Site Update
Yes, I've been particularly busy lately and just now getting caught up. Thanks for being patient and checking back.
And thanks to those of you who took the time to respond to my query about problems viewing the site. I made an adjustment to the template that will hopefully clear up the “overlapping columns” problem that some of you have been having. If the problem's not fixed, let me know and I'll try something else.
Biomedical Engineering's Role in Hospitals Shifting
The other day I received a GE “SmartMail” announcing a new training program for biomedical and clinical engineers. The training focuses on PC hardware (components, ports, booting and setup), operating system basics, and lots on networking. The cost is $3,500 for the 5 day class. In reviewing the course outline, this looks like a great course. One thing that's not clear from the syllabus is which operating systems are covered – the descriptions sound Windows oriented, and there many Linux and Unix based medical devices out in the world. Perhaps someone from GE could enlighten us.
The mission, training and organizational structure of biomedical engineering was conceived when analog circuits made up most medical devices, and biomeds could get in and make component level repairs. Today, medical devices are mostly computers with signal processing front ends, and the “smarts” are found in the software rather than analog circuits. Many biomeds are sharp electrical engineers but lack an equally deep understanding of information technology – not that you'd want to hack medical device software.
The days of making component level repairs are fading. Advancing manufacturing technologies have not made it any easier for biomeds, as more products become single board devices with little or no opportunity to service the devices themselves. Besides testing equipment for safety and doing periodic maintenance, biomeds have been reduced to swapping broken devices with manufactures.
Biomeds are an undervalued and underutilized resource in hospitals today. The real value in biomeds lies not in their electronics service expertise, but in their understanding of what it takes to ensure patient safety when using medical devices for therapy delivery or surveillance – this is something that IT can't even begin to imagine. Biomeds pass JCAHO muster every time the hospital is reviewed, they deal with FDA regulatory issues, and most importantly, they understand clinicians and the care delivery mission. Biomed's patient safety expertise needs to be extended into IT and fully absorbed – and quickly.
Health care IT folks are pretty sharp too, but as they encroach into the clinical domain with CPOE, EMRs and other applications, it is clear they have lots to learn from biomeds about patient safety. The “scream test” does not work when clinical devices are involved. (The scream test is when one disconnects something, say a network patch cable, and waits to see if anyone complains – this is a not uncommon IT maintenace procedure.)
The days of the biomed as medical device repair man are fading, and working for the facilities VP no longer fits. One of the trends in hospitals is to put biomedical engineering “under” IT. I think that's a mistake. While biomeds don't command the budgets that IT does, their mission is every bit as important as IT; in fact, you can run a hospital without IT, but you can't run one (at least for long) without biomeds. In recognition of their true value, perhaps biomeds should report into Risk Management or some other patient safety department in the hospital.
The real question is if biomed and IT were dogs and cats, which would be the dog? I know that both biomeds and HIT types read this blog, what do you think?
UPDATE: I got a great email with some additional information about the GE training course mentioned above:
developer for Essentials of Healthcare IT, or the training program you mentioned [in this post]. The course is built around Windows
XP. At this time we do not discuss Citrix or Linux, since the focus of the
course is building and troubleshooting 4 types of networks. Discussions of
multiple flavors of OS would push the class beyond one week.
mentioning the class, I happen to think it IS great. As a 19 year veteran of
field service in operating rooms across the US, I realized our industry was
falling behind, very behind, in the networking game.
Thanks.
New Wireless Implant Delivers Drugs on Schedule
Startup MicroCHIP has developed technology to dispense drugs over extended periods using microchips that are implanted in the body. Unlike other passive internal drug-delivery methods already on the market, MicroCHIP can control time of delivery and dose externally. The MicroCHIP active drug delivery system uses a wireless device to control the release of
multiple drugs from a single implant in the abdomen, while also
monitoring drug levels and adjusting dosing accordingly.
The secret sauce here is their proprietary micro arrays that can release any combination of drugs or biosensors stored within the reservoirs. Preprogrammed microprocessors, wireless telemetry, or sensor feedback loops can provide active control of the opening of the reservoirs to initiate drug release or expose enclosed biosensors.
While not listed as an investor, Medtronic does have a seat on the board of MicroCHIP.
Mayo Clinic Competes with Medical Device Vendors
Many of today's large university teaching hospitals have established track records of new technology innovation. The Mayo Clinic has traditionally licensed new technologies to other companies for them to productize. With all the outsourcing that's available today, the Mayo decided to do it themselves this time (press release). That's the product, pictured at right.
and hand and decided they could produce a better-quality device — and
do it faster — if they didn't pass the idea off to somebody else. So
Mayo hired IBM (IBM
) to help design and manufacture the product, and relied on medical equipment giants GE (GE
), Siemens (SI
), and Philips to market and sell it.
