Wednesday, April 30, 2014

Right almost by accident: The JASON Report on HealthIT Infrastructure

I've finally read through the most recent JASON report on A Robust Health Data Infrastructure.  It fails to live up to one pundit's description as the "Son of PCAST" widely reviled by the Healthcare industry.  It failed to live up to its billing.  The PCAST report had some semblance of professionalism about it.  However, the JASON report is amateurish by comparison, full of outdated references, pedantic writing and unjustified opinions.  Even so, its more right than wrong, but probably for the wrong reasons.

My favorite quote from the most recent JASON Report is:
Innovation in health care appears to be frozen by a deluge of overly ambitious, insufficiently practical, and often conflicting advice
Never have I seen such a well qualified self-referential statement statement in a report before, and I have to completely agree with it.  In this case, I'd also have to add confusing advice.

The report relies on out-of-date references in citing the slow growth of adoption of electronic medical records by healthcare providers.  We are in the midst of a technology revolution in terms of adoption.  In four short years, the US has emerged from the bottom of the pack of all nations with respect to Health IT to being near the top, and if the pace continues we will shortly be at the top.  With such rapid change, there is no question that there will be fits and starts and growing pains, and that we won't have gotten it right in the first try.  The critique that we aren't moving fast enough:
The level of interoperability set forth through the CMS Meaningful Use criteria, as a result of the HITECH Act, is too low to drive meaningful progress
Fails to indicate what pace is fast enough, and how we could expect to achieve that pace.  Meaningful Use is starting its second cycle this year.  Systems conforming to the 2014 criteria are being deployed and used and haven't even been in place for half a year.  And yet somehow, we've already learned enough to state that Meaningful Use stage 2 is also a failure.
The criteria for Stage 1 and Stage 2 Meaningful Use, while surpassing the 2013 goals set forth by HHS for EHR adoption, fall short of achieving meaningful use in any practical sense. 
I would like to see what the authors define to be meaningful use and how they can expect a program that has been in place for three years has failed so significantly.  The first two-year stage jumped the national adoption level of Health IT from 30-40% to nearly 80% according to the CDC.  In four short years we've bent the adoption curve faster than any first world country.  And this is too slow.

And once again, we need "scientists" to tell us:
With respect to data formats, the current lack of interoperability among the data resources for EHRs is a major impediment to the effective exchange of health information. ... However, simply moving to a common mark-up language will not suffice. It is equally necessary that there be published application program interfaces (APIs) that allow third-party programmers (and hence, users) to bridge from existing systems to a future software ecosystem that will be built on top of the stored data.
But who fail to recognize that Blue Button Plus is in fact a published API to do just that.  No, to them it's a business model that has yet to succeed (see page 16 of the report).

And what is Blue Button Plus based on?  A standard API (called FHIR) that allows third party programmers to bridge from existing systems developed by one of the hundreds of standards organizations that could only be mentioned by name ONCE in the entire report (HL7), and not even in the context of organizations developing standards.

And while the report remarkably makes mention of connect-a-thons (using the NFS spelling), the authors remarkably make no mention of any current such events occuring (such as a couple of weeks ago in Vienna), or next week in Phoenix (although some might argue that's not quite a connectathon yet).

Of course the report is full many other opinions and absolutes such as:
Current EHR systems do not interoperate at all, and in many cases are unable to even exchange data between hospitals running the same system from the same vendor.
I could spend the entire day ripping the rest of the report apart, but I won't, because there are two things it got right (perhaps for the wrong reasons).

The JASON report fails to make a distinction between the lifetime longitudinal Electronic Health Record, and the electronic medical record systems, payer databases, care management and other Health IT systems from which the EHR will become an emergent property.  But, it properly recognizes that:
EHRs should not be things that one buys, but rather things that evolve through cultural change aided by technology
But fails to distinguish between the components of the EHR (the Certified electronic medical record systems confusingly called EHRs in the Meaningful Use program), and the emergent property that will be the EHR supporting patient care, population health and clinical research.

And secondly:
The architecture must be based on open standards and published application program interfaces (APIs) and protocols. 
There is work led by HL7, and supported by IHE, DICOM and others under development over the last three years, known as Fast Healthcare Information Resources (FHIR) that supports the API that JASON is looking for.  If only someone on the advisory committee had mentioned it to them (or perhaps they did, but it was remarkably unreferenced in the JASON report).

I beg ONC that when they next commission a report, be it from PCAST, JASON or whomever else, that they include on the team of advisors someone who can point the team to current standards work so that it can at least be evaluated, and also give them some up-to-date references, so that they don't embarrass themselves by referring to data that is woefully inadequate.

And, yes, ONC please continue to support HL7 efforts on FHIR, because I truly think that's the open API that you are looking for.

1 comment:

  1. Yes, definitely an example of throwing a lot of mud and some of it sticking. As an Application Architect, that particular section and the models caught my attention. Pretty old hat to anyone who's been designing distributed software applications over the past 15 years based on the Microsoft Patterns and Practices - but maybe new to many in the health sector? Either way, rather than looking for a single 'silver bullet' API, maybe mapping frameworks such as FHIR and openEHR to this architecture might facilitate better take-up of interoperability standards from the world-wide eHealth application development community. The nearest thing to this that I've seen has been RIMBAA - but that stands, or fall, on the RIM...