Reference Driver: Not Enough

The core software component of a Wi-Fi product is the device driver for the Wi-Fi radio that operates in the device. That driver provides the interface between the device’s operating system and the radio. Intel, Atheros, Broadcom, Marvell, and other silicon providers may be known for making Wi-Fi chips out of silicon, but they employ teams of software engineers that develop device drivers for APs, routers, laptops, and other devices that use the radios with those chips inside.

While drivers from silicon providers (often called reference drivers) are sufficient for mainstream client devices such as laptops, they are not designed for mobile medical devices. For starters, a driver may not run on a medical device because the driver was written for a different operating system than the one that runs on the device. Even when it runs on a mobile medical device, a driver may not address the requirements of that device, especially requirements for reliable connectivity when the device is in motion. (more…)

A lot has changed in almost 30 years - I mean besides feeling older.

The nirvana that was the 1970s came to an abrupt end on February 27, 1998 at 2:17 pm, when, “WFAA-TV channel 8 television began broadcasting on digital TV channel 9 and continued until 10:35 p.m., shutting down transmission a few times to allow a tower crew to work on the antenna.” This and subsequent tests of digital television broadcasts by the Dallas broadcaster, knocked Baylor University Medical Center’s (BUMC) telemetry off the air. Fallout from this intentional (and completely legal) interference resulted in the creation of the new WMTS frequencies for use by telemetry monitors. Between that fateful day in 1998 and 2006, BUMC has spent $6.6 million shifting frequency and upgrading the telemetry systems at their hospitals. (You can read about BUMC’s ordeal reprinted from the AAMI publication Biomedical Instrumentaiton and Technology Journal story on this FDA web page.) (more…)

Because it promises throughput as much as 10 times greater than that available with current Wi-Fi® standards, the forthcoming IEEE 802.11n standard is generating tremendous interest among users of wireless LAN (WLAN) products. 802.11n throughput rivals that of Ethernet, and so availability of 802.11n may cause some organizations to use WLANs as the primary means of network access for typical computer users.

Although the 802.11n standard will not be finalized and ratified until 2009, it is easy to find laptops, home routers, and other products with radios that are based on a draft of the standard. The Wi-Fi Alliance, an industry association, is performing product interoperability testing and certification based on the draft standard. Should makers of medical devices be racing to add 802.11n to their devices?

(more…)

Wi-Fi radios adhere to a set of IEEE and industry standards that define how the radio interoperates with a wireless LAN infrastructure. Devices that bear the Wi-Fi CERTIFIED^™ seal have passed a set of interoperability tests defined by an industry association called the Wi-Fi Alliance^®. A medical device that is Wi-Fi CERTIFIED should interoperate with any wireless LAN infrastructure, but there are no guarantees that operation will be flawless or that connections will be reliable. That’s because Wi-Fi interoperability testing uses access points (APs) from only a few vendors and doesn’t include such things as roaming from one AP to another.

What Is CCX?
(more…)

FierceHealthIT notes a new CDC study on ED overcrowding - it's getting worse.

The 32 page report is fuel for the American College of Emergency Physicians lobbying efforts to get congress to, “create a commission to study the ED overcrowding problem. Under the
terms of the ACEP-backed bill, hospitals would have to report to HHS on
how many patients are boarded in the ED, and how long they're boarded.” [Patient “boarding” is the practice of placing patients in hallways, usually in the ED, where they wait for an inpatient room to become available. Patients commonly wait for hours, and sometimes more than a day, on a stretcher parked in a hallway.]

Ambulance diversion data is tracked by hospitals, regional and state hospital associations, and sometimes the state. This data is not available to the public or most state health agencies. Given how bad ED diversion is, I'm not surprised hospitals want to keep this data private - especially the worst offenders. Data on patient boarding is tracked less often by hospitals and to my knowledge, is not tracked across hospitals by associations.

Public reporting of both diversions and boarding would provide an important customer service metric and patient safety indicator and should be available to prospective patients. It is too bad that such a requirement must be forced on the industry by government.

You can download your own copy of the CDC report here (pdf).

