3 ms·

Does the calculus change based in how many people you'll be giving this to? A 1:1M chance is much different to me if you give it to 3 people or 2 Billion

• sokoloff 6y ago
  I don't think it's different based directly on that, but rather based on something which is strongly related to that. If you're giving a chemotherapy drug to patients with advanced cancer, 1 in 10⁶ is perfectly acceptable. But it's acceptable because the do-nothing course is probably worse. Even serious, chronic pain relief I'd consider 1 in 10⁶ to be a good tradeoff. With vaccines, you're taking healthy people with no acute condition and exposing them so the do-nothing course is much better than the cancer case. Those differences of course drive the "do we give it to 3 people or 2 Billion people?" decision, but it's not directly the breadth but the underlying reason for the breadth that make me think differently about the acceptable risk level. If 2 Billion people had cancer or serious, chronic pain, I'd be in favor of giving them a drug which would cure 999,999 of them for every 1 that it killed.
• dillondoyle 6y ago
  Along that line i'd be interested to hear insight here from a stat person or someone who does these studies. At those low %s what are the chances the actual clot rate say doubles, quadruples as we keep getting larger audiences. With such low numbers random chance would not likely show up in small studies. So what are for instance the statistical chances of seeing a 10x increase in clotting going to 250mm population?