As radiology departments nationwide begin the transition to a filmless environment, print networks will continue to play an increasing and sustained role in the ability of physicians to read and evaluate radiological studies. However, as many administrators have discovered, not every department is created equal when it comes to print needs. Before creating a network and installing printers, the administrator has to evaluate an array of criteria, all of which have to be balanced against future expansion and the move to picture archiving and communications systems (PACS).

John Muir Medical Center

Five years ago, when Jim Grosskopf, administrative director of the Department of Medical Imaging at John Muir Medical Center, Walnut Creek, Calif, and his staff were setting up its printer network, they evaluated all the variables associated with the printer needs of the 18-camera radiology department. Among the criteria were the size of the printer and how it would impact the space limitations of the department; power requirements of individual machines-whether they required a 110-volt or 220-volt power source-which would add to the cost of installation if the outlet voltage had to be changed; cost of the film media in relation to operating costs; and reliability. Grosskopf and his team asked for references from the manufacturer and investigated how long the printer was on the market, its durability, and maintenance requirements. “One of the most important criteria is films per hour; what’s the speed of delivery?” asks Grosskopf. “If you have a printer that’s going to give you 20 films an hour in a busy department, that’s not going to cut it.” Grosskopf’s team set a 55-60 film per hour minimum as the printer benchmark.

During the network planning phase, Grosskopf and the technology team-which included administrators, radiologists, and the information technology services department-developed a 5-year strategic plan. “To get to that final document was a long process; it took 9 months to 1 year,” Grosskopf says. “We looked at volumes per year, we looked at volumes by room, we looked at work flow. There were countless interviews not only with radiologists but other physicians about their expectations, what they needed, what they saw for the future. We weighed all the costs to do business, and then we did the forecasting in terms of where we thought we should go in the future and how we should get there. And the first step in that process was the print network.”

Among the work-flow issues Grosskopf and his colleagues had to consider when designing the network were the physical placement of the printers, the ability to print the benchmark 55-60 studies per hour, and the ability to print to multiple machines. “So these were the things we were thinking about,” explains Grosskopf. “Do we have to walk further, do we have to walk less? Do I have to tweak the XYZ to make the image optimum whereas before I did not have to? With computed radiography [CR] you can change some parameters on your image before you actually print it, thereby decreasing your repeats. What was our repeat rate before we started the print network and this whole venture into CR and what might we save?”

Today, the print network is made up of five dry laser printers: one is placed in the outpatient department, two are located in the main imaging department, and one each is in the MRI and CT sections, which are located apart from the outpatient and imaging departments. Grosskopf says that the machines’ physical placement was determined by work flow, productivity, and efficiency. “[They’re] where the processing has been all along, since the inception of the department,” he says. “It’s where light boxes are located so technologists can view those films and quality control and critique them. You have got to locate them to lessen the steps, so people do not have to run around wasting time.”

Grosskopf says that the wet printers were removed because of issues relating to chemicals and maintenance. “Dry media are just as good in terms of resolution and sometimes better [than wet processing], so we just went away from wet chemicals and went to dry processing; it’s fantastic,” he says. The printers can print any type of study the department does, from fluoroscopy to magnetic resonance to nuclear medicine.

The radiology printer network, which is separate from the rest of the 326-bed acute care hospital’s network, has already been upgraded from its original 10 megabit per? second speed to 100 megabits. Grosskopf’s dream is to move to PACS and away from the necessity of printing film. Apart from the cost savings of not printing film, the advantage of PACS, says Grosskopf, has to do with the fact that studies cannot be lost and can always be retrieved. Once the radiology department goes to the PACS, radiology studies will be housed in a central archive that will be networked into and available to the entire system-which is comprised of John Muir Hospital and Concord, Calif-based Mt Diablo Hospital.

When a printer goes down, Grosskopf can call on the services of an in-house technician. He also has service contracts with various printer vendors who can troubleshoot many problems via the Internet.

In the immediate future, Grosskopf hopes to expand the network to the emergency, operating, and trauma rooms. “This organization wants to go to an electronic medical record-no paper-which is great, and I’d like to do that too with imaging, and that’s why we want to do images and reports electronically,” he says. “If you have a printer network, then I think the advancement in imaging electronics can fit right in.”

Though Grosskopf discourages the printing of studies, he knows that even with PACS there will still be a need for films to be printed, ensuring the future of the printer network at John Muir.

North State Imaging

Don Hubbard, controller at North State Imaging, a 17-radiologist group in Chico, Calif, had a different challenge than Grosskopf when setting up his printer network 4 years ago: how to get the most features for the best price.

North State Imaging’s network is made up of two dry printers and two wet printers, and does about 170 studies per day. Because of the low volume of studies that are performed, printer speed was not an issue, but reliability, ease of use, service support, and price were. The printers are spread throughout North State Imaging’s three Chico locations.

Studies can be printed on any of the printers, but for logistical reasons-the printers being in the same physical location as the camera and radiologist doing the study-fluoroscopy, general x-ray, ultrasound, CT, and MRI are networked to the dry printers, and mammography and the department’s older fluoroscopy room print from the wet. “We’re printing where it makes the most sense to keep track of the film,” says Hubbard. “Our primary objective was redundancy so that if one of the printers failed we had a way to keep printing. And that has happened. We had a minor glitch, and we had a printer down for awhile and we didn’t miss a beat. We just printed everything to the other printer, walked across the street, and picked it up. And then when it came back online, we went back to our original designations.” As his budget allows, Hubbard plans to replace the wet printers with dry laser printers.