A medical device is an instrument, apparatus, implement, machine, contrivance, implant, in vitro reagent, or other similar or related article, including a component part, or accessory which is: …intended for use in the diagnosis of disease or other conditions, or in the cure, mitigation, reatment, or prevention of disease, in man or other animals…

He noted that the medical device connectivity market has changed dramatically in the past 5 years - with the advent of “smart” pumps and WiFi adoption, for example. They have a pretty extensive background in this topic, with the following past projects:

• Central fetal monitoring system and Labor and
• Delivery Decision Support (1998)
• EKG (1998)
• PACS (1998)
• Endoscopy (1999)
• Radiation oncology record and verify (1998)
• Patient Monitoring networks
• EEG and video monitoring (2001)
• INET (ongoing vital signs/EMR integration project) (2005)
• Stress testing (treadmill and vitals) (2006)
• Simulation Lab (2007 and beyond) – partnering with Vendors to provide tomorrow’s technology today

Steve described the Nachi/Welchia worm that hit the net on August 13, 2003 as a wake up call for their hospital. The worm had a catastrophic impact on the hospitals networks, especially medical devices running Windows 2000/XP. Any unprotected devices were infected - none of their medical devices were protected. There was a 2 day network outage and it took weeks to recover and get all their medical devices back online.

Many changes were made as a result of the Nachi/Welchia worm. The formed partnerships with device manufacturers to architect solutions to this problem. Siemens and Baystate Health teamed up to design an acceptable antivirus solution running on imaging equipment - this took 2 years to negotiate, design, and implement. They also worked with their network architect and implemented a VLAN ACL design to isolate medical devices on the network.

Another key lesson learned from Nachi/Welchia was the need for software virus protection. Baystate got a lot of push back from vendors at first. Ultimately though, no vendor ever provided them with a legitimate reason not to use anti virus (AV) software with active protection. Excuses included: the FDA won’t allow us : WRONG! It harms the patient data: WRONG! I will slow the system down: WRONG! Proper design and implementation of the AV software will prevent any anticipated problem. He recommended that hospitals partner with vendors to find the solution, and noted that you might have to twist some arms along the way. Baystate uses McAfee VirusScan and ePolicy Orchestrator to document and enforce security policies.

Steve went into good details regarding network architectures and management policy. Much of this was similar to the VA’s Medical Device Isolation Architecture Guide.

Numerous important reference documents were mentioned - documents that readers have seen mentioned in this blog. Included in his presentation were the FDAs guidance documents on Off the Shelf Software in Medical Devices, and General Principals of Software Validation. [You should also check out their draft guidance on wireless medical devices.] Also mention was the VA’s recommendations for networking medical devices in their Medical Device Isolation Architecture Guide. Their network security is build mainly on access control lists (ACLs). Architecting a proper ACL is key to protecting medical devices, and is based on the following principles:

• Only allow the bare minimum permission to talk
• Communication with vendors is required to design an appropriate ACL
• Vendors need to start doing a better job documenting ACL requirements

Finally, Steve described the skill sets required of clinical engineers involved in medical device connectivity. He also emphasized the need for rigorous documentation, including: project plans, specifications (requirements) for each component in a system, status updates, and change control.

Questions: Does INET project use an Aware gateway. It’s a VLAN by IT with biomed as first responders to users. Virus protection policies, that includes definition updates specific, are established for each medical device system. The IT department hosts the virus software update sever, but biomedical engineering initiates the push of updates from the server in IT to the target medical devices. They are a Cerner shop and use their IMDB connectivity software to integrate their Philips monitors.

• Multi-functional purpose of networks is accelerating growth of wireless LANs
• Wireless network priorities: Robust security, Upgrading to 802.11g, Reliability and Quality of Service (QoS)
• IT Department will “own” shared networks
• Regulatory certification with specific network hardware
• “Wireless LAN best-practices” need to be supported by vendors
• Challenges
  • Device companies need to invest in keeping products current
  • New support issues result from shared networks
  • Integrating products with untested, existing hospital networks

New requirements make deploying medical devices on shared networks, specifically the need to simplify implementation and testing of devices on common network infrastructure through industry-wide best practice. Device vendors need to use device design and development strategies to accelerate new standards and new customer network requirements.

