When selecting a cohort, I make use of Hesin and primary care data checking for ICD codes that have been issued to the participants. Checking for different diseases, they always had a code with up to 3 digits, i.e. X00.0, both in hesin and primary care. UKB also states that they are using 3 digits only to report first occurence in category 1712 (https://www.ukbiobank.ac.uk/enable-your-research/about-our-data/health-related-outcomes-data; under First occurrences of medical conditions").
However, looking for M05, M06 (rheumatoid arthritis) I discovered ICD10 codes with 4 digits. Is this a glitch or can one apply for having ICD10 codes extended for certain diseases?
I did grouping of digit for different category (first letter) for three ICD10 fields that are in UK biobank data, see attached table. 4 digit codes are present for some categories but not for all.
Comments
6 comments
It seems it has been the case: https://biobank.ndph.ox.ac.uk/showcase/field.cgi?id=41202
Or do you say that the 4 digit codes are not available?
Yes, exactly. But it's strange that it's so arbitrarily for M and not for the other codes
The ICD10 structure is also explained here: https://www.webpt.com/blog/understanding-icd-10-code-structure/
Is this 4 digit codes somehow blocking you when performing your analysis?
No, I wish I also had it for other diseases
I did grouping of digit for different category (first letter) for three ICD10 fields that are in UK biobank data, see attached table. 4 digit codes are present for some categories but not for all.
Thanks, this is really insightful!
Please sign in to leave a comment.