Jess Dalehite, MD (left), director of radiology, and Cary Vanley, PACS coordinator, rolled out PACS at Medical Center Hospital, Odessa, Tex, in phases.

When some hospitals and imaging centers deploy picture archiving and communications systems (PACS), they begin with enterprise-wide systems that make film use all but disappear when PACS use begins. Other institutions find it less daunting to ease themselves into PACS use with a modest, very localized implementation at the start of implementation. This is followed by a lengthy period of gradual upgrading and expansion that lasts until, eventually, all radiologists and referring clinicians involved are accustomed to the technology. During this process, film vanishes slowly but inexorably, and with the less potential to create discontent during the transition.

Medical Center Hospital (MCH), Odessa, Tex, stands as a prime example of an enterprise that opted for the gradual form of PACS rollout. "We didn't like the idea of suddenly saying, OK, here's PACS and, starting today, we're totally filmless," Jess Dalehite, MD, director of radiology at MCH, explains. "The conversion to PACS is such a huge paradigm shift that you have to let everyone get used to it over time. Users have to put a toe in the water, not just get pushed into it."

MCH planned a transition to PACS that would take no less than 2 years and no more than 4 years to complete. "When we added up how much time would be involved in acquiring the equipment, deploying it, and changing the culture sufficiently for the technology to be well accepted, we were looking at somewhere between 24 and 48 months," Dalehite says.

By all measures, the strategy of deploying PACS in stages worked well for MCH. Cary Vanley, the hospital's PACS coordinator, says, "The comments we've gotten from referring physicians, surgeons, and other users are, 19 out of 20 times, very positive about what's been accomplished. They tell us they're glad we did this in the way we did."

ACCESS ANYWHERE

Although committed to a modest start, MCH planned a PACS deployment on a grand scale that would extend the technology's reach to all corners of the hospital campus and then stretch beyond that to several affiliated rural hospitals and the offices and homes of radiologists and clinicians. "We wanted referring physicians to have access to images any time, from anywhere," Dalehite says. "We also wanted our radiologists to be able to do primary reads off campus. Our west Texas locale is somewhat off the beaten path, which makes it more of a challenge to recruit radiologists. The scope of our envisioned deployment of PACS would allow us to outsource some of our work to radiologists of like mind in private practice outside our service area in the event that our imaging demand and staffing requirements collided."

Opened in 1949, community-owned, community-governed MCH offers an array of health care services. These include maternal-child programs, rehabilitation therapy, diagnostic testing and treatment, and cardiac rehabilitation. Its emergency department is staffed be physicians 24 hours per day, and the hospital offers a level-III neonatal intensive care nursery. In addition to conventional inpatient surgery suites, there is also an ambulatory surgery unit for outpatient procedures. MCH, with nearly 500 beds, serves as the primary teaching hospital for Texas Tech University and claims a staff of more than 225 practicing physicians representing 75 specialties and subspecialties, supported by 1,500 employees in clinical and administrative capacities.

Radiology at MCH is a full-service department. Modalities in use include two multislice CT scanners, two MRI units, a digital radiofluoroscopy machine, a digital angiography system, three ultrasound machines, two multiloader and four single-loader computed radiography systems, and two digital nuclear medicine cameras. All of these are currently connected to the PACS; they will be joined in the coming months by several digital radiography units, which the hospital is in the process of acquiring for the emergency department and for an outpatient center now under construction.

Imaging volume reaches approximately 105,000 examinations annually at the MCH main campus alone. Counting images from the affiliated rural hospitals, the seven radiologists and 50 technologists who constitute the MCH radiology department handle more than 150,000 studies per year.

EAGER PACS USERS

MCH dreamed of installing PACS as long ago as the early 1990s, Vanley recalls. In 1995, the hospital took its first tentative steps toward implementation, but decided to wait after concluding that the technology was not yet sufficiently well developed to be worth the investment (and PACS, at that time, was a very expensive proposition). In 1998, with the technology's capabilities improving quickly and costs declining, PACS champions at MCH (who included radiology administrative director Carol Evans) once again began contemplating an acquisition. The same year, the hospital purchased its first diagnostic workstation and an image-storing disk jukebox. "This was a rudimentary system, merely something we could use to see what PACS was all about and get our feet wet," Vanley says. Nevertheless, some radiologists flatly refused to use the system. "We expected that, which is why we didn't make a big deal about their choice, and why we made no attempt to force them to try working with PACS," Vanley adds.

Perhaps one reason that those radiologists failed to use the system was that its archiving computer was unstable and prone to frequent crashes. Realizing that this made for an untenable situation, MCH soon replaced the faulty computer with a product from eMed. "The replacement system was very stable," Vanley says. "Its performance was so dependable and impressive that it helped convince our reluctant radiologists that PACS was the way to go." That new unit featured a 150-slot jukebox capable of accommodating 5.2-gigabyte magneto-optical disks. In 2000, MCH purchased (from eMed) a split-system archive for long-term storage that is, Vanley says, "basically a big, powerful archive, allowing our earlier, smaller archive to become a host that maintains the database of the jukebox and handles image transmission between the jukebox and the archive."

