In June, a report stated that radio frequency identification (RFID) tags used by hospitals to track devices such as IV pumps and prescription medications can interfere with medical equipment. The study — published in the Journal of the American Medical Association and conducted in a controlled nonclinical setting at Vrije Universiteit in Amsterdam — detected 34 instances where electromagnetic interference (EMI) had occurred; of those, 22 were deemed hazardous. In light of this, the authors cautioned hospitals to conduct onsite testing, and called for updated industry standards.
However, despite the potential ramifications, the study was quickly dismissed by manufacturers of real-time location systems (RTLS) and RFID-based equipment. In the days after the results were made public, dozens of vendors posted open letters on their Web sites and distributed messages either defending their particular brand of technology, discounting the findings, or in some cases, both.
With the RFID market experiencing such rapid growth, it is certainly understandable that companies would want to defend their turf. According to an end-user study by Spyglass Consulting Group (Menlo Park, Calif.), 76 percent of larger healthcare organizations in the United States have adopted RFID solutions, primarily to track assets, patients and staff. These numbers, says Spyglass Founder and Managing Director Gregg Malkary, represent an increase of 204 percent from three years ago.
Clearly there is significant interest in the technology, and with millions of dollars at stake, the desire for vendors to clear their products' names is understandable, says Fran Turisco, a principal in the Waltham, Mass.-based Emerging Practices area of CSC Corporation (Falls Church, Va.). However, she says, the fact remains that the study results are indeed relevant, and even if the data doesn't apply specifically to every RFID user, she still hopes the results will serve as a wake-up call to the wireless industry, acting as a catalyst to improve safety.
“This is real — this is not a fluke,” says Turisco. “I read the study and then I read the responses from some of the clinical engineering folks who claimed that there was no interference, but it happened. It does happen. And there are probably other studies that show it.”
What's most important, she says, is how hospitals and health systems — even those that aren't deploying RFID — react to the findings. One important way is enacting policies and procedures to deal with a potentially hazardous incident. The study, she says, can also prompt important discussions about best practices for conducting assessments, the importance of CIOs being proactive rather than reactive in terms of testing, and the need for improved standards across the industry.
A better plan than ‘plug and pray’
Spectrum Health, a seven-hospital network based in Grand Rapids, Mich., began piloting RTLS this past spring at its Butterworth Hospital. When the JAMA study was published, Senior Vice President and CIO Patrick O'Hare received statements from vendors providing details about the tests they had performed measuring the levels of interference in different care settings. O'Hare and his staff, however, still had their own work to do.
“We conducted our own detailed testing and risk assessment on our high acuity equipment, including ventilators and other equipment in the ICUs based on the technologies that we're piloting,” he says. O'Hare's staff placed RFID tags in close proximity to different medical devices at varying levels of power and frequency and monitored the output of the equipment over a period of time to assess the relative interference and risk. “I don't think it's a perfect science by any stretch, but as long as you have vendors that are willing to stand behind and do the testing on their part, making sure that they're doing their level of diligence, and have a fairly consistent risk management program in place, you're at least putting yourself in a better position,” he says.
At Spectrum, assessments start long before the implementation — or even the pilot, for that matter. According to O'Hare, technologies are meticulously monitored and evaluated prior to the start of any trial period. With RFID, for example, O'Hare and his staff studied the Hype cycle from Stamford, Conn.-based Gartner (see sidebar below), and researched its track record in the industry long before making any moves.
According to Turisco, this kind of planning is critical when it comes to wireless solutions. “A lot of places say they don't do a ‘big bang’ implementation. Instead, they try it in one unit or several units and then roll it out. That's the time you would have to do the testing,” she says, adding that a technology as diverse as RFID comes with a host of different variables.
Not only can it come in the form of either active or passive, but RFID can be utilized for various functions and paired with different medical devices. Add to all this the fact that there are so many models available from such a large number of vendors, and it becomes even more crucial that the needs of each facility are addressed individually. “The onus has got to be on the hospital. You can't put it on a vendor,” she says. “You can ask your vendor what they've done for testing and what the risk is with their technology, but ultimately, it needs to be the hospital's responsibility.”
For ‘wired’ hospitals, running the types of comprehensive risk assessments suggested by Turisco isn't just a precautionary measure; it's a best practice. At Spectrum, as soon as the staff starts to discuss implementing wireless technologies, a thorough spectrum analysis is performed, along with RF frequency distribution studies. These studies determine the appropriate levels of access points and wireless technology needed to support the infrastructure, according to Scott Drazen, director of enterprise technology services. “We have to physically account for all the interference that might exist within a space from high-end magnets as part of the MRIs or cellular devices,” he says. “So we have a fairly good and strong methodology for approaching those situations to mitigate that risk.”
