STATVIEW
Wireless LAN technology is coming of age. More rugged mobile devices, longer battery life, quicker return on investment and product improvements driven by increased vendor competition are making wireless an attractive technology for healthcare.
Wireless LAN technology involves sending and receiving data without using cables--usually by radio frequencies (RF), sometimes with infrared (IR). The data flows between healthcare enterprise computer networks and client devices that include mobile (and stationary) computers, medical monitoring devices, beepers and pagers, mobile telephones and other gadgets. Sounds dull, just like electricity sounds dull: It’s what you do with wireless technology that makes all the difference.
For example, wireless can link LANs in separate buildings on a geographically dispersed healthcare campus. Wireless bridging replaces cables between contiguous buildings and eliminates costly T-1 links between distant (up to 25 miles) buildings. Your savings on thousands-per-month T-1 rentals will pay for the transceivers on each building in a few months--big and rapid return on investment here.
Wireless can link desktops to LANs without running cable. Without being tethered to the LAN, wireless medical monitoring devices (such as cardiac monitors from Marquette Medical Systems, Annapolis, Md.) can transmit data to the LAN from wherever the patient is--even as they move within the facility. Patient records can have up-to-the-nanosecond information for more accurate data, more rapid decisions, more timely care and more efficient use of resources.
Doctors and nurses can update patient conditions at bedside. This means, first, that the paperwork gets done--no interruptions, end-of-shift confusion or delays. Second, no "double documentation"--first to paper chart, then to computer system--so less wasted time and fewer opportunities for error. The same personnel can care for more patients, more efficiently.
Roaming healthcare workers (respiratory therapists, anesthesiologists, and so forth) can find their patients in the facility more easily. They can download past patient conditions and upload new information from bedside.
Up-to-the-moment checks of patient condition and the latest physician instructions can mean faster, more certain and more accurate medication dispensing. The legal folks like that.
Push and pull modes
Wireless technology from companies like Data Critical Corp., Oklahoma City, can intercept alerts from medical monitoring devices and notify the right person immediately. (These systems have Class 3 certification for use in life-threatening situations.) As Brad Harlow, vice president and general manager of Data Critical indicates, this alert notification can happen in "push" or "pull" modes.
With "push," the system pushes the alert out to a special pager, notifying the nurse in charge. The nurse can see an actual picture of EKG and other graphical information, not only text data. Currently, someone has to constantly monitor these devices from a central station, then when an alert occurs they must either find and notify the nurse in charge, or find someone else--who may know a little about the patient. Wireless means fewer workers needed to monitor, faster response time, clearer information and improved quality of care. With such real-time and accurate notification, the nurse can conceivably monitor more patients.
With "pull," the user can pull the information off an intranet Web site from anywhere on earth, as they need it. They can even phone into the facility remotely and review the latest patient information, viewing graphical as well as text data on special telephones. This lets personnel handle more patients, at more facilities, than before.
One system including wireless technology at Maimonides Medical Center in Brooklyn, N.Y., reported advantages including: fewer duplicate medical tests (reducing costs); swifter arrival of routine medications at patient areas; decrease in problem medication and medication discrepancies; lower re-admission rate; reduced patient length of stay; greater capacity for servicing new patients; and improved overall patient satisfaction.
Radio waves
How does wireless work? It’s basically the same principles in your car radio, applied to computer networking. Traditional LANs send data around as electrical signals over wires. But with wireless, stationary "access points" turn LAN data into radio signals and back again. Client devices--laptops, monitoring equipment, pagers and so forth--communicate by radio with these access points, sending and receiving data.
(Infrared is another wireless technology, but since it requires line-of-sight between sender and receiver, it can’t penetrate walls. Radio can. Still, radio won’t penetrate some places either--X-ray rooms and elevator shafts, for example. That’s why you do a "site survey" before installing your system, to place access points to achieve optimal coverage. The range of each access point is several hundred feet (indoors), often enough to cover a single building.)
