Home Use of Medical Devices Challenges FDA
Medical Devices Today has a good post on potential safety and effectiveness issues surrounding medical devices where home use by patients is outside the scope if intended use. “Although manufacturers need FDA approval to market a device
over-the-counter directly to a patient or specifically for home use,
there are few restrictions on whether a physician can send a patient
home with a device that is not specifically labeled for use in the home.”
One of the biggest issues revolves around older devices that have been replaced and become hand-me-downs for home health. Concerns include:
- Neither devices or directions for use not designed for patient use
- Hand-me-down devices frequently missing directions for use
- Patient purchased devices, e.g., after the reimbursed rental period, end up getting “recycled” through eBay – with no safeguards for proper maintenance, operation or directions for use
- Proper distribution channel safeguards to ensure safety and effectiveness when sold retail or through other resellers
Besides contemplating new regulations or legislation, the FDA is making the following responses:
Meanwhile, in the next six months FDA hopes to invite manufacturers
to participate in its new online labeling repository for home-use
devices. The voluntary repository, which will initially focus on
infusion pumps, will give consumers access to the most up-to-date
instructions for specific models.
The agency is also preparing to launch a sub-network of its MedSun
adverse event reporting program called HomeNet. Participating home
health agencies will be encouraged to report to FDA adverse events,
including close-calls, related to devices used in the home setting.
Be sure to read the whole thing. Pictured right is a Baxter Syndeo pump – perhaps something that might be used in the home?
Thoughts on RFID in Hospitals
Yesterday I noted that the Biomed Listserv has moved to new digs. In addition to a new hosting service, there is a message limit of 450 words – sadly, only enough for one or two points for me. I learned about this limit when my message bounced, so I’ve posted it here.
James asked about RFID in hospitals: who’s got it, what they’re doing with it, recommended vendors, and problems people have had. The following are my two cents.
Besides infant abduction systems (which tend to rely on monitored choke points rather than positioning throughout the unit), there are only about 200 hospitals that have deployed active RFID in the US. Of that 200, most are pilot deployments – limited to specific areas or departments. There’s currently about 10 to 20 house-wide RFID installations. This is not the rush to adoption that many (especially some of the vendors) expected.
There’s been a bit of a shake up among RFID vendors over the past year, and some very interesting new vendors have come to market. One of these vendors, RadarFind, introduced their system at the last AAMI conference in Boston. (You can read a bit about them here.) All in all, the vendor situation is unsettled.
Things to keep in mind:
- Different applications require different levels of performance, especially positioning accuracy. Positioning accuracy varies considerably among the different RFID technologies on the market. A thorough needs assessment to identify positioning applications and requirements for each application is the first step to selecting the right vendor.
- The details around knowing where someone or something is at a given point of time vary with different applications. If you’re looking for that Newport vent that just got recalled, knowing the general vicinity is fine. If you’re trying to improve utilization in your operating rooms, location data is just the beginning – you need an application that uses location data to optimize a much bigger set of processes. Thus, the workflow to which your positioning data would be applied is also an important requirement to gather.
My view of current RFID offerings is that the technology has not matured to the point where you can treat RFID as a generic multipurpose piece of infrastructure that can be leveraged by most any application that uses positioning data.
Let me know if you’d like help with an analysis of vendors, needs assessment, or planning/vendor selection.
Pictured right is the back of an Awarepoint positioning receiver – they plug into 110 volt wall outlets.
Sensor Productization Challenges and Potential Solutions
Simon Aliwell, Director of the Sensors and Instrumentation Knowledge Transfer Network, National Physical Laboratory, in the UK, has a piece in the latest issue of MDT magazine. His outfit is a network of excellence supported by the UK Technology Strategy Board to develop innovation in sensing. Check out their web site here.
In his story Aliwell suggests that regulatory hurdles represent the greatest market barrier for developers of new sensor technology.
well defined testing and approval process before they can be adopted.
This is the case even when technologies have proven their effectiveness
in other fields and applications. Although this is to patients'
benefit, it does delay the return on investment for companies and makes
the path to commercial success longer. This, in combination with the
significant upfront investment required to develop, manufacture and
trial new devices, may mean failure for many small companies as they
wait months or even years for returns.
Certainly starting a new medical device company represents a major investment. And certain sensor technologies, like the Oncoprobe chemotherapy sensor that Aliwell mentions must be tested for each clinical application (i.e., or type of cancer in the case of Oncoprobe) – a heavy regulatory burden. A bigger burden I think is creating the “whole product solution” for senors.
In a perfect world, the value of a novel sensor technology is recognized and eagerly licensed and incorporated into medical device vendor's products. In the imperfect world of health care, almost all sensor vendors must create and sell their own product to create the market demand that drives existing vendors to incorporate new sensor technology. There are few Fields of Dreams (build it and they will come…) in health care.