It looks like Mayo made the right decision. Within eight months of its
initial conversations with IBM, the tech giant had delivered the new
devices, Mayo Clinic MRI coils, for use in Mayo's own medical
facilities. A second version, begun in January, 2004, was readied for
sale by GE four months later. A more typical time-to-market span in the
medical-device field is 16 months to 24 months, according to “How Fast
is Fast?” a product-development benchmarking study by IDEO, one of
America's leading contract design firms.
That's 8 months to the first manufactured prototype, and total time to market of less than a year. Wow. A typical time to market for a medical device is more like 2 to 3 years (or longer if you screw up), especially if it is truly new and not just “reskinned” or upgraded.
This radical new business model has been around for some time. There's books about it, and several published profiles on companies who've done it – here and here. It all makes the head spin, doesn't it? Another great book on this topic can be found on the Connectologist's Reading List, called Value Migration, by Adrian Slywotzky.
With this new business model the Mayo retained greater control and ownership of their
intellectual property, they got a much shorter time to market (with a
quality product), and they stand to gain more financially – they've
sold 100 so far. I wonder what being reduced to a distribution channel
will do to GE, Siemens and Philips' business models. Will the big
medical device vendors offer contract manufacturing to go along with
their outsourced distribution? Design services? In an industry with a serious case of “not invented here” syntrome, I'm sure traditional
medical device executives would be outraged – but it may not be such a
bad idea.
Of course, if it was easy then everyone would be doing it. Competitive barriers are the biggest threat to innovation in health care. Outsiders might think FDA regulations are the big barrier – I don't think so. The biggest barrier comes from group purchasing organizations (GPOs). But the Mayo clinic sidesteps that nicely by using GE, Siemens and Philips (who all have GPO contracts for a variety of things) as their distribution channel.
What other examples of innovation and new business models are changing the health care market? Let me know!
Brigham and Women's Combines Wireless Monitoring and Patient Location in ED
Under a trial funded by a $3.1 million grant from the NIH, Brigham and Women's Hospital is using 10 waist packs (that sounds so much better than “fanny packs” doesn't it?) that patients will wear containing sensors, transmitters, and
tracking gear. The packs will allow medical staff to constantly monitor
patients' heart rates and blood-oxygen levels while they await
treatment. From the description, there is no arrhythmia analysis provided, a major shortcoming. The project is called SMART for Scalable Medical Alert and Response Technology.
The technology deployed includes a 3 lead ECG and a finger sensor connected to a PDA that transmits the signals to a server that provides surveillance. The real time positioning system (RTLS) uses ultrasound from Sonitor Technologies, rather then traditional RFID or infrared technology. There was no description on the fanny pack used.
This story has gotten pretty good press, but is far from ground breaking. From what's been reported, you could even say that Brigham and Women's has recreated commercial products with their $3.1 million grant. Wireless patient worn monitors are currently available off the shelf; the most advanced are the Micropaq (3 or 5 lead ECG, heart rate, SpO2) and the Propaq LT (non-invasive blood pressure, respiration, 3 or 5 lead ECG, heart rate, SpO2) from Welch Allyn. Both of these monitors are ruggedized and water resistant, and I'll bet they will fit Brigham and Women's fanny packs.
Wireless patient worn monitors are great, which is why vendors make them. What's really needed is nurse-carried alarm notification. Putting patient waveforms on a central station that nurses may glance at as they rush by has limited value. Greater value is gained by sitting a monitor tech in front of the display (they use paramedics at Brigham and Women's), but that comes at a cost – and more than a little uneasiness from the caregivers who depend on them for alarms. A really ground breaking project would integrate alarm notification with patient location; if you included support across vendors and different devices you would reach nirvana.
The RTLS portion of this system highlights that fact that each of the different RFID/RTLS technologies has inherent strengths and weaknesses – there is no one “best” positioning technology for all applications. Brigham and Women's chose ultrasound for one reason: walls. When a patient alarms, caregivers need to know where to find ambulatory patients. (Of course, location should be communicated via a nurse-carried device as well.) Because ultrasound does not penetrate walls, you always know what room or hallway – or which side of a wall – the patient is in. This room-accuracy comes at a price; ultrasound receivers must be placed in each room and hallway where patients are to be tracked. The leading RTLS vendor in the room-accurate market segment is VERSUS – they are the vendor of choice for nurse call system vendors who offer an RTLS option. It would be interesting to know if VERSUS was considered by Brigham and Women's and why they went with Sonitor. Pictured right is the Sonitor ultrasonic tag.
It seems that the hype about 802.11-based RFID is subsiding, and folks like Brigham and Women's are selecting technology based on a thorough needs assessment.
I'd love to hear from anyone who can comment on this deployment – I'm sure there's much that never made it into the stories.
