Michael J. Cannavo

Current estimates are that only about 5% of US hospitals now have a digital picture archiving and communications system (PACS) in place or are starting to deploy one. But that figure is expected to jump to perhaps 20% in the next 2 years, and some optimists believe 40% of all institutions are currently looking into it.

Why? PACS by itself does not speed up the distribution of text reports to referring physicians. That remains perhaps the most sensitive issue facing radiology departments. And while PACS can make images quickly available to the referral base on office personal computers (PCs) with a Web server, four out of five physicians are unimpressed by this capability. They have no clue as to what they are looking at, with the exception of a few key specialties, including neurology and orthopedics.

Gary Reed

One reason for the keen interest in PACS is that many institutions believe a PACS investment will produce substantial savings by eliminating film and file room technicians, and that it will make radiology more efficient by solving once and for all the vexing problem of lost and misplaced prior images. This ability to increase throughput will, in turn, enable them to take away profitable imaging business from their competitors.

Another reason PACS is center stage is that many hospitals have unspent funds earmarked for new capital equipment-money they embargoed awaiting the outcome of their efforts to become Y2K compliant-and are looking for new technology in which to invest, experts report.

PACS vendors, looking to get a foot in the door and assure an ongoing relationship, are aggressively pushing hardware with price tags that range from hundreds of thousands to millions of dollars at prospects who are, at best, often ill-informed and, at worst, naive.

Michael Gray

Not surprisingly, many PACS consultants are counseling caution, pointing out that in 2 years’ time technological advances are likely to make some components of today’s turnkey installations as redundant as the inevitable use of the word system after the acronym PACS. To help hospital administrators, radiologists, and information technology (IT) management better understand the financial and other aspects involved in launching a PACS, Decisions in Axis Imaging News has sought answers to some of the more commonly asked questions from four leading hospital consultants: Michael J. Cannavo is president, Image Management Consultants, Winter Springs, Fla, specializing in the evaluation, design, and implementation of PACS, teleradiology and telemedicine networks; Stuart C. Gardner, president, SG&A Consulting Inc, specializes in PACS project plan development, operational analysis, strategic planning of clinical information systems integration, archiving strategies, and project management including deployment; Michael Gray, principal, Gray Consulting, San Francisco, has been consulting on PACS exclusively with health care institutions for the past 4 years; and Gary Reed is president and founder of Integration Resources Inc, Lebanon, NJ, a company specializing in electronic management of medical images and patient data.

Stuart C. Gardner

IMAGING ECONOMICS: What makes up a PACS configuration?

REED: A PACS requires a RIS [radiology information system] interface, an operating network, an image archive, a minimum of one modality interface, and a diagnostic workstation. Most PACS will include CR [computed radiography] and DR [digital radiography], modality interfaces, workstations, servers, gateways, an archive with disaster recovery, RIS and HIS [hospital information system] interfaces, paper printers, and operating and clinical software.

CANNAVO: The traditional PACS includes some kind of file server, some kind of display workstation, some form of archive, and some form of network. An RIS is also an integral component. You can go in 16 different directions from there.

GARDNER: A PACS provides the ability to acquire, display, communicate, manage, and archive an imaging file. Many hospitals are starting with a mini-PACS where they just take ER [emergency room] or ICU [intensive care unit] film, or a cluster of CT [computed tomography] or ultrasound units, and tie them into a single file server with some display stations and a small archive.

GRAY: Most people think of a PACS as being made up of interfaces to acquisition technology and the HIS/RIS, along with server sub-systems to manage the archive, and a Web server to distribute images to diagnostic displays or to PC monitors in referring physician’s offices. When I build a PACS, I include two more elements-a voice recognition reporting system and special workstations for the technologists to scan all the associated paper into the PACS. The reason: most physicians want a text report, not images. And it is not enough just to get rid of the film, you also need to get rid of the jacket.

IMAGING ECONOMICS: What benefits do hospitals expect to get for their PACS investment?

CANNAVO: The new sub-second multi-slice CTs that can do 700-800 slices in 5 minutes have increased film costs dramatically, and it takes the radiologists longer to view them in hard-copy form. That makes PACS a major contender. But the real name of the game with PACS is getting information to the referring physician faster.