There are two technical schemes for using radio frequency wireless. Frequency-hopping uses up to 83 different channels, each channel about 1 MHz wide, to compensate for interference. If some equipment--or, more likely, another mobile user--is hogging one of the frequencies, the signal moves to other unused ones and the data flows.
With fewer interference problems, frequency-hopping is more attractive for enterprises with many users. Since it is more forgiving of interference, installation can be easier and more flexible. However, all this frequency hopping does come at a price: more up-front processing to decide which channel to send a piece of data through. Furthermore, the narrowness of each channel limits the throughput of the system. Frequency-hopping systems will probably max out at 1-2 Mbps (megabits per second)--which may well be fast enough for most text-based applications.
The other scheme is direct sequence, using three channels, each channel 21 MHz wide. Because of the wider channels (and a new physical layer in the 802.11 standard), direct sequence will allow data rates up to 10 Mbps by the end of 1998. (That’s the standard data rate for many cable LANs now.) These higher data rates in turn will allow more powerful applications, supporting more graphical data. (Speaking of 10 Mbps, RadioLAN already achieves that using direct sequencing in the 5-plus GHz band.) The disadvantage of direct sequence is that it’s more vulnerable to interference. Furthermore, installation is trickier, because you need to choose a channel to avoid interference.
Naturally, the two schemes don’t talk to each other, so you must choose either frequency-hopping or direct sequence. If you have many users in a small area, or problems with interference or security, frequency-hopping is the way. But if data speed or range is, or will become, important to you, you’ll want direct sequence for the promise of things to come. Most facilities will install a 10 Mbps wireless backbone--to be available by the end of 1998--that will support faster applications and be backwards-compatible with 1-2 Mbps apps.
Choose wisely. If department A makes one choice and department B makes the other, you are never going to unify them: One will eventually have to change. If two facilities with different systems merge, again you have to choose one over the other.
Barriers to wireless
There have been problems with wireless LANs, but new technologies are overcoming them. In the past, the antenna on the laptop was a two-piece job, connected by a wire vulnerable to the rough and tumble of medical environments. Now, most of the electronics are safely tucked into a Type III PCMCIA card within the laptop, with the merest nub of an antenna protruding. This makes the wireless client as durable as the laptop.
Industry leader Proxim, Mountain View, Calif., has developed a Micro ISA board smaller than a business card. Such a small and low-power device--containing its own transceiver--could be part of a patient identification badge. Products using the new board should appear by the end of 1998.
Batteries have been a bane, too, rarely lasting through a nursing shift. But several technologies are combining to change that. Newer lithium ion batteries provide nearly four times the voltage of nickel-metal-hydride (NiMH) batteries, but at a greater cost. Newer battery technology will keep improving voltage and life. The laptop’s structural design can help too. Some laptops let you swap out modules rarely used in healthcare--like floppy drives, modems or CD-ROM drives--and add extra battery packs instead. More recent laptops and operating systems also provide better power management. Many hospitals have developed a low-tech--and effective--solution also, installing the laptops in carts that carry medications, other supplies and extra batteries.
Cost can be an issue if the analysis is superficial. Wireless solutions may cost three to five times what a wired solution would cost. Only when you look at the activities do the costs make sense. Is it more cost-effective for a nurse to be at bedsides checking vital signs, or back at the central station, transcribing paper notes into the computer? Are you paying a nurse or a data entry clerk? Besides, as Ken Kleinberg, research director at GartnerGroup’s IT Healthcare, points out, you receive many benefits by having your information in electronic format right from the beginning: Many applications become possible down the line. Paper is an obstacle, and wireless is a way around it.
Formerly, wireless would start out as a pilot program, usually by a single department (like respiratory care). These days, it’s more usual to embrace the whole facility. The system may still start out for a single intended use, but--once installed--other uses find the system.
If you build it, they will come. The most growth is in inpatient care. Nurses rapidly realize how it saves steps, simplifies and improves documentation, and strengthens medication and dietetic programs. Plus, all that information is accessible where and when you need it. Still, as Kleinberg notes, less than 10 percent of healthcare enterprises are using wireless.
