Guest Post: Two on a match; the problem of CDI and coder DRG mismatches

by Linda Renee Brown, RN, MA, CCDS, CCS, CDIP

Recently a question was posed in ACDIS CDI Talk about establishing CDI/coder DRG match rates. (Read the related Blog post.)  Shameless plug here; if you do not visit CDI Talk, you’re not taking full advantage of your ACDIS membership. It’s the best national networking opportunity you’ll have outside the annual conference.

Anyway, being the shrinking violet I am, I jumped right in with my two cents (before taxes):

“I do not have a CDI/coder DRG match metric in my program. While I do expect the CDI specialists to understand MS-DRG and produce a working DRG, and I believe fervently in the importance of reconciling the DRG so that the most accurate codes are reported, I don’t actually care if the working DRG matches the final DRG. I do care that the CDI specialist understands how the DRG is impacted by the documentation and how changes in the documentation may impact the DRG. I do care that the CDI specialist and the coder come to an agreement that the final billed DRG is the correct representation of the documentation. But for me, evaluating a CDI specialist on his or her ability to see into the future leads to a lot of wasted effort on the coding process that should be spent on the documentation process, and generates a lot of competition between coding and CDI that should never exist in a collaborative environment.”

When I was a CDI specialist, the match rate stressed me most. I’m not a coder. Our differences make us both uniquely awesome and a great complement to each other. My CDI team includes both RNs and coders and that’s the way I like it. Most CDI specialists are nurses and will never be coders. And even though some CDI specialists are coders, that’s no longer their role.

The role of the CDI specialist is to identify and correct existing documentation issues and to prevent future documentation issues. It’s not to code the record. As a CDI specialist, I assigned many, many codes, and calculated many, many working DRGs. For me, this served to clarify how the documentation would impact the coding, and how a query would impact the coding, so I thought the process was important. In the end, though, the DRG I assigned was irrelevant; the final DRG was the coding specialist’s determination.

We didn’t always match. Sometimes documentation would come in after my final review that affected the DRG. Sometimes it was a matter of picking between two diagnoses that both met the definition of principal diagnosis; I picked one, they picked the other. Sometimes I just saw things differently than the coding specialist did.

Sometimes I accepted the coding assignment as a valid alternative to mine. Sometimes I would disagree with coding’s determination, and then began the process of reconciling the difference.

Sometimes they agreed with me, sometimes they didn’t. The reality is that it was their name, not mine, going on the coded record, and they would have to stand behind their coding assignment, so they always had the last word.

Every record that I couldn’t or didn’t try to convince them to change to my working DRG would go against my match rate. Because of this, I spent a grossly inordinate number of hours struggling over DRG assignment when I could have been moving on to another chart and finding gaps in documentation. I recall looking up a lot of complex procedures on the Internet to try to find the exact code so I could get them into the right DRG, when none of the operative reports had any actual documentation issues.

Because I knew it negatively impacted my metrics when the coder didn’t select my DRG, I spent significant time taking some cases personally and trying to prove they were wrong. My ability to duplicate their work really had nothing to do with my ability to effect clinical documentation improvement. Resenting a coding decision does not generate CDI-coder collaboration. Ever. And if we need anything in this job, it’s CDI-coder collaboration.

If a working DRG differs from the final DRG, by all means, as the CDI specialist, I would investigate why they differ. Errors can be caught and corrected. The discrepancy may afford an educational opportunity for the CDI specialist or the coder.

If the CDI specialist never matches the final DRG, they probably need to review principles of MS-DRG. But as the director, I never plan to include my team members’ ability to predict future documentation or read someone’s mind in my assessment of their effectiveness as CDI specialists.

Editor's note: Brown, at the time of the article's original response, was an independent CDI consultant based in Carrollton, GA. With experience in critical care, nursing education, disease management, case management, and long-term care, she has worked as a CDI specialist, educator, director, and consultant. She is a frequent writer on topics involving clinical documentation and published her own "The Case Manager's Quick Guide to Diagnostic Related Groups" in 2013.