[i_cannot_hack]

You are almost certainly using the wrong numbers. According to https://en.wikipedia.org/wiki/Cutter_Laboratories#Cutter_inc... and https://www.nejm.org/doi/full/10.1056/NEJMp048180, the faulty production applied to 120.000 doses of vaccine, which lead to 56 cases of paralysis (0.046%) among vaccinated children. These sources further mention "the exposures led to an epidemic of polio in the families and communities of the affected children, resulting in a further 113 people paralyzed and 5 deaths", which tells me that your number of 250 probably includes cases caused by such transmission. This is of course reasonable when discussing the total damage caused by the vaccine, but this number can not be compared to the incidence of paralysis for a single infection, as you do.

You would also have to include further transmissions caused by wild polio if you would like to make such a comparison.

> still patently false. pfft. Perhaps false under some assumptions, definitely not "patently false".

Even with your own calculation, which is inflated by also including transmission within the community for the vaccine but not for wild polio, the vaccine was very much comparable to wild polio (0.125% versus approx. 0.1%). Without vaccination, polio would usually infect virtually all children. So even with your own inflated assumptions it's false that the vaccine was worse than polio.

So under which assumptions would it not be false?

[beagle3]

> You are almost certainly using the wrong numbers

Take them with wikipedia, not me. I will note, that Wikipedia talks about both Cutter and Wyeth having problems, so it is possible the numbers are a sum of wyeth+cutter which would be compatible with your numbers being cutter only, but in the context of safety vs. polio wouldn't matter.

> You would also have to include further transmissions caused by wild polio if you would like to make such a comparison.

All wild polio numbers include those numbers by default - there actually is no way to get them otherwise, I guess you are saying "you should probably debase by more than 200,000 because of secondary infections". I don't know how to do that, exactly, but it will likely still be similar.

> Even with your own calculation, which is inflated by also including transmission within the community for the vaccine but not for wild polio, the vaccine was very much comparable to wild polio (0.125% versus approx. 0.1%).

With this calculation, the vaccine was 25% worse than the wild type. No error bars, but, that's easily worse, even much worse.

> So under which assumptions would it not be false?

How kind of you to drop "patently". If you didn't mean "patently" false earlier, why did you use that word? twice?

From[0]: "Patently: in a way that is so obvious that no one could disagree.". It's nice of you to finally bring sources, but even these sources don't make it "patently" false, given that they disagree with the sources I gave (which, I concede, reference more than just the safety of the Cutter incident which I originally mentioned, but which are definitely the subject matter under discussion)

[0] https://www.macmillandictionary.com/dictionary/british/paten...

[i_cannot_hack]

> Take them with wikipedia, not me. I will note, that Wikipedia talks about both Cutter and Wyeth having problems, so it is possible the numbers are a sum of wyeth+cutter which would be compatible with your numbers being cutter only, but in the context of safety vs. polio wouldn't matter.

Wikipedia is not wrong, but you are using their numbers wrong. Your source is talking about the total number paralysis cases that occurred as a result of the vaccine ("250 cases of paralytic illness had occurred"). This includes secondary infections. If you follow the sources listed on Wikipedia page you reference, you will find two sources: my source [1], and [2] (via [3]), both which declare around 60 cases of paralysis in vaccinated persons, and then a larger number (in the order of 200) of total cases. The numbers you reference are referring to only the Cutter polio vaccine, by the way.

And you cannot use the total number of cases in the comparison that you did.

> All wild polio numbers include those numbers by default - there actually is no way to get them otherwise, I guess you are saying "you should probably debase by more than 200,000 because of secondary infections". I don't know how to do that, exactly, but it will likely still be similar.

If by debase you mean divide, then yes: you would need to divide by more than 200,000.

Your number is {how likely am I to get paralysed by the vaccine polio + how likely am I to cause further paralysis via secondary infection chain}, and you are comparing it to simply {how likely am I to get paralysed by the wild polio}. The first number is inflated a lot by the addition of secondary infections.

The numbers are not at all likely to be similar. As you can see in my (and your) sources, secondary infections accounted for more than double the number of paralysis cases, and therefore there it is likely a lot more people got sick via the secondary infections than the number people who was vaccinated. This causes the big discrepancy between our numbers.

Wild polio causes secondary infections as well, but this is not included in the number you are using for comparison, since it only includes the individual risk.

> With this calculation, the vaccine was 25% worse than the wild type. No error bars, but, that's easily worse, even much worse.

