I came across an article on FierceHealthIT which has a really fascinating quote from Karen DeSalvo, National Coordinator for Health IT. Here it is:
“What seems that it would have been helpful is if we had agreed as an ecosystem–the government, the private sector–that we would have a set of common standards that would allow us to have more seamless sharing of basic health information,” she said. “We’re moving toward that with the industry, but I think what that’s created is a complexity and aggregation of data … In hindsight, maybe some more standardization, or a lot more,” was necessary.
Is lack of a standard what’s keeping healthcare from being interoeprable?
I personally don’t think that’s the biggest problem. Sure a standard would help, but even with the best standards in the world if organizations see data sharing as contrary to their best interest then no standard will overcome that view. It’s been said many times that we have an issue of desire and will to share data. It’s not a technical problem. Sure, a standard would be helpful once there is a will to share data, but if organizations wanted to share data they’d figure out the standard.
Later in the article, CommonWell Executive Director Jitin Asnaani said “Standards are not standards because we say they are; standards are standards because everybody uses them.”
This is the problem. People don’t want to share health data and so no standard is being used. I still wish they’d blow up meaningful use and use the rest of the money to incentivize organizations to start sharing. People went bat crazy implementing an EHR as they chased government money. I’d love to see healthcare organizations go bat crazy becoming interoperable as they chased the rest of the government meaningful use money.
As much as I like to beat up on ONC’s numerous MU failings, I think that standards are better addressed by NIST. It’s not a regulatory agency. It works with stakeholders to develop consensus around standards and does a first class job of it.
When ONC is gone, and no one knows how long it will be around, NIST will remain because US businesses see value in what it does.
Even so, we don’t need some standards, we need a spectrum to cover HIE’s range. These also need to have a practical, extensible basis.
Finally, someone in government agrees that MU has not had a focus on the right things.
It’s amazing to me, coming from the industrial process control domain where interconnectivity across multiple systems and applications has been routine for 70 years, to see healthcare fussing over interconnectivity.
Their is no need to standardize.
Each set of trading partners the publisher needs to format data for easy posting to a generic data exchanger and the subscriber needs to be able to read data at the data exchanger for easy import to the subscriber environment.
The design criteria are that each partner be allowed to read/write using their own native data element naming conventions (i.e. I post “abc”, you want to be able to read it as “def”, a 2nd subsriber may want to read “abc” as “ghi”.
Of course, a long name is required per publisher data element so that subscribers are able to figure out what they are subscribing to and the other requirement is that a publisher be able to share on a need-to-know basis.
And, yes, since the usual setup will be “pull” instead of “push” each subscriber needs to be able to retain a cursor position at the exchange so they know the last line item they read.
Agreed. Standards are useful and necessary, but not the real goal. They are a surrogate, a proxy – like LDL or A1c. If one defines the job as standards, one runs the real risk of diving down a long, dark rabbit hole and might end up with great standards and no sharing.
Define the job as sharing. Incentivize sharing and punish barriers to sharing. Standards are necessary for sharing, so when the industry sees it needs to share and will be paid to share, it will develop the tools and standards to share.
Peter Elias,
A perfect description of what I believe. I love how succinctly you describe what needs to be done.
I read this tweet a few weeks ago from @dsmorgan77: “There are two kinds of people in this world: those who have data and those who want to standardize it.”
This clarified for me the inherent problem in a healthcare data exchange standard and settled it for me: there won’t be one in our current system. The data is too complex and there are too many parties. There’s no scenario where we define a standard, all EHR vendors adopt it and all end-users populate the data in the same way.
Long term, out of all the ideas I’ve heard, I think a de-centralized model like what UnPatients is describing has the best chance to succeed. That 10+ years out.
Near/mid-term I think we need what Karl describes above. It sounds a lot like EDI and there is no single standard for that. Orgs that want to trade might use it as the standard language for communication, but they still need to create point-to-point relationships and maintain them. I’ve worked with industry groups that created there own “sub-standard” on top of EDI and then everyone talking in that “sub-standard” could get up and running more quickly, but even those arrangements failed to deliver on the promise and we still ended up having to tweak and maintain to satisfy each new partner (i.e. because the data means something slightly different to them).
In principal, I agree with Peter that the industry should be left to figure it out. That’s going to happen with value-based payment, but it still won’t lead to a single standard. We’ll find the path of least resistance to be similar to those industry sub-standards and we’ll do them with the groups we want to share with most often. It’ll be very regional and ultimately we’ll look to the regional HIE. People think I’m nuts when I say that, but I’ve been doing data exchange since ’97 in many industries and given healthcare’s goals this is the only plausible path to making it all work in the next 5-10 years.
Thanks for sharing your perspective Don Lee. I take from your comment that it’s going to take time, there’s not going to be one standard, and we’ll follow the path of least resistance.