The importance of conducting thorough assessments and launching a proper response in the event of an incident is something the staff at Spectrum learned firsthand. Strangely enough, the experience also taught the staff that interference doesn't always stem from internal issues. A few years ago, unusual activity levels were detected in one of Spectrum's facilities. According to O'Hare, an in-house investigation revealed that the source was a signal being emitted from a nearby building that was interfering with the hospital's connections and causing drops in wireless coverage. Upon this discovery, the staff immediately took action and began negotiating with the other party to devise a solution.
The lesson was an important one, as the issue has resurfaced in recent months. According to Drazen, the staff has picked up on interference that is impacting the Lemmen-Holton Cancer Pavilion, a facility that opened its doors in June. This time, he says, the health system is tapping the FCC for assistance.
“We've had some spectrum graphic analyzers looking at the signal patterns trying to understand what might be the source and where it might be coming from, because it seems to be inconsistent in distribution in terms of the facility that's experiencing the problem,” says Drazen. “It's not our internal services or technologies that are causing the problem.” Spectrum's executives are consulting with the FCC for assistance in determining the source of the interference, which has resulted in intermittent activity with various wireless devices, including laptops and tablets, phones, voiceover IP, and nurse call technology.
Test, and test again
While knowing how to manage an incident is crucial, having the foresight to conduct vigorous analysis at all stages of implementation and remain vigilant is just as important, says Tim Zoph.
“It's a very active environment, and it takes very active management,” says Zoph, vice president and CIO at Chicago's Northwestern Memorial Hospital, an 897-bed system that utilizes a wireless antenna infrastructure from MobileAccess (Vienna, Va.) in all its patient care facilities. As someone who monitors an infrastructure that can support several robust applications — including wireless fetal monitoring, telemetry, X-ray transmission and more — Zoph can attest to the importance of conducting signal strength analysis before, during, and after technologies are installed.
Contrary to what some may believe, the latter part is just as critical as the first two steps. “Once you put equipment and people in a physical environment, it certainly does impact your coverage,” says Zoph. “We did a lot of signal strength analysis after we opened to make sure that our coverage maps were current, and that we were not seeing any dead spots in the building.”
This brings up another critical point, says Zoph, stressing that facilities need to guard against oversaturating the wireless environment. This crowding, he says, can inadvertently create interference, particularly with technologies like voice communications that are more susceptible to disruption.
Another key advantage in analysis is being able to monitor the source of interference, which can have a very unlikely culprit, he says. After running a routine test, Northwestern's staff found that the microwave ovens installed in the kitchen were impacting the facility's wireless signal. Thanks to constant monitoring, a crisis was averted. “It's a sensitive environment that takes proactive management and monitoring,” says Zoph. “You've got to be receptive to what your users are telling you about their experience, and then respond to it.”
The wireless environment, he says, needs to be constantly reanalyzed, both for its coverage capabilities and to check interference levels. “The environment, especially with people, process and technology inside of it, behaves differently from a wireless perspective than it does on its own, just in an isolated environment,” he says. “You need to come back and verify, once the environment is fully built up and operational, that the system is performing as planned. And be prepared that you've got some adjusting to do.”
Perhaps most importantly, CIOs need to judge what type of situation requires a yellow flag — meaning an implementation continues while the technology is closely monitored — and what type of situation merits red. Even the most wired hospitals can experience difficulties or fall victim to an incident such as interference, and in these cases, it's important to have a plan. Says Spectrum's O'Hare, “We have various practices now where if we think we're under a certain level of threat that puts an implementation at risk, we take action and mitigate those risks, including, if necessary, cutting ourselves off.”
While these types of incidents are rare, they do exist, and it's much better to be ready than to attempt to devise a response on the fly. “We haven't had to do this in years, but if we perceived a particular threat, we'd bring down our connections if we needed to,” says O'Hare.
A wake-up call for the industry
While the findings from the JAMA study have prompted important discussions from both vendors and hospital executives, one group Turisco wants to see more response from is industry associations.
According to Turisco, the Association for Automatic Identification and Mobility (AIM Global, Warrendale, Pa.) has issued a call to arms to the vendors encouraging them to collaborate with industry associations to devise better methods of testing. This, she says, could go a long way in providing hospital executives with some piece of mind. But while the statements issued by AIM Global and other associations are certainly a step in the right direction, she wants to see less talking and more action.
“They need to take it up a notch,” says Turisco. “There has to be an industry-wide response to this, as opposed to letting individual hospitals figure it out. These results raise important questions that need to be answered.”