It's time for another fisking! This time, the website medical devicelink has an interesting story called IT Showcase from the Business Planning & Technology Development section of MX magazine. The story explores the integration of medical devices with information system technology (aka, connectivity). While this article touches on areas in and outside the hospital, my comments will be limited to the hospital environment. Let's get started.
Even among healthcare professionals who are commonly the earliest adopters of new technologies, black boxes filled with zeroes and ones don't have a lot of appeal. So it's easy to understand how some users might be tempted to disregard some of the most complexâ€”and even revolutionaryâ€”enhancements currently being made in the world of medical
Market adoption is a funny thing in hospitals (frustrating is frequently a better term). And yes, the times they are a changin'.
But little by little, device manufacturers are beginning to incorporate high-end information technologies (IT) into their products. And in doing so, they are gradually creating a world of next-generation healthcare products with features and functionality that users care about very much. Whether the need is for easier transmission and archiving of diagnostic imaging, quicker delivery of patient information to healthcare professionals on the move, or advanced monitoring of patients in alternate-site healthcare settings, device manufacturers are framing new solutions using advanced IT systems.
Of course connectivity is not as new as it sound here, vendors have been integrating devices with computers to add value and differentiate for more than 20 years. As medical device markets mature, products become less differentiated and end user requirements change. Requirements shift from the creation or quality of data to the management of data. I've written on these changes before, here and here.
Common connectivity applications include automated data analysis, report generation, data storage retrieval and query, workflow automation, surveillance, alarm notification/management and event logs.
While medtech manufacturers have seen device interfaces, communication standards, and networking topologies come and go over the years, most recognize that medical technology is now firmly on the path to convergence with information and communications technology. In response, they are increasingly designing and manufacturing their products to be fully compliant with industry standards.
The pressure to integrate IT is independent of whether or not standards exist to make that integration easy. It seems to me, that there has been little change in device interfaces, communications standards or networking topologies in device connectivity. Serial ports are ubiquitous on most medical devices, and have been for years. Serial interfaces require device drivers, an area with no standardization whatsoever -- serial drivers and even cables vary across the same vendor's product lines and even across different versions of the same product.
After some confusion in the 1980s (when ArcNet, IBM's Token Ring and Ethernet all gained early adoption), Ethernet became the de facto networking standard. Network cabling did evolve, from coax cable to twisted pair and optical cable (used for long runs orhigh speed backbone). The latest wrinkle, introduced in the 1990s is the WLAN based on the IEEE 802.11 standard.
While this technology migration has been relatively stable and predictable it has added a significant cost to vendor's sustaining engineering costs, as they have worked to incorporate current technology into new product releases. Because medical devic e life cycles are longer than computing components, vendors can be faced with doing an unplanned product release to keep up with technology. Minimizing this sustaining engineering cost and the cost to support different generations of technology are major vendor connectivity challenges.
A common engineering and IT joke is that the great thing about standards is that there are so many of them. Not only are there numerous established standards, but there are also de facto standards and what I call "wannabe" or pseudo standards. The only broadly adopted health care standards in this area are HL7 and DICOM. Other technology standards in broad use, like Ethernet and 802.11 don't come close to "plug and play.". An up and coming technology replacing RS232 is the Universal Serial Bus, or USB. These technology standards form a core set of de facto standards for medical device connectivity.
The Medical Information Bus, or MIB, is an example of a failed standard. Started in the late 1980s, progress towards plug and play connectivity (meaning that everyone's serial device driver operated the same way), never got adoption. Now referred to as the IEEE 1073, I believe that this standard will merge with HL7, resulting in a higher level XML based standard that rides on a variety of connections.
Pseudo or wannabe standards include things like WMTS, or Wireless Medical Telemetry Service. WMTS is simply a radio frequency spectrum allocation to replace the previous VHF spectrum that the FCC has reallocated to other uses. Vendors using WMTS are positioning it as a "protected" standard. Unlike 802.11 which specifies both frequency and protocols, WMTS does not specify any protocols. Consequently, WMTS implementations are proprietary because a vendor's WMTS infrastructure only works with their devices, creating a closed system. Traditionally the only devices using WMTS were telemetry monitors. Datascope has extended WMTS in their new Panorama system to include multi parameter patient monitors. Other WMTS users (GE, Philips, Spacelabs) limit WMTS to telemetry. I would not be surprised if some vendors attempt to leverage WMTS to create proprietary networks that connect medical devices beyond telemetry. You can read more about WMTS and ISM here and here.
Of course buyers want a very limited number of broadly adopted standards. Vendors, if they're smart, want limited standards too so they can concentrate on their core competencies (building a better mouse trap). Frequently though, vendors end up spending considerable dollars (both theirs and their customer's) on proprietary technologies in an effort to erect high switching costs to preserve market share.
