Don't look at any double-blind studies for cancer medication!

Of course it can, as long as the sample is truly random, you just need a really strong result with a population this small.