Media figures, including many physicians and public health professionals, are attacking the new Chair of the ACIP committee, Dr. Kirk Milhoan, for some of his comments made about the rate of miscarriage among women who received the mRNA injection.
Dr. Milhoan is reported as saying in the past that mRNA vaccines caused an 80% miscarriage rate among women who took the vaccination. The citation, according to media reports, that he referenced is a New England Journal of Medicine (NEJM) article from CDC employees, which is a reference to a study entitled "Preliminary Findings of mRNA Covid-19 Vaccine Safety in Pregnant Persons"[1]. This study was misleadingly used in hundreds of news reports to claim that there are no worrying safety signals for mRNA injections in pregnant women. In fact, Dr. Milhoan is correct in noting that the data presented in that paper upon its initial release, indeed reflected over 80% rate of "spontaneous abortion" after receiving the injection.
This paper is remarkable in its deceptive use of data, which resulted in an extremely substantial and public correction by the NEJM, which published a rare admission of "inappropriate" representation of data from the authors, but an inadequate and confusing follow up explanation. In fact, this paper and particularly the changes to table 4 is a textbook example of how to employ statistical tricks in peer reviewed publications to spread a message of "safety" of a particular drug or injection.
The researchers reported on outcomes for 827 women who received the mRNA shot while pregnant, which they reported as the denominator. Among these women, they found 104 spontaneous abortions and claimed the 12.6% rate was around the "normal range" for pregnant women. When readers inspected table 4 however, they noticed that exactly 700 of the 827 women were in fact injected with the mRNA vaccine in their third trimester—a very short period of observation for any kind of reaction. More importantly, during the third trimester, spontaneous abortion is not possible, but is rather a stillbirth, by their own definition (death of fetus 20 weeks or later in pregnancy). Only 1 stillbirth was reported. CDC leadership deceptively included 700 women in the denominator who could not, by definition, experience the result they were looking for in the numerator (spontaneous abortion).
The authors note in the original table 4 below (see annex at end of article) that 96 of the 104 spontaneous abortions occurred before 13 weeks of gestation, which leaves 8 spontaneous abortions for the time period between 13 weeks and 20 weeks.
Therefore, many readers drew an extremely reasonable conclusion, based on the authors' definitions, that 104/127 women who took the mRNA shot experienced a spontaneous abortion before week 21. That rate comes out to 81.8%. Dr. Kirk Milhoan is correct in his interpretation of the original, published results in the NEJM.
Dr Milhoan is not alone in noticing that sleight of hand used to get to 12.6% instead of 81.8%. A PhD researcher from Belgium wrote in, pointing out that they have used the 700 women inappropriately in the denominator[2]. The paper was then "corrected" by simply saying a rate could not be found and they replaced "104/827" with only the number 104 (see "current table 4" below) with the authors saying that they tried to follow up many of the women and couldn't establish a true rate of spontaneous abortion so there simply is no denominator. To "correct" for this, the authors also created a separate "life-table" study that used some of the original participants and added thousands of new subjects, and again claimed the rate of spontaneous abortion was low. But the detailed data from this follow up has not been made available.
But this desperate effort, despite their claims, is unrelated to the original, peer-reviewed CDC-authored publication in the NEJM that has not been retracted, but rather only manipulated to hide the disturbing results. The fact that the Journal did not retract the paper, despite its primary findings being rendered "misleading" and "inappropriate", is an example of political capture of the peer-reviewed journal industry. In fact, the NEJM allowed the CDC authors, yet again, to obscure the horrifying original results in a different way. For these reasons, this paper is the subject of a current Senate investigation relating to the potential destruction of their follow-up study data under Tom Shimabukuro and his colleagues at the CDC.[3]
Media figures and commentators attacking the new ACIP Chair should issue Dr. Milhoan an apology immediately, as his only crime here appears to be that he believed peer-reviewed CDC data, published in the most "prestigious" journal in medicine, and that he remained unconvinced by the obvious manipulations of that data that came later.
Table 1. Authors' presentation of original table 4 data:
| 0-12 weeks | 13-20 weeks | 21-27 weeks | 28 weeks + | |
| first dose received | 127 | 700 | ||
| spontaneous abortions | 96 | 8 | 0 | 0 |
| stillbirth | 0 | 0 | 0 | 1 |
ANNEX
ORIGINAL TABLE 4 Suppressed, nowhere to be found on NEJM site, but can be found at (https://web.archive.org/web/20210422005941/https://www.nejm.org/doi/10.1056/NEJMoa2104983)
CURRENT TABLE 4 (suddenly there is no denominator, and normal rate is suddenly "not applicable"):
[1] https://www.nejm.org/doi/10.1056/NEJMoa2104983
[2] https://www.nejm.org/doi/full/10.1056/NEJMc2113516
[3] https://www.aol.com/cdc-doctor-monitoring-bad-covid-172316191.html
