Unfortunately there have been many, many deaths and also populations left disfigured and disabled from reckless medical experimentation. That is why there are such strict rules now.

Phase I is just safety ( and a bit of dose range finding ) - does the drug/treatment/vaccine harm or kill people on it's own - small number of people, with varying doses - ideally starting at the low end. You will do some tests to see if you get the expected biological response as a surrogate for efficacy.

Phase II is where you start looking at if it works ( while still looking at safety- always looking at safety ), here you are also trying to typically work out what is the right dosing regime ( how much, how often ).

Phase III is large enough numbers to get population level stats for safety and efficacy, given the treatment regime you have decided on.

Also you don't just need willing people to take the vaccine to do a proper trial, you need the medical infrastructure around it.

People to administer it, keep track of the patients - monitoring them for any signs of adverse reaction - reliable people to take blood samples and handle them properly, proper note and record keeping etc.

The other thing to consider in trials is that roughly half your patients don't get anything ( more accurately a placebo ) - and in the case of vaccine trials, you need to control group to get infected!!! Otherwise you can't see any protective value.

You'd need to choose you control group to carefully reflect the test group - if you had your vaccine test group in New Zealand and your control group in the US - nobody would believe your result.

At the moment people doing trials in America have an advantage in that the virus is still infecting lots of people, so you can more easily demonstrate protection. If you are trying to trials in China - pretty hard.

I don't think lack of volunteers is the problem. If you want to do it ethically, you have to do it incrementally.

Are you referring to the subjects injected with the experimental vaccine? A phase I trial's purpose is to see if the drug is safe for humans and get an idea of what the best dose is, not if it's effective. As to deliberately exposing people to the coronavirus, that's an ethical mess. The parent blog's author has another post that covers this as well [0].

tl;dr (the blog in general is a great resource though):

* We're not at a point where we need to deliberately expose people, as there's enough people being exposed in their communities.

* The normal standard for those deliberate tests also only recommend it if there's known effective for treatments for the diseases, which (obviously) there is not for COVID-19. Hell, we're still figuring out what exactly it's doing in the body.

* As a sibling comment brings up, history is riddled with unethical medical experimentation. We don't want to repeat those mistakes.

0: https://blogs.sciencemag.org/pipeline/archives/2020/07/02/ch...

So, easy now. The phase 1/2/3 process has been developed over a long period of time, and has given good results. There are doubtless a lot of ways to do it worse. Deciding to make up a new process on the spot in the middle of a crisis, with a lot of high-anxiety thinking, may not actually deliver a quicker/better result. Plus, the manufacturing is going to be started in parallel with the testing, but we don't have the capacity to manufacture every one of the dozens of vaccines that are in development right now, so which ones do we make "risk starts" on? Well the best way to guess would be to look at the phase 1/phase 2 results, and see which seem most promising.

There are a lot more ways to get it wrong, than to get it right, and we want the quickest, good result, not just the quickest result. Aside from the ethical consequences of experimenting on poor people (and I don't think those are something we should put aside), coming up with a new process on the spot is not even probably the best way to get a good result fast. Would the people who step forward to do that be unusually healthy (hence willing to take the risk), and so throw off the results? Would they be disproportionately people who have already had Covid-19, for the same reason? Would they be different in some other way? Let's not try to wing this. We've got a time-and-battle-tested method for developing a vaccine. I would prefer we use that.

Why should only the financially desperate be asked to sacrifice for their species? What a repugnant stance you've adopted.

You can randomize it in an authoritarian way if you reject here the free market, the way of selection matter far less than the discussion of how much lives could be saved.

What a repugnant stance you've adopted. Thinking like you think is consenting to mass murder, do you realize it ?

Here's my calculus: The phase I trial has approximately lasted 2 months. I checked the number of total deaths from 01 may to 01 July -> https://www.worldometers.info/coronavirus/worldwide-graphs/#...

This give us 277,145 human deaths in this trial period. According to https://en.wikipedia.org/wiki/Mortality_due_to_COVID-19#Mort... mortality rate for a ~20 years old should be at least lower than 0.1%.

With all those data in mind, here's the reasoning: The metric of effectiveness for this Moderna vaccine candidate is epistemologically LOW. They use fuzzy heuristics of matching approximately the immune response of naturally recovering patients as a claim of being effective. So firstly this reasoning is simplistic especially since researchers do not understand precisely how the working immune response work. So there is a risk of the vaccine to be a disappointment. If it is, those 277,145 human lifes would have been wasted and far more will die until we find the true vaccin. But what you need to understand is that researchers are condemned to match the existing immune response as otherwise, without testing on infected young humans, they cannot predict the vaccine expected performance. This alone is far enough for defending that they should have tested on a few dozen of infected patients the effectiveness. If they did: two possible scenarios: (lets suppose they tested 30 humans in their twenties that would make 0.3% of chance of deaths (comparatively to 277,145 human deaths) and the likelihood would be actually far lower as they would be actively monitored) scenario A) the vaccine doesn't work and it shows empirically, it save them weeks of evidence and therefore make them reconsider another paradigm for the virus. That alone would save at the very least thousands of lives by allowing the true vaccine to be found earlier. So you must realize that your "repugnant stance" of inaction is responsible in this likely scenario of far more deaths, sadly numbers quickly become unproportional to the emotional empathic response for a homo sapiens, this kind of brain failure honestly give a taste of vomit in my mouth, but yeah everybody is victim of this cognitive bias, even myself so no offense. In the second scenario, the empiricism would show them that their vaccine works: here again it would accelerate the delivery, and crucially the data would give them insights and allow for far better fine tuning of how the vaccine works and the optimal dosage. Is is not unreasonable to think that such speedup would allow to save a similar order of magnitude than 277,145 human deaths but could actually be far more (imagine 6 months of work saved y avoiding empirically invalidated or fine tuned research directions ? Even if it were to be lower, even if the data sped up the research by only one day (an insanely pessimistic view that I only show to defend your reasoning to the extreme, to its limit) would still save more than 5000 lives which is indeed far bigger than the likely lower than 0.3% of death rate for ~30 persons, so even in the ridiculous limit we talk about 5000+ vs 0.3 death.

What is the error rate of the immune system response currently used, compared to actual immune response? If it is small, then the value of infecting humans directly goes way down. Further, what is the probability that the control group and the vaccine groups will not be exposed naturally and accidentally to Covid19 during the Phase II, Phase III studies? If they are likely to be accidentally exposed, then purposefully exposing them has even less value. Given the current infection rates in the US, this likelihood seems pretty good.

The 277,145 number is meaningless. The vaccine was not known to help in any way 2 months ago. For all we knew at the time, if we had started giving this to people on mass 2 months ago, we would have had 300k deaths from the vaccine by now.

Then, the vaccine is not yet known to be effective. We only have proof that it is safe, but we don't have rigorous control studies showing that it is effective yet (we need the Phase II & III trials for that). Before that, there is only a "hunch" that it will help at all, so taking death numbers for the next few months is also going to be meaningless.

Finally, your "even if the data sped up the research by only one day" calculation is pure sophism. Why not assume that the data could slow down the research? What if one of the patients dies because of complications after being infected, and that affects team morale, slowing down the research? What if one of the researchers accidentally gets sick themselves because of improper handling of the live virus, again slowing down the research?

The final point is: we are already rushing this. There are good reasons we have stopped testing drugs on volunteers, we have a horrible history behind us of the effects, often without much better results (and often much worse results) than current methods. There methods we have for estimating actual immunity based on immune system response have been studied for some time and are considered good. Putting people's lives in danger is needless complication, more likely to slow everything down than to speed it up.