From left, Mike Gressel, BSRTR; Lisa Martelli, BSRTR; David Briden, VP, CIO

It is no longer uncommon to find a community hospital’s radiology department using a picture archiving and communications system (PACS). A community hospital with an enterprise-wide PACS is still infrequently encountered, however, and one with such a deployment accessed through an electronic medical record system is even more unusual. At Exeter Hospital, Exeter, NH, complete integration of the PACS and the electronic medical record system has existed for barely a year. Already, though, officials of the 100-bed facility report that the pairing has led to substantial gains in staff productivity and has permitted faster, earlier clinical interventions, paving the way for better outcomes, shorter stays, and greater satisfaction among patients.

This was a bold step and it required a significant capital investment, of course. More important, it demanded vision. Michael Gressel, director of diagnostic imaging, says, “We have always been a forward-thinking hospitala step ahead with technologyand this was no exception. From the very start, it was felt that this ought to be an enterprisewide endeavor that would benefit all departments. Because that is the approach that we took with it, PACS (along with the electronic medical record integration afterward) became an easy decision.”


PACS surfaced as a topic of discussion around Exeter Hospital in 2000. One reason for this was that the diagnostic imaging department (and others, from referring physicians to top administrators) had begun to worry about the impact of imaging volume growth. In 1998, for example, Exeter Hospital conducted some 41,600 imaging studies, counting all modalities. A year later, the number rose to nearly 47,400. By 2000, it had reached 52,000, representing a 2-year increase of 25%. Moreover, in each of those years, there was a spike in the quantity of multi-image studies performed, meaning that imaging output, expressed as individual pieces of film, was increasing even more dramatically. Exeter Hospital’s acquisition of more and better imaging equipment, coupled with the addition of extra staff (bringing personnel totals to five for radiologists and 45 for technologists), made such growth possible. These enhancements enabled the hospital to begin satisfying demand for imaging services sparked by the area’s population increases.

Another reason that PACS appealed to Exeter Hospital is that it was seen as an ideal remedy for the problems of misplaced film and, owing to the growth of CT and MRI multi-image study volume, a swift disappearance of file-room capacity. The timing for stepping up to PACS could not havebeen more propitious.

“As it happened, the hospital was entering into a period of major building expansion, and that included a plan to locate the diagnostic imaging department between two sites on different parts of the campus,” says Lisa Martelli, PACS system administrator. “Under this plan, our women’s imaging center would be moved to an outpatient care facility that the hospital was planning to build. Once relocated, the women’s imaging center would be outfitted with relevant modalities and a full-time dedicated radiologist. That gave rise to the question of how that radiologist would be able to respond rapidly to a request to look at images generated in the emergency department on the other side of the campus. We felt that if we were going to be reorganized in this way, PACS would be essential.”


Although Exeter Hospital staff quickly fell into accord on the need for enterprise PACS and its linkage to the electronic medical record, the process of transforming vision to reality took nearly 3 years to complete. A sizable portion of that time was devoted to planning, which required the participation of representatives from the departments of information services, emergency services, cardiology, orthopedics, and general surgery, among others.

Most pivotal among these players was David Briden, vice president and CIO. During an 8-month period in 1999, while the hospital sought to fill several management vacancies, Briden had line responsibility for radiology; it was through that experience that he gained an abiding sensitivity to the department’s needs and concerns. Led by Briden, the information services department and the diagnostic imaging department spearheaded the PACS endeavor.

“Hospital leaders recognized, early on, that a medical imaging system is really no different from a hospital-wide information system, in terms of how it must be managed,” Briden says. “For that reason, the hospital looked to the information services department to help it take the lead on this.” The diagnostic imaging department was comfortable with this arrangement, due in large part to their familiarity with Briden. As vice president and CIO, he was uniquely positioned to champion PACS in meetings with top management. “The radiologists knew that I would be looking after their interests in the search and selection process, which greatly aided the effort to gain their support,” he says.