But the lack of error bars means that the calculation is meaningless.

You are assuming that "one in 1000 cases" means exactly one case per 1000 cases, and translate this to 0.100% with three decimals of accuracy.

It is clear from context that "one in 1000" is a rounded number for convenience, and they could very well have runded up from 0.8 or down from 1.4. You simply cannot conclude that the vaccine was 25% worse from your data. But we can conclude that they were in some way similar under your false assumptions, as they both would round to 1.

If you want to conclude anything else except "they are both around 0.1%", you would have to find a source that specifies at least 1 decimal of accuracy in the number incidents per 1000 cases.

Of course, this does not really matter, since you would still be comparing the wrong numbers.

Your argument here is "It's ambiguous and inconclusive when using clearly wrong assumptions that inflate the difference". Taking that argument into consideration, I'm still willing to confidently call it patently false.

Although I agree that I should not have said "Even with your own inflated assumptions it's false that the vaccine was worse than polio." What I meant was "Even your own inflated assumptions does not support that the vaccine was worse than polio", but I clearly worded it badly.

> How kind of you to drop "patently". If you didn't mean "patently" false earlier, why did you use that word? twice?

I meant it when I used it, and my question had a purpose. If you cannot give any reasonable assumptions where it wouldn't be false (which you haven't been able to do), then it would be patently false to me. The reason it does not seem patently false to you is because you have several misunderstandings in your reasoning and your reading of the sources. But I'm not really interested in discussing this terminology further.

> given that they disagree with the sources I gave

They don't. As I said, you are simply misreading your sources.

But seriously, even if you would manage to find some source which would refute my sources and back up your original claims, I have a bigger point to make now:

During this discussion, you have made numerous mistakes in many posts, beyond my criticism of your faulty reasoning:

(1) Mathematical mistakes

(2) Misreadings of the sources

(3) Inability to clear up ambiguities by looking at the referenced source or secondary sources

Even if by some happenstance you would happen to be right (even a broken clock, etc), you clearly are not confident enough doing this kind of analysis to be lecturing people about the specifics of vaccination risks. You are just as likely to mislead yourself and others as you are to educate.

Of course, this probably won't stop you, since you are not likely to respect my opinion. But hopefully it will still be in the back of your mind next time you approach this topic.

[1] https://www.nejm.org/doi/full/10.1056/NEJMp048180 [2] https://www.cabdirect.org/cabdirect/abstract/19642705083 [3] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2928990

[beagle3]

Thank you for your thoughtful answer. I pointed the lack of error bars myself, as well as mentioned that I don’t know what other number to divide by, as you noted. I am not trying to be intellectually dishonest.

And I may agree with your analysis - on phone now, can’t really do the reading.

However, we have two sources you claim are only for the cutter incident, whee one quotes 200,000 doses and another 120,000; as you note, the 1/1000 might well be 0.8 or 1.3 in 1000; it may be false but it is definitely bot “patently false”.

On the same note, it’s the internet, and for all you know I am a dog, but I have saved a family members from years of agony and wrong treatments due to wrong diagnosis after being told “it is patently obvious” that my analysis was wrong. And I have apology letters from department heads at two of the world’s highest ranked hospitals after it turned out I was right and they were wrong.

And you know what? I’m quite sure I probably had a few mathematical errors along the way back then as well. But what they thought was two orders of magnitude more probable evidence towards one direction, turned out to be higher probability in the other (by less than an order of magnitude) and turned out to be what the thought was improbable beyond consideration. (And, I found a few mathematical errors in a three peer reviewed papers they were relying on)

I take issue with people who claim “patently obvious” about things which have about factor of 2 or so (if I take your numbers) without error bars (120000-200000 is a large difference) and without supplying sources, which you didn’t bother until I did.

(And I may not have time to delve into this further - this is merely of historical interest to me, not life and death like that other event).

Also, you may notice this sub thread is basically the only one that actually discusses numbers, others use arbitrary determinations like “most people’s living memory” to discard concerns. Despite advocating vaccines myself, I have a problem with religious zealotry around vaccines, which is what most pro-vaccine people practice - I have a child who could not get vaccines for medical reasons for many years, and I have to explain that, no, vaccines are not perfectly safe on an individual basis, and have not historically been perfectly safe on a population basis (Swedish Pandemrix), even though on a net population basis they are a net positive.

I thank you again for your thoughtful discussion.