By Sheldon M. Stern
Republished with permission from Decisions in Imaging Economics, March/April 2000
PACS from the ground up
One image/many eyes
Circuit-riding radiologists
A scaled-down PACS
Avoid costly mistakes
Seeking smaller solutions
Once limited by cost and technological considerations to a handful of university-based research and teaching centers, picture archiving and communications system (PACS) networks now operate in hospitals and institutions of varying size and healthcare models. The proliferation of PACS vendors has helped make acquisition more affordable, as have advances in software and data communications technology. Companies that once produced only turnkey systems now offer components and services across a broad range of prices and configurations.
Experts emphasize the importance of working with knowledgeable information technology consultants and PACS pros when planning an imaging network. Configuring a system and integrating it with radiology and hospital information systems, digital imaging modalities, and high-speed data transmission networks has never been more attainable and affordable. They also emphasize the need to plan as carefully for future growth as for present needs.
Although it is easy to spend millions of dollars developing a filmless department supported by high-speed, local- or wide-area-network-based PACS and teleradiology systems, a facility does not have to spend that much to implement a good-quality system capable of expanding with changing needs. In fact, the greater affordability derived from competition, manufacturing efficiencies, and new technologies is making it feasible-and profitable-for freestanding facilities to implement PACS solutions of their own.
A growing number of non-inpatient healthcare organizations are beginning to appreciate that their profitability -- if not long-term viability -- may hinge on their ability to effectively leverage their radiology services with next-generation PACS technologies.
While few would argue against the benefits of PACS implementation, the cost remains substantial enough to warrant careful analysis before moving ahead -- especially for enterprises that operate outside the relative financial security of a large hospital. Examples include multisite radiology groups, freestanding imaging center networks, and multispecialty health care groups that need to share images among themselves as well as with referring physicians.
Determining whether a freestanding facility would benefit from PACS and financially justifying such a decision is not necessarily as difficult as it might seem, according to experts who evaluated their options and made the decision. Bob Burns is vice president of information technology for the Kelsey-Seybold Clinic in Houston, and Robert Gallup is executive director of Northwest Radiology Network in Indianapolis.
PACS from the ground up
Among the pioneering out-of-hospital healthcare providers to implement PACS, the 50-year-old Kelsey-Seybold Clinic could be the largest. Owned by Caremark and MedPartners during the 1990s, Kelsey-Seybold was sold last year to a joint venture partnership owned by two Houston institutions -- St. Luke's Episcopal Hospital and The Methodist Hospital. According to Burns, "Not only is Kelsey-Seybold the biggest PACS installation outside a hospital, but the fastest by many orders of magnitude."
Prior to the organization's sale by MedPartners, Burns and radiology colleagues Dr. William Fisherman, Dr. Sylvia Trumble, and radiology manager Robert Honeycutt were in the process of constructing a new central campus to accommodate their growing imaging department and other patient care operations. Realizing that PACS represents the future of radiology, they intuitively recognized that constructing a new facility represented a golden opportunity to cost-effectively solve many problems.
By installing dual-function network infrastructure to support both business and clinical services, they realized that a new facility equipped with all-digital radiology technology and full PACS support would cost less than upgrading all of their aging light boxes, film processors, and file rooms. As they further assessed their options, the feasibility of implementing a PACS solution to address the problem of locating and moving images throughout the 330-physician multispecialty outpatient medical practice with a central campus and 26 satellite clinics looked increasingly good to them.
At the time, the main campus provided radiography, CT, MRI, ultrasound, and nuclear medicine. Radiography, mammography, fluoroscopy, and ultrasound studies were performed in 16 of the outlying clinics as well -- representing approximately 220,000 examinations per year. By installing high-speed network capacity, they could economically expand to an all-digital department on a foundation of computed and direct radiography and PACS.
"One of our big problems was that all 10 of our radiologists read and interpreted images in the central campus because that was the only place that had the volume to warrant on-site radiologists," Burns says. "A picture would be taken at a satellite clinic and the ordering physician would look at it and make some diagnosis but would not get a radiologist's interpretation for an extended period of time-perhaps days. We knew that we had to speed up that turnaround time," he explains.
One image/many eyes
Trumble and Fisherman agree with Burns' assessment as well as the solution.
"The problem with hard copy when you have multiple sites and multiple physicians is that there is only one image, and if the orthopedic surgeon has it, I cannot issue a report," Trumble explains. "If a pediatrician has it and wonders if there is a fracture, the orthopedic surgeon does not have it. So our main reason for going with PACS was for the benefit of patient care in a system that has so many sites and physicians. As it turned out, it also happened to be a bargain because our costs were less with PACS than with film."