This line between OEM supplier and medical device vendor is a thin one, and must be approached carefully. One one side, your sensor-based product must be compelling enough to win end-user sales and fuel market demand. On the other you need a rapid time to market at a cost you can underwrite. So what might a “whole product solution” include?
In addition to productizing your novel sensor technology, a company has to package it so it can be used in its intended clinical setting. The relative maturity of other medical devices at the point of care – patient monitors, infusion pumps – create expectations among clinicians for certain capabilities. These expectations include:
- Some degree of wireless operation – an increasing number of devices are battery powered and wireless connectivity is de riguer for portable medical devices, battery powered or otherwise.
- EMR integration – any hospital with paperless charting will insist on the ability to electronically move sensor data into the patient's electronic record.
- Monitoring – sensors provide a monitoring function – alone, in conjunction with other sensors, or as feedback to therapy. Monitoring requires surveillance (data display), alarms, and retrospective event review.
- Point of care workflow – all these capabilities must be implemented in a system design that supports the care delivery process – where caregivers circulate between the bedside, central station and elsewhere. The system must enhance caregiver productivity and patient safety, rather than the opposite.
These expectations define the system that supports the practical use of your sensor in a real world customer environment. Along with these system features, there is a whole range of business requirements to properly sell, manufacture, install, service and support the whole product solution.
Conventional product development strategies for these system features entail developing most all of the features from scratch, a very time consuming and expensive approach. While we are still a ways from being able to license these features from a medical systems OEM vendor, there are innovative new development strategies that can greatly minimize time to market and R&D costs.
Rome wasn't built in a day, and a new medical device system will not be full featured at first release. In fact, you'll need some system capabilities to support sensor development, like data acquisition and analysis. Part in parcel with your development strategy is the product roadmap. The roadmap lays out the progression of system features over time, and must align with your sensor product development plan and your market entry strategy.
Most novel sensor technology requires that clinicians adopt a new use model different from that used with conventional sensors. Getting anyone in health care to change the way they do things is a challenge, and this has an impact on market entry strategies. Because market entry strategies drives both system requirements and product roadmap, the whole process is rather iterative, with each variable impacting the others until a balance is achieved.
So there you have it, an overview of productization issues for novel sensor in health care.
Pictured right is some Nexense sensor silicon.
Biomed Listserv Is Moving
Mike Kauffman’s labor of love, the Biomed Listserv has moved. Once hosted by AOL, Mike’s snapped up an Internet domain (www.bmetsonline.org) and is adding some new features to extend the email listserv.
If you’re already subscribed to the Biomed Listserv, Mike has already moved all the subscribers to the new system (thanks Mike!). If you want to subscribe, click here. Be sure to check out the FAQ first. Starting January, 2008 Mike will be charging list subscribers a modest annual fee – well worth the investment.
UPDATE: The Biomed Listserv is now called Biomedtalk, and is sponsored and administrated by the ECRI Institute. Here’s a description of the transition from Mike Kauffman to the ECRI Institute from a press release:
ounded in 1997 by Mike Kauffman, assistant director of facilities for the Reading Hospital and Medical Center in Reading, PA, Biomedtalk became an active forum for discussion across all members of the biomedical and clinical engineering profession.
The listserv now has over 1,700 members, having attracted over 100 new members since March 1, 2010, when ECRI Institute took the reins and discontinued the membership fee.
See FAQ number two on this web page on ECRI’s site to learn how to subscribe to the listserv. There’s a searchable archive too, but you have to register (here’s how).
AT&T Jumps on RTLS Bandwagon
These days phone companies like AT&T are more like network services providers than old time POTS (plain old telephone service) monopolies. Most, if not all of their land-line voice traffic runs over IP networks and the wireless unit is moving to full IP networking for wireless voice when they move to LTE in a few years.
That said, the whole “public utility” thing is becoming more a liability than the license to print money that it once was. In response to this transition, AT&T has been beefing up their networking and IT hosting services. Now they’ve taken the plunge and announced an RFID managed service designed for hospitals. From this RFID Journal story we learn that AT&T will design, install and manage your WiFi network. For an extra fee, they’ll throw in some AeroScout RFID tags and provide Asset Visibility. It is not clear whether they’re using the Cisco or AeroScout positioning engine.
I can certainly see hospitals using AT&T’s wide area network services, and maybe even NOC (network operations center) services. As a relative newbie to hospital wireless network, I’m skeptical. Ditto for any near term success selling RFID. If you’ve had any experience with AT&T networking services, let me know.
[Hat tip: StatCom newsletter]
