OHDSI Home | Forums | Wiki | Github

Charlson comorbidity index

(Mollie McKillop) #1


Does anyone know of resources or has previously mapped SNOMED codes to the relevant conditions needed for calculating the Charlson comorbidity index? Most resources and examples for calculating the index use ICD codes, but I am reluctant to map OHDSI SNOMED codes back to ICD codes as OHDSI, as I understand, will map multiple ICD codes to a SNOMED code, which creates ambiguity if mapping back to ICD codes. Thank you for your thoughts.

(Martijn Schuemie) #2

Hi @mmmckillop!

We have an implementation of the Charlson index here in the FeatureExtraction. I didn’t create it myself, but as you can see in the SQL code it uses the OHDSI Standard concepts.

(Christian Reich) #3


The Charlson Index is defined by the prevalence of the diseases, not ICD codes. Unless you want to go awry and calculate it using source codes for some reason I would take @schuemie’s version.

(Mollie McKillop) #4

Thanks yes I agree I was hungup on how to chose relevant codes (e.g. Moderate to severe CKD) and previous sources I had looked at in the literature all used ICD but since it has been done here I will forget the source codes and use these OHDSI codes.

(Mollie McKillop) #5

Hi Martijn,

This is a great help thank you so much.