I attended the HIMSS23 in Chicago taking part in the new IHE SDPi medical device ecosystem standardisation. As always, I was primarily interested in interoperability as it is deployed now and how it is evolving independent of organized standards bodies.
One of the main trends I saw was a core use case called the Silent ICU that attempts to tackle the big issue of too many alarms coming from medical devices causing problems like nurse fatigue. Research suggest that 86 to 95% of alarms are not useful clinical information and are just an additional burden to the medical personnel.
I was especially impressed with the showcase because of the focus that was given to how the family of the patient in the intensive care unit would feel - without and with the Silent ICU. Imagine having a son or a daughter connected to life saving devices and having loud alarms going off every few minutes. We often focus on the efficiency aspect of interconnected devices, but from my perspective it is the other capabilities – reduced stress, reduced errors, analysis, etc that are far more valuable in most cases.
Medical interoperability standards such as HL7 often miss this type of use case. With this particular use case, IEEE 11073 service-oriented device connectivity family of standards that defines a communication protocol for point-of-care medical devices - the SDC in short (also used in the IHE SDPi profiles) accommodates these needs even better.
“SDC is a web services-based architecture that enables interoperability amongst point-of-care medical devices and data exchange between point-of-care devices and HL7 compatible clinical and hospital information systems.”
But there is another aspect of the SDC that I think is worth further discussion and that I discussed with other participants of the conference. It is the fact that the standard externalises control over the medical devices to the so-called SDC network (a set of connected medical devices around an ICU bed), thus making them in a sense a dumber version of themselves.
Of course, the dumbing down of medical devices can also be seen as making them more robust. But in fact the main issue is that if a medical device gives external control to some other device in the SDC network, this suggests that other device needs to be a medical device of higher level classification to still comply with the e.g. FDA or MDR regulation. So if, for example, a patient data monitoring system that is usually present in hospital wards and allows monitoring of patients hooked on different medical devices, wants to be able to send remote commands to the medical devices which SDC focuses on, then this patient data monitoring system is actually becoming capable of impacting the patient. For that risk to be mitigated appropriately, the patient data monitoring system needs to become a medical device of the same class as the medical device itself.
In practical terms this means that a simple monitoring system that previously only displayed clinical data from medical devices now needs to become itself a medical device, too. This inevitably means an increase in number of registered medical devices and their respected classes which adds additional burden to medical device manufacturers.
A natural question arises how well the data monitoring systems manufacturers, or any other systems that would want to take part in the new SDC network, are going to be ready to invest in increasing the medical device classes. As this is not a trivial task to achieve, and due to ever increasing regulatory pressure in the form of the MDR, it certainly will have a great impact on the adoption of SDC.