REED: A major benefit is the ability to put more clinical information at the radiologist’s fingertips. This includes prior radiology reports, laboratory reports and images, pharmacy reports, medical history, and other relevant information. The essential piece of patient data, however, is the most recent prior images from the same modality. While prior films initially may need to be digitized or hung on a light box for comparison with workstation images, within 6 to 12 months after the archive is installed, it will contain about 90% of the required comparative studies.

GARDNER: Greater efficiency. Radiology is a very expensive component. It is very capital intensive and its operating costs are very high. One of the benefits of deploying a PACS is decreasing the turnaround time on reports and making them available to referring physicians in a much more timely manner than they are today.

GRAY: If a hospital thinks they are building a PACS to please the radiologists, they are all wet. The radiologist is never going to be any more efficient reading soft copy than he is reading film. If you are running a hospital, you are not concerned about whether the radiologists are reading film or not. If you deploy a PACS, the film savings will only offset your service costs. How do you offset the cost of the hardware? Not by replacing four FTEs [full-time equivalents] in the file room. The only way you are going to come close to breaking even on a PACS is if you make radiology so much more efficient that you can go out and take another 10% worth of business away from everybody else without any additional staff or hardware.

IMAGING ECONOMICS: Where is the pressure to deploy PACS coming from?

GRAY: The majority of calls I get come from hospital administrators, followed by IT managers. The radiology department is a distant third.

REED: There is no one decision maker. We are clearly seeing a shift from a radiology-centric decision process into a combined radiology/IT process. IT and the hospital board are the key decision makers.

GARDNER: The movers and shakers are the IT people and the radiologists, along with the board and the CEO. They all get involved in making the final decision.

CANNAVO: Usually it is the radiologists. Then the IT people.

IMAGING ECONOMICS: What is the biggest mistake a hospital can make in embracing PACS?

CANNAVO: Moving forward without some sort of plan. Ninety percent of those that do, fail. If your volume has stayed consistent or you are looking only at a 5% annual growth, I would not mess with PACS.

GRAY: Locking themselves into a single PACS vendor, and moving ahead to the deployment of an advanced system before they have valid reasons to do so.

REED: There isn’t just one. There are so many opportunities to mess up that almost all PACS don’t operate up to their capabilities. If I had to pick one, it would be not having a well-defined specification.

GARDNER: Not having a thorough understanding of why you are deploying an imaging network, and not having an archiving strategy. The archive is what you are legally liable for.

IMAGING ECONOMICS: Are buyers looking for turnkey or best-of-breed systems?

GARDNER: It depends on what the institution is capable of supporting, and how much money it has. Not every institution can support a best-of-breed environment, and if you have problems, there can be a lot of finger pointing. But the flip side is this: Will a single vendor maintain product development and support in your best interests? If a better workstation comes along, can you integrate it into the environment you have with your turnkey vendor?

The quick answer is that it is easier to deploy a PACS with a single vendor. But that can make you captive to that vendor for a long time to come. And that is not always in the institution’s best interests.

CANNAVO: In an ideal world you would choose the best components from the best vendors, but if you do that you can wind up with systems that are not optimally configured for one another. While there are a few companies that do best-of-breed integration, virtually all my clients prefer a full turnkey solution from a single vendor.

REED: We are seeing more offerings from integrators, but turnkey solutions are favored as finger pointing happens very quickly in a multi-vendor environment. PACS is moving toward plug-and-play, but it is not there yet.

GRAY: I favor a low-risk deployment strategy. It involves installing a digital archive and building up a library of priors for 2 years. Buy at least one CR unit to digitize the plain film from the ER and portables. Implement voice recognition technology, and put in a Web server to distribute reports to your referral base, as well as images to the 20% of your physicians who know what they are looking at. The radiologists would keep reading film while learning how to use the Web viewers.

You can buy all these pieces from a bunch of different vendors and they all plug together, and there are no serious workflow issues. It is just business as usual. And you don’t have to make a decision on who your PACS vendor is going to be for 2 years, when it is likely that viewing stations and Web server infrastructure will cost substantially less.

IMAGING ECONOMICS: Does a hospital have to spend $5 million to deploy a PACS?

GRAY: You can spend that much. But I prefer to initially spend about $2 million to determine if spending another $3 million on full deployment makes economic sense. If it doesn’t and you don’t proceed, you at least have put together a modern radiology department that can handle its information more efficiently and does not run the risk of losing it.