By this point, MCH had four diagnostic workstations. The hospital also had installed an architectural feature known as On-Demand Service that greatly accelerated retrieval of images from the archive. "We could pull up a 100-slice CT over the network in about 2 seconds," Vanley says. In 2000, PACS was also extended to the general care floors, the surgery suites, and the emergency department.

The following year, PACS acceptance and use had grown sufficiently common among referring physicians for MCH to stop printing film for CT, MRI, ultrasound, and nuclear medicine studies. Film was printed after that time only when it was requested by a physician for use outside the range of the PACS (for example, for use during surgery). At the beginning of 2003, the PACS was deployed hospital-wide and beyond, with full integration of the PACS with the radiology information system (RIS) and a digital dictation system. The few modalities that had still been generating film then converted to digital output only.

One of the key features of MCH's PACS is its powerful web server. To optimize transmission of images, the hospital operates a pair of auto-routing servers using eMed proprietary wavelet compression methods to reduce most images at a 6:1 ratio (found by the hospital to yield the best blend of image quality and file-size manageability, Vanley indicates).

"We send our information over the Internet using either Secure Sockets Layer (SSL) or Virtual Private Network (VPN) technology," Dalehite says. "In both cases, the information is encrypted. SSL is what the banking industry uses to secure financial data being transmitted over the World Wide Web. VPN is even more secure. Either way, this more than meets the requirements of the Health Insurance Portability and Accountability Act, which are simply that data be encrypted. There is no mention in the rules as to how that encryption is to be done."

For network infrastructure, MCH is relying on hubs rated at 100 megabits and tied into a 1-gigabit fiber backbone for images carried inside the hospital. Outside, there are two T-1 lines to satellite facilities.

LOW-HANGING FRUIT

In launching its transition to filmless operation, MCH started with what Dalehite calls "the easy-to-reach, low-hanging fruit: those modalities already in digital form, such as CT and MRI. Since these cross-sectional images read better on-screen, as opposed to on film, we felt that breaking into PACS with these would ensure a more favorable reception from radiologists getting acquainted with this way of working. It also allowed us to build a backlog of archived images for comparison purposes easily and quickly. At no point did we digitize old films for comparison (that would have been too labor intensive). Our goal was to have 1 year of archived CT, MRI, ultrasound, and nuclear medicine studies, but we ended up with 2 years' worth of those because of delays related to a concurrent deployment of a RIS. The problems we encountered with that RIS deployment occupied enough of our attention that we couldn't manage going live with PACS within the originally planned time frame."

Fortunately, no such difficulties emerged when it came to the PACS deployment. Indeed, so accommodating was the system that testing prior to going live proved a comparatively tame affair. "Our vendor so thoroughly tested the equipment before shipping it to us that virtually all we had to do was install it and start using it live," Vanley says. "Of course, just to be on the safe side, I did test the system by pulling images and performing every function possible. When I felt comfortable that everything was working correctly, I'd turn it over to the radiologists and other users. Often, though, it wasn't possible to test prior to going live because we had only the live system to work with, but we never had a bit of a problem because of all that preshipping testing. We were able, in those instances, to skip testing, yet have a good deal of confidence that everything would work as planned." The hospital, nevertheless, took the precaution of arranging to have eMed provide on-site, next-day technical support. It was never needed, Vanley notes.

Also unneeded after the final stages of the transition to PACS were film stock and chemicals, and the personnel required to process hard copy. Even though the demand for film-library staff fell precipitously, no one was laid off; instead, positions that became vacant as employees moved on were never refilled. This attrition process eventually winnowed the staff to a skeleton crew that largely occupies itself burning images onto CDs for patient use and reading reports to referring physicians who don't have web access.

Within a few years, all stored films will have been destroyed, as allowed by law, and the file room will stand empty. Plans for the area have not yet been made, but the likeliest use will be conversion to suites for extra modalities, Dalehite reports.

Promoting Use

Vanley believes that the biggest challenge to MCH's PACS deployment was the task of overcoming the objections that referring clinicians raised to the new way of viewing images. "When we introduced the web server, few of the physicians wanted to look at images that way. They didn't even want to consider it," he says. Over the span of about a year, however, radiology's PACS advocates persuaded most of the objectors at least to give it a try. "When a referring physician who wasn't willing to use the web server visited our department to ask a radiologist questions about an image, the radiologist would make it a point to pull up the image via web server," Vanley explains. "The referring physician would then see the quality of the images and the speed at which they were available on-screen. The radiologist would casually mention that these images were available in the same quality and at the same access speed from the nursing floor, operating room, or physician's home. The referring physician would then start getting intrigued. That's when he or she would realize that PACS wasn't so bad after all."

An effort also was made to make certain potential users of PACS comfortable with the technology before it was deployed in their areas. "It became a requirement for new physicians brought on staff at MCH that they receive training in the use of the web server," Vanley says. "This way, as soon as they started working on the floor, they would know the technology, they'd know it was available to them, and they wouldn't be reluctant to use it. I don't have much to compare it against, but I would have to say that our PACS deployment was about as painless as it probably could be for all concerned. We conducted our deployment in an intelligent fashion, and no small amount of credit for our success must be given to our decision to use the best available technology."

Rich Smith is a contributing writer for Decisions in Imaging Economics.