In health care delivery, the network is a medical device. Okay, not always, but when devices like patient monitors or infusion pumps are connected to a network that carries data critical to patient surveillance or alarm notification, the network is part of the regulated medical device. Now the FDA has been kind enough to mostly look the other way regarding wired and wireless local area networks. The FDA would be well within their legal jurisdiction to tell network vendors to follow the Quality System Regulation (QSR) and get premarket approvals for their products in applications that are regulated medical devices. To date (with a few minor exceptions) the FDA has been content to let people who really know what they’re doing (i.e., medical device vendors) manage networks as off-the-shelf technology incorporated into their regulated devices. Device vendors follow the QSR so vendors like Microsoft and Cisco don’t have to.

As networked medical devices and the resulting distributed systems of devices, servers and clients become more complex, the need to actively monitor and manage those networks becomes increasingly important. In days past, vendors could create their own proprietary island of information, and by creating a tightly controlled and isolated system they could reduce performance variables (especially pesky external variables) to the point where a system that is certified upon installation will pretty much run reliably on its own – until some major component, like a network switch or something, fails – which becomes a simple break/fix task.

Hospital IT has been breaking down islands of information, forcing them into enterprise solutions, for as long as I’ve been in the industry – over 20 years. That day is dawning for medical devices. Virtually every patient monitoring and infusion pump vendor has a gateway server product that converts their proprietary protocols to HL7 so they can export patient identifiable clinical data to EMRs and such. But many of these vendors still rely on private networks, their own VLAN, or some other inconvenient and obtrusive method for isolating their system from the customer’s enterprise infrastructure. Now most of these vendors will tell you that that’s the way it has to be, and will mumble “FDA” or some other such higher power to justify their position. Draeger’s OneNet was the first patient monitoring system designed to run across the customer’s enterprise infrastructure – and even supports the customer’s choice of wireless network cards. This radical departure was not made without special tools to ensure safety and efficacy – in this case the Packeteer Quality of Service network appliance.

With all this in mind, a recent deal between NextNine and Apparent Networks caught my eye.

Networks have become the lifeline of all IP communication systems and business applications. With the integration of NextNine Service
Automation and the Apparent Networks AppCritical Technical Support Edition, support organizations can proactively maximize system availability by accurately identifying performance issues in their IP networks and resolving them quickly and effectively in an automated fashion. Designed specifically for remotely assessing and troubleshooting converged networks, AppCritical provides end-to-end visibility into any live network and identifies and outlines all conditions and faults that will impact network quality.
NextNine’s intuitive Virtual Support Engineer checks the equipment on-site, correlates this information with network data and rapidly resolves the identified problem, all without the need for field visits or customer assistance.

NextNine Service Automation Ecosystem Edition automates support processes that allow vendors and support providers across the service ecosystem to deliver efficient, superior, scalable services while lowering costs. NextNine’s innovative, proprietary Virtual Support Engineer can either be deployed at the customer’s support organization for 24X7 proactive support or downloade on demand to resolve issues when required. The company’s products have been deployed by global leaders such as GE Healthcare, Allscripts,
Openwave, Comverse, and airwide solutions.

Apparent Networks AppCritical Technical Support Edition provides network-dependent vendors the ability to instantly pinpoint problems on their customers networks and discover why their applications are not working properly. With its focused and intuitive remote network troubleshooting capabilities, AppCritical provides unobtrusive access to customers networks beyond their firewalls. Its ability to instantly identify whether the application problem rests within the network or not enables an accurate and targeted approach to problem resolution, therefore saving time and money and boosting customer confidence. Its ease of use encourages broad
levels of adoption within the support organization and the unequaled depth of knowledge present within AppCriticals expert systems ensures non-network experts are able to confidently troubleshoot and resolve problems.

The press release mentions GE Healthcare, Allscripts, Comverse, Openwave and airwide solutions as users.

Network design and management is a lot like managing RF in a hospital – there is no magic “one way” that one follows when throwing up a network or deploying wireless devices – this is complicated stuff, and diligent proactive planning and management, plus better tools are needed. Putting life critical applications on enterprise networks is not always the best choice – redundancy has advantages – but having the option to run on enterprise networks is important to the market. Since they have released NetOne, Draeger has won a number of high profile sales against Philips in Europe. Draeger has little market share in the US and has not greatly impacted “private network” vendors here. But vendors like Spacelabs and Welch Allyn are making similar, if not quite so dramatic, moves as Draeger.

Pictured right is Welch Allyn’s new nurse-carried alarm notification product, introducted at HIMSS07 (pending FDA approval). I’ve much more to say about this later, but for now note that upon release (later this year) it will run on an Aruba WLAN – and a follow-on version will run on Cisco infrastructure.