The archetypal wireless medical device is the telemetry monitor for measuring electrocardiographs. First introduced in the 1970s, cardiac telemetry systems were pretty straight forward. Analog signals were transmitted with each telemetry transmitter/receiver using its own dedicated channel. Medical device vendors placed ceiling mounted antennas connected with coaxial cable back to central radio frequency (RF) transmitter/receivers in a wiring closet. There were no other wireless medical devices. Nor were there any wireless LANs - or even wired local area networks, for that matter.
A lot has changed in almost 30 years - I mean besides feeling older.
The nirvana that was the 1970s came to an abrupt end on February 27, 1998 at 2:17 pm, when, "WFAA-TV channel 8 television began broadcasting on digital TV channel 9 and continued until 10:35 p.m., shutting down transmission a few times to allow a tower crew to work on the antenna." This and subsequent tests of digital television broadcasts by the Dallas broadcaster, knocked Baylor University Medical Center's (BUMC) telemetry off the air. Fallout from this intentional (and completely legal) interference resulted in the creation of the new WMTS (what FCC called Wireless Medical Telemetry Service) frequencies for use by telemetry monitors. Between that fateful day in 1998 and 2006, BUMC has spent $6.6 million shifting frequency and upgrading the telemetry systems at their hospitals. (You can read about BUMC's ordeal reprinted from the AAMI publication Biomedical Instrumentation and Technology Journal story on this FDA web page.) So the new WMTS solved all our wireless medical device problems, right? Although some may differ, the bottom line to the foregoing question is a definite "no."
About the only thing WMTS has going for it is that the designated spectrum is "protected." A "protected" frequency is one where you can make someone generating intentional interference cease and desist - once you've successfully identified the offending source of interference, made the appropriate legal requests, and perhaps responded to rebuttals. As you might guess, this whole process can take weeks or months, which is a problem is your wireless medical devices are unusable in the interim. The incidence of intentional interference where a hospital has greater rights than the interfering party is almost nonexistent. By far, the greatest source of RF interference in WTMS (or almost any other band) is unintentional, resulting from bad brushes in a hair dryer motor, faulty fluorescent light ballasts, noisy paper shredder motors, and a myriad of other sources.
There are numerous flaws with WMTS.
- First the frequency bands are not contiguous (608-614 MHz, 1,395-1,400 MHz, and 1,429-1,432 MHz) which adds to the cost and complexity of developing and deploying products using WMTS.
- The bandwidth available in WMTS is just 13 MHz - barely enough to deploy a few hundred channelized patient monitors in a large hospital. This was sufficient for current requirements in 1999, when telemetry transceivers were the only wireless medical device in use, but not in today's hospitals.
- The WMTS band specifies frequency only; there are no provisions to ensure reliable coexistence and maximum utilization of the bandwidth between vendors. Consequently, it has taken years for most coexistence problems to be worked out between vendors. One can argue whether the solutions reached to date actually maximize WMTS spectrum.
- The 608-614 MHz portion of WMTS is susceptible to co-channel interference from near by digital television broadcasters. This adjacent channel interference is perfectly legal and hospitals have no recourse but to narrow their use of that portion of WMTS (that represents almost 40% of all WMTS bandwidth).
About the same time FCC designated WMTS, the IEEE ratified the first standards for wireless networking. These new standards included 802.11 (also known as 802.11FH or frequency hopping), 802.11a and 802.11b. Some medical device vendors, looking to develop next generation wireless medical devices at the time, evaluated both WMTS and the new IEEE 802.11x standards. One vendor that chose to go with 802.11 was Protocol Systems (acquired by Welch Allyn). GE launched Apex Pro using WMTS, while using 802.11 for their patient monitors. Spacelabs, Datascope and others also ran their telemetry on WMTS, putting 802.11 in their wireless medical devices. The exception here is Philips, who operates all of their patient monitoring devices - telemetry, patient monitors, defibrillators - on WMTS. (An exception might be their EKG carts which I believe use 802.11x, while all other Philips factories have built 802.11x into their medical devices.)
Vendors with existing wireless telemetry products (notably GE and Philips) rushed revisions of existing products that supported the new 608-614 MHz WMTS frequency. Referred to as "re crystaled" these upgrades and new products simply shifted the old frequencies used by BUMC by incorporating the appropriate RF components. Late 1999 and 2000 were big revenue growth years for the patient monitoring market, as many hospitals looked to upgrade or replace telemetry systems in response to unexpected interference from digital TV stations. Upgrades were less expensive than outright replacement with non-WMTS technologies, and most hospitals went this route. This significant turnover in telemetry systems was leveraged most effectively by Philips Medical Systems, and the resulting sales cemented their position as the number one patient monitoring vendor in the U.S. - a distinction they've maintained since.
