Although the purpose, as per disease, is common in some cases of tests, in many others it was not as specific as to the purpose of the test and it was to get a number for something that was generalized..
And with that in mind, if you take the number, lets say, 1000, and divide that in half, then you administer a test, and you use the outcome to determine the impact on the entire population of the planet, I doubt that would be a reliable outcome, why?, because when you deal with large numbers the variables are very unstable, just to take that a step further, if the numbers were actually taken to infinity, the results would be moot, you could take a group of people from the other side of the globe, and do the same test and get totally different results due to many many reasons.
So I still am not sure about the parameters set for "Placebos" and the results they give us.
Just some thoughts from beside the bay on a nice morning

The link below is an article on the Placebo Effect
http://skepdic.com/placebo.html