Trapeze Partners with Nortel on Mesh Networking
Now that Cisco's acquired AireSpace, Nortel Networks is partnering with Trapeze Networks to offer mesh networks and asset tracking. AireSpace was Nortel's OEM partner until acquired by Cisco. Nortel has also kicked in some cash for Trapeze's latest VC funding round.
For the near future, Trapeze will focus on mesh networking, Vogt said, making Trapeze access points work in mesh networks as well as letting the company's RingMaster management software detect and manage meshes—including the mesh equipment Nortel already offers. Mesh networks dynamically route packets from node to node. Only one access point needs to be connected directly to the wired network, with the rest sharing a connection over the air, though in large mesh environments several access points may be connected.
You can read about Cisco's mesh network plans here.
By year's end, Trapeze plans to add more client location capabilities into its management software, Vogt said. This is a growing industry trend, serving customers who need to track anything from notebooks to IV pumps. Recently, market leader Cisco introduced the 2700 Series Wireless Location Appliance, which can track thousands of Wi-Fi clients on a corporate network.
I've written on 802.11 asset tracking before. While it may seem intuitively obvious that a system that uses existing 802.11 access points is better than a system using different tag/receiver technology, that's not necessarily so.
Study: Medical Errors Persist Despite CPOE Systems
iHealth Beat reports (registration required) on a new study about the impact of CPOE on patient safety.
Researchers looked at 937 admissions and found 483 clinically significant adverse drug events, Reuters Health reports. Nine percent of those errors caused serious harm, and 33% resulted in further monitoring or interventions. Sixty-one percent of errors took place during ordering, and 25% occurred while monitoring (Reuters Health, 5/23).
Remember, it's just a tool…
UPDATE: …and the tool may need improvements. According to Fierce Healthcare:
The conventional wisdom in health IT is that e-prescribing cuts down on preventable medical errors. A study done at the Salt Lake City VA Hospital appears to cast some doubt on that conclusion.
The study authors conclude that the CPOE application used lacks certain capabilities that would reduce the number of errors that are occurring. The abstract from the Archives of Internal Medicine is here (full text $12).
Report from AAMI 2005
Thanks to Patrick Lynch, from Northside Hospital, Atlanda, for this report from the show:
It was the best AAMI I've been to in a while. Being an older biomed, all of the equipment vendors were getting stale – same ole, same ole. But this year there was lots of new tension. IT, indoor positioning, who are we gonna work for next year, consultants galore, and a huge presence by all the big employers – GE, Philips, Siemens, Aramark. GE was even giving away a t-shirt that said “Biomeds Rule.” One company even had a purple Corvette to push their purple blood pressure cuffs. The giveaways seemed better this year. My personal favorite was a calculator (from MedEquip) which, when the button was pressed, opened up in a lengthy dual-retracting process I can only liken to a Transformer. The most puzzling giveaway was from Aramark – it was a nice metal tin, with a level inside. Exactly what I need a level for, I;m a little lost, but it was a nice gesture. Zoll was doing CPR on a stuffed bear.
Overall, I think there is a lot of uncertainty inthe field,and lots og BMETs wilbe going to computer and networking classes in the next year, if only as a defensive measure.
Great update Patrick, thanks. Anyone else out there with a report from the field? How about you vendors?
Biomed-IT Convergence: Support Contracts
There was some discussion on the Biomed listserv today about support contract pricing that vendors are charging. Comparisons were made between the percentage of the list price for annual support contracts for hardware (8% to 12%) and software (20% up to 35%). This is another area where IT and biomed are merging.
Buyer expectations for medical devices are warranty (hopefully an extended one) and the ability to repair broken devices themselves — taking it so far as to consider second source parts availability. Installation and training is frequently negotiated “away” as part of the purchase price, i.e., bundled rather than priced separately.
IT regularly pays the percentages noted above, and they pay gladly. For that they get updated software releases and software service and support (bug fixes, cleaned up or fixed files, reconfigured systems, etc.). Interestingly, there is a “second source” market for health care IT, usually vendor refugees who provide customization, systems integration and other services on their previous employer's software for less than the OEM. Software installation and implementation are rarely negotiated “away” and priced separately.
Sadly (from the vendor perspective), no one's getting rich with either scheme. A typical medical device will have a factory margin around 70%, with gross margins (the factory margin minus the cost of marketing, sales, installation, support and company overhead) running in the thirties to fifties. Software is a very high margin product. If the vendor capitalizes their development (and most of them do), they have a “factory” margin of at least 98%. However, few vendors have the luxury of selling just the software. Hospitals want a “whole product solution” that includes everything (hardware, software, integration, installation, training, support) for a turnkey system. Once a vendor's done all that, they can be getting somewhere in the teens or low twenties at the bottom line.
The most uncomfortable position is the medical device vendor with a software product. Medical device software products include smart pump servers and telemetry central stations. Sold as a “medical device” the vendor's service contracting adoption rate is much lower than for software products. The service contract itself is priced like a medical device, usually not much more than 12% of the purchase price. It costs a vendor $500,000 to turn a software release with bug fixes and a few minor enhancements; annual R&D for an existing product is typically $1 million or $2 million. A major product release, either a new product or a rewrite of an existing one, runs $4 million to $6 million. The box business executive looks at a medical device that cost $5 million to develop and virtually nothing from a sustaining/support perspective vs. a software product that sucks up a substantial amount of his R&D budget every year — with no direct incremental sales increase — and wonders, “what's wrong with this picture.
One of the biggest differences between an embedded device product like a patient monitor and application software is that the software is never finished. There is a competitive “feature race” that is run in most software markets, where new features and performance/reliability enhancements are released once or twice each year. This continues until the underlying product architecture becomes outdated and too expensive to update; then it's time to rewrite the whole thing.
UPDATE: John Walsh reminds me, via the Biomed Listserv, that there are device vendors who oversell the free upgrade benefit of their software support contracts. Many devices include applications of such narrow scope and sufficient maturity that there are no reliability problems nor any real features to add to the product.
