Friday, February 17, 2012

On the ICD-10 Delay, ICD-11 and SNOMED CT

I was asked what I thought about the ICD-10 delay on twitter yesterday, and it is the topic of an #HITsm chat today. Tom Sullivan over at Health IT News writes about how this might set the stage for use of ICD-11. I wrote about ICD-10 and ICD-11 and SNOMED CT a few years ago.  In response to that post, Chris Schute at Mayo, who is also the ICD-11 Steering group chair told me about how they planned on ensuring that ICD-11 would be linked to SNOMED CT.  Details are in this 2008 report to NCVHS. You can even read about it on Wikipedia:

SNOMED CT is currently used in a joint project with the WHO as the ontological basis of the upcoming ICD 11.

Now I have to tell you, I'm of two minds about the whole idea.  Because of the timing for ICD-10, Health IT vendors have invested quite a bit of effort into it already.  A lot of that work would need to be tossed out and redone.  Not so much for field widths, but dealing with mapping, lookup, et cetera.

On the flip side, using an ontology for billing that is strongly linked to one that is used for providing clinical care will (I believe) really improve care.  It would also provide for much greater interoperability between clinical (e.g., EHR and CDS) and revenue cycle management applications.  And there's the whole "quality measurement" set of issues regarding EHR vs. Abstraction workflows that we were discussing yesterday in the Clinical Quality Workgroup that could be dramatically improved too.

The idealist in me shouts "Yes! Let's use ICD-11".  The pragmatist in me says (more quietly) to move ahead with ICD-10, perhaps a bit more slowly because of all the investments made by organizations thus far. Right now, the idealist is winning, but my guess is that he's likely to be disappointed.

The benefit of being of two minds on this topic is that no matter how it turns out, I'll win.  The disadvantage is that no matter how it turns out, I'll also lose.

2 comments:

  1. What a terrible thing to consider when making policy (although, sadly the stark reality): what's the investment companies have made in it.

    In business they'd call this a sunk cost and choose the best path forward. Unfortunately politics doesn't quite work that way and you're right that it's quite unlikely that the jump to ICD-11 will happen based on this very reason.

    Great analysis.

    ReplyDelete
  2. Hi Keith, Tom Sullivan here. I'm siding with the idealist in you. When I started covering ICD-10, I was a proponent. But given that ICD-10 really is antiquated -- as Dr. Chute explained to me on Monday it's based on thinking about medicine and technology from the 1980's -- it seems a shame for the U.S. to spend literally billions getting there, only to reach what is essentially a classification system at least approaching the end of its lifecycle, or perhaps what ought to be the end of its lifecycle rather than the beginning. And it's not just the linkages to SNOMED that make ICD-11 attractive (though they certainly do); a system that is semantic web-aware, with rich information spaces, definitions of terms, genomic underpinnings, a foundational fabric, and I'm sure there will be more in there, just makes so much more sense than ICD-10.

    At any rate, I view the delay as an opportunity to re-evaluate our approach and maybe, just maybe, put politics aside and opt for the system that has the most potential to improve public health with data.

    So perhaps adopting ICD-11 is not really idealist, after all, but pragmatic!

    ReplyDelete