BBI Contributing Editor
ATLANTA, Georgia — The turnout for this year's Healthcare Information and Management Systems Society (HIMSS; Chicago, Illinois) conference was a very respectable 18,750, comparing favorably with the slightly over 19,000 who attended last year's gathering and demonstrating that attendance at medical conferences is beginning to return to normal. Interest is high anytime there is rapid market growth, and health care information technology (IT) is such a market, projected by the Gartner Group (Stamford, Connecticut) to increase from around $40 billion in 2000 to more than $60 billion by 2004.
In spite of being delayed until April 2003, the Health Insurance Portability and Accountability Act (HIPAA) was the topic nearly everyone who flocked to the Georgia World Congress Center in late January was either talking about or seeking to learn about. There also was strong interest in wireless network infrastructure — perhaps due to the FCC-mandated exodus from the current UHF and VHF frequencies due to new operating rules announced in June 2000 — as well as the increasingly wireless components of hospitals' IT infrastructure. Realizing all of the interest in the latter subject, the HIMSS steering committee structured a special HIPAA track and held a day-long pre-conference briefing on wireless infrastructures, as well as sessions sprinkled throughout the four-day conference itself. HIMSS even featured sessions that discussed the role of the new "privacy officer," the person responsible for assuring that health care enterprises comply with HIPAA privacy requirements.
Linda Kloss, president of the American Health Information Management Association (AHIMA; Chicago, Illinois), announced a $500,000 project to study the needs for future medical record professionals, based on the changing skill sets required by new technology being adopted. The study, to be coordinated by the State University of New York-Albany (Albany, New York), will run until early 2004 before completion. AHIMA also has been involved in redefining what constitutes the medical record, working on consistency in coding and some privacy issues.
Attending the HIMSS sessions, it seemed clear that many IT folks do not yet have a solid grasp on the scope of the wireless issues or needs in hospitals. Today's hospital must accommodate wireless patient-worn monitors, portable monitors with both a cabled and wireless local area network, wireless patient, staff and equipment location/tracking systems, wireless paging, voice communications and PDA networking, and are faced with a variety of bands and systems that do or do not comply with industry-wide standards. Many attendees seemed unaware of how poorly what they have already chosen to provide wireless voice and data communications would work for wireless vital signs monitoring. Such monitoring, with its need for no drop-out and accommodation of patient roaming, poses special challenges to IEEE-802.11b, the most widely deployed infrastructure in hospitals.
The U.S. hospital has become a complex wireless environment, but in only a few cases is it a well-planned environment. It often is more of a conglomeration of whatever has been installed in various clinical areas, usually without regard to an overall wireless infrastructure plan. This reflects a lack of vision on the part of hospitals, but also the marketing prowess of a few large vendors of computer and voice infrastructures such as Cisco Systems, Lucent Technologies, Intel and others. These firms have been pushing ISM-band, 802.11b-compliant wireless infrastructures into hospitals. IEEE 802.11b is a standard that specifies direct-sequence (DS) transmission at a maximum rate of 11 million bits per second (Mbps), but it is only one of a family of possible protocols for wireless LANs operating in the 2.4 GHz and other RF bands. For applications that must work over a large building with several floors, the difficulties in providing uniform coverage and the limitation of essentially using only three channels make such installations troublesome for stationary-to-stationary wireless networking, which is much simpler than making them reliable for applications where there needs to be communications between multiple points that roam throughout the space at will. In such situations, IEEE-802.11b almost never achieves its 11 Mbps data rate and remains dropout-prone in some areas.
Mostly overlooked by hospitals so far are 802.11 frequency hopping (FH) LANs. Such systems operate at 1.5 Mbps data rates. And 802.15 (Bluetooth) wireless is even slower. A variety of proprietary transmissions also are used, such as Proxim's (Sunnyvale, California) older RadioLan technology and its newer, recently FCC-approved HomeRF technology, a FH LAN that operates at 11 Mbps and is positioned to better compete with 802.11b networks, but using a frequency hopping rather than direct-sequence approach to wireless networking. Frequency hopping LANs have an advantage, as they have 79 channels to hop over, and multiple access points can be added to increase throughput as needed in a particular coverage area.