The printer network was an expansion of North State’s preexisting technology. “We had a single purpose communications network for our radiology information system that allowed a remote server to support?machines in our clinics, and our remote server was in our administrative offices,” Hubbard says. “We also had a point-to-point teleradiology network that we used to serve the nine hospitals in our area after hours. And what we did was put in a wide area network that blended both of those applications onto one backbone. That was our starting point, and subsequently we built a new office building; when we did that, we put in a fiber-optic backbone to give us bandwidth for the today and for the future. Instead of having dial-up telephone links between facilities and our host servers, we put it all on one area network with a fiber-optic backbone. Prior to that, it would have been impractical for us to do an imaging network like we’re doing right now.”

To improve work flow, the printers were moved to the center of the clinic. “From a networking standpoint, our only issues are the flow of data and making sure there aren’t collisions on the network, and there aren’t because we’re using newer switching equipment, and we’ve put in all new wiring,” he says. “There is no splicing going on and no hubs other than the switch itself and we can monitor that.” Hubbard adds that the network has yet to be taxed by an overflow of data.

If any of the printers go down, studies can be printed to another printer. According to Hubbard, one of the strengths of the network has been that the printers have experienced almost no downtime. “It’s almost worth the price of admission,” he says. “It’s definitely an improvement because we’re doing 50 or 60 MR and CT studies on top of the 170 general x-ray, ultrasound, and mammography studies, so if you start backing up and are unable to print them and get them out to referring physicians, it causes havoc.” If there is a problem, the printer manufacturer can resolve some problems via the Internet. Problems that require more expertise are handled by a company technician.

Though its network is small, North State Imaging is not behind the technological curve, according to Hubbard. “We have already upgraded several times within the last 3 years from where we started, so we’ve never been afraid to make the commitment to stay current,” he says. “And as long as we can cost justify the moves we’re making, then we’ll continue to do it. Our problem is that we’re in the lower tier, we’re not a high volume [department]?we don’t have the resources to spend what it now costs to put in PACS. But those price points are improving. So as PACS becomes more cost-effective, we’ll migrate there.”

Besides price, there are two other stumbling blocks for North State Imaging in the move to PACS, says Hubbard. “It’s cultural-switching the doctors to reading off of the display terminals rather than off of film-[but] we’re beginning to do that now,” he says. “The problem from our standpoint is keeping the thing running. It’s even more important to have network support. In our neck of the woods it’s hard to find good networking people you can afford. We’re working on developing a couple internally. That’ll take us awhile, but when they’re ready to go, we’ll have the wherewithal to support a PACS network.”

University of Michigan Health System

For Steve Ramsey, image management and computer services administrator for the Department of Radiology at the University of Michigan Health System, Ann Arbor, PACS is an ever-present concern. Ramsey instituted the department’s 14-printer network about 10 months ago and is in the process of converting the radiology department’s one campus and six off-campus locations to PACS-a process that is about 30% complete and is scheduled to take another 2 years to finish.

The printers, which are evenly split between dry and wet, are scattered over seven sites throughout the university health system-the farthest is 25 miles from the main campus-and are not modality specific. The radiology department offers all modality imaging, including CT, MRI, ultrasound, and nuclear medicine. The radiology network is part of the larger health system’s printer network. The university does between 350,000 and 400,000 radiology examinations per year.

When the network was being set up, the criterion for how many studies the printers had to do was more than 50 per hour. The printers also had to be reliable and durable. An important aspect of Ramsey’s network is redundancy. All of the printers have to be able to print from all of the cameras, necessitating that each printer be able to print both clear base and blue base films.

Because of the ongoing PACS conversion, the printers had to be compliant with the digital imaging and communications in medicine (DICOM) standard. “Because of the PACS, the only way they can print is through DICOM,” says Ramsey. “The [DICOM] print is the standard. Workstations can communicate with printers in a standard language, and it allows the ability to have vendor A for your PACS and a different vendor for your printer.”

As PACS is introduced, wet printers will be removed from service and the remaining dry printers will be consolidated in a few central locations-Ramsey says it could be as few as five in three locations. “One of the assumptions is that you are going to print less when you are running a full PACS system and everybody in the institution has access to those images,” he says. “But we still realize that there’s going to be a need for film. In the beginning, we aren’t able to send images out to everybody in the institution, so we still need to do quite a bit of printing. Later on, any requests to have film will be sent to the file room. The file room will print those images on film, and then send them out. And we also need to have a backup plan, and that includes film printing in case your PACS is not available for some reason. That will most likely mean that in certain key areas, we will want to have film capability.”

If a printer goes down, studies can be routed to another printer. Ramsey says that the network was set up as a redundant system, so there is always a backup printer available. Service is provided by in-house technicians. The off-site machines are handled through service contracts with the manufacturer.

As Grosskopf, Hubbard, and Ramsey have learned, in order to be successful, printer networks must be able to deliver an optimum number of studies in a reliable and cost-effective way-all the while being open to future expansion.

Chris Wolski is associate editor of Decisions in Axis Imaging News.