Different organizations have different security concerns. If you’re concerned about all that confidential data flying through the air, you’ll be glad to know that vendors are too. Products can require user authentication and password protection at each client device. The data itself is routinely encrypted and scrambled before transmission.
The outlook on wireless is promising. As a First Consulting Group white paper by Jerry Mourey points out, perhaps the most important aspect of selecting a wireless solution for a healthcare environment is choosing a proven and reliable technology. With companies rallying to the 802.11 standard, you can expect lower prices, more competition, better interoperability, and greater variety of products. New medical monitoring devices will include wireless as a feature. For once, you have a technology that is a safe bet. If wireless isn’t already part of your plan, it’s time to rewire that plan.
Standards in Wireless
THE BIG NEWS IN WIRELESS LAN STANDARDS is IEEE 802.11, ratified in June 1997. The 802.11 standard specifies the physical (PHY) and medium access control (MAC) protocols. PHY defines one infrared and two radio frequency options (frequency-hopping and direct sequence: but they still don’t talk to each other). Medium-independent MAC lets wireless components talk with ordinary wired components--like Ethernet or Token Ring--seamlessly, an important consideration.
Both PHY and MAC boost performance. The tight PHY specification means conforming products must perform better, with less variance in quality among vendors. Devices typically have a wider range, tolerate interference better, produce less interference and use the bandwidth better. The interaccess point protocol (IAPP) specification determines how one access point hands off a mobile user to another access point.
Phil Belanger, co-author of the MAC spec (and vice president of marketing for Aironet, Fairlawn, Ohio), notes that the MAC protocol had received criticism for its 500 pages. But the complexity is worth the result: The MAC protocol is very efficient.
The standard uses the same FCC-unlicensed industrial scientific medical (ISM) bands: the 2.4 GHz (gigahertz) area, the 900 MHz (megahertz) area that cordless phones also share, and the 5-plus GHz and above range.
Unlicensed means less bureaucratic interference, but possible electronic interference, since anyone can use those frequencies.
The new standard is already having effects on the industry. Previously, every company could have its own proprietary scheme--like every railroad having its own width between rails. Now companies can stick with their own scheme, move to 802.11, or support both. Most companies are putting out 802.11 products.
For users, 802.11 could be a boon. You no longer have to lock yourself into one company’s proprietary scheme, depend on their products alone and fret about their existence. As long as components satisfy 802.11, you should have interchangeable parts. As Belanger points out, implementors know they will be able to buy compliant products, like buying Ethernet-compatible products. For administrators, this lowers the risk of getting into wireless.
Testing for the truth
Unfortunately, sometimes compliance is in the eye of the vendor, and one GartnerGroup report suggests that true interoperability will remain elusive.
Recent testing by the Tolly Group (www.tolly.com) indicates significant differences in throughput, performance, reliability, sensitivity to interference and robustness of connection by similar products from different vendors under identical conditions: Try before you buy.
Organizations like the Interoperability Lab at the University of New Hampshire help keep things honest. Michael Froning, marketing manager, explains that the lab tests conformance (does the product conform to the standards?) and point-to-point (do two--or more--products interoperate correctly?). Currently, vendors pay the lab a fee, provide their products and receive the test results. Froning says this system will change: Unfortunately, some vendors aren’t reporting test results accurately. So, the lab will soon post information on its own Web site (http:// www.iol.unh.edu). Not only will users see precisely how compliant each vendor’s products are, but it will help protect the lab’s integrity too. In a similar spirit, the Wireless LAN Alliance is a vendor consortium to educate users about the benefits of wireless.
What’s next?
Competition among 802.11 vendors should drive prices down and quality up. Furthermore, expect more innovation and variety in products. With an open standard, any company can identify an unexploited niche and dive in with a new product.