During the late 1990s new wireless medical devices came to market, notably "smart" infusion pumps. The past few years have seen the advent of wireless point of care testing devices from Abbott and Johnson & Johnson. Vendors like iSirona and Capsule Tech have launched wireless modules to connect legacy medical devices. Expect all point of care testing devices that are carried to the point of care to eventually go wireless. With the recent news that CMS plans to end reimbursement for ventilator acquired pneumonia in 2009, we can expect to see ventilators go wireless too.
The fact is, that 802.11x has become the defacto standard for wireless medical devices. There are two basic reasons 802.11x has come out ahead of WMTS. First, 802.11 has proven to be safe and effective after years of experience. Patient monitors using 802.11 have been shown to be more reliable than channelized telemetry (regardless of frequency used) and virtually as reliable as wired Ethernet. The second reason for 802.11x's ascendance is that the technology is much less expensive to develop and build in to medical devices - not to mention being less expensive for customers too.
In fairness, I should mention another advantage of WMTS: placing some medical devices on a separate wireless infrastructure does eliminate a single point of failure. If WMTS fails and you lose telemetry (and possibly other patient monitoring stuff), all your wireless medical devices on 802.11x will probably still be operational. Planning enterprise architecture to minimize single points of failure is a good thing, although you don't have to use WMTS to accomplish this objective.
There are two big advantages to vendors using WMTS. First, by running on a separate infrastructure service and support is greatly simplified. In most installations, the medical device vendor has WMTS all to themselves and doesn't have to worry about pesky variables introduced by third parties (including the customer fiddling with things). Finally, requiring a dedicated infrastructure for your solution increases customer's changing costs (even if it doesn't necessarily raise the acquisition cost too). This factor comes into play when the buyer (inevitably) gets into a tiff with the vendor and wants to replace them, not only must the customer replace the medical devices they also need to replace the wireless infrastructure. Given that 802.11x is a shared infrastructure this is more a matter of when a buyer may make certain investments into their wireless LAN, rather than having to make infrastructure investments that are dedicated to a specific medical device or vendor.
That's not to say that there aren't costs involved in deploying medical devices on 802.11x wireless networks - there are, and they're getting ready to go higher.
To bring us up to the present state of the art, you must read Medical-Grade, Mission-Critical Wireless Networks, by Steve Baker and Dave Hoglund, in the March/April issue of the IEEE Engineering in Medicine and Biology Magazine. You can buy an electronic copy of this peer reviewed paper for $35. I'm usually pretty critical of journals that sell the published results of research funded by my own tax dollars, but that is not the case here. Baker and Hoglund wrote their paper based on years of industry experience developing and deploying wireless medical devices - no tax dollars were harmed in the making of this journal article - and the content is well worth the cost.
Deploying 802.11x medical is not simple - how the hospital controls access (and security) to its wireless network is critical along with how it wants to create its “virtual lans” that support these devices are critical issues. Also it is not always vendor neutral - ask Welch Allyn about Aruba Networks vs. Cisco. One is supported and as of the writing of this comment, one is not. This is true of other medical vendors too.
What about the people who were early adopters of the inital wireless networks who used frequency hopping access protocols now that the frequency hopping access points are no longer sold (technolgy moves on)? They are left behind and forced to upgrade their networks too in a much shorter timeframe than some WMTS solutions. What does that say about current 802.11a/b/g vs .11n vs whatever alaphabet soup of wireless network “standards” are around the corner? Most medical vendors still use lowest cost 802.11b, correctly pointed out in the reference article as a speed bottleneck when it comes to access point utilization. Ask a vendor about g or a and you get some sort of mumbled “incorrect bus speed” answer.
The point is that although 802.11x has become one kind of recognized standard - the implementation of it requires, at the current time, a lot of work, and generates a lot of confusion in the process.
802.11 has some major issues, and it is important to balance these when considering medical applications.
The 802.11 standards provide no guaranteed access slots for devices trying to reach a wireless access point. In an 802.11 network there is little that can be done to control the latency of critical data over the air.
In addition to competing with other medical devices connected to the access point you are also competeing with other services that may have been rolled out on the 802.11 network including patient data systems, and even VOIP services.
Many vendors use trafic shaping to prioritise data on the network, but this is only effective on the wired side once data has reached the AP.
802.11 was also designed to be slow moving. The standard does not handle hand over between access points resulting in further latency and retries if a device is moving.
802.11 resides in the ISM frequency bands. These are shared by a huge range of devices from Bluetooth headsets to the nike plus transmitter in your shoe. It can become a busy frequency band, further delaying your data.
So WMTS and 802.11 both have issues to consider.
