Modern healthcare is full of complex forms of data, from images to clinical codes, and from monitoring signals to lab results.
Networking medical devices through the IoT will bring great benefits in patient monitoring and treatment, but it will also add a large amount of structured data to this already stressed system.
What decisions do healthcare organizations need to make when implementing IoT connectivity so that they get the best value out of this data without imposing excessive costs on busy physicians and nurses?
The current healthcare data situation: stressed and overloaded
A few years ago Bridgehead Software
estimated that the average US hospital has around 660 terabytes of data, much of it in the form of unstructured images. This will soon be over a petabyte per hospital.
There is a wide lack of data interoperability in healthcare, the legacy of multiple vendors, separate developments in different clinical departments, and regional differences. There has been a lot of progress in Health Information Interchange (HIE), but problems persist.
When it comes to expansion of data collected and made available, physicians and hospitals tend to be suspicious and dubious, the legacy of electronic health record (EHR) rollouts over the past decade or so, with their malfunctions, inadequacies, and ease-of-use problems. Healthcare workers feel that an inferior system was imposed on them by vendors who, no matter how good their IT skills, lacked an understanding of healthcare workflows. Hospital staff has been known to rebel
against what they perceive as inadequate EHR implementations.
Anyone seeking to integrate IoT data into healthcare will need to bear these lessons in mind.
But, no matter what, the IoT is coming to the hospital
There are currently a vast number of sophisticated devices operating in every hospital, including a wide range of monitors, medication pumps, and physiological supports. Most of these are still not connected to a network.
In 2015, for example, less than a quarter of physiological monitors were connected to an EHR
, and less than a tenth of vital signs monitors.
Inevitably all of these devices will become networked, and will become part of the IoT. They will produce a vast amount of potentially useful and lifesaving data. In order for that potential to be realized, that data must be managed properly.
Nurses, who perform most patient care, can spend nearly two and a half hours per shift
on documentation activities, the largest single category of time use. Because of the hectic pace on a hospital floor, many nurses do not chart as they go, but later, in a batch process that sometimes requires them to stay past their shift—and which can introduce errors.
Using the IoT to connect the sensors and treatment devices, and thus measure and enter their data on EHRs in real time, has the promise of decreasing that part of a nurse’s workload, while ensuring up-to-date EHR data.
But healthcare workers fear, and not without reason, that they will soon need to face yet another layer of noninteroperable data, rather than a seamless augmentation of their capabilities.
How can the IoT be integrated effectively into hospital workflows?
There are several important things for hospitals, provider networks, and anyone else seeking to leverage big data for improving patient care and outcomes to keep in mind.
Evaluate each connected medical device as to the actual clinical value of its data stream. Just because something generates a lot of data, does not mean that that data is necessarily useful for making treatment decisions.
In doing this, hospital will need help. The IoT market is moving so quickly that it’s hard to know what’s available now, much less planning for the future. Like players in other industry verticals, hospitals and health systems will need to partner with IoT platforms to get the necessary support.
Develop a clear and robust data governance policy. According to a 2014 HIMSS Analytics survey
, forty percent of hospitals had no formal EHR data governance policy, many years after EHRs are in routine use.
Focus on the interface. How can the overworked physicians and nurses who directly deal with patients gain the best overview of the new data, with minimum impact to their workflows? To be accepted and used, IoT data must support patient care, not interfere with it. This both the most difficult and the most important part of a healthcare IoT strategy.
Don’t assume that enthusiastic speeches at kickoff meetings and cheery posters at the nurse’s station will do your work for you, as far as acceptance and use on the floor is concerned. These people have been through this before. Genuine collaboration with frontline staff is essential for turning raw data into information, understanding, and action.
Accept the inevitable effectively
Hospitals have been forced to accept specific technologies and processes before, and scrambled to catch up. They need a plan on how to incorporate the IoT.
Photo by Piron Guillaume on Unsplash