currently the only company to offer an Apple Macintosh-based PACS/RIS solution.

Hazari notes that Health and Human Services has delineated four main goals it believes must be reached sequentially: informing and interconnecting clinicians, personalizing care, and improving population health.

“The changing economics of healthcare require optimizing efficiency as an ongoing means of controlling costs and reducing erosion in the standard of care,” he says. “In today’s current market, [new] technologies are often key enablers for enhancing efficiency.”

And although they always want to protect their previous investments, facility administrators are more willing than ever to keep their options open for embracing inevitable enhancements. The solution for keeping the best from both worlds, according to Hazari, lies in deploying open, adaptable systems that will allow for accommodating legacy applications within the new, often rapidly changing workflow needs. The end result needs to appear seamless, even transparent, to the end user—wherever that user is within the enterprise.

Collaborative Distribution
“Medical information management systems will generally evolve by catering to the increasing demand for collaborative systems and applications,” Hazari says. Anticipating and providing for such “continuity of growth” from the outset is going to be a significant factor as PACS makes its way into the full enterprise—not just institutions and localized healthcare systems but regional healthcare information organizations (RHIOs) and the eventual national healthcare information network (NHIN).

Hazari notes, “The notion of a unifying NHIN presents more potential interactions and services and richness of capabilities than merely the flow of EHRs.” He poses the example of an independent, standalone imaging center in today’s volatile environment. “If and when a viable NHIN comes into effect, and the EHR is a routinely flowing envelope for quanta of health information, why would a ‘standalone’ facility not want the benefit of incorporating [each] patient’s EHR into the path of diagnosis, just as a larger hospital may do? [Doing so would both] improve the healthcare outcome and reduce the professional liabilities of the healthcare provider.”

He continues, “In addition, imagine that [same situation] with the advent of an NHIN and with the establishment of EHR as an industry standard. For example, imagine being able to [coordinate] a hospital admission or an outpatient elective procedure from [the referring physician’s] office without any human in the path.”

Expanding the Departmental Model
While many in the healthcare industry cautiously share the greater enterprise vision Hazari outlines, it’s no news that fiscal considerations heavily tip spending decisions for healthcare technology. Recent moves toward reining in imaging reimbursements could mean, he says, “the [already delayed] incremental modernization of the healthcare entity gets postponed yet again.” The immediate result for most institutions is likely to be even more emphasis on sharing as much as possible, especially information, but sometimes also hardware, software, and even staff resources.

Meta Fusion’s scalable PACS and information management systems act “like piping and wiring,” says Hazari, to provide a reliable infrastructure for just that kind of distributed collaboration. For example, being able to count on such connections allows the same basic solution to serve, “a single radiologist doing night reading for one or more hospitals or a single modality imaging clinic providing Web access for referrers and a single reading radiologist,” all the way up to integrated multiple care entities with multiple sites, such as a metropolitan healthcare system or a geographically scattered RHIO.

And because “a mixed environment was contemplated as the norm from the outset of our design,” says Hazari, Meta Fusion RIS and practice management systems are multiplatform. They can run on a number of popular operating systems (including Windows, UNIX, Linux, and Solaris) and accommodate a wide range of DICOM workstations, viewers, and other end user tools. This ensures that clients aren’t locked into a single vendor or single system solution, a key to eventually supporting varied PACS usages throughout the larger network.

Distributed intelligence helps system administrators deal with a number of critical issues that inevitably crop up as local and regional health information networks make their way to the NHIN. In Hazari’s words, “Business cannot stop and wait for prospective legally mandated reform” to resolve ongoing concerns about data security and access authorization; data location and storage; transport; mapping and delivery; and audit and authentication services, to name only a few obvious problems that come with extending the network. Meta Fusion’s intelligent infrastructure is also self-restoring across the network. “If communications or accessibility of information are temporarily impaired at one or more facilities, the systems gracefully self-adjust and provide continuity of operations for individual entities and clusters of entities … until the whole is restored,” says Hazari.

Influences Outside Radiology
A number of factors—not all directly related to imaging needs per se—will most likely influence PACS spending and implementation decisions in the future, says Hazari. He cites “institutional learning from prior systems, better IT resources with greater domain knowledge, core emphasis on quality and reliability, and objective evaluation of value.” In particular, he says, as reimbursements tighten while service demands grow, institutions will put progressively greater importance on realizing objective, measurable efficiencies.

The current interest in adding PACS coming from specialties such as cardiology, pathology, orthopedics, and emergency medicine is creating new opportunities for vendors and users alike to develop distinctive applications, says Hazari—often derived from or enhancing existing imaging programs yet retaining the original operating characteristics. He anticipates many advances springing from user-defined requirements and even user experiments to make existing PACS products do things they weren’t intended to do. And as the emphasis on vertical applications catering to specialties strengthens, it “creates the future possibility of uniformity of PACS management tools across vendors,” Hazari says—although that day may still be fairly far off.

— J. K. Bucsko is a freelance healthcare and technology writer based in Westville, N.J., and frequent contributor to Radiology Today.


For More Information
www.bycast.com

www.hhs.gov/healthit/goals.html

www.metafusion.net

www-03.ibm.com/industries/healthcare/doc/content/resource/insight/1684579105.html



Copyright © 2007 Great Valley Publishing Co., Inc.
3801 Schuylkill Rd • Spring City, PA 19475
Publishers of Radiology Today
All rights reserved.