Edit: ICD Revision Number, Cause of Death

[Jim Hofferkamp]

ICD Revision Number, Cause of Death (SEER IF37)and ICD Revision Number, Cause of Death (NAACCR)
-The only difference between these two edits is the NAACCR version allows blanks and the other version does not. The NAACCR version is only in the CINA edit set. The other version in the NPCR Edit Set and the Combined Edit set.Both versions of these edits are working correctly.

Don’t be surprised if you find some invalid cause of death codes that weren’t there last year when you run your submission file through the call for data edits. The tables for this edit were updated just after last years call for data. The updated edit will catch invalid codes that were allowed to pass last year.

The updated edit was included in the v15a metafile, but if you haven’t run older cases through edits since last years call for data they wouldn’t have shown up.

One other thing related to a statement in the description that says:

This version of the edit allows 3-character ICD-10 codes to be coded with either a 9 or a blank in the fourth character.

Most ICD 10 codes are 4 characters long. However, there are some that are only 3 characters long. They don’t have a 4th character. The edit allows the fourth character to be blank or 9 for these codes that are only 3 characters long. If the code is 4 characters long, then the edit will not allow a 9 or blank in the 4th character unless the code truly ends with a 9.

For example:
The ICD-10 code of K809 and J699 are not valid codes and will fail the edit. Both K80 and J69 have fourth digit definitions but ‘9’ is not a valid code for the fourth digit. The ICD 10 code of J80 is only a three-digit code – there are no fourth digit codes defined – in that case the edit would accept a blank or nine in the fourth digit.

I’m going to take the confusing language out of the version of the edit in the NAACCR metafile. However, the NPCR edit set has already been released so it will still be there.

NPCR is aware of the issue.

[Recinda Sherman]

When Jim and I were investigating some cases that generated errors this year, there were a few plain incorrect codes. However, we found a number of correct ICD codes but they were procedure codes, not diagnosis codes. I suspect that the edit was updated so that procedure codes, regardless of whether they are valid ICD codes, are not allowed in the cause of death fields. Only diagnosis codes should be used in the cause of death field.

[Author]

Posts