There's a new drug for African-American heart patients (found at Volokh). There's something right about this, and there's something a little dangerous. The drug had been tested and abandoned in the 80s due to little success among most people. At some point someone figured out that it had a better effect among black people, and studies to confirm this showed good results. What's most likely going on is that a genetic trait common among Americans of recent African descent (one suggestion has to do with nitric acid levels) allows this drug to be more successful with this particular kind of heart condition, just as some other kinds of medication are less successful in the same population. It doesn't seem related to the genes for pigmentation. Skin color may be a good guide to seeing whether someone has the relevant trait for the drug, but enough race mixing has gone on in the history of this country, and enough immigration of people who look enough like Americans of African descent but who are genetically very dissimilar, that prescribing such medicines according to racial identification is a very bad idea.
We've encountered other problems in tendencies to diagnose according to race. Ethan had an anomaly in his hemoglobin readings that, because of his black parentage, the doctor assumed might have something to do with sickle cell. Now Sam's family had never heard of sickle cell until coming to this country. The African slaves who were taken to Barbados apparently didn't have the mutation protecting from Malaria, or there were few enough of them that it's an extremely rare trait there. They wouldn't have assumed this except for her skin color, though. It led them on a wild goose chase trying to make sense of some conflicting information, and it turned out to be something rare but completely harmless, as far as I could tell.
This isn't a serious problem, but it was an assumption based on skin color that was unwarranted given that Barbados isn't in the high-risk group for sickle cell. When it comes to which heart medication to prescribe for someone, you better have more to go on than a statistical correlation between most Americans with a skin color and which medication works better, especially because African-Americans in Georgia are probably genetically closer to me than they are to many immigrants who look more like them but come from a part of Africa that American slaves weren't taken from (never mind other dark-skinned peoples, such as aboriginal Australians).