In 2001, the PACS development team was ready to propose a plan formally to the hospital’s CEO and CFO. Under the plan, PACS would be implemented as an enterprise solution. Seemingly without hesitation, the top decision-makers approved the endeavor. “They committed to going forward with this, and we had not yet, by that point, even started looking at vendors,” Martelli recalls. “All we could give them was a ballpark figure of how much this would cost, based on input that we had received from our consultant on a 5-year expansion plan. That was sufficient to convince them that this was doable, assuming we stayed within the dollar confines identified by the consultant.”

Following CEO and CFO approval, the PACS plan still had to survive the scrutiny of the hospital’s board of directors. The board was heavily weighted with local business leaders who did not hesitate to approve the initiative. “The board understood exactly what we were trying to accomplish because they had been digitizing information as a way to achieve new efficiencies in their own companies for years beforehand,” Briden says. “They did not even need to see a detailed return-on-investment projection to recognize that this was a necessary and valuable undertaking.”

Ready, at that point, to begin choosing a system, the hospital invited proposals; from those, it selected six vendors (two with strong PACS software, two with a reputation for excellent hardware, and two for which film management was the strong suit). “Each vendor delivered a 2-hour presentation to our PACS search team,” Briden says. “We then narrowed the field by half. The three finalists returned, each on different days, with demonstration systems. They set these up so that anyone within our organization who was interested could try them and develop a genuinely well-informed opinion about their product offerings.”

Of all the presenters, Agfa came closest to matching Exeter’s expectations and vision. “They were not taken aback by the fact of the acquisition being driven by the information services department rather than by radiology, nor by the fact that this was to be an enterprise solution,” Briden says.


The Agfa system, installed in 2003, makes use of a longterm archive system from EMC Centera that features a software-driven storage architecture specifically designed to address the unique handling requirements of fixed content,such as radiographs, CT scans, and MRIs. It is a more cost-effective storage solution that also provides redundancy by automatically duplicating every piece of stored data. As configured for Exeter Hospital, the archive has 5 terabytes of capacity (of which only 0.75 terabytes have been used to date), Martelli reports. “We saw the Centera archive system as allowing us to do more than just store images,” she says. “For example, we can use it to store written orders that we scan in from referring physicians. The idea was to have versatile technology affording us opportunities to maximize its utilization, giving us more bang for the buck.”

Deployed around the hospital in conjunction with the PACS installation are five diagnostic workstations for radiologists’ use only. Each station comes with two flat-panel monitors. In addition, there are two clinical review stations, one in the emergency department and the other in the diagnostic imaging department’s front office. Exeter Hospital might have ended up with more workstations, had Agfa not suggested the use of its Web1000 browser product instead. Martelli says, “Agfa conducted a review of our work flow to determine what the staff needs to see and how those needs can most costeffectively be addressed. Other vendors wanted to put workstations here and there, which would have been very expensive and would have eaten up a lot of precious space. Agfa’s analysis of how our people work revealed that we did not need all those extra workstationsthat those people could be well accommodated by letting them access images online via the hospital’s electronic medical record through their own office-based desktop and laptop computers.” Soon, Exeter physicians may also be able to access images from their computers at home via Internet. The hospital is testing an enhancement to the Web1000 browser that will let physicians log onto the World Wide Web through a Secure Sockets Layer connection and use a special password-protected access point, Martelli reports.

Once the PACS was installed and running, responses from radiologists and referring physicians alike proved overwhelmingly favorable. Gressel says, “Their delight and acceptance is completely understandable. First, through PACS, we have images always immediately available. This reduces the potential for delays in surgery and delays in the initiation of other interventions. Moreover, referring physicians are able to consult with each other and share images without physically having film in hand. An intangible, but no less substantive, benefit of PACS is the way that it signals that Exeter Hospital is committed to providing highest-quality service. Referring physicians tell us that their patients are quick to pick up on that.”


