Six years ago I worked on a program to implement an advanced system of clinical terminology for
the National Health Service in England. The exciting new terminology was called SNOMED-CT. The “old” terminology was ICD-10. I remember the terminology advisory meetings with clinicians and administrators about the limitations of ICD-10 and how SNOMED-CT would improve health informatics in the NHS. I have deja vu when I see similar arguments being made for implementing ICD-10 in the US,(see the HIMSS Open Letter to HHS published Feb 7th), The push for standardized terminology is important for our healthcare system, but it is essential that this nation-wide mobilization around terminology infrastructure be able to deliver the promised benefits. It is becoming increasingly clear that ICD-10 is not the breakthrough we need.
I returned to the US in 2006 and worked with GE Healthcare and its development partners to develop more advanced knowledge solutions for health care. In my consulting practice I have focused on patient engagement and this has given me a new perspective on the need for actionable and expressive clinical terminology. This is a time to carefully consider the consequences of our choices.
I believe that the federal mandate to convert to ICD-10-CM by October 2014 will be a step backwards. Here’s why-
- Credibility,once lost is hard to regain. The advertised benefits of ICD-10-CM migration are grossly overstated. Just as many of us experienced in January 2000, the recollection of the pain to continue business as usual beyond Y2k was like having recovered from the flu. After the pain of implementing ICD-10, no one will want to touch clinical terminology infrastructure again for a generation. The hyped ICD-10 benefits of eliminating waste, population health, cost savings, patient experience will be revealed as empty promises. The arguments in favor of keeping the October 2014 deadline stated in the HIMSS letter are an example of wistful thinking and will not stand the light of day once the real pain of the migration begins.
- ICD-10 is not worth it. The April 2012 Health Affairs article by Chris Chute, Stan Huff, John Halamka,and others presents compelling evidence that “the practical ability of ICD-10 to capture content typically contained in clinical records is not measurably better or worse than that of ICD-9-CM”. They conclude that “ICD-10 CM conversion is expensive, arduous, disruptive, and of limited direct clinical benefit”. The authors of this analysis are among the most respected health informatics thought leaders in the world. The calls for delaying the implementation fo ICD-10 is not a rear-guard action by change-resisters but an appeal to reason by some of the most effective health change agents of our time.
- ICD-11 will be better because of its intrinsic alignment with SNOMED-CT. The investments made by the US healthcare industry to date in preparing for ICD-10 are not entirely wasted if transformed into an infrastructure that supports ICD-11 prior to its deployment. Let’s avoid having to do another big terminology upgrade in a few years- or worse, being stuck for another 30 years with a labor-intensive administrative classification ontology in a world that needs clinical relevance. Please, let’s not populate patient databases with ICD-10 terms that only add complexity and provide minimal benefit to patients. A migration to ICD-11 and SNOMED-CT, properly orchestrated with ACO’s, payers, clinicians, and especially patients, offers a more credible approach that will spark innovation. Let’s take it.
That is why I believe the ICD-10 terminology conversion should be re-envisioned with an additional two or three more years so that it becomes worthy of the investment. With all due respect to the hard work of those who have been implementing ICD-10, let’s get it right this time. We may not have another chance for a generation.


