Drug relevant basis here is defined by whether or not the color of one's skin alters the efficacy of a particular drug.
Sickle-cell anemia is perpetuated by a trait that appears with strong correlation to the traits that create dark skin pigmentation. IE) Black people are more likely to carry the sickle-cell anemia trait. However, the color of a person's skin does not contribute to the sickle-cell anemia pathology in any way, it is merely correlated to the presence of sickle-cell trait. A light skinned person with sickle-cell anemia responds to treatment in the same manner as a dark skinned person, the dark skinned person is merely more likely to develop the pathology due to genetic correlation.
The same logic applies to this case-study, if it were not compounded by various other idiosyncrasies. The color of a person's skin (what we use to empirically determine race) does not change the effectiveness of most drug agents, and therefore is not biologically relavent in a pharmacological context.
Sickle-cell anemia has very little to do with race and everything to do with protection against malaria - http://www.sciencedaily.com/releases/2011/04/110428123931.ht.... That people with darker skin live in areas where malaria is present is just a coincidence.
Try telling that to people with sickle-cell anemia.