Through December 1998 -- when Kelsey-Seybold acquired the first phase of its PACS solution -- the organization used the unit-record approach to film management. Each patient's images were jacketed at a home site. If the patient made an appointment to be seen at a different Kelsey-Seybold clinic, however, the films would be shipped to coincide with the visit. The clinic used couriers and a fleet of trucks to move images and supplies between its many sites so the radiologists could remain within the central campus. Films were often misplaced in transport, and repeat studies were often performed -- at substantial cost that could not be passed on to payors.
On April 12, 1999, barely four months after purchasing the initial phase of its PACS equipment, the Kelsey-Seybold Clinic went live with the first phase of its PACS.
"We had deadlines," Trumble explains. "The building was going to open and we had to have everything ready by that date, so we found a PACS vendor who could move like lightning and get us ready in time.
"We were at our main facility on Friday, April 9th, reading film everywhere," she recalls. "On Monday, April 12th, not only were we 100% filmless except for mammography, we were using voice-recognition technology as well. We have reduced report delays to the point that we are practically doing examinations in real time." The average turnaround time was recently calculated at 20.2 hours, according to Honeycutt.
Circuit-riding radiologists
In contrast to the couriers driving film jackets throughout greater Houston, most of the 26 radiologists of Northwest Radiology Network were driving themselves 30 to 40 minutes between seven sites in greater Indianapolis in order to review images, according to Gallup. Like Burns, Gallup wanted to streamline report turnaround time, improve productivity and efficiency, and minimize the need for circuit-riding radiologists.
What helped Burns and Gallup cement the decision to implement PACS solutions within their organizations was the serendipitous fact that both organizations were constructing new facilities: a new central campus in Houston and a new freestanding outpatient center, respectively. The advantage to de novo construction is that high-quality infrastructure can be built into the facility from the outset, rather than as a more costly retrofit.
"Our infrastructure included IEEE [Institute of Electric and Electronics Engineers] category 5 wiring out of multiple IDF [intermediate drop facility] closets throughout the building," Burns explains, "and fiberoptic backbone back to the primary MDF [main drop facility] closet-the computer room. We installed high- and medium-speed fiber backbones so we could support some of our older business equipment as well as direct digital imaging equipment," he notes.
Trumble and Fisherman note that T1 lines had linked the Kelsey-Seybold satellites to the main campus and each other for data transfer long before PACS-which required additional lines. A 100-megabit per second Ethernet backbone within the main campus also connects the facility to one of its larger satellites, an off-site film storage library, and the cardiology department, which still resides on the old central campus, via a wide area ring.
"For short-term storage, we utilize a multi-system array of RISC- [reduced instruction set computing] based Unix processors," Honeycutt says. "We have approximately 686 gigabits of short-term storage on hard drives. This allows us to store about 10 or 11 days' worth of images. After that, they are switched to a tape storage system for long-term archiving. Each tape holds 40 gigabytes of data. We estimate that we have about 14 terabytes of long-term storage in our jukebox, enough to last about 5 years or until we begin doing digital mammography. We expect that to use about 1 terabyte per year."
Kelsey-Seybold radiologists are more than satisfied with the system's retrieval speed. A 152-image CT study of the chest can be retrieved from long-term archive in as little as 1 minute and 45 seconds. A typical lateral or posteroanterior image of the chest can be retrieved in 45 seconds.
In calculating the expected costs of going with PACS versus those of the conventional film-based department, Fisherman, Trumble, and Burns argued that substantial savings could be achieved in image storage space as well as elimination of film processors, film and chemicals, and printers. Productivity was another key component.
"Our cost was approximately $10 per procedure with film," Fisherman calculated for a presentation to the Radiological Society of North America (RSNA) meeting last November. "Since going filmless last April, we are leveling off at about $8 per procedure. That is a cash savings of $2 per examination times more than 200,000 studies per year. We have also eliminated lost films and costly repeat examinations as well as film and chemistry costs," he adds.
A scaled-down PACS
Although the situation at Northwest Radiology Network is different, Gallup's goals and expectations for his mini-PACS network are much the same as Burns': faster report turnaround times, increased productivity, more efficient patient care, reduced image storage space requirements, and lower procedure cost.
Northwest recently partnered with Seaton Healthcare Corporation of Central Indiana -- the for-profit arm of St Vincent's Hospital of Indianapolis -- to create St Vincent's/Northwest Radiology, LLC, a joint venture that will operate an imaging center in St Vincent's new physician office building now under construction.
Northwest Radiology Network -- which includes 26 radiologists -- owns seven facilities, the joint venture owns one, and the ninth will be housed in the new professional building, according to Gallup. Unlike Kelsey-Seybold, which already had much of the infrastructure in place when the decision to implement PACS was finalized, St Vincent's/Northwest Radiology has opted for a more phased approach.
"We are going to install PACS first in the new facility for several reasons," Gallup explains. "First of all, we have only a limited amount of square footage there and we do not want to spend all of it on film storage. Second, we want to help our radiologists maximize their efficiency and productivity by ultimately transmitting images to them instead of having them drive to where the images are."
