23

Steve Kirsch claims in Should you get vaccinated?:

82% miscarriage rate in first 20 weeks (10% is the normal rate).

Is this claim that the miscarriage rate is significantly increased through COVID-19 vaccinations backed up by data or bogus?

Laurel
  • 30,040
  • 9
  • 132
  • 118
Christian
  • 33,271
  • 15
  • 112
  • 266
  • 1
    The quote by the author is misleading a bit. The link is pointing to letter to editor where it is pointing out a potential mistake. Even if this letter we 100% accurate it would still only apply to 1st and 2nd trimester mothers. – pinegulf Jun 17 '21 at 12:35
  • @pinegulf : The first 20 weeks is the the 1st and half of the 2nd trimester, there's no claim about events after the 2nd trimester. – Christian Jun 17 '21 at 13:04
  • Mayo clinic says the miscarriage rate is at least 15%, and more not reported. "First 20 weeks" is needlessly confusing, since that's when most miscarriages occur. It's listed as "10-20% or higher", which is probably where he gets his 10% from. – Owen Reynolds Jun 17 '21 at 21:06
  • A pregnancy can end by miscarriage, abortion, rarely brute violence, and most of the time with childbirth. If you check after 20 weeks, the percentage of pregnancies that ended in childbirth is practically zero. So miscarriages are about 4 times higher than abortions. Not surprised. – gnasher729 Jul 03 '21 at 15:46

1 Answers1

64

The referenced claim misuses statistics from a study of the vaccines effects based on V-Safe data. The study finds that of 824 completed pregnancies of vaccinated women, 104 of the pregnancies were completed due to miscarriage. The letter to the editor claims that in 700 of the pregnancies, the vaccine was administered after 20 weeks when women are at much lower risk of miscarriage; if you exclude those from the results, you get the supposed 82.5% rate of miscarriage.

What this analysis ignores is that the vast majority of women who received the vaccine before 20 weeks and did not miscarry had not yet carried their pregnancies to term at the time the data were collected and thus were not included in the statistics of completed pregnancies. In fact, if we assume that vaccinations were distributed evenly across the term of pregnancy, you would expect that about 700 women would be vaccinated before 20 weeks if 700 were vaccinated after. Since only 126 of the pregnancies were completed, that would mean there were 574 non-miscarried pregnancies not accounted in the statistics. This indicates that the methodology used by the original researchers, counting miscarriages against total completed pregnancies, was the sound one.

antlersoft
  • 2,147
  • 1
  • 13
  • 16
  • 4
    I don't think it is fair to say it *ignores* the still-pregnant women. "We acknowledge this rate will likely decrease as the pregnancies of women who were vaccinated <20 weeks complete but believe the rate will be higher than 12.5%" That's fairly weak argument by the authors, but it would help if you could find someone at least as expert as the authors who expresses an opinion on that, so this isn't a "He said, four experts published (???) in a journal said" argument. – Oddthinking Jun 17 '21 at 15:21
  • 62
    Not published, no (it's been "sent"), and so obviously idiotic from a statistical perspective that it never will be. It's equal to saying that 100% of all pregnancies end in miscarriage because at 5 months gestational age, all the completed pregnancies are miscarriages. – CJR Jun 17 '21 at 15:31
  • 38
    VAERS data is not data and no studies can or should use it even as a proxy for real data ( https://vaers.hhs.gov/data/dataguide.html ) their whole page about it is a very polite way of saying that it is untested, unverified anecdotes. Useful for *surveillance* of possible problems but that is it. – Yorik Jun 17 '21 at 20:10
  • 1
    @Oddthinking Does someone with a Bachelor's writing a letter to the editor of a journal actually count as an expert "publishing" something? It seems like the original actual published paper in NEJM would count as experts expressing an opinion that their methodology was the correct one, not the nonsense in the letter to the editor. For reference, the original paper was published in April. Vaccination only began in late December. Almost no one vaccinated at < 20 weeks gestation prior to that time would have completed a pregnancy normally, so their whole argument is meaningless. – reirab Jun 18 '21 at 23:27
  • @reirab: On one hand, we have three people, two of which are MDs, the other with a Bachelors. On the other, we have an anonymous software developer (?) on the Internet. Now, I think the random Internet denizen in is probably right, but it would be nice to have a reference to show that his calculations are supported. – Oddthinking Jun 19 '21 at 02:23
  • 4
    @Oddthinking My point, though, was just that, in addition to the anonymous software developers (?) on the Internet, we also have [the original study](https://www.nejm.org/doi/full/10.1056/NEJMoa2104983) from experts published in one of the world's most prestigious medical journals. Incidentally, it turns out that their data collection ended on Feb 28 (having begun on Dec 14.) So, anyone vaccinated < 20 weeks would have necessarily been < 31 weeks at the conclusion of the data collection. Full-term is minimum 39 weeks and < 37 weeks is preterm. – reirab Jun 19 '21 at 05:24
  • There is one error in this answer, though, and it does make both the analysis of the original paper and that of the letter to the editor make more sense: the data in question was from v-safe, not VAERS. As mentioned in previous comments, VAERS contains only reports of adverse medical events after a vaccination (whether or not they are believed to be related to the vaccination.) v-safe, on the other hand, is a voluntary program specifically for the Covid vaccines that anyone who is vaccinated may join and researchers attempted to contact everyone who reported as pregnant in v-safe. – reirab Jun 19 '21 at 05:31
  • 1
    VAERS data is also discussed in the original paper, but Table 4 (and the entire sub-section in which it is located) is about v-safe data. – reirab Jun 19 '21 at 05:41
  • some more data showing *lower miscarriage rate for vaccinated*, the sample is very small however, https://www.nature.com/articles/s41577-021-00525-y/tables/1 – Roah Jun 24 '21 at 10:11