Friday, January 30, 2015

The Interoperability Roadmap

On Thursday, CMS announced their intent to update the Medicare and Medicaid Electronic Health Record (EHR) Incentive Programs for 2015 and beyond, reducing the reporting burden on providers, while supporting the long term goals of the program.

On Friday, ONC released the Interoperability Roadmap and the Standards Advisory.

A busy week.  What does it all mean?

The CMS announcement includes planning for a Spring 2015 rule that:

*Realigns hospital EHR reporting periods to the calendar year to allow eligible hospitals more time to incorporate 2014 Edition software into their workflows and to better align with other CMS quality programs.
*Modifies other aspects of the program to match long-term goals, reduce complexity, and lessen providers’ reporting burdens.
*Shortens the EHR reporting period in 2015 to 90 days to accommodate these changes.

It’s a positive sign that HHS leaders are listening and responding to stakeholders.    Meaningful Use Stage 2 contains numerous goals that require an ecosystem/marketplace to develop first.   The HIE marketplace is just developing and usable apps for patient view/download/transmit are still 6 months away.   The new timeline gives us the flexibility we need to do these projects right, leveraging  the market foundation that is developing.

The draft Interoperability Roadmap released by ONC describes the current market and issues in interoperability.

It aligns well with the work of the Jason Task Force, recognizing the role of the private sector and the importance of market-based networks.   It supports the move to modern internet interoperability conventions (including RESTful APIs like FHIR)

It wisely suggests  “non-government governance” for health information exchange rather than trying to create a single top-down nationwide governance entity.

It focuses on the importance of clarifying HIPAA to reduce confusion and misconceptions about HIPAA restrictions and enablers.  For example, does everyone know there is no such thing as HIPAA certified software and there are no restrictions on giving patients access to their own data?

The Roadmap does have a few areas of concern.   It suggests that States become more active in the area of interoperability.   We need to be careful with this approach or else we’ll create 50 interoperability silos given variation in State laws.  The new economic incentives of accountable care organizations will motivate vendors to address health information exchange needs based on business cases, not geography.

The Roadmap needs to describe a few more concrete steps that government should take to support the listed goals.  Increasing value-based purchasing and having federal agencies accelerate standards-based interoperability are very good, but there are other levers to consider such as creating reusable components at scale (i.e. a nationwide provider directory), aligning quality measurement programs/aligning quality measure reporting with modern interoperability standards, and harmonizing other health care regulations with interoperability conventions (i.e. community health center, nursing home, and home health reporting and regulation).

ONC is currently accepting public comments and key commitments on the draft Roadmap for approximately 60 days which will end at 5pm ET on April 3, 2015.

The “Standards Advisory”  companion document does specify some standards that are mature or becoming mature such as HL7 2.x, CCDA, and FHIR.   However, it also includes standards that are not likely to achieve maturity based on the  objective criteria outlined by the HIT Standards Committee.   I would advocate for leaving some of the boxes blank, since no mature or becoming mature standards are available for them.

Overall, I give ONC kudos for their articulate summary of where the federal government would like the industry to focus.   The market is making tremendous progress at this point, and the roadmap is a useful directional guide.


Medical Quack said...

I had to laugh at the comment about letting states add their input with interoperability, and you are correct as how much more input and how useful would it be. I agree it always important to get input by all means but you do reach a saturation point where some becomes just noise.

What I really fear though are the new talks on perficient medicine and of course I don't have any issues with genomic data to help diagnose but they've left open this little gray area which has not yet been defined and everyone wants some of that data input from the web and that's a danger zone too. Right now I got stuck on a clinical trials list that is sold for people who take blood thinners and I can't get it off as I have been repackaged and sold and can't the originator. So, when folks find that information about me out there in error I certainly don't want any garbage from this web world out there to impact any sound clinical or genomic data about me.

You can read this informal campaign I have about licensing data sellers so we know who they are, one to be able to find them and get flawed data off the web and second with the fear of this web data in whatever format it arrives is not stolen data either. As we see counterfeit drugs doing this, hacked data, the same thing could happen.

Maybe there's a better answer but with the rapid speed we have going out there today, data sources and credibility origins are becoming more important as they don't look at the data, they just import as it can be massive.

Again it's something to me of a concern as HIE start doing apps as well. It's not supposed to get messed up out there but it does. There is no way I would ever want this erroneous information about me being a blood thinner taker anywhere near my clinical data. If one day I did start taking blood thinners it might not be so bad, but now I don't:) I know what the clinical trial phone calls that are blind asking me if I take certain drugs, etc. and those we're stuck with but when they say they have records of me taking blood thinners on file, then I start to worry.

Again too I like what Zoeticx is doing with their platform with interops too in the fact that everyone goes to the party together and then when collaboration is done, they all go home so there's only one spot where the records permanently reside without an HIE. I think right now they are a bit in hibernation mode as they have something cooking and developing with authentication and some new use of Apple Kit coming up. Again it may require more work but if the scenario for HIEs goes away with interoperating on the fly, what a lot of money to be saved and the platform can also use and HIE as a data source as well. I read somewhere that both FHIR and non FHIR interops options were on the table to look at for methodologies, so who knows what we will end up with:)

Walter Sujansky said...

John, you make some excellent observations about the strengths and limitations of the interoperability roadmap. As you point out, the roadmap makes the case that the incentives to interoperate need to be increased and the costs decreased, but doesn't provide much specifics on how to do the latter (or, at least, how to do it differently than the approaches that have already been attempted for the past 10+ years). The roadmap does call for generally "simplifying" interoperability standards, but I think some stronger medicine is called for. I've sketched this out in a recently posted opinion piece ( and would be interested in your thoughts and those of your blog-readers thoughts on those prescriptions. Thanks for summarizing the interoperability roadmap for us!