During ICD-9 to ICD-10 codes conversion (using ICD-9 to ICD-10 crosswalk provided by CMS), I have found multiple ICD-10 codes for single ICD-9 code. Since there is no way to confirm which ICD-10 code represents best, should I use any one ICD-10 code for a particular ICD-9 code (probably the first one to be consistent)? Or I should drop the observation from my analysis? Please suggest...
Depending on what you are going to do with your data, you could just work with the lower levels of classification where there should be less multiplicity.
Thank you James Leigh . Yes I am using lower levels of classification known as CCS (clinical classification software) codes. However, still there are some ICD-9 codes I cannot decide which ICD-10 code should I consider. To avoid debate, I would like to remove those observations from analysis. Does it sound good?
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence