Today was the last day of the AACN/NTI. I used the last 3 hours to meet with some key vendors. At Draeger I talked networks with Ray Wardell. Their contrarian position is that it is easier and less expensive to deploy one network encompassing administrative applications and real time monitor data rather than separate networks (which I suppose is true -- the question is real life execution). They call their approach Infinity OneNet.
All the other monitoring vendors insist to varying degrees that "life critical" traffic be physically separate from administrative traffic. All the vendors will cave on this separation and support VPNs if the customer insists. Draeger also utilizes a VPN-based shared infrastructure. They use a Packeteer network appliance to provide Quality of Service (QoS). They also work very closely with Cisco. Draeger even lets the hospital choose whatever WLAN card they want , as long as it is WiFi certified. This is an interesting strategy that appeals immediately to the CIO, and could cause IT to prefer a Draeger solution. Given IT's growing influence this could give Draeger a chance to compete where they might otherwise be shut out by clinical buyers.
The life critical/mission critical dichotomy is driven by different requirements. Life critical networks move real time monitor data for surveillance and life threatening alarms. Briefly, five nines (99.999%) up time is required, as is low latency and minimal packet loss. There's no reason why hospital IT can't move life critical data on their network, they've just never done it before and lack a complete understanding of the performance and management requirements for life critical networks. Hospitals that have tried running both monitoring and administrative apps on the same network have experienced significantly lower levels of reliability and outages. (In fairness to these hospitals, vendors do little to help IT transition into this new role.) Typically Biomedical Engineering provides 27x7 support for private life critical networks, backed up by the vendors who designed, installed and validated their networks -- and know what to expect. Most vendors don't have a business delivery system that includes supporting life critical data in a general purpose network environment that can vary substantially from site to site. Private networks also let vendors get away with providing only limited network configurations, like fixed IP addresses and factory configured routers and switches.
Infinity OneNet installations must be validated by Draeger and the customer is required to purchase a support contract to cover the costs of the inevitable glitches. Actual network management can be done by the hospital or outsourced to a third party. According to Ray, most hospitals want to use the "best" WLAN technology, and they can help hospitals select the best fit. They usually recommend 802.11b; performance is adequate and it is the lowest cost WLAN technology available. As device density goes up, a hospital will want to migrate to 802.11g or a for greater capacity. Because the hospital can use any WiFi certified WLAN card, they can swap radio/clients in their monitors when they make a change with their laptops, tables and PDAs.
Like Philips and GE, Draeger sells telemetry on WMTS and multi parameter monitors on WiFi (802.11 a/b/g). Of the three designated WMTS bands, Draeger supports the 608-614MHz band. They use 25KHz channel spacing for their channelized digital radios, providing about a 115 monitor capacity. Additional telemetry monitors can be supported if they are sufficiently distant to allow channel reuse.
I saw another Hospira demo (my first one was at HIMSS). MedNet, their smart pump application and server works with the single channel Plum A+ and PCA 3. The 3 channel Plum (6 in/3 out) will support MedNet later this year.
I also caught up with Matt Likens of GMP Wireless Medicine (the LifeSync folks). GMP has some very critical intellectual property (IP) that directly impacts wireless connectivity. Called the Besson patent, after inventor Marcus Besson, you can read the full patent document here. Numerous vendors are currently selling systems that overlap GMPs IP. These vendors will have to either discontinue selling products that infringe GMP's IP, negotiate a license, try to come up with an approach that does not infringe, or fight GMP in court (in which case they could lose and have to choose one of the first three options). This is important stuff for both customers and vendors because both could be liable if they are found to infringe. Here are three high level criteria to determine if there is possible infringement:
- The wireless connectivity must be digital and two way
- The radio must be on the body (i.e., strapped, in a pocket or pouch, or carried by the patient)
- Error correction of some type must also be used
I'll be digging into this some more with more details in the future.
Finally, I came across a great paper in the GE booth: "Effects of Acuity-Adaptable Rooms on Flow of Patients and Delivery of Care." This was one of the topics for which they were providing online CE credits. The paper was published in the American Journal of Critical Care. I'm going to add this to the bibliography on Resources tab above at the top of the page.