A few weeks ago I reported that GE was going to be releasing a new Dash patient monitor. I've heard that the monitor sports an embedded 802.11b radio. Kudos to GE for going with 802.11 rather than being tempted to use WMTS. The most ubiquitous flavor of Wi-Fi in hospitals is 802.11b; but you
don't have to be Alvin Toffler to see a day when 802.11b (and 2.4 GHz
for that matter) could be too congested for continuous monitoring of
critically ill patients. Some hospitals are installing WLANs that support 802.11 b/g and a. Rapidly changing IT requirements (at least compared to medical devices) can result in unantcipated product development projects for vendors, and unbudgeted upgrades for providers.

Using a widely adopted industry standard like 802.11 is appealing
to medical device vendors. Because of its broad adoption, Wi-Fi
technology costs less than a proprietary solution. The availability of
chips, sub assemblies and finished modules reduces development cost and
time to market. Hospitals like it too; almost 80% of have adopted
wireless networks. The rub is that by using general purpose wireless
LAN technology in medical devices, they must coexist within a
hospital's broader IT infrastructure. GE is going to be feeling this
rub with their new Dash monitor.

One of the advantages of running medical device systems (like patient monitors/central stations or smart pumps/servers) isolated on private networks is that vendors and hospitals can put them in time warp. This time warp creates a more tightly controlled environment allowing vendors to greatly reduce variables that could complicate post-sale support and minimize IT features that have to be developed. For hospitals this time warp extends the medical device's useful life beyond IT products to 7-12 years or more, and avoids costly upgrades to keep up with information technology. The problem with time warps is you get a proliferation of point solutions, each with their own networks, their own set of interfaces to the outside world, and each imposes their own quirks on your IT infrastructure. Another factor is the increasingly wide deployment of wireless medical devices - infustion pumps are used house-wide, patient monitors are increasingly being deployed outside critical care units - it's hard to time warp a deployment that covers most of your hospital. What starts out as an "out of sight, out of mind" convenience becomes an operating headache once these systems reach a certain point.

As medical devices become more integrated with information systems two different scenarios could occur. Tightly integrating devices with their associated systems reduces their useful life as they're more rapidly replaced to keep up with changes in information technology. This scenario would keep device/system integration proprietary and within the domain of the device vendor. Vendors would like to see the buy cycle (the average time between purchases) reduced, but the increased R&D burden will reduce their gross margins. Systems do represent a potential source of additional revenue for vendors, but also a considerable expense and business risk.

The other scenario would entail medical device vendors adopting the standardization efforts of IEEE 1073, the IHE PCD and others, to isolate medical devices from most IT changes through the use of a standard. This is the path that's been taken by the diagnostic imaging market with DICOM. The use of DICOM provides both diagnostic modalities and information systems the isolation to develop independently, which is good for both vendors and buyers.

UPDATE: I got a download on the new wireless Dash at HIMSS 06 - you can read the new post here.