OHDSI Home | Forums | Wiki | Github

Mapping EGFR - male & Female

Implementers
, ,
#1

Hello Everyone,

I had a look at our hospital database and found out that we have the below measurements

Estimated Glomerular Filtration          ml/min    
Estimated Glomerular Filtration Rate     ml/min    
Estimated Glomerular Filtration, Female  ml/min    
Estimated Glomerular Filtration, Male    ml/min

Am I right to understand that terms with Male and Female are being very specific or the doctor/data entry was made in such a specific manner.

Whereas, if we look at Estimated Glomerular Filtration or Estimated Glomerular Filtration rate they are no different from Estimated Glomerular Filtration, Male or Estimated Glomerular Filtration, Female. Meaning they all are exact same measurements and difference is only in gender value.

So, even if I map all of this eGFR, that should be fine?

We can any find the gender info from the patient demographic.

Am I right to understand this? It’s just the level of info that was keyed in the EHR system?

#3

Hello Akshay Kumar!

Yes, you’re right. Mapping to plain Estimated Glomerular Filtration or Estimated Glomerular Filtration rate will be OK.

#4

What do you mean by “they are all exact same measurements and difference is only in gender value”? I would be careful about mapping all of these values to the plain “Estimated Glomerular Filtration” without ensuring that these values aren’t adjusted based on the sex modifier. Commonly, eGFR values are based on creatinine and c-cystatin measurements and adjusted based on age, sex, and race.

Depending on your data source, you may have multiple of these values populated for one lab test, not all of which are adjusted appropriately for the individual.

For instance, in our dataset, all eGFR labs exist as two values, both returned each time the lab was drawn, regardless of the patient’s race:

  • concept_id 37061620: Glomerular filtration rate/1.73 sq M.predicted among blacks [Volume Rate/Area] in Serum, Plasma or Blood by Creatinine-based formula (CKD-EPI)
  • concept_id 36303797: Glomerular filtration rate/1.73 sq M.predicted among non-blacks [Volume Rate/Area] in Serum, Plasma or Blood by Creatinine-based formula (CKD-EPI)

I would recommend figuring out which adjustment formula is being used for your sex-specific eGFR values (CKD-EPI, MDRD, or other), and mapping the “Male” and “Female” values to the appropriate LOINC concepts.

It is also worth noting that the use of race-based adjustments for eGFR is controversial: https://jamanetwork.com/journals/jama/article-abstract/2735726

For this reason, we retain both records in our warehouse and make explicit on a study-by-study basis which values are used for which patients. I am currently looking into other ways that we can handle these values in our data warehouse to provide more accurate and equitable estimates.

#5

Sounds right, @Akshay.

t