> Its almost impossible to find out what happens after long time, say a decade, with the current approach. Most vaccines are much older than a decade. We know

> Its almost impossible to find out what happens after long time, say a decade, with the current approach. Most vaccines are much older than a decade. We know who gets vaccines; it's in their medical record, if not centralized anywhere. Longitudinal studies aren't impossible. So any random hospital could do a longitudinal retrospective data-collection study on the consequences for their own patients of either having received a given vaccination decades previous or not, no? > What kind of nonsense is that? I understood what they meant, and they're right. They're making an incomplete-information decision-theory argument. To rephrase them with more precision: Selecting a person from the population at random; and taking the category of "vaccines that will confer absolutely no benefits to that person specifically", and selecting a vaccine at random from that category; that person will, with high probability, incur more QALY-risk from paying for the vaccine (because they'll then have fewer dollars to pay for unplanned expenses, which can be quite problematic for the median member of the population), than they will from any iatrogenic illness induced by getting the vaccine itself. Outlier vaccines do nothing to change this, any more than those few poisoned bottles of aspirin did anything to increase the non-contingent risk profile of taking aspirin. Especially because, like the poisoned aspirin, outlier vaccines aren't on the market for very long compared to safe vaccines.