By far the most widely installed ISM band technology in general business settings, as well as in hospitals, is the 802.11b DS network. But the characteristics of DS spread-spectrum transmissions make them vulnerable to being interfered with by any of the newer, frequency- hopping modalities — of which the number grows daily. For a time it appeared that HomeRF — an 11 Mbps FH technology — might be the contender that would come forward and challenge 802.11b DS networking, but Intel's withdrawal from the HomeRF vendors' alliance last year was fatal to the new approach, which is now shepherded mostly by Proxim, one of the smaller network suppliers. Unless some industry heavyweights weigh in to support HomeRF aggressively, this new approach is doomed to stillbirth. That is unfortunate, as any FH approach in ISM is more resilient and reliable than the DS approach, particularly for many of the wireless applications that hospitals have yet to implement. To succeed, Proxim will have to stand alone and convince the entire health care industry that HomeRF makes sense — a daunting task for any vendor to face single-handedly.
The premature death of HomeRF eliminates the only high-speed, frequency hopping contender to 802.11b in the ISM band. There is still the 802.11 FH originally specified, but its 1.5 MHz data rate has caused most vendors and applications to shun it, and as a result the sales of 802.11 FH access points and transceivers in all market spaces is tiny (at least in the health care market) compared to 802.11b access points and transceivers. It is not clear how long remaining suppliers such as Symbol Technologies (Holtsville, New York) and very few others will continue to support it, or more importantly, develop new devices for 802.11 FH applications. The dismantlement of Symbol's health care specialist support, an event for which that company did not prepare its vendors, raised concerns about its commitment to the health care segment and 802.11 frequency hopping technology. Symbol previously had enticed such health care suppliers as McKesson HBOC (Alpharetta, Georgia), Picis (Arlington, Virginia), GE Medical Systems Information Technologies (Milwaukee, Wisconsin), Welch Allyn Protocol (Beaverton, Oregon), Criticare Systems (Milwaukee, Wisconsin) and others to base critical vital signs monitors that require wireless networking on its almost proprietary IEEE-802.11 FH platform. Some of these OEMs can possibly change to direct-sequence 802.11b, but for others such a change will be a major effort and performance compromise. Of the various Symbol health care partners, none is potentially more negatively impacted than Welch Allyn Protocol. Vertical market specialists from Symbol made joint presentations with Welch Allyn Protocol to hospitals in 2001, heralding the unique benefits that 802.11 FH had for roaming wireless monitoring applications such as ambulatory patient-worn monitors. That campaign was launched to convince prospective buyers that Welch Allyn Protocol's Micropaq patient-worn monitor was a better choice than migrating UHF and VHF-based telemetry into the newly created WMTS (608-614 MHz) band, the approach virtually all of its largest competitors were pitching to hospitals.
When one simply compares the operating characteristics of the two bands (the new WMTS band and the existing ISM band using 802.11 FH LANs), the ISM band offers voice and data vs. data only in the WMTS, and offers bi-directional transmission which, while available in WMTS, has only been implemented by GE Medical to date. The FH ISM band approach, with its 79 channels and hopping, made a system that is much more immune to interference — and therefore reliable for patient data — than using an 802.11b DS approach, with its three broad bands of frequencies. The joint campaign was effective, creating the largest backlog in Protocol's history, and contributing to the strong third- and fourth-quarter revenues that company booked in 2001. Yet just as momentum was mounting, Symbol abandoned the health care vertical market, leaving Welch Allyn and many others wondering if 802.11 FH will remain viable and if any new development will happen in the frequency hopping LAN space. Even if Symbol continues to produce FH, will hospitals be comfortable supporting both an 802.11 FH and an 802.11b DS infrastructure in the same physical space, particularly as new technologies emerge? This question becomes more important with the IEEE-802.11g proposals to increase the data rate of DS to several times its current 11 Mbps rates. 802.11g is being driven to compete with the 802.11a (5.7 GHz) LANs, which also offer higher data rates, but require yet another wireless infrastructure be installed in hospitals. If hospitals begin to think of 802.11 FH as an archaic technology, as they already do for HomeRF, a major component of the Welch Allyn value proposition to hospitals is tarnished.
