23 October 2012
The challenges for medical device interoperability
Hospitals have complicated and connected technology ecosystems. Few places have such a diverse array of machines, skills and information, all attempting to coexist in a stressful decision making environment. Each department has an array of devices and a staff with highly specialised skills.
Information on patients and their treatment must flow seamlessly through this ecosystem, across more than one hospital and, potentially, between hospitals and the home. And this needs to be sustained throughout the patient's life for treatment and care to happen in an effective way. Modern hospitals are filled with complex machines and the bar is set high for their safe and effective use. Taking a device to market is an increasingly complicated task.
One example is the potential for an ambulance to download data. Ambulances feature ever more equipment to allow patients to be treated quickly, but it would be even better if information on their condition could be relayed to the hospital. Unfortunately, these devices were not designed with this in mind and retrofitting them can be difficult and expensive.
Interoperability could be defined as 'the ability of medical devices and healthcare systems to interact seamlessly with each other to enable better outcomes'. The goal is not only to do things faster, but also to achieve better outcomes for patients.
This can be achieved by the adoption of standards, so different devices from different manufacturers can communicate. Devices must connect securely across the hospital's ecosystem and through telecoms systems into the home, allowing information to flow. The goal is to reduce the variability in workflows for a given treatment so a positive health outcome can be achieved at the most optimal cost.
The data under consideration is largely that representing the patient. This data not only needs to be accurate, but also to be in the right place at the right time, delivering the required detail for a person or another machine to understand it and act upon it. Data is ubiquitous, but the lack of interoperability is making it hard to take advantage of this information and improve patient safety.
For example, it has been demonstrated in intensive care that the ability to synchronise xrays with the patient's breathing cycle improves image quality. Unfortunately, interconnecting and synchronising these devices automatically is not possible today.
Another example is a safety interlock and alert system that could stop the flow of pain medication from an infusion pump if a patient showed signs of respiratory distress. Sadly, we are locked into a risk averse world, where interconnecting these types of devices is considered to be an additional integration burden that no device manufacturer wants to tackle.
The list of stakeholders is long: governments and providers, physicians and device manufacturers, standards organisations and patients. Each can benefit from greater device interoperability, but there are many barriers to adoption.
Although efforts are being made to regulate medical devices by health delivery organisations and regulators, there is no strict enforcement for interoperability. While there are necessary guidelines and standards, lines are drawn as to where the box starts and ends and therefore interoperability is not strictly regulated or enforced. Companies, therefore, can get away with the lowest common denominator of interoperability.
Another important dynamic is the network effect, whereby a small network that connects to a large network gains massively and disproportionately. Incumbents in healthcare systems do not have the incentive to interoperate with other companies' devices as it may erode their market share. A few players dominate the market and, in specific hospital chains, even fewer players dominate the device ecosystem.
Thirdly, most medical device companies have been vertically integrated, so the concept of standards and componentisation has been restricted to their environment. Intervendor interoperability means pushing interoperability a level down – to their suppliers.
Interoperable software standards should be implemented no differently than basic communication protocols. The same holds true for connected components – companies should encourage their suppliers to buy components that are prefabricated and integrated one level down in the value chain.
Progress is being made, despite these barriers. Health delivery organisations are starting to realise the huge costs incurred and are pressing vendors for interdevice interoperability. The US FDA, along with other standards bodies, is pressing suppliers and hosting medical device interoperability workshops that encourage greater collaboration in the testing of devices. Also, government health agencies and large healthcare providers are working with some telecom and networking companies; and telecom providers can be highly effective at enforcing standards.
There are many interoperability standards bodies which, at a very high level, can be separated into regulated (hospital) and unregulated (consumer or home) types. As an example, the Continua Health Alliance has defined IEEE11073 medical device communication standards for personal healthcare, such as individual personal telehealth devices ranging from blood glucose meters to thermometers. More than 70 devices have been identified under this umbrella standard, which allows compact devices to connect via USB and Bluetooth, with the goal to monitor patients reliably and remotely.
Other standards address different environments or ensure messages are formatted in the right way, so an alert on an infusion pump gets to the intended recipient. The goal is always to get the right data to the right place at the right time in the right level of detail, so it can be acted upon.
Vertical integration can work in narrow markets or where the entire device has IP that adds value. But fewer companies can make this approach work. In the health market, most of the electronics and control architecture that go into devices ranging from low power insulin pumps to complex imaging devices can be designed outside and bought off the shelf. A classic example is the TCP/IP stack; 15 or 20 years ago, this would have been written internally and the standard interpreted in a way for easy implementation. Today, most companies will obtain a stack from an OS vendor, including all the appropriate required certification such as IPv6.
To ensure interoperability, three forces need to come together. Enforcement, regulators, healthcare delivery organisations and standards bodies must come together and increase the pressure on companies to comply with the standards. Secondly, there needs to be an incentive system through which a device manufacturer can 'score points' for designing an interoperable system that works with those of competitors. This is an area where CTO organisations within large companies have been proactive and are starting to insist on interoperability. Thirdly, companies need to stop thinking vertically. Interoperability is not where the innovation exists; this happens at the application layers and it is where revenues lie for device manufacturers. Companies should also insist their suppliers implement the appropriate standards.
Wherever they are in the chain, people will pay for better outcomes. So there is a significant market opportunity in creating the middleware to implement these evolving standards, enabling innovation in treatment applications. The recipe is simple – find the ecosystem, find the suppliers and partners with the same or better quality metrics and push the standards, focus on solving new problems and actively seek out partners. Collaboration is the new competition – particularly in the connected health ecosystem.
Ken Herold is Wind River's lead medical engineering specialist.