I’m confused why at times this difference can even be an order of magnitude or greater. (I.e 1000 patients receiving the new treatment vs 100 receiving nothing or a placebo.) What am I missing?
Wouldn’t this just lend a more favorable outcome for the drug being studied and help a manufacturer meet the statistical criteria for clinical significance? This, if true, seems unethical. Why not just make them as equal as possible from a representative sample?
In: Mathematics
You’re looking for a variety of possible outcomes, side effects, and possible adverse effects from a drug trial, but generally speaking we know what placebos do to people and don’t need a big group of people not getting the drug to prove that not getting it doesn’t give noticeable results. Having a big active pool means you’re more likely to get any given random side effect, even if it’s not very common. Having a placebo group means you have a clean sample of the same sample background to compare your active results to.
There is no reason that having more people would lead to a more favorable result. More people just leads to a more accurate result (smaller variation). So yeah more people does not mean your result will be better.
But the big reason this is done is because the “treatment” is often not just one thing. Typically they are testing many different treatments at once. In the example you gave it might be something like:
* 100 got a placebo
* 100 got a 5mg dose once a day
* 100 got a 20 mg does once a day
* 100 got a 5mg dose twice a day
* 100 got a 10mg dose twice a day
* and on and on…
Latest Answers