The presentations at HIMSS left little doubt where ISM band wireless development is headed. In the preconference presentations by Helmuth Orthner, PhD, of the University of Alabama (Tuscaloosa, Alabama) and others of his group, there was virtually no discussion about 802.11 FH regarding future network development. The talk was about 802.11b (current DS), 802.11g (faster DS), 802.11a and 802.15 (Bluetooth). Like HomeRF, 802.11 was simply missing from the presentations. That same issue was discussed during the conference itself in a standing-room-only session on the future of wireless PDAs in health care. Despite one speaker being a Symbol vendor employee and another the former vice president of engineering at Welch Allyn Protocol, the discussion revolved around 802.11b more than 802.11. While that session dealt with PDAs and not patient-worn monitoring, the silence about 802.11 was deafening. It didn't appear in many slides, some of which showed conceptual diagrams of future Welch Allyn Protocol product concepts. Symbol products currently using 802.11 infrastructure were shown, but any commitment to frequency hopping was generally unspoken. One attendee asked the panel to speculate on the longevity of HomeRF, 802.11, 802.11b and 802.15 networks in the ISM band but the issue of 802.11 was gracefully ducked by all four panelists. HomeRF was identified as dead, Bluetooth (802.15) was acknowledged as emerging and 802.11b was cited as the platform upon which Symbol was developing a number of new products. While there was speculation about 802.11g as an alternative to 802.11a, nothing was said specifically about 802.11 FH. The HIMSS exhibit floor highlighted many 802.11b DS-based products. Nearly all of the point-of-care computerized patient record or electronic patient record vendors were showing Palm or Windows CE-based (also known as Pocket-PC), PDA-enabled applications using 802.11b networking, but aside from the Symbol booth, we couldn't find one that had a new 802.11-enabled product.
How PDAs would be used in health care was a matter of discussion. There was general agreement that physicians would have their own PDAs and simply bring them along as they entered and roamed around the health care enterprise. For nurses, however, the model was not as clear. Some thought PDAs would be hospital property, checked out and turned in by nurses on each shift, while others envisioned nurses also owning them. Others envisioned additional caregivers, such as respiratory therapists, pump technicians, dietitians, monitoring techs and others using them. The use of a device that can access patient records by more than one caregiver raises questions about user validation and audit trails. The theft of such devices raises questions about controlling unauthorized access to patient data. All of this will ultimately be sorted out by the addition of third-party software, such as Trust Digital's (Fairfax, Virginia) PDA-Secure and new biometric ID or other authentification technologies to comply with HIPAA. One key to securing patient data in PDA applications is to use thin-client/server approaches that provide screens, but no data storage on the PDA itself. The next application for such devices for nurses would be as pager replacements to visualize patient data during alarms, and to provide more robust interaction with medical devices and other staff. Given that these devices will use either 802.11b DS or 802.15 Bluetooth, it seems clear that there will be ISM-band infrastructures installed in the hospital of the future. It will probably be tested first in members of the Leapfrog Group, a consortium of 600 hospitals that explores the use of new technologies and applications to improve health care.
Passing reference was made at HIMSS to the proposed ultra-wideband (UWB) communications technology that consumes the entire RF spectrum up to 6 GHz. The sense of the conference, however, was that, given the Department of Defense objections to it, the FCC would kill it. In subsequent action, the FCC left the issue somewhat unresolved, saying it would conduct its own tests to assure interference with global position satellites and military communications were avoided. XtremeSpectrum (Vienna, Virginia) said it was pleased with the FCC actions, yet the exact future role of UWB communications remains to be clarified. More may be known by next year's HIMSS gathering.
Of the two PDA OS choices — Palm's (Santa Clara, California) Palm OS or Windows CE/Pocket PC — by far the greatest number of devices and applications shown were on the Windows CE platform. The Gartner Group has estimated that physician use of handheld PDA devices will reach 20% within two years, but judging from what we observed at HIMSS, all of the physicians that have already adopted them must have been there, because when asked about their use at one session, virtually every hand in the room went up. This makes wireless security an issue of major importance given HIPAA compliance requirements. There were some interesting middleware products to bolster the limited security of the 802.11b DS networks. One Palm-based product we noticed was IQMax's (Charlotte, North Carolina) IQ Speak, a daily work-flow manager that pulls patient demographic profiles from practice management systems for each scheduled patient appointment and displays them on the PDA.
There were also middleware products for PDAs themselves, including PDA-Secure from Secure Trust. That company also has a Secure family of Products that sit between the OS file system and the application, creating an encrypted directory that the application reads and writes its files to and from. PDA-Secure dynamically encrypts all application data. If the application handles patient data, that data would be encrypted. Thus, the patient files in the encrypted subdirectories would be unintelligible to any other application accessing data in this same directory — in essence, the patient data is protected by strong encryption. Couple this with biometric application startup security (to the applications that are able to decrypt and use the data) and you have HIPAA-compliant security. In terms of bringing actual PDA applications to the medical vertical market, MercuryMD (Durham, North Carolina) was showing its Mdata Enterprise System, a system conceived to use handheld devices and existing network infrastructures to provide work flow enhancement. Physicians looking to adopt handheld technologies may find MercuryMD's integration services helpful.
(Next month: More from the HIMSS exhibit floor.)