BBI Contributing Editor

CHICAGO, Illinois – The annual meetings of the Radiological Society of North America (RSNA; Oak Brook, Illinois) are always thematically cacophonous, as more than 2,000 researchers and 650 vendors vie for attention from the 60,000 attendees. RSNA 2000, held in the gigantic McCormick Place exhibition center along the Lake Michigan shoreline the first week of December, was no exception. No single theme stood out, unless it was the onrush of conversion of all imaging to digital methods. But a new theme has emerged among the large imaging equipment and film vendors, who are transforming themselves from suppliers of equipment and materials to suppliers of digitally-based services.

Traditionally, the large equipment companies – GE Medical Systems (Waukesha, Wisconsin), Hitachi (Twinsburg, Ohio), Marconi (Cleveland, Ohio), Philips Shelton, Connecticut), Siemens (Iselin, New Jersey) and Toshiba (Tustin, California) – have seen themselves as suppliers of "big iron," that is, equipment they could sell and install and then leave alone except for occasional maintenance and upgrades. Similarly, the film companies – Agfa (Ridgefield Park, New Jersey), Fuji (Stamford, Connecticut), Eastman Kodak (Rochester, New York) and Konica (Wayne, New Jersey) – saw themselves principally as suppliers of a material – X-ray and digital recording film – and related equipment such as developers and printers. The advent of picture archiving and communication systems (PACS) – electronic image management systems – has changed all that. For a long time, the equipment vendors tried to treat PACS as if they were just another kind of equipment, while the film companies treated PACS as if they were just an alternative to using film. For as long as they maintained these views, they were ineffective players in the PACS arena, leaving success in the field largely to homegrown systems and to companies that specialized in PACS.

The last few years have seen a sea change. The large suppliers have recognized that PACS require more continuous attention, starting with installation and being sustained throughout the systems' lifetimes. Unlike, say, computed tomography (CT) or magnetic resonance (MR) machines, which are self-contained and have as a rule at most one connection to other systems, PACS are connected to multiple machines and to other information systems, typically provided by a variety of vendors. Making these machines and systems exchange data reliably and consistently is one of the major challenges of a PACS installation, and, as everyone familiar with computer networks knows, it is an ongoing challenge that seldom is solved once and for all. Furthermore, installation of PACS entails substantial changes in workflow and work patterns, and institutions adopting PACS must accommodate these changes.

For these reasons, suppliers have retooled themselves to offer more complete services, including planning, installation, interfacing, facility management, using the Internet, remote storage, maintenance, upgrades and ongoing training, and financing alternatives such as pay-per-study. Notable among these new configurations are Application Service Provider (ASP) models, where vendors offer to store and manage images off-site using the Internet or other communication lines so that the hardware and software necessary to operate a PACS on site is less expensive and requires less maintenance attention from the owner. Positioning themselves to deliver these services has led to a number of partnerships and acquisitions, as summarized in Table 1.

Probably more important for Siemens than gaining experience with ASPs, however, is gaining access to a large installed base of hospital information systems (HIS). As all hospital records become digital, the concept of the electronic medical record (EMR) embracing all kinds of data – alphanumeric, tracings, and images in color or grayscale – makes increasing sense. HIS contain some of the information for the EMR, and PACS contain another part; integrating them is a step along the road to the EMR. Starting from the HIS is probably easier than starting from PACS.

Nevertheless, there is more than one way to fulfill a recognized need like the integration of PACS and HIS, and partnerships or mutually beneficial cooperation are also promising, as the list of partnerships in Table 1 testifies. Because there already is a large installed base of HISs and a growing installed base of PACS, the stance that vendors of either kind of system have to take is that they can interface their system satisfactorily to any other vendor's system.

This is the point of view embodied in an initiative repeated for the second year at this year's RSNA meeting: Integrating the Healthcare Enterprise (IHE). As it had last year, IHE demonstrated that the 32 participants were able to exchange images and other data between HIS and PACS in the course of carrying out several predefined scenarios where such interchanges are necessary. For example, one scenario concerns images made of a patient who arrives at an emergency room unconscious so that an arbitrary name like John Doe must be assigned to his records and images. Subsequently, when the patient's real name is determined, all files, including image files on the PACS and any duplicates that have been created, such as caches at workstations, must be updated with the correct name.

IHE is following the same path that led to widespread adoption of the DICOM standard, namely, allowing vendors to demonstrate publicly that their system intercommunicates with others under established standards, though IHE does not create new standards but builds on existing ones – DICOM for PACS vendors; HL7 for HIS vendors. Companies are reluctant to ignore the demonstration, since prospective purchasers are likely to want proof that vendors can connect with other systems. Presumably, IHE will promote integration of PACS and HIS, but it appears that the integration is taking place without additional stimulus.

While the renewed vigor of the major equipment and film companies is an important development, smaller companies are still very much a part of the PACS world. eMed Technologies (Lexington, Massachusetts) has had the ASP model in place for a couple of years, using the Internet to deliver images and reports to referring physicians. InSite One (Wallingford, Connecticut) and Emageon (Birmingham, Alabama) are two companies that provide remote storage and management for images. Stentor (San Francisco, California) plans several ASP products in collaboration with IDX Systems (Burlington, Vermont). Stentor's initial product was a system to distribute images to referring physicians using a unique method for reducing network traffic by matching the amount of data transmitted to each user's need. The unforgettably-named WAM!net (Bloomington, Minnesota) offers similar services, mainly as an OEM to Stentor and other providers of off-site image storage. Vendors of ASP products typically offer their services on a pay-per-study-archived basis at prices in the range from $5 to $10 per study.

There is no question about the interest in ASP products from the vendors' side, but so far the market for these products has been limited. Prospective purchasers worry about reliability, security and ownership. It does seem that all these worries will gradually diminish as experience grows, but for the moment they are rather high hurdles, especially in the light of the new Health Insurance Portability and Accountability Act (HIPAA) regulations, which put in place some hard-to-meet requirements for security. It is said that large hospitals are likely to be interested in off-site storage of images for the sake of reliability through redundancy, but such organizations can probably arrange their own off-site storage just as easily and more inexpensively. It seems more likely that smaller hospitals and imaging centers that lack their own information technology personnel will be better customers for ASP offerings.

According to a study about to be published by the Concord Consulting Group (Concord, Massachusetts), the U.S. market for PACS hardware and software in 2000 was $534 million. About half of this market is accounted for by full-scale PACS, which allow an enterprise to operate almost filmlessly. The other half comprises mini-PACS of all kinds (PACS confined to one or two modalities), on-call teleradiology, systems to distribute images throughout the enterprise, systems to bring images to the radiology department for interpretation and radiology information systems. According to the report, this market is destined to grow about 10% per year for the next 10 years. Including the revenue to PACS suppliers from maintenance and upgrades substantially increases the market. Estimated market shares for full-scale PACS are shown in Table 2. While small companies play a role, the market is dominated by the larger companies, and it seems likely that this pattern will only intensify.