all 4 comments

[–]thefirststone 2 insightful - 2 fun2 insightful - 1 fun3 insightful - 2 fun -  (3 children)

Note that this article is from July using data collected in March.

Here's a study with 2,456 participants who all took at least one mRNA shot before the third trimester, from December to July (I think it incorporates some of the other study's data, which is referenced), and cuts off second trimester at 20 weeks to fit the SAB defintion: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8366802/

The results (detailed here) are:

Among 2,456 pregnant persons who received an mRNA COVID-19 vaccine preconception or prior to 20 weeks’ gestation, the cumulative risk of SAB from 6–19 weeks’ gestation was 14.1% (95% CI: 12.1, 16.1%). Using direct age standardization to the selected reference population, the age-standardized cumulative risk of SAB was 12.8% (95% CI: 10.8–14.8%).

They say that's within the expected rate, controlled for age, but they admit that 65 of the younger participants didn't follow up, so their expected rates are skewed. They admit they didn't control for race (it was 78% self-reported "non-hispanic white", so maybe it should have been lower than they expected). They also didn't control for obesity (almost 20%, which is lower than the general population, so maybe it should have been lower).

I'm not sure what to make of this. It could be that 700 in the third trimester note on Table 4 was wrong. Surely somebody would have noticed in real life instead of just months-old studies if the rate were that high, right? I think the presentation on the first study was botched, and this article only went by that. You'd have to go to the numbers to confirm it was wrong. The second study fixes this in several ways. I don't know if it's too slick to believe, but it's written the way the first one should have been.

[–]FlippyKing 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (2 children)

Surely somebody would have noticed in real life instead of just months-old studies if the rate were that high, right?

It is not like there are armies of scientists and statisticians waiting to jump on every study and pick it apart or try to reproduce it. Someone has to pay them to do it or approve their research. When that "someone" does not exist, the "armies" of the first sentence is replaced by people dismissed as cranks or conspiracy theorists (often fairly) and whose work is then dismissed outright or the errors in their work is not treated as kindly as similar errors when committed by those with acknowledged authority. The issue of how un-reproducible results are problems in all sciences is above/beyond the problems of: 1 how money influences what studies are done, which get scrapped or altered mid-way through, or which are not done; or 2 what editors of all kinds (funders of studies or journal editors or lead authors or "shadow" lead authors where a professor will get a student to be a "lead" and put their words in there mouths so the becomes an appearance of multiple voices towards a conclusion) exert over science.

[–]thefirststone 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (1 child)

un-reproducible results

It's inherently unreproducible because it took place in a unique time, and they've done everything they can to eliminate the control group (unvaccinated). If anybody made a loud stink about the veracity of the first study, it wouldn't matter because it's a voluntary self-reported phone study, and it'd get torn to shreds as fake or useless. In other words, the OP's article doesn't matter, because it won't be used for talking points. Only lies and statistics matter.

Ideally the data would come from physicians, OBs, and hospitals, even if not every woman bothers to report miscarriages. One wonders where those figures are, and how they compare to these. But I only care to the point that it can be used to convince emotional women of which specific thing they should be afraid of.

[–]FlippyKing 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (0 children)

I agree 100% Self-reporting studies in science are at best hints for where someone might do real research when the are not customer satisfaction surveys and they should be taken as seriously as customer satisfaction surveys.