A third reason for placing a PACS network near the hospital is that St Vincent's is seriously considering the acquisition of its own PACS in 1 or 2 years.
"Northwest Radiology and our partners at St Vincent believed that this would be a good opportunity to test-drive a PACS on a much smaller scale to see what types of problems and issues would arise," Gallup adds. "It was also perceived as perhaps not as costly." The new installation will operate five digital modalities in addition to general radiography and mammography, he notes. Only the digital equipment will be tied into the PACS network.
Gallup anticipates the completion of construction and PACS installation by the end of June. Until then, Northwest radiologists will continue to multitask among its owned and operated sites in addition to serving seven more sites owned by St Vincent's and other hospitals.
Where Kelsey-Seybold enhanced its PACS system with voice-recognition technology to streamline report turnaround time and save transcription costs, Northwest Radiology equipped its PACS with the ability to tag the significant images so that when transferred to a Web server, referring physicians can access and review the selected images via the Internet.
"We are still looking for a workstation and a short-term archive, and we have yet to determine whether or not we are going to do a PACS broker to interface with our radiology information system," he explains. "Our decision will be based on the dollars available -- currently around $300,000 -- and the imaging volume we anticipate.
"We are hoping to base our long-term archive at St Vincent's so that comparison studies can be pulled up there or in the new center that will be finished in June," he adds. "The hospital opted to install a teleradiology system now rather than a PACS. We are still selecting a vendor for the mini-PACS in the new facility that the joint venture will operate. We are also investigating how to interface the St Vincent's radiology information system with our own."
Avoid costly mistakes
Gallup, Burns, Honeycutt, and their radiologist colleagues agree that making site visits independently of a PACS vendor and having information technology or information services people on the PACS planning committee are important steps that can help any facility avoid costly mistakes.
"I think the most important thing is to get a consultant who has done it before or has done some level of implementation before," Gallup says. "If you do not have a person who understands the technology and the process of radiology, everything becomes fairly complicated.
"You can get into a box really quickly if you focus just on the dollars," he warns.
"One of the big thorns in my side operationally speaking," Honeycutt says, "was the digitizing of our old films for comparison. I highly recommend as a project cost that when you purchase a PACS system you calculate the cost of hiring a company to come in and pull the previous year's films and digitize them and have that ready. Otherwise, you utilize an excessive, and often unbudgeted, number of FTEs (full-time equivalents) -- having them do it on an as-needed basis. That was a huge, huge expense to Kelsey-Seybold, because at one point, we had almost 30 people digitizing films since we had to go from film to filmless over one weekend."
A third option is to adopt a phased approach, archiving new examinations while continuing to read hard copy for 1 to 2 years.
Honeycutt points out that having an accurate examination count is critical for multisite organizations.
"Get an accurate examination count per day. Get an accurate per film per examination statistic," he cautions, "because that is what will help you calculate your storage needs. The last thing you want to do is turn on your PACS and not have enough storage."
Burns and Gallup also agree that physician buy-in and training are crucial to the success of any PACS installation. Information technology can take matters only so far because it is the physicians, technologists, and nurses who have to be able to call up and read soft copy, send images to referring physicians, and manage the radiology information system aspects of any operation.
Seeking smaller solutions
"I know we are not going to be able to get down to a $5,000 or $10,000 PACS solution," he admits. "What I am saying, however, is that I do not think we really need a $1 or $2 million installation for an outpatient office. With input from information technology staff, radiologists and other physician-users, technologists, and an outside consultant, we should be able to select and purchase only the features that we actually need. Our goal is to ensure that every square foot of space produces revenue."
The experts at Kelsey-Seybold and Northwest Radiology Network advise PACS planners to be alert for technology advances that can help them cut costs as well as enhance their archiving and communications solutions. Fisherman and Trumble point to the new computed radiography (CR) system they acquired for several satellite clinics.
"An advance in technology by our vendor just before RSNA last year made it possible for us to buy a CR system for what it would have cost to buy a digitizer," Fisherman says. "The cost to go filmless was cut by about one-third for each one of our satellites because this lower-cost technology became available at just the right time for us. That is why we were able to do what we did last year that we could not do a year earlier. There may be technology coming out that will let people do things next year that we were not able to do with our PACS and digital imaging systems," he states. "This is a rapidly changing industry."
Many lessons can be learned from the Kelsey-Seybold and Northwest Radiology examples. PACS technology can help solve problems of image and report availability as well as image quality, productivity, and profitability. In the not-too-distant future, bringing PACS to a freestanding facility could become less of an option and more of a necessity.
Sheldon M. Stern is a freelance medical writer in Irvine, CA, and is a contributing writer to Decisions in Imaging Economics and AuntMinnie.com.
Let AuntMinnie.com know what you think about this story.
