Service providers gain favor as budgets tighten

Hospitals continue to tighten their economic belts. The quest to keep costs under control persists unabated, even as consumers' healthcare demands spiral. PACS is increasingly viewed as a means of meeting this goal through gains in productivity and efficiency, but justifying large capital expenditure on a new software system is far from easy. Business models offering large cost savings through shared purchase or scalable fee-per-service arrangements are therefore attracting considerable interest.

The application service provider model offers hospitals all the benefits of having PACS, without the initial capital outlay. Clients instead contract a third party to deploy and maintain an appropriate data management, distribution, and storage solution. Fees are scalable according to the level of service provided, including the volume of data handled, and they can be budgeted as operating costs rather than capital expenditure.

ASP-based PACS is a relatively new concept, and issues of ownership, security, and reliability have yet to be resolved. Few healthcare institutions have signed up for the entire package. Most hospitals tempted by the funding model are considering an ASP to look after their long-term archiving needs rather than daily data management, according to Dr. Osman Ratib, vice chair of information systems in the radiology department at the University of California, Los Angeles.

UCLA is one institution to have taken this step, Ratib told delegates in his opening plenary lecture at September's EuroPACS/MIR meeting. Medical imaging data generated on campus are now archived offsite at a dedicated center. The data warehouse is linked to UCLA's PACS via a secure, redundant high-speed wide area network.

"The vendor has to support the system, guarantee uptime, disaster recovery, and security, and provide unlimited storage. We have to be sure that they have the proper resources for maintenance and are not subcontracting to someone who does not have these resources," Ratib said.

The solution's scalability is a major plus. UCLA aims to archive all its imaging data with an ASP eventually, and it will buy more space as it becomes necessary. This avoids having to make accurate forecasts on data load in advance, possibly purchasing storage that may be obsolete by the time it is needed. Data from UCLA's preexisting radiology archive have already been moved offsite, a procedure Ratib is not eager to repeat.

"We had years and years of old data we wanted to migrate to the outside vendor. This turned out to be more of a challenge than we had expected, not only because of the size of the data, but also because of the content, which had to be 100% clean," he said.

The financial case for ASP-based archiving is generally more clear-cut for small and medium-sized healthcare institutions, whereas larger institutions may find it cheaper to buy the components and build an in-house solution, according to Ratib. He recommends that institutions weighing outsourcing against ownership construct a detailed financial model showing the likely costs of each route. The capital purchase model should include estimates of upgrade costs and maintenance as well as initial outlay. Consideration should be given to the number of years the investment is expected to span. Models of ASP costs should indicate whether contracts specify a fixed fee per examination (linear model) or allow cost per examination to decrease as the total volume of studies increases (nonlinear model) (Figure 1).

But even if the figures add up, institutions should remain on their guard when signing on to a long-term service agreement, Ratib said. Negotiations should cover details of data migration procedures in the event that either side terminates the contract.

"It is good to outsource, it has many advantages, but you are taking a risk. Right now the risk is which vendor on the market has the technical expertise to provide a guarantee for five or 10 years and is financially stable. You don't want a company that has 10 years of your data to go bankrupt," he said.

ECONOMIES OF SCALE

ASP-based solutions to PACS can also have economic benefits for large-scale and even regional projects, according to Hanna Pohjonen, a healthcare IT consultant with Rosalieco Oy in Espoo, Finland. Calculations have shown that in Finland, for example, it is at least four to five times more economical to take the ASP route to PACS in an installation generating 500,000 examinations per year, when compared with hospital-wide purchase of a PACS used for 10,000 to 20,000 examinations. When the hospital district of Helsinki and Uusimaa decided to go digital, it opted for a giant ASP archive to store date from 21 separate hospitals.

"There are clear economical benefits from thinking big. Common components and shared infrastructure can be used, common service and maintenance contracts can be made. You can really gain benefits from megasize projects," Pohjonen said.

ASP-based funding is not the only alternative to capital investment in PACS. The so-called managed imaging service model fits between the two extremes, Pohjonen said. PACS components are bought as an investment, but a third party is used to integrate and take responsibility for the imaging service. This option reduces any risks regarding ownership of patient data or vendor suitability. Hospitals must still pay for equipment upgrades, however, and they manage any data migration if systems become obsolete.