Scott referred to the recent FDA draft guidance on wireless medical devices. Specifically, “FDA believes the more critical the medical device function and information passed via RF technology, the more important it is the wireless connection be robust.” He suggests that ISO 14971 is a great risk management tool for wireless medical devices. Vendors with wireless medical devices should also refer to IEEE 802.15.2 for coexistence risk management and mitigation, and IEC 60601-1-2:2001 describes in detail testing for unintentional electro magnetic interference - the latest draft of this standard is intended for vendors and users. Users are increasingly falling under standards like this because it is only in actual use that certain factors like interference are ultimately experienced and resolved. The draft IEC 80001 is another standard for both vendors and end users, that describes standards for the deployment and support of medical devices in a network - this standard also includes a risk management process.

Next up is Dave Hoglund, representing Andrew Corporation. Dave started with a survey of the many wireless technologies deployed in hospitals. Hospitals are the worst “Multi-Path”environment due age of construction of hospitals and the variety of construction materials and methods. Nothing is standard! Dave’s presentation also includes a lot of great reference data on wireless interference, power levels and other technical issues. Dave touched on the perennial hot issue for biomeds, cell phones in hospitals, suggesting the risk is overblown in consideration to the value derived from cellular technology.

Best practices were suggested:

1. Perform an initial RF survey
2. Perform periodic RF update surveys
3. Establish, publish and enforce a “spectrum policy”

An industry issue that Dave mentioned was the indirect distribution channel that network vendors use. Buyers must not only select the best network vendor, but they are also highly dependent on the ability of their VAR (value added reseller) who actually sells, designs and installs the network. Another key point is the importance of having your own tools and using them. In fact, hospitals should have multiple tools in some cases because they do things in different ways, providing a more complete picture of your RF environment.

Questions - bandwidth consumption of wireless medical devices on an 802.11 wireless LAN? 10-20 kilobits per second - relatively low bandwidth utilization, but sensitive to latency.

Question inspired by experience with a smart pump vendor who released their product with only one kind of encryption - not the kind used by the hospital. The vendor took 9 months to reverify and revalidate their product using the new encryption method. Takeaway: vendors need to provide sufficient flexibility and choices so they can conform to customer’s network environment.

Suggestion from audience that Biomed could assume responsibility for the wireless LAN. The past approach has been reactive, and the big point of the presentation is to be proactive. Of course, wireless LANs can be monitored from any department or location - even more than one.

Another suggestion is for medical device vendors formulate best practices and submit them to wireless LAN vendors, so they can address new requirements in future versions of the product, and can train their VARs on device vendor’s best practices.

The last session of the day was Managing your Wireless Spectrum: Realizing Your Full Potential,” presented by Dave Hoglund. In this presentation Dave got deeper into RF best practices. Before you can talk about managing, you have to know what it is you’re going to manage. As new wireless applications are adopted - wireless VoIP, patient monitors, alarm notification, point of care barcoding, etc., you must reevaluate your network design - and may need to redesign your network. Dave reviewed wireless technologies with emphasis on things seeing increasing adoption like ZigBee, cell phones and distributed antenna systems (Mobile Access, InnerWireless, Andrew). After a review of unique challenges and requirements for the hospital RF environment, Dave launched in to best practices. The underlying theme here is being proactive.

Another area of focus includes the tools that are used to monitor and manage RF environments. Dave described the basic requirements for a network operations center that would provide planning, monitoring, and trouble resolution for the RF side of networking.

Questions: RF monitoring that extends down to 402 MHz MICS band. Recommendations have been made to Cognio to extend the RF monitoring range of their system. The larger wireless LAN vendors will be building this monitoring capability into their APs.

An IT department trash talked wireless medical devices because their 802.11b radios would pull down the speed of 802.11g clients already on the network. This is true for older APs, but all the current APs support independent speeds for 802.11a and 802.11g clients on the same AP at the same time.

Can you do a lower cost abbreviated site survey with single channel wireless LAN vendors (like Meru or Extricom) or should you do a complete site survey. The recommendation is to do a full RF site survey that covers all the RF spectrum that is in use in a hospital. Consequently, a single channel infrastructure may not result in a lower cost site survey.

According to this story in the New York Times, “More than half the doctors from New York State, New Jersey and
Connecticut who responded to a survey conducted in April by the
American College of Emergency Physicians said that boarding had
increased significantly in recent years.” Boarding is the practice of treating patients in hallways of busy Emergency Departments, frequently while waiting for an in-patient bed to become available.