When Exeter Hospital was reviewing the field of PACS vendors, a question that it posed to each candidate was whether the proposed installation would be capable of interfacing with the facility’s electronic medical record system, which was installed in 1999. “Whatever PACS we ended up buying would have to be able to make images available through the electronic medical record,” Gressel says. “Our physicians already were accustomed to accessing medical reports, laboratory values, and other text information on their patients through the electronic medical record. We sought to capitalize on that familiarity simply by adding to the electronic medical record screen a camera icon that would pop up next to any report for which there was one or more associated PACS images available.”

The integration of PACS with the electronic medical record was completed in September 2003, nearly 4 months later than anticipated. Maryann Serafin, senior information services programmer-analyst, says, “The delays were due less to technological issues than to logistics. Agfa and the other vendor were capable of interfacing with one another’s products, but this wasthe first time that they had formally worked together on a PACS and electronic medical record interface.”

With the interface achieved, the next step called for strenuous testing before the enterprise-wide transition to PACS. “We tested every function by throwing a lot of what-if scenarios at it,” Jennifer Mulholland, information systems senior analyst, notes. “For example, the system permits users to edit orders and reports. We wanted to find out what would happen if an order for image acquisition was placed on the electronic medical record side and then the user decided to cancel or alter that order before the images were delivered by the PACS side. Similarly, we wanted to know what would happen if the order was associated with a report and then subsequently dissociated from it. As it turned out, there really was nothing that caused the interface to experience problems.”

Despite not having the PACS and electronic medical record integration ready for use when originally planned, Martelli was never disappointed. “We put the time to good use, making sure that we had everything else up and going (for example, verifying that all our PCs were up to date and able to connect),” she says. “We could have pressured the vendors to move faster but, in retrospect, I am glad that we did not. If we had rushed everything last year, we would now be working our way through a lot of unpleasant ancillary issues, such as those stemming from trying to connect with the outdated PCs.”

Hindsight also confirms, for Martelli and her colleagues, that the PACS and electronic medical record integration was an effort well worth their while. Gressel says, “More referring physicians look at images now than ever before, thanks to their easy accessibility through the electronic medical record. We do not have solid figures to characterize the increase, but our sense is that it is substantial. At the same time, we have fewer referring physicians coming to the department for face-to-face consultations.”


Martelli believes that Exeter Hospital is the first community hospital in New Hampshire with an enterprise-wide PACS. “There are a few others in the state with PACS, but not on the scale of ours,” she says. Why haven’t more community hospitals embraced PACS? Martelli speculates that hospitals may have been taking a wait-and-see approach. In putting off the move to PACS (and enterprise-wide deployment in particular) facilities may be hoping for further technology improvements and/or cost decreases.

Gressel, meanwhile, theorizes that misplaced priorities are to blame. “The state of health care in the United States is so riddled with holes that many community hospitals just do not know which ones to plug first,” he says. He suggests that many community hospitals would love to adopt PACS, but will continue to hesitate, perhaps because they still do not recognize or appreciate the value of the technology. He also predicts, however, that a confluence of pressures will drive many to reconsider. “There is the medicolegal risk of losing patient records. There is the expense of housing patient records. There is the challenge of keeping up with your competition,” he says. “Granted, today you do not find a lot of community hospitals with enterprise PACS, but 10 years from now, yours will be odd man out if you still have not embraced it.”

The Whole Picture: A Single Repository Vision

The major work on Exeter Hospital’s PACS and electronic medical record solution is largely complete. Still to be done is the integration of a few radiology modalities, including echocardiography, bone densitometry, and mammography, along with several nonradiology image sources, such as endoscopy and pathology.

Eventually, the fruits of this effort will be reaped, giving referring physicians access to virtually every type of information and image associated with any given patient.

Lisa Martelli, PACS system administrator, says, “Let us say that a patient comes in and there is a question of cancer. The physician will be able to sit in front of the computer and conveniently look at the biopsy, the pathology results, the endoscopy images, and everything else, all residing in one place: the PACS archive.”

David Briden, vice president and CIO, adds, “Our enterprise vision is to have a single repository of patient information so that everything that a physician needs to know about a case will be available from one place. It is vastly more efficient to do it this way. It is vastly more cost-effective as well.”

Rich Smith is a contributing writer for Decisions in Axis Imaging News.