With WMTS a device designer can develop a system that is optimised to the requirements of the application. For continuous ambulatory monitoring you cant beat the optimised performance that can be achieved.
However the system is proprietary, so you cant reuse the infrastructure for sending emails..
Craig and Paul have both raised excellent points. 802.11a/b/g for wireless medical devices is not a panacea for easy to design, deploy-and-forget connectivity. But then neither is WMTS.
To my way of thinking, WMTS made more sense back in the day when the only wireless medical devices were telemetry packs used in just one nursing unit. Now there are many types of wireless medical devices, and they’re deployed enterprise wide.
WMTS has 2 very serious fundamental limitations:
1) With just 13 MHz of bandwidth, WMTS lacks the elbow room to support the growing number and types of wireless medical devices. Deploying all your wireless medical devices on the same infrastructure should make designing, deploying and managing the infrastructure a more manageable process.
2) WMTS has no standards. Sure ISM and 802.11a/b/g are not perfect, but these standards help ensure coexistence and interoperability. They also maximize capacity of the available bandwidth and provide powerful testing and monitoring tools. Every vendor’s implementation on WMTS is proprietary, and given the limited bandwidth and proliferation of wireless medical devices running everything in a more simple and inexpensive channelized fashion is too inefficient. So not only must device vendors create the wireless radios and receivers, they must use sophisticated technologies to maximize capacity - all while building their own monitoring and system management tools. Even when vendors appropriate other commercial technologies for use in the WMTS band, like Philips did with DECT, monitoring and management tools are lacking. Yes, Wi-Fi can be crowded and complex but the resulting commercial ecosystem has resulted in ongoing innovation to improve and monitor performance, helping to ensure safe and effective communications.
The commercial ecosystem around ISM and 802.11a/b/g standards has created better performance and manageability - at a much lower cost - than is possible with the proprietary technologies required for WMTS.
Deploying wireless medical devices regardless of band or technology is a daunting task. And with the advent of IEC 80001, it is going to get more daunting still.
I have deployed both VHF/UHF/WMTS and 802.11 within the patient monitoring domain. Also, was the architect behind OneNet from Draeger. To this, I have also worked for Symbol, with Welch Allyn, and am intimately into Cisco LWAPP, Aruba, and Meru.
WMTS will be around, but not longer term. Proprietary radios and networks are a thing of the past. Tools are available for real time spectrum analysis, network management, and QoS in the 802.11x realm. This will never be available for WMTS simply because economics of scale of the technology envelope. The high cost of WMTS is also questionable when 802.11x costs are continually coming down. However the medical device industry does need to drive some improvements to the 802.11 standards, not unlike 802.11x to ensure QoS and interoperability were added in response to needs from the enterprise IT market.
I was wondering if someone had an opinion on the advantages/disadvantages for smaller setting hospitals and outpatient services, 200 beds or less, when considering WMTS vs. 802.11 networks for patient monitoring devices?
Thanks in advance. This is a great community of ideas!
Here’s my perspective on your question.
1. Splitting off patient monitoring on a separate network (rather than using Wi-Fi) does eliminate a single point of failure. But then it’s another network that needs to be designed for high reliability and actively managed to ensure continued performance.
2. The claim that WMTS is “protected” is moot for two reasons: 1) the vast majority of interference is unintentional interference (where having a protected frequency does you no good), and 2) Wi-Fi standards, and the systems that use them, are designed to facilitate coexistence among many different vendors and devices in the same environment - thus rendering the “protected” claim of WMTS in the face of intentional interference moot.
3. While Wi-Fi is based on industry standards with cross vendor interoperability, every vendor’s implementation of WMTS is proprietary. Switch vendors, switch out all the WMTS infrastructure. This increases “switching costs” when hospitals want to change vendors. Vendors like this because it tends to lock in customers.
4. WMTS comes in two flavors, a low cost / low capacity system and a higher cost / high capacity system. A trend in hospitals for the past few years (and continuing) is to want to monitor patients in broader areas (even house-wide). The number of patients being monitored is also increasing. The issue here is to really think about your current and future coverage and capacity requirements. If you outgrow the low capacity system, you have to replace most or all of the infrastructure to upgrade.
5. WMTS specifies a frequency band and provides nothing for coexistance. Consequently, there can be coexistence issues between different vendors products. Currently GE and Philips have things worked out, but a new product release from either vendor could upset the apple cart.
While using Wi-Fi for your patient monitors is more complex, there are many excellent monitoring and trouble shooting tools available. This is not the case with WMTS - since these systems are proprietary, anything that is available must be developed by the vendor (GE or Philips).
The good old days of throwing up an analog antenna system for trouble and maintenance free wireless patient monitors are gone.
