When doctors talk shop about medications it is frequently in statistical terms.  I remember over hearing one such conversation about an extremely common post-heart surgery drug many years ago that was stopped when statistically it was found to be far more deadly to give it than to not.  The surgeon was saying "Now I don't know that it killed anyone, but I gave it to so many patients it must have."

They use terms like "We must give this drug to c number of people to cure one" and if c is low enough, or cheap enough, or if the "we must give this to k number of people to kill one" has a high enough k, then the drug might be deemed worth giving.

Vaccinations can kill.  Make no mistake.  Hell, just poking the skin with a syringe, statistically speaking, has killed someone.  And of course if that infinitesimally small number comes to rest on your loved one, no amount of charts is going to kill the pain.  But the evidence is pretty solid that the risks are small and the risks of failing to act are dire.

In fact, we have to reset our measles clock for the first time in twelve years.