Skip to content Skip to navigation

Will ICD-10 Complicate Research?

February 7, 2014
| Reprints
How will ICD-10 impact research given older records are not likely to be converted?

I had a conversation with a vendor this week regarding pricing for ICD-10.  The focus of the conversation was around how to price the change to ICD-10 in a Radiology Information System (RIS)/Picture Archive and Communications System (PACS) environment.  While I don’t portend to be an expert on ICD-10, the discussion did raise a real concern for me: What is the impact of the change on the ability to use ICD coding information in research?

Clearly companies have had to invest time and effort to make the change within their applications, and need to recover some costs.  At a minimum companies should expect to recover the cost of implementation in terms of project management and implementation resources involved in software updates.  What I learned from this vendor is that there is no plan for any form of data migration, which means all the old reports will continue to be coded as ICD-9.

That got me to thinking about the implication for general searches, research, or clinical trials.  One of the advantages of ICD-10 is purported to be that it is more granular in terms of diagnostic code accuracy.  If that is the case, then new reports going forward will not match with older reports using ICD-9.  This will limit comparisons to the use of the old ICD-9 codes, which means research comparisons won’t be as granular as they could be.  True, going forward they could be, but wouldn’t it be ideal if comparisons could be made to the ICD-10 codes?

Clearly it would be a monumental task to convert all ICD-9 to ICD-10 codes, and beyond economic feasibility.  But perhaps there are some shortcuts that could be taken, such as only selectively updating codes for specific studies.  Could there be algorithms generated that would enhance the ability to convert codes?  Or, am I fixating on a non-issue? 

Another issue which a colleague raised to me is the learning curve for ICD-10.  Many physicians today can rattle off ICD-9 codes at will.  What will be the learning curve to remember ICD-10 codes, and what impact will they have on physician productivity if they have to look them up? 

I’d like to know what you think.  Should there be some analysis of cost/benefit within research facilities to substantiate conversion?  Does this have other implications or applications beyond research?  Is there an opportunity for ICD-10 learning aids or lookup applications?  I welcome your thoughts.