The majority of European hospitals with PACS have traditionally invested in the hardware and software themselves, but this situation is beginning to change, Pohjonen said. Scotland and northern Finland are both introducing a managed service approach to PACS. Early adopters in areas saturated with investment PACS, such as Norway and Sweden, are now looking at ASP-based archiving as part of a replacement solution.

At least 11 ASP PACS projects are already up and running in Europe. Most are based in Nordic countries, where widespread availability of broadband networks has favored offsite archiving, Pohjonen said. She predicts a new market for ASP PACS to emerge in Eastern Europe as well, as new entrants to the European Union invest heavily in technological upgrades. Estonia, for instance, is planning to introduce a national PACS centered in its capital city of Tallinn.

"I work with these new EU countries a lot," she said. "Some will make a big technology jump from modality-based miniPACS straight to nationwide ASP solutions."

LEGACY LINKAGE

Countries moving their hospitals en masse from hard-copy workflow to regional ASP projects have certain advantages over early adopters of digital imaging. In England, for example, plans for nationwide PACS deployment and large-scale regional cluster archiving could rest on efficient integration between the newly acquired PACS and legacy systems in neighboring hospitals.

The ambitious RIS/PACS rollout in England is part of the National Programme for Information Technology (NPfIT), a centralized structure for procuring clinical IT systems. Just 25 of England's 178 acute hospital trusts have a PACS at present, and government officials would like to see the rest adopt digital image management solutions.

The proposed schedule has 80% of these new installations completed by the end of 2005. The remaining 20% should be operating by the end of 2006.

NPfIT has an estimated budget of Euro 7.5 billion to spend on healthcare IT over the next decade, with most of it directed toward so-called core IT systems such as electronic patient-centric booking and prescribing services. Confusion reigns as to whether PACS or RIS is core, and debates about their funding are ongoing. Nonetheless, NPfIT remains in charge of the entire RIS/PACS procurement process, in a bid to keep prices low through mass purchase deals.

NPfIT's mass procurement strategy involved dividing England into five regional clusters, each serving around 50 million people. Officials then awarded contracts for RIS/ PACS supply on the basis that individual vendors would be serving all (or most) major hospitals within a certain cluster. Hospitals seeking to purchase a new RIS/PACS under this program must use a designated local service provider to receive the discounted deal. Freedom to opt out from the scheme and negotiate a one-to-one deal with a different RIS/ PACS vendor is allowed but discouraged.

Institutions buying their RIS/ PACS through the NPfIT-approved route are expected to receive 12 to 15 months of local archive storage as part of the package. Small hospitals will be managed as satellites of their larger neighbors. All hospitals within a specific cluster will send data for long-term archiving to a single, centralized data repository. Information stored in this archive should eventually become available to a national electronic clinical record service.

"They have been talking in terms of petabytes being stored in these cluster archives," said Dr. Keith Foord, a radiologist at Conquest Hospital in East Sussex, U.K. "The archives will be mirrored so there is a secondary data set available."

Foord used the example of his own health authority, part of the NPfIT's southern cluster, to illustrate the issues involved in regional RIS/PACS deployment (Figure 2). The hospital group covering the Conquest Hospital and Eastbourne Hospital already has an Agfa RIS/PACS. Radiologists at Conquest use Agfa CR equipment, while those at Eastbourne have sourced CR from Philips Medical Systems. One other hospital in the same health authority has installed Agfa's PACS and CR solutions but uses iSOFT's RIS. These three hospitals will all be encouraged to start using the cluster archive to upload and retrieve images.

None of the remaining three hospital groups (six sites) and three individual hospitals in the same health authority are running PACS at present. All are operating CR equipment, though sourced from three different vendors (Kodak, Ferrania, and Agfa). These hospitals are expected to purchase a GE Healthcare RIS/ PACS from their local service provider. This RIS/ PACS solution should incorporate long-term archive servers for the cluster archive.

The project's goal-to have every hospital in England using PACS by 2007-is undoubtedly ambitious, Foord said. If it fails, however, responsibility will go straight to the door of the service providers, not the hospitals, vendors, or government officials.

"It is certainly a very bold project," he said. "Financially and politically, I doubt it will be allowed to fail. But if it does, the very commercial structure of the contracts will not point the finger of blame at the politicians, but straight at the cluster providers who have signed up to some huge and very rigorous contracts."