The field of imaging has long been in the vanguard of information technology applications for medical equipment...While other sectors of the device industry are at square one in figuring out how to develop technical standards that will enable them to take advantage of IT-related opportunities, imaging companies already have more than a decade of experience at implementing such standards.
Taking advantage of this experience, a number of imaging companies have begun to turn the corner into the field of health information management, building new business units that may expand their presence in a variety of clinical environments. Companies in other sectors will have to move quickly to catch up with the imaging sector, but they can also learn a lot from the experience that imaging companies have already gained.
They are correct that imaging is a very mature connectivity market, perhaps the most mature. Once medical device markets go through the connectivity phase, the value proposition for that market shifts. This shift causes companies to adjust their strategies to maintain growth and recapture value in order to maintain margins. It's been interesting to watch imaging vendors as they struggle to maintain growth and value -- there has been a wide variety of strategies and results.
As hospitals prepare to implement end-to-end, enterprise wide information systems, integration and interoperability are the guiding watchwords. The electronic health record (EHR) will become the key focal point of clinical information systems. All electronic medical devices and equipment will be required to share and exchange information with EHRs, which are expected to become ubiquitous over the next 10 years as part of a federal government initiative to streamline the delivery of healthcare services.
The biggest driver to medical device connectivity at the point of care (POC) is the electronic medical record (more on this here). Hospitals face several challenges getting medical device data into clinical information systems. Most currently installed medical devices are several years old and lack connectivity capabilities beyond an RS232 serial port. Networked medical devices that are out there are "islands of information" running on private networks, many using outdated network infrastructure. On the vendor side, no one vendor makes all the devices that will eventually need to be connected (patient monitors, IV pumps and ventilators) so a single vendor solution is unavailable.
The traditional connectivity model for therapeutic and surveillance devices (pumps, vents, monitors, central stations) is based on the ICU. The ICU environment is characterized by gorked out patients, bolted down monitors and other devices that were wheeled up and remain with the patient much of their stay. As a consequence, the patient monitor became the data hub, using data concentrator products like GE's DIDCA. Monitors are networked, but other medical devices plug into the monitor via RS232 serial cables. EMRs got their start as ICU charting systems, and this hardwired, monitor-centric approach worked fine. Companies like Capsule Technolgie and HEI sprang up to handle all those messy proprietary serial interfaces. This approach also relieved most devices of the complexity of establishing and maintaining patient context; "dumb" devices do not know the identity of the patient they're connected to, they rely on the monitor, data hub or central station to know the patient (device x is plugged into port y, which is in room z, occupied by patient Jones).
As EMRs start to penetrate general patient care areas there are new environmental considerations. Multi parameter patient monitors are replaced by spot check vital signs monitors, and other medical devices (especially IV pumps) are used with no continuous patient monitor to act as a hub. There are no gorked out patients, they get up to go to the bathroom or take a smoke break and are encouraged to walk around as soon as they can. Now it becomes much more cost effective to wirelessly connect medical devices instead of installing data concentrators and device specific serial cables in each patient room. Because few patients outside the ICU are monitored, medical devices must connect on their own. And because they're wireless there must be a way to reliably associate data from the device with a specific patient and maintain that association as patients move throughout the hospital, going in and out of coverage.
Patient acuity is rising in care areas outside the ICU, resulting in more aggressive therapeutic interventions and a greater need for surveillance. Variable acuity units are being adopted by more hospitals to relieve patient flow bottlenecks (and avoid considerable cost) in ICU and step down units. These trends take connectivity requirements even further from the ICU model. In response, vendors are stepping up with more sophisticated connectivity solutions. Some, like Sensitron's or Care Fusion, offer systems targeting legacy devices with serial ports. The new Welch Allyn Propaq LT patient monitor (pictured at top) has state of the art connectivity built in; patient context is established at the monitor and connects wirelessly to a server with an HL7 interface that connects to the EMR. Stinger Medical sells a spot vital signs monitor that has all the connectivity built in as well.
By its nature, healthcare is a mobile profession. Whether making rounds at a hospital or going from patient to patient in a busy clinic or private practice, physicians cannot be tethered to fixed computer workstations to remain connected. Handheld devices, typically built around a personal digital assistant (PDA) or Tablet PC, are increasingly finding their way into a wide range of point-of-care medical devices. Typically running either the Palm or Microsoft Pocket PC operating system, these off-the-shelf devices can be readily adapted to both medical diagnostic and reference applications. Clinical data can be readily uploaded and downloaded from a PC using industry standard wired or wireless interfaces.
The mobile nature of health care is right on. In the hospital at least, focusing exclusively on the physician is a mistake. Caregivers have a much more direct impact on patient safety and the ability to deliver the most appropriate care in the lowest cost setting. Regardless of which user you're focused on, the best market research I've seen on mobile computing in hospitals are the reports from Spyglass Consulting (who has reports on both physician and nurse mobile computing). Microsoft has won the mobile computing operating system battle in the hospital; Pocket PC and Windows CE are used extensively in both embedded medical devices and client software for PDAs. Palm has an installed base of physicians based on reference apps like Epocrates, but wireless interactive applications are being written for Microsoft rather than Palm.
