Medicine circles tout the holiness of RCTs. Small differences can be picked up, they say, and you can design drugs and vaccines much better. While the Math behind them is sound, RCTs are hardly the gold standard they are touted to be.

If you do RCT of insulin injections on the whole population, RCT will show it is a great idea, because some people (the diabetics) really will be helped by insulin. But such an RCT will now suggest EVERYONE be given insulin injections.

If you don't target your specific subgroup (diabetics), you basically overdo your therapy. It is statistically valid, but subjects a lot of people to unnecessary treatments.

The central problem is that small differences can be statistically significant, and an RCT will pick them up. It will then recommend a therapy for EVERYONE. While insulin injections will do no harm to the non-diabetics, they are completely unnecessary for them.

Most vaccinations are for diseases which are rare. Instead of targeting the small number of people who are vulnerable and susceptible to a disease, vaccines are given to the whole population. This is if you believe virus/vaccination theory to be true (which I don't anymore, by the way). It is just statistical trickery passing off as science/medicine. In common parlance, you are burning the house if you find a rat in it.

Whenever only a small percentage (less than 10%) of a group has a disease, it is unfair to subject 90% to a treatment or vaccination. There is a cost of doing interventions, of hassling a large fraction of the population to save a few. One must try to figure out with is common in this subgroup, and do the treatment or vaccine trial on this subgroup. When it is less than 1%, it really is bad medicine to subject the 99% to something which they will not even be effected by. Many vaccines are approved for just 0.1% or 1 in 1000 benefiting. You give the vaccine to 999 people to save 1.

If you cannot identify a group where at least 10% of the people come down with a disease, you really need to work on your research more. The presumption of innocence in medicine translates to presumption of not being vulnerable to a disease, to be immune to it. You must prove that the person has a good likelihood of being vulnerable (in this case more than 10%) before making them take a treatment, specially vaccines which are supposedly preventative.

This is in addition to the safety profile of a treatment or a vaccine-where you may actually be harming the non-vulnerable by your treatment. Basically don't assume that insulin injections do no harm to non-diabetics.