Some might think it not ethical to withold medicine from a sick person.

Well, maybe I should have put "ethical" in quotes, and replaced it with legal. Whether or not you think certain actions are ethical carries a lot of baggage that we don't need to dredge up here. But there are copious rules and restrictions that govern who is and is not eligible to participate in clinical trials for experimental drugs which will introduce biases. Some trials have age restrictions, some restrict giving placebo treatment to people whose disease has advanced beyond a certain stage (particularly in cancer trials), etc. . .

If it were completely random then all members of the population would be equally likely to be chosen. To select only those who apply through the paticipating clinics would introduce sample selection bias. Everyone would have to have equal probability of being chosen AND would not be able to refuse selection because again there would be sample selection bias. Therefore, if you were chosen then a doctor or nurse would have shown up at your door three times over the year to give you an injection of either the medicine or a placebo whether you protested or not. That is an ethical issue.

JabairuStorkQuoteCertain groups are more or less predisposed to genetic disorders of a particular type, but there is no evidence that this is true for infectious agents such as virusesI have heard of one group of people, staying in marshy backwaters of kerala, india. These people have sickle shaped Red blood cells, as against nice round ones in most other humans. These people never get maleria. It seems maleria cannot breed in sickle shaped RBC.Shailesh

I have heard of one group of people, staying in marshy backwaters of kerala, india. These people have sickle shaped Red blood cells, as against nice round ones in most other humans.Except of course black ones...Sickle cell anaemia is a depressingly common illness in black people. The malformed cells cause intense pain and death.

But the condition is as common as it is in West Africa precisely because the majority of sickle cell carriers are asymptomatic and resistant to maleria.

My understanding is that if you inherit from one parent, it is not life threatening, but two spells trouble.This is possibly going to become a eugenics issue. The african variant clearly evolved where malaria was such a problem that the cost worthwhile in evolutionary terms. However, people of recent african origin are by no means constrained to malarial countries, and of course are breeding with people who have no recent connection with africa at all. Thus the number of people with the gene is appratently going up rather quickly, even in areas of the world where malaria is unknown.

The original study is an example of data mining rather than flawed design. The effectiveness on blacks was based on 9 of 111 infections for the placebo group versus 4 out of 213 for the group with the vaccine. That is significant at the 2% level. But in a study of 5,000 you expect many subgroups of this size to show significant results.VaxGen claims (a) the subgroups were prespecified and included in the significance calculation and (b) there is biological supporting evidence (higher antibody levels in vaccinated blacks). But the company has not released details to support these claims.It's not implausible that there will be racial differences in effectiveness, although a 20 to 1 ratio is highly improbable. Also, the average person classified as "black" has 20% caucasian genes, making things even more muddled.The real test will be the FDA filing at the end of March. But the upbeat press release based on what is at best a hopeful indication in a bitterly disappointing trial overall, was bad public disclosure to investors and bad medical practice.