I understand your position on WiFi, but have problems with the comment on interference:
“The claim that WMTS is “protected” is moot for two reasons: 1) the vast majority of interference is unintentional interference (where having a protected frequency does you no good)”
The issue with the 2.4GHz band is that there are just many more active transmitters - even in a hospital environment.
People with WiFi and Bluetooth running on their phone and using headsets, wireless keyboards and mice, and even Nike plus foot pods. Some fire alarms use this band for smoke detection. 2.4GHz is a very crowded space.
Would you suggest that system developers look at the higher 802.11 frequency bands?
Paul, the ISM band is indeed crowded - yet it never seems to get too crowded. With the increasing adoption of 5Ghz, the available bandwidth in ISM has much room for growth.
Any wireless deployment, whether in the ISM band or WMTS, must be proactively planned, designed, and managed after installation. Admittedly, this is a more complex task in ISM. It seems to me that this complexity is a small trade-off for
And of course, WMTS remains a legitimate choice for medical devices - if only to avoid putting all your eggs into one RF basket.
Such decisions just need to be made with eyes open and an awareness of all the implications.
Not only GE and Philips are in the telemetry market. Spacelabs was the inventor in the 70’s with Appollo program. Draeger is also in the TLM field.
I enjoyed reading the article and excellent comments, and considered two low-rate protocols which are being considered in academia.
Is ZigBee a valid alternative to 802.11a/b/g in scenarios that include a small healthcare facility, non-critical patient monitoring (for instance, post-op, emergency room and recovery) and out-patients?
What about Bluetooth in association with 802.11a/b/g?
Thanks in advance.
One of the biggest challenges facing health care providers is the increasing number of wireless devices. This is good because wireless better supports the types of workflows in health care, but it is also bad due to the increased complexity of managing a more crowded wireless environment.
Most of these wireless devices are in the ISM band, and the vast majority are based on 802.11 standards, in other words wireless LANs.
In the enterprise, both ZigBee and Bluetooth can produce coexistence challenges. Many providers are poorly equipped to resolve these coexistence problems when they arise. In the home environment, or outside of both the home and enterprise, these technologies are much more easily deployed.
Given the broad picture above, both Bluetooth and ZigBee are better suited for low powered wireless sensor radios than as an alternative to WiFi for enterprise communications.
The care delivery areas you mention are all points along a broader care delivery continuum, and as such are better served with an enterprise-wide technology. As you note, bandwidth is an important requirement. A typical multi-parameter patient monitor generates about 12 kb/s of data. It does not take too many monitors to overwhelm the bandwidth available in ZigBee. Running Bluetooth at a power level high enough to serve as an alternative to WiFi would likely result in coexistence issues.
The bottom line is that WiFi is the accepted enterprise wireless infrastructure in health care. There are roles for ZigBee, Bluetooth and other technologies, but they are typically in applications other than connecting devices to enterprise networks.
Great Website. My situation is this, I am currently using quinton QRS telemetry system in my outpatient clinic. I monitor up to 8 patients an hour on my current system. Can I integrate a seperate system to a satelite office and be able to store pt info on the main system?
What about wireless effects on the human system? I had read somewhere that any frequency above 2.8GHz is harmful to the human body.
Also what about an architecture where UWB is used for short distance, high bandwidth communications, to a gateway. Once a gateway is reached multiple wireless techniques could then be used to transmit the data to remote access point.
WiFi I feel is getting very crowded. I could be wrong but in spite of all the co-existence build in, when it comes to critical time bounded data we would need to look at alternatives.
I find this very informative and interesting. I have learned A LOT here. I am a tele tech watch 50 channel(which is too much safely for one person) in a cardiac unit. I have been doing this for 13 years and I have encountered ALL kinds of “drop out” with our Phillips system and even WORSE with space labs. I actually wish we had our old out dated Phillips system back cause Space labs is horrible and NOT user friendly what so ever. Space labs calls their drop out “scwelch” which I find to be HOG WASH as the send spectrum analyzer after analyzer only for us to be told, you may wanna try a different patch. I call bull on that! I would assume our hospital took the cheapest route possible because we too had to reroute wires through out the house. I will check into what GHz we are @ since I and another tech or siting with cancer. Dually noted on the band jumping since we have two other telemetry systems in house, datascope in ER and Phillips in OB/PEDS. My secretary sits with an ear piece…. I would assume is blue tooth about 3 feet away from the central system. Could this be causing our drop out.
Philips is using ‘smart hopping’ at 1395-1400 and 1427-1432 MHz as per their MX40 Service Guide. Bluetooth uses 2.4-2.5 GHz. It’s unlikely the bluetooth earpiece is an issue, but why not have her use a wired one just to see if it is an issue.