Not everyone shares this rosy view of 802.11. For example, users of the OpenAir standard (based on Proxim’s RangeLAN2) probably account for 80 to 90 percent of the hospital market. This large user base makes OpenAir important. And that’s the point: a standard doesn’t matter much without a base of support. Without knocking 802.11, Scott Lucas, industry marketing manager for Proxim, points out, "It remains to be seen how successful 802.11 will be."
Indeed, even if 802.11 succeeds, extensions to 802.11 may ensure its eventual demise. The Wireless LAN Interoperability Forum is using Proxim technology to create beyond-802.11 capabilities. Another extension, called HyperLAN, would use above-5 GHz frequencies, pushing 23 Mbps. There is no reason to expect that will be the final limit.
CASE IN POINT
Pitt County Memorial Hospital, Greenville, N.C.
"If we took away their wireless, they’d revolt," says Bruce Hedreen. Hedreen, network systems analyst for University Health Systems, is speaking about the clinical users of the wireless LAN system at 800-bed Pitt County Memorial Hospital in Greenville, N.C.
Originally, some clinicians saw a wireless system elsewhere and wondered if their facility could do something similar. The project began in 1996 by examining the market to see what was available, reading reviews, talking to other facilities and trying out some systems.
They began winnowing the candidates using their facility’s own concerns. (Naturally, the concerns are different for every enterprise, but their decision process is instructive.) They wanted the widest bandwidth; at the time, this was 2 Mbps, eliminating systems with lower bandwidths. Rightly concerned about power consumption of client devices, they scrutinized the power features of the remaining choices, narrowing the list further.
They wanted the wireless LAN to run the same protocols as the rest of their network, including TCP/IP, IPX and a little AppleTalk. Again, some systems couldn’t make that cut. The size of the access points was important. If you’re sticking access points on ceilings and in closets, you want them as small as possible.
They chose frequency-hopping wireless because they had interference concerns. They found that direct sequence had interference problems and was a power drain on client devices.
The resulting short list included Aironet, BreezeCOM, Proxim, Symbol Technology and WaveLAN. Their final requirements were crucial. First, they wanted to minimize the cost of installing access points--in particular, about providing AC power to the access point. Distributing a few dozen access points over several buildings could mean extensive electrical rewiring, possibly requiring conduits. They wanted access points that could run on DC power over twisted pair wiring.
Finally, they wanted to manage the access points centrally. Then they could reconfigure an access point without visiting the access point itself--a real concern on a geographically widespread multibuilding facility. Not many solutions offer such central management.
Their final decision was to go with Proxim solutions. The AC/DC problem is a nonissue, since they can run either way. But perhaps more important, Proxim offers Web-based central management. Administrators can visit a management intranet from anywhere on the planet and manage the system remotely.
On the air
Their current system includes about 20 access points (with another 30 on the way, says Hedreen). Advancing technology has meant they need fewer 500 mW units than the 100 mW models they had originally planned. Client devices include Dell laptops with large high-resolution screens, Fujitsu Point 510 pen devices and Hewlett-Packard Windows CE-based handhelds.
The battery problem they handle with the laptop battery itself (lasts two to three hours), on a cart with a UPS (uninterruptible power supply) for two to three hours more. When the carts aren’t in use, they plug the UPS into any wall socket and juice up again. They anticipate eventually handling hundreds of client devices.
Their pilot program was taking admissions information with wireless client devices. Next came clinical applications, taking charts at bedside, completely paperless. Soon to come will be the emergency department for rapid treatment of trauma cases. Next will be labor and delivery and the birthing centers. Users love the system.
Besides all this intrabuilding wireless, they have an interbuilding system, from Solectek. With options from 2 Mbps to 10 Mbps, they have room to grow. And a range of 25 miles.
Hedreen points out that they are proceeding gradually but steadily with their wireless program, keeping pace with demand from the various user communities, and upgrading their system to meet and anticipate those demands.
Edmund X. DeJesus, formerly a senior editor at BYTE, is a writer in Norwood, Mass.