REED: No. But it might have to spend $10 million, depending on its size, what it is trying to accomplish, and what it currently has in terms of technology and networks. You can spend less but probably will not realize benefits that the cost justification is based on. There is no specific minimum PACS that will deliver archiving and distribution for everybody. To a large extent, you get what you pay for.

GARDNER: The cost will vary depending on the application. Is the PACS for the ER, ICU, or the radiology department? Today the average initial expenditure for a starter PACS is in the $500,000 to $750,000 range. Not very many exceed $1 or $2 million.

CANNAVO: You can get systems that will address baseline problems for as little as $75,000. That buys a workstation and a network connection. The typical entry point for PACS is somewhere in the range of $250,000 to $300,000. Most of the installations we are seeing right now are starting in the $600,000 to $900,000 range. I estimate the number of PACS sales in 2001 that will be over $2 million will not exceed 12. Ultimately, the hospital that gets into PACS for $300,000 will wind up expanding it and spending $4.5 to $5 million.

IMAGING ECONOMICS: Is there going to be a role for the application service provider (ASP)?

GARDNER: The ASP concept is to have your archive in a main service center off-site that is supporting many other hospitals; every time you use it you are charged on a per-click basis. But there are a lot of unresolved issues, such as who actually owns the archive. The confusion is very, very high. You can talk to six different vendors and get six different answers. When you look closely, you find an ASP is not exactly cheap and not necessarily in your best interest.

REED: There is a lot of interest in ASPs, but a true ASP has not yet been demonstrated. There are firms that offer disaster recovery services to radiology departments and call that ASP. Others offer PACS archive service online and call that ASP. Still others are offering the very same PACS they would sell you on a capital equipment basis, but restructuring the financial parts and calling it an ASP.

With technology obsolescence, high capital costs, and the lack of IT professionals, it makes sense to get your PACS under a true ASP model. It is a good model for someone who does not have a lot of money to spend on capital equipment. However, at this point it is unproven and a high-risk solution.

GRAY: When you run an ASP cost model, you find that at $6 to $10 a study there is no way it is going to be less expensive than if you did it yourself. However, an ASP can make sense for those hospitals that have otherwise already convinced themselves that it is economically feasible to do PACS and come close to breaking even but don’t have, or can’t recruit, the IT resources to run yet another project. If I could find an ASP to do archiving for the first 2 years for about $2 a study, I’d jump on it. That, in my mind, is about what it is worth.

IMAGING ECONOMICS: Why do hospitals need a PACS consultant?

CANNAVO: Anyone can put a spread sheet together and look at the specifications and pick a vendor. But there are a host of issues that need to be looked at and evaluated. The real issue is prioritizing the client’s problems and evaluating whether PACS is, or should be, the No. 1 priority. They may be looking at PACS to solve problems that PACS cannot solve. People think it is a magic wand. At two out of every three institutions we talk to, we wind up telling them they don’t need PACS.

GARDNER: As a consultant, I look at what the client’s needs are, and what his capabilities are. We want to thoroughly understand why you are deploying an imaging network and what problem you are trying to solve before making any recommendations. Clients need to understand that if they make a wrong decision they can be in deep trouble. A consultant may cost you 4% to 7% of the total cost of your deployment, but it is an insurance policy.

REED: A good PACS consultant can save you significant amounts of money, reduce time to implementation, and greatly increase the probability for a successful outcome. Look for consultants that will challenge your opinions.

GRAY: You could solicit peer recommendations, but you have to wonder if someone who has done it wrong really wants to talk about it. You need to find somebody who is fairly neutral and has nothing to gain to help you talk it through. I take the position that I will not build for somebody something they want to build unless it makes sense.

Why are you doing this? When I ask that question, I find the radiology department in some cases is hiring a consultant to help them figure out how to bamboozle the administration. The administration hires me to try to slow down radiology because it doesn’t have the money anyway. And when the IT department hires me, it is because the other two guys have no idea what to do and so have dumped it in their lap. So in many cases I wind up as the arbitrator-the guy who comes in and has to try to make some sense out of it.

Mike Cannavo: [email protected]

Gary Reed: [email protected]
Micahel Gray: [email protected]
Stuart Gardner: [email protected]