We've come a long way since 1921, when Fred R. Barnard—in the December 8 edition of Printer's Ink—coined the phrase "one look is worth a thousand words," which became the adage, "A picture is worth a thousand words."¹ Today, medical-imaging equipment can capture thousands and even millions of bits of information via one "picture," enabling physicians to successfully diagnose and treat countless patients. With its great speed, efficiency, and omnipresence, PACS continues to tie it all together.

During January and February, we surveyed Medical Imaging readers about their facilities' PACS solutions via our Web site (www.medicalimagingmag.com). We asked everything from vendor, solution, and training to decision-making, service calls, and satisfaction. About 300 industry professionals completed the survey in full, representing about a 7% response rate of our targeted audience.

What follows is a selection of pie charts and a summary of our findings in the first-annual Medical Imaging PACS Usage Survey. The findings were both expected (89% of survey respondents' facilities use a PACS, for example, and only 27% began with a mini-PACS) as well as surprising (46% of respondents' facilities have required more than 20 service calls since installing PACS, and only 2% based their final PACS selection on price). So, how does your facility compare?

About the People and Facility

Before learning about the solution, we wanted to know more about the individuals who completed our survey. Not surprising, 31% of respondents were radiology administrators, directors, and managers, and 21% were PACS administrators. A handful were chief technologists/technologists (12%) and PACS/RIS administrators (10%). And even a few radiologists (6%) and chiefs of radiology (4%) took the survey. (For a complete list of job titles of respondents, see Figure 1.)

Most—and that's 84%—of our survey-takers work in a hospital setting, with 11% in a freestanding imaging/surgical center and just 2% in a group practice. Of those facilities, the modalities offered range quite a bit, with most offering CT (14%), X-ray (14%), MRI (13%), nuclear medicine (12%), and mammography/women's health (12%). (See Figure 3 for all modalities offered.)

Just 11% of respondents' facilities do not currently have a PACS. However, 39% of those plan to purchase a PACS in less than 6 months. Also, 24% plan to purchase in 6 months to 1 year, 24% in 1 to 3 years, 9% in 3 to 5 years, and 4% in more than 5 years.

More than half (56%) of respondents were involved in their facility's PACS purchase. When asked to explain, respondents noted:

• "active member of the selection committee, including evaluating and going on site visits";
• "arranging various demos from selected OEMs; then participated and voted on vendor selection";
• "a member of the core research group when we began strategic planning several years ago";
• "performed on-site reviews and data collection"; and
• "wrote the RFP, went on site visits, chose the vendor, and presented my proposal to the board".

One respondent explained, "We formed a PACS team, including front-end staff, managers, and directors from the imaging and information solutions departments. We also promoted a PACS coordinator from within, who did a lot of the legwork. The team met regularly for about 6 to 9 months and discussed what we wanted out of our PACS and evaluated the vendors. Several of us went on site visits once the selection team narrowed the field to three vendors." Most respondents, however, said they were part of the research committee, part of the PACS committee, or one of the on-site visitors.

In fact, 48% of respondents attended the on-site visits of the PACS vendor's other customers before purchasing the system. (And almost all—89%—respondents' facilities took advantage of on-site visits.) Respondents said that the site on-visits were incredibly helpful and added:

• "We were able to see each system in use at a real facility, ask questions of users with and without the vendors present, and establish contacts for help later."
• "We were able to learn from other facilities' mistakes and evade some of the early pitfalls that each facility seemed to experience."
• "Site visits allowed us to see the product in action, to see if the vendor's claims were real or imagined, and to see how even the small aspects of the product would fit with our unique workflows."
• "It was helpful to speak with radiologists and technologists who work with the system; it was also helpful to see layouts of reading rooms and PACS equipment rooms."
• "It helped to see workflow, use of software, configuration of hardware, and the IT involvement."

Finally, and not surprisingly, most respondents' facilities (84%) recognize the important role of a PACS administrator and have a PA on staff.

About the Vendor

One thing is true about PACS vendors: There's a lot of 'em. In the Medical Imaging Buyer's Guide, we list almost 60 PACS manufacturers—and that doesn't include dealers and distributors, or the accessories and supplies providers. The market is big, and it can be tough to make a decision on the best vendor for your facility.

As "Informatics Report" columnist Michael Mack always advises, don't evaluate too many vendors; in fact, he recommends keeping the number down to three. We asked survey-takers how many vendors their facility evaluated, and the majority (32%) followed Mack's advice. However, 23% evaluated four vendors, 17% evaluated five, 10% evaluated only two, and 8% evaluated more than seven. (For the complete breakdown, see Figure 6.)

A number of aspects about the vendor were important to respondents' facilities in making a PACS purchase. The top reason: service and technical support (14%), followed by RIS integration capabilities (13%), training (12%), installations at similar facilities (12%), previous experience with the vendor (11%), and the number of years the vendor has been providing PACS (11%). The vendor's geographical location, personal recommendations, and OEM partnerships were not as important. (To see all of the important aspects, see Figure 7.)

Almost every facility (92%) received on-site training from their PACS vendor. For the most part, the training included 2 weeks or less of on-site training with several days of off-site training for particular users, including PAs, radiologists, and technologists. One respondent explained, "On-site training consisted of four days of Applications for PACS and two days of Applications for CR. The training was very helpful. It eased the migration for our radiologists tremendously. Follow-up was conducted two weeks post go-live and consisted of two more days of Applications. At that time, things were progressing smoothly enough that only one of the follow-up days was utilized; the other was kept in reserve in the event of future issues requiring more in-depth training."