Device selection must be based on use cases. Mobile computers at the point of care must be rugged, operate for a shift without recharging, and able to be wiped with disinfectant. Frequently barcode readers are required as well. The PDA leader in this space is Symbol, with Hand-Held Products coming on strong.
Mobile apps fall into two categories, data intensive activities like order entry or charting, and messaging apps delivering alarms, alerts, worklists and other time sensitive data. Data intensive apps require lots of screen real estate while alarms, messaging requires a device that's small, light weight and easy to carry. There is no one device that meets both sets of requirements without seriously compromising one or the other. Messaging devices must be multi purpose because caregivers will only carry one, not one per vendor. Accessing multiple applications on a tablet, laptop or computer on wheels (COW) is similar to setting up a desk top client for the same purpose. I think the long term winner for messaging apps will be a VoIP phone with a graphics display (this Cisco phone is a good start).
Accessing multiple point of care messaging applications is much more complex, the device is less "general purpose" than a desk top PC and applications can be much more demanding. Messaging applications include alarm notification (life-critical FDA regulated), data capture, patient context management, results reporting alerts, and caregiver worklist management. The vendors involved include traditional HIT vendors, smart pump vendors, those making patient monitors (both spot vitals and continuous), and enabling vendors like Sensitron and Care Fusion. What a mess.
The rising costs of hospital-based healthcare have brought about a move to provide medical services at alternate sites, including the patientâ€™s home. Reflecting this trend, homecare is now one of the fastest growing sectors of the medical industry.
As patients continue to be discharged only partially recovered, the need for home care will continue to increase. Better management of the chronically ill, especially those with comorbidities, will also improve. Growth in this area has never been held back by an absence of technology, but rather a general lack of reimbursement.
As manufacturers increasingly commit to integrating network connectivity and enterprise wide interoperability into their electronic medical devices and equipment to meet the competitive demands of the healthcare marketplace, what forces will shape the next generation of medtech equipment? If the electronic health record is to fulfill its promise of capturing clinical data at every point of contact, won't every medical device have to become an electronic device?
Just to pick a nit, "every medical device" is in fact "an electronic device." And yes, they will all end up connected. Smart IV pumps are just the latest in this wave of true connectivity integration -- and most of the IV vendors have just started with the easy stuff like formulary management and the 5 rights. True interoperability (remote control of the device by another system) and life critical features like alarm notification represent a huge step beyond current capabilities. The use of wireless continuous patient monitors outside the ICU will continue to be adopted. Future spot vital sign monitors will also incorporate wireless connectivity and patient context -- Datascope's Duo is probably the last spot vital signs monitor to be introduced without wireless connectivity.
Vendors face important strategic decisions about connectivity that will directly impact their long term success. (More on vendor challenges here.) The installed base of current connectivity products is small, but as adoption increases hospitals will struggle with the integration of multiple point solutions and the inevitable migration to an enterprise approach to connectivity.
Many common medical office devices like temperature probes, stethoscopes, scales, and blood pressure measuring devices already feature digital circuitry, but are typically designed as stand-alone devices. Expect a major shift to wireless connectivity for these and numerous other devices in order to accommodate the networking requirements of electronic health records.
Vendors that play in both the private practice or ambulatory care market and the acute care market face additional challenges. While both markets share many requirements, the markets themselves and their environments are quite different. Developing a strategy that address both markets and minimizes unnecessary duplication is not trivial.
For medtech manufacturers, this is a world full of promise. But they're not alone. With such unparalleled opportunities in healthcare markets, IT hardware and software manufacturers are also taking notice. The healthcare divisions already established by those companies could become medtech's next competition.
Vendors in this "world full of promise" will find themselves struggling to operate outside their medical device comfort zone, while fighting a rear guard action against both health care IT and enabling technology vendors looking to separate (and pocket) the incremental value provided by connectivity from the device. The smart vendors will develop strategies to carve out their portion of this bigger pie, meeting customer requirements and holding these new competitors at bay. There is no doubt, however, that some market segments and product categories will be commoditized; new players will emerge and some current vendors will go into decline.
The risks are no less for hospitals. Traditional organizational silos (nursing, biomed, IT) must be breeched to bring a cross functional approach to the next big leap in hospital automation, the point of care. The absence of a potential single vendor solution means that hospitals will have to navigate the connectivity issues discussed here and plot a course that will support organizational goals on schedule and within a proscribed budget. There will be casualties on the hospital side as well, with system implementation, budgets and time lines all compromised by poor choices and unforeseen circumstances.
The Chinese adage, "may you live in interesting times" certainly applies to today's connectivity market.
UPDATE: At the AACN/NTI meeting Philips announced WMTS support for their multi parameter patient monitors, specifically models MP20, MP30, MP40 and MP50.