When you make an argument like this, please show your work. Let’s think of the cases
1. So we deploy an investigational vaccine that doesn’t actually confer immunity or strong immunity. If people think it provides protection, & take more risks MORE people get sick and die.
1. So we deploy an investigational vaccine that doesn’t actually confer immunity or strong immunity. If people think it provides protection, & take more risks MORE people get sick and die.
2. Imagine vaccine provides some protection, but a subgroup of people who access it outside of an RCT have a serious adverse reaction. If people think the vaccine is harmful, fewer take it or enroll in trials and MORE people die.
3. Why should we believe that outside access to an investigational vaccine and speed at which a clinical trial completes are independent? If outside access slows accrual in a RCT MORE people could die in the long run since (a) pre-approval production rate will be smaller than...
4. ...post-approval production rate plus (b) there will be less hesitancy to use a vaccine supported by clear evidence of both safety and efficacy.
5. Why think that what we do with a COVID-19 vaccine will have no impact on the way that people think about vaccine safety and efficacy more generally? Widespread use of a vaccine that does have adverse effects could exacerbate vaccine hesitancy, already public health crisis...
5.b. If adverse events from widely used investigational vaccine lead to more vaccine hesitancy, then MORE people could die as a result of this dangerously bad idea.
6. This proposal is self-defeating as one role of a large phase-3 IS to provide access to the vaccine but under controlled conditions that allow us to detect efficacy and safety. Deploying ANY vaccine outside that context can increase uncertainty and delay wider use.
7. Obviously @StevenSalzberg1 is contemplating the best case scenario—that the vaccine is safe and confers significant protection to have a positive public health effect. What should our prior be on this? 90% of new interventions are never approved for any indication...
7.b. ...half of drugs that make it to phase 3 fail. Nobody spends a billion dollars to bring a drug to a phase 3 trial because they think it’s going to fail. Yet, fail they do. Often for lack of efficacy, sometimes for harm.
8. In response to 7 above some have pointed out that the overall probability of success for vaccines is higher than other interventions, as high as 39%. But that number should be taken with some caution IMHO. Why?
9. consider this brand new analysis by Lo and colleagues: nber.org/papers/w27176. Look at their breakdown of the data. Green highlights represent industry taking advantage of existing knowledge. But yellow paths are more difficult nuts to crack. 
10. Note that the probability of successfully moving an #hiv vaccine from phase 1 to phase 2 is 65%. Things look good. Probability of moving from phase 2-3 is 36%. But no vaccines have proven safe and effective.
11. A more realistic figure would look at our past track record for vaccines for novel viruses, some of which are using novel platforms—I doubt 30+% captures the probability of success for that reference class.
