Feature Request for ICD-9 about drgpy HOT 5 CLOSED

Thanks for the interest, @drrckln! I probably need to look at the details of the data files to assess the level of difficulties, but allow me to suggest a quick and easy solution for your problem.

You can convert ICD9 diagnosis/procedure codes to ICD10 using General Equivalence Mapping (GEM). Please see

• ICD9 dx to ICD10 dx
• ICD9 procedure to ICD10 procedure

After converting ICD9 codes to ICD10 codes, and then you can apply this drgpy to get the corresponding DRG? Let me know if this solves your problem. Best,

from drgpy.

Ah, that could work. Thanks! I saw this quote at some point:

"From a recent study conducted by 3M comparing the ICD-9-CM grouper to the ICD-10 grouper based on recoding patient records, they estimated that slightly more than 99% of the cases showed no change in MS-DRG when coded in ICD-10."

But I cannot recall which document I pulled this quote from, and it seemed to be from around the ICD-9 to ICD-10 switch, eg ~2014. It looks like CMS took pains to keep MS-DRG groupings stable then, but it's unclear to me if that's still the case 5 years later.

If I take 2010 dx/pcs codes, convert them, and apply drgpy, do you know if it's still 99%+ likely to have the same MS-DRG assigned as it was originally in 2010?

from drgpy.

I discovered that CMS's freely released MS-DRG grouper software applies the correct grouper version based on claim date. It runs on Windows, but is usable in interactive or batch mode.

from drgpy.

Yes, you mean the Java application: https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/MS-DRG-Classifications-and-Software.html
I think there are a couple of other interfaces that you can download as well.
This package is developed to easily integrate into other Python analytics stacks. If you are okay with running as a separate application, I recommend using those applications out there.

Regarding your question, I don't know if it will maintain 99%+ match with the conversion with drgpy, unless I have some data to compare to; sorry for not providing much insight on this.

from drgpy.

Closing the original issue as it can be addressed with other packages.

from drgpy.