In general, respondents found the training to be adequate. Having several vendor trainers on-site for a week or two was helpful, especially when troubleshooting became an issue. Also, sending certain users to the vendor for tailored training also was beneficial. Added one respondent, "Three weeks of training for radiologists was the difference between a successful and an unsuccessful installation." And too much can never be said about following up. Said one respondent, "Training was extremely helpful. Our vendor came out multiple times. The first time was for several weeks when the system went live. The second time was about a month later, when users had a chance to use the systems and needed follow-up. And the third time was several months later, when users looked for more advanced training. We've also had additional vendor training following a major system upgrade."

However, not all respondents were happy with their training. The dissatisfaction stemmed either from a poor trainer or from the training not lasting long enough. "The training was too brief, and it glossed over too much material," one respondent said. "The idea was to present as much as possible in as little time as possible. It was geared as a ‘get started' type of training, and the vendor expected that we would take more training off-site at one of its facilities." Another respondent didn't find the vendor training helpful, but found a work-around: "Our on-site training was not very efficient or helpful, but a PACS administration course six months in advance was helpful. You really need basic training up front, allowing time to get acquainted with the product and tools, and to develop a question list for vendor training and customization requests." In general, the more training the better, and follow-up training was very beneficial.

Once installed, facilities had a surprising number of service calls: 46% of facilities needed more than 20 service calls since installing its PACS (see Figure 8). Noted one respondent, however, "We have had our PACS for 14 years, so 20 service calls isn't bad." Only 9% did not require any service calls, 24% required one to five service calls, 13% required six to 10 service calls, and 7% required 11 to 20 service calls. The nature of the service calls varied, most of which were minor, and respondents explained:

• "Early on, we had major problems with power interruptions, which caused complete interruption of the system. Several parts had to be replaced due to the outage."
• "We have only been down once in 10 years for more than a day. The rest of the issues relate to a specific workstation or archive devices that are taken care of without the system being compromised."
• "Failed paper scanners, replacement of monitor, failed hard drive, software issues, interface issues, etc."
• "It's a $7 million system, so of course we've had hundreds of service calls and an on-site service engineer. PACS is not like a CT scanner. Ongoing maintenance and support is critical."
• "Only two in which someone had to physically come on-site. Most all other issues are handled remotely and in a matter of minutes."
• "We have hundreds of PACS devices. Our service contract included a full-time employee from the vendor on-site. We made sure the RFP and contract included a 99% uptime requirement—but even that would not be acceptable at 1 hour out of 100. Still, we are pleased with very little downtime in the years we have been using the system ... very pleased."

About the Solution

Attendees at the upcoming annual meeting of the Society for Computer Applications in Radiology (SCAR of Great Falls, Va), which takes place April 27–30 in Austin, Tex, could spend all of their time in the exhibition hall watching product demos of different PACS solutions. Yes, there are that many options.

When survey respondents named the features of their PACS solutions that were important in making the final product decision, the top answer (15%) was compatibility with the RIS and other systems. Other top features included scalability (14%), being Web-based (14%), remote access (12%), customizability (11%), the solution's time on the market (10%), and a disaster-recovery system (9%). Features that were not as important were cost/price (2%), technological capabilities and functions (1%), and ease of installation/use (1%). (To see all of the important features, see Figure 9.)

So, which PACS solution are most respondents' facilities using? Centricity from GE Healthcare (Waukesha, Wis) was the most common (15%). Other popular solutions include Synapse from FUJIFILM Medical Systems USA (Stamford, Conn) with 14%, Horizon Medical Imaging from McKesson Provider Technologies (Alpharetta, Ga) with 13%, Impax from Agfa HealthCare (Ridgefield Park, NJ) with 8%, and iSite from Philips Medical Systems' Global PACS Business Unit (Brisbane, Calif) with 8%. (For a complete list of respondents' PACS solutions, see Figure 11.)

Overall, most respondents are very happy with their PACS solution (see Figure 10). In fact, we asked survey-takers to rate their facility's overall satisfaction using a 1 to 5 scale, with 5 indicating very satisfied. It seems that 40% are, in fact, very satisfied with their solution, 39% are satisfied, 19% are somewhat satisfied, 2% are dissatisfied, and just 1% are very dissatisfied. Comments were quite varied, from good to bad and everything in between, and they include:

• "For our first venture into PACS, we certainly made a lot of mistakes and learned a lot. We also found out that just because it is a name-brand PACS and very expensive, it isn't necessarily the best for the application. We have since looked at other solutions for other facilities."
• "In the past year, we have had many issues that the vendor has had difficulty in diagnosing and/or correcting. The vendor is a reseller of another vendor's PACS, and the relationship between the two has not been good for the past few years. We, the end-user, have suffered as a result."
• "It works! Referring physicians love it, uptime is excellent, we have instant verbal reports, and it's the best thing we ever did."
• "Our system is not as robust as initially portrayed by the vendor."
• "The radiologists did not get the PACS that they wanted; the purchase was decided by IT. It was a difficult training period after installation."
• "The system is not as fluid as we'd like. Different browsers are used for the system, the Web access, and the exams that are burned to a CD, which means needing to know three systems to view the same exam, depending on the locale of the referring physician. Customization is limited, too, due to the high degree of proprietary issues we've run into as well."

In general, most respondents shared a similar stance—that minor tweaks need to be made—but on the whole, their PACS solution was a good purchase. One survey-taker summed it up beautifully: "We are very satisfied with our PACS. Physicians within our community love the system and could not imagine going back to film. Now is our time to bring everyone together so that we all give the same message and demonstrate to the physicians who are not using PACS how great the system is and how much easier their lives would be. The best thing about the system is that everyone wins." We couldn't have put it better ourselves.

Andi Lucas is the editor of Medical Imaging.

References

1. Wikipedia: the free encyclopedia. A picture is worth a thousand words. Available at: en.wikipedia.org/wiki/A_picture_is_worth_a_thousand_words. Accessed March 19, 2006.