OHDSI Home | Forums | Wiki | Github

Mapping NHS ethnicity codes

Hello!

I was wondering if anyone here has mapped the NHS ethnicity codes to OMOP (we can’t be the first ones):

White - British
White - Irish
White - Any other White background
Mixed - White and Black Caribbean
Mixed - White and Black African
Mixed - White and Asian
Mixed - Any other mixed background
Asian or Asian British - Indian
Asian or Asian British - Pakistani
Asian or Asian British - Bangladeshi
Asian or Asian British - Any other Asian background
Black or Black British - Caribbean
Black or Black British - African
Black or Black British - Any other Black background
Other Ethnic Groups - Chinese
Other Ethnic Groups - Any other ethnic group
Not stated
Not known

E.g., should I map the first two to ‘European’ and the third one to ‘White’ or should I map all of them to ‘White’?

Any suggestions are welcome.

Cheers,
Bert

1 Like

Hello, @bertoverduin
Definitely not first. I will refer you to this discussion

I believe what you have would be considered races in OMOP: chech these codes

Edit: I believe you can map values 1-3 to white. Should not break anything

@bertoverduin:

We are preparing a release of the new race-ethnicity combo vocabulary. It can contain any proposed vocabulary with those concepts, including yours. SNOMED actually might have a lot of them. Since these concepts aren’t externally defined, but socio-biologic concepts self-assigned, we won’t be able to deduplicate them. It will be up to the analyst to pick whatever is deemed relevant to a research question.

“Not stated”, “Not known”, “Prefer not to say”, “Leave me alone” and all those are flavors of null and will not appear.

Works?

Are we keeping the current, standard race_concepts_ids? Or will everyone need to remap this domain?

They will be there. Except we will exchange the concept codes. Apparently, we used the wrong ones (with hierarchy in it or the other way around). But the concept IDs and of course the concepts themselves will stay the same.

Correction. I was mixing them up with the CDC codes. NHS codes have not made it in the Vocabs so far, we would have to add them. It’s not that many.

Hi @bertoverduin,

Are you mapping for an NHS trust?

I am part of a team that is converting our data in prep for the trusted research environment currently in development. We are keen to find out who else in the NHS is moving over this the OMOP CDM to facilitate their TRE (or other) work.

Especially to help us/each other build our business case/funding bids when demonstrating improved future interoperability across regions when a federated data network is developed by NHSE/Digital

Dale

Hi @Dr_Dale

Do note that a number of EHDEN certified SMEs are working with UK Data Partners on ETLs - for instance have a look at https://www.ehden.eu/business-directory/, and Data Partners – ehden.eu

Our 7th and final Data Partner Call will be from next month, and not promoted yet, but we’ll be collaborating with HDRUK on this for UK Data Partners.

Let me know if you want to know more about any of the above points (nhughes@its.jnj.com)?

Nige.

1 Like

Hello @Dr_Dale,

We are doing the mapping for Great Ormond Street Hospital (GOSH).

Cheers,
Bert

1 Like

Hello @Christian_Reich,

It would be great if they could be added. The list above is the complete list of NHS ethnicity codes. See also: ETHNIC CATEGORY.

Cheers,
Bert

Hello @zhuk,

Yes, I think I will map to “White”, “Asian”, and “Black or African American” where possible. I’m not sure what to do w/ the “Mixed” ones though, and the “Any other ethnic group”.

Cheers,
Bert

I am afraid that you have to map it to 0.

Ha. @zhuk is right, except there is no such a thing as a map to 0. You just don’t have a map.

Here is what we will do: We will add these codes as standard concepts, so no mapping whatsoever (except to themselves). We will not deduplicate, because it is not possible for the lack of clear definitions. These attributes are self defined, and there is no way to determine if “Black” is the same as “Black or Black British” or “Black or African American”. Since they are constructs based on phenotype, language, culture, social status, place of origin of ancestors and religion, most likely they are not.

Thanks @Christian_Reich, much appreciated! Just for my information: when can I expect this approximately to be done?

Сouple weeks. That’s for the proposal. It then has to be ratified by the health equity WG, and then it has to go and be released into Athena. Does that work?

@Christian_Reich Yes, that should be fine. We are just starting the mapping, so actual implementation will be several months away …

Hi Nige,
I would be very keen for information - I have infact already contact EHDEN via the website asking for details on the next Data partner call as we are focused on ensuring the sustainability of the projects by multiple means and routes.

Thanks for the directory link - I didn’t realise you could get partners contact details via the site I was only aware of the organisaitons that have already partnered.

I’ll follow up via via email too - Thank you

t