[skissane]

Do any of them really cluster genetically? Compare a person whose ancestors come from Senegal to a person whose ancestors come from Mozambique – Americans will call both persons "Black", yet I would not expect those two people to have many recent ancestors in common, or much "genetic clustering" at all.

[sseagull]

You are thinking too literally. It's just a question that opens the door to more discussion. It's another piece of information, not the end of the discussion.

[skissane]

But does it always "open the door to more discussion", or do some medical professionals just make judgements based on which box was ticked on the patient form? Presented with an African immigrant patient, is a US doctor going to realise that the patient may have rather different genetics from most African-Americans, and hence information about health risks for African-Americans may not be relevant to them? I'm sure an above-average doctor would realise that and take it into account, but would a below-average doctor (of whom there are very many) do it?

[sseagull]

No not always. But overall more information is better than no information, especially when spread across an entire medical system.

[skissane]

If more information is better, isn't that a good argument for replacing "race" questions which expect coarse-grained answers with "ethnicity"/"ancestry" questions which expect more fine-grained answers? (i.e. "Chinese" or "Filipino" not "Asian", "African-American" or "Kenyan" not "Black", "English" or "Polish" not "White", etc.)