I was perusing, as one does, the newly released “Connecting Health and Care for the Nation – A Shared Nationwide Interoperability Roadmap“, issued by the Office of the National Coordinator for Health Information Technology, and a few thoughts came to mind. Nothing too alarming, but I think pertinent to ask. The roadmap is a 10 year plan that is part of a much larger, ongoing effort to streamline the industry’s use of IT in healthcare and at the same time make it more effective.
As the document says, “[…] [I]nteroperability is defined as the ability of a system to exchange electronic health information with and use electronic health information from other systems without special effort on the part of the user.” Sounds worthy enough. Quite succinctly the document suggests that there are four critical actions for near term wins and progress to this worthy end:
- Establish a coordinated governance framework and process for nationwide health information interoperability
- Improve technical standards and implementation guidance for sharing and using a common clinical data set.
- Advance incentives for sharing health information according to common technical standards
- Clarify privacy and security requirements that enable interoperability.
This is where some of my thoughts and questions pop up. First a note: the key is “common clinical data set”. So I would assume that the idea is that all individual systems will carry on using the data that they do. These disparate systems will all therefore need to map to the ‘common clinical data set’ or embed the use of that common set into their own information systems. Thus the shared data model will operate as part of the solution that supports interoperability between the systems. This means there is much less impact on the remote systems involved and more a need to a) govern the central model, and b) have all edge apps modified to connect in some way to the central framework. This is a good thing – I think.
The third item concerns me – and the answer impacts the second. In my experience there are only two methods to change behavior, which is the carrot and stick. Regulation and compliance can be used but it is not really a positive motivator. If there was a benefit to sharing and maintaining this data, as in how I am recognized, or how that data is needed to do my job, then there is room for a positive motivator. So with the lack of either, why do we need to “advance incentives”? Are they not self-evident? If they are not then perhaps we don’t need to share the data? In my information governance travels the more we know about the real stake holder and ultimate, final (business) outcome, the more likely we can unearth an effective incentive. We can’t, and should not, manufacture them. I admit I just don’t yet know enough about the use cases here but the value accruing to them is the key to determine the metrics framework to bring the use case about.
The second item, once the third is addressed, tends to become easier. If the third item is not resolved, the second item becomes a bonanza for IT vendors. We only need to focus on the data and tools related to actually sharing it – nothing else. So I would not worry about the technical issues yet until and after the governance is set up, the concept of the data set is at least tentatively defined, and the set of impacted stakeholder and their incentives are identified. If you go straight to technology before passing go, you will pay a whole lot more than $200.
Finally I think privacy and security will inhibit interoperability, not enable it. They are barriers that will tend to slow down the freedom of information thus they are policies to be taken into account in the first item – the governance board.
As I read more of the document I spied the following gem: “There are also many aspects of health IT beyond interoperability that are important and will be critical to a learning health system, including technology adoption, data quality, documentation and data entry, usability and workflow. However, these topics are out of scope for this Roadmap and deserving of separate, dedicated attention.” Um, OK, yes dedicated attention. However, what will the governance board not do if not understand the role of data quality on its outcomes? Other industries have tried large-scale interoperability models – look at GS1 and Global Data Synchronization. They did not take account of data quality (going into the infrastructure) as a standards and governing board and their lofty goals for GDSN fell far short. It would be like setting up a national telephone infrastructure service and not checking to make sure the users actually all speak the same language.
The use of data, its quality, is central to information governance. Without (and we see this every day), the governance meeting’s start off fine: lots of coffee and donuts and lots of business people attending. After a few months and a few decisions, the attendees dry up. “It wasn’t relevant to me”, they all said. This is because information governance work is seen as an IT issue for data people to work with; it should have been couched as a business issue for business people who are concerned with processes held hostage to bad data. The metrics, up front, should have focused on process metrics and how data impact them. By leaving data quality out at this early juncture, I wonder if this will lead to challenges later? Perhaps there is previous document that covered data quality – I will keep looking and let you know.