You can chalk up this survey as another example of the increase in reporting of patient safety and outcomes:

Ms. [Julie] Lloyd [a spokeswoman for the American College of Emergency Physicians] said the survey was the first by her group, or any other she
knew of, to try to assess the number of deaths resulting from boarding.
She said in an e-mail message that a key point in legislation before
Congress “is to collect boarding statistics, which heretofore have
remained the province of the hospitals.”

Public reporting of boarding seems to be in our future, nation wide.

Don't kid yourself, boarded patients have died, in New Jersey and beyond. There are no statistics on this because hospitals don't code the cause of death to reflect the preventable adverse event that killed them like, “death from inattention due to boarding.” Harsh? Perhaps, but this is a long standing problem in which most hospitals still wring their hands and ask, “what can we do?” Here's what they did at Stony Brook hospital:

[Hat tip: FierceHealthcare]

Tony Chen at the Hospital Impact blog has a great post on the tsunami of recent deals in health care. These deals are changing the health care industry. Outsiders in the form of private equity investors and insider hospital M&A are gobbling up failing organizations or those weakened by market changes for which they have ineffective responses. Newly recast ventures that are successful will reward effective change which will drive additional change.

The change required is not the simple head count reduction, and trying to figure out how to do the same old thing with fewer people. What is needed are fundamental changes to the way that care is delivered. These are the changes that will reduce length of stay and improve both patient safety and outcomes. Hospitals today are stuck between a rock and hard place - the rock is the inexorable reduction in reimbursement, and the hard place is the increased visibility and transparency surrounding patient safety and outcomes.

Here are a couple quick examples of the change I'm referring to. Almost every hospital boards patients in the emergency department (ED) to a greater or lesser degree. These patients, shuffled off to out of the way halls awaiting admission, are cared for by off-service nurses (ED nurses) at ever worsening - and sometimes hair raising - nurse to patient ratios. A few heretical hospitals have started boarding those patients awaiting beds up on the floors, on their service. Consequently, each nursing unit could have 1 or 2 extra patients - receiving appropriate care (because they're in the appropriate unit) and at a nurse to patient ratio that is only slightly impacted by a couple extra patients. To someone from another industry this makes admirable sense. From my peers, I've heard excuses ranging from “it's jut not done” to “we can't do it because of fire codes (or department of health regs, etc.)”.

The other example is variable acuity care delivery. Hospitals are organized into units that are specialized based on the level of care, nursing vigilance, therapies that can be delivered, and staff training. This regimented organization was used in manufacturing 20 years ago. Like manufacturing in general, manufacturing healthy patients is not a steady state process. Consequently, census in units varies wildly from unit to unit and day to day. Some units are habitually over capacity, and represent the most common patient flow bottleneck found in hospitals - those over capacity units are critical care units with patient monitoring. Variable acuity units are an analog to manufacturing clusters or pods where the physical environment is quickly modified to adjust to new requirements. Rather than transferring patients because of needed patient monitoring, or a more sophisticated therapy, those patients are kept in their unit and those resources are brought to them. Monitoring patients outside traditional monitored units is a growing trend - half of the telemetry transmitters in hospitals are used on non-cardiology patients. Implementing variable acuity care units is not easy; it requires some pretty fundamental changes.

So, how do you accommodate falling reimbursement and demands to improve patient safety and outcomes? The opportunities to improve operations are many, almost as many as the excuses used to avoid substantive change. As private equity and M&A roils the health care industry, smart people will ask the hard questions with increasing fequency, and change will come.

Pictured right, “resistance is futile.”

The California Healthcare Foundation has underwritten a study looking at ambulance diversions across the state. Findings in this phase 1 report showed that state wide emergency departments (EDs) were on divert an average of 10% in 2005. Emergency department closings to ambulances continue to confound hospital administrators. You can read about a recent survey showing patient flow as an increasing problem here.

ED diversions also frustrate policy wonks and consultants due to the almost total absence of public reporting. Certainly hospitals (and their state associations), emergency response districts, along with some local and state governments, know their emergency room diversion statistics - they just aren't publicly available. Even though access to data is improving, organizations like the CHF have to pay to collect much of their data.

As expected, the study finds that ED diversions occur mostly in urban areas. Increasingly districts and hospitals are implementing “no diversion” policies - in other words they simply take the patients that they used to refuse - even though in most cases little or nothing has been done to improve emergency department overcrowding or patient flow.

You can download your version of the first report here.