The MHRA and COVID vaccines surveillance

Something more doesn’t add up.

The  MHRA preprint on the Implementation and results of active vaccine safety monitoring during the COVID-19 pandemic in the UK is complex: 55 pages, 27 tables, 14 more tables and five figures in six appendices. 

The MHRA and COVID vaccines surveillance

So this time, we focussed on the issue of pregnant women.

Of 5,689 women of reproductive age in the sample (Table 8), 3845 pregnant women reported an Adverse Drug Reaction (Table 14), mainly in Pfizer and Moderna vaccine recipients, as these were “preferred” vaccines.

After one dose, 2,156 pregnant women reported 750 ADRs. However, for dose 2, the number of women dropped by 46% (n= 991) to 1165, reporting 271 ADRs. After that, the number of recipients is tiny (25) in Table 14.

If we assume Doses 1 and 2 recipients are the same women (given that the last time we looked, pregnancy lasts 9 months), that adds up to 2,156. Doses 3 and 4 presumably would have been given to a separate cohort of newly pregnant women, taking the total to 2,555 - 139 women more than the total reported in Table 13. We cannot explain this discrepancy. 

In addition, 995 vaccinated women reported 1,934 ADRs (top of page 20); however, there is no report for the 1,421 women who we assume were still alive and chose not to respond for their own reasons. As far as we can understand, the text does not give details and makes the usual assumption that non-response = no ADRs.

What we have described tallies with the following sentence on page 30 of the document

Multiple pregnancies are possible but rare in the 25 months of observation, corresponding roughly to the follow-up from November 2020 to 31st December 2022. And then the text goes on to state:

The TTE translation services think “challenging” means lost to follow-up. If this were a randomised controlled trial that failed to follow up roughly half its participants, the trial would be considered a failure. It would certainly not allow anyone to make inferential statements such as the ones the authors have made.

To be clear, a loss to follow-up could be due to death or serious disease, but you have no way of identifying the possible cause of non-response.

Therefore, the statements in the Results section of the abstract 

does not represent the YCVM data, and the authors point to passive surveillance in the discussion as the only source of harmful data:

“Inclusion of YCVM data has contributed to the safety review for COVID-19 vaccines in pregnancy and breastfeeding, which concluded the YC data did not raise any safety concerns with COVID-19 vaccine exposure during pregnancy. However, as with passive spontaneous reporting, the completeness of reporting is unknown.”   

So, to sum up, we do not know why 1,421 out of 2,555 enrolled pregnant women did not respond, but hey, all’s OK. Carry on, please.

Wait a minute, instead of being so critical all the time, why do you not take a look at what the trial in pregnant women tells you? You know there is one:

Exploring regulatory data sets of the Comirnaty vaccine - 7

Yes, and it was completed on 15 July 2022. The only problem is that two weeks ago, on 30 October, BionTech posted: ‘Results will be submitted; however, please note that data are not yet available for all serology outcome measures.’

The delay is contrary to the FDA Act, which requires the submission of results information no later than one year after the clinical trial's completion date (referred to as the "primary completion date"), which is defined as the date of final data collection for the primary outcome measure (81 FR 64983).

The FDA can impose penalties exceeding $10,000 per day when companies violate the law. The trial was completed 858 days ago, and the results should have been published 493 days ago, as required by law. The accumulated fines now amount to nearly $5 million. However, the FDA has yet to enforce the law while vaccines are being recommended for pregnant women despite the delay in the evidence from the only randomised trial conducted to date. 

This post was written by two old geezers who are staying put.

Readings

Jenny Wong, Katherine Donegan, Kendal Harrison, Tahira Jan, Alison Cave, PhilTregunno. Implementation and results of active vaccine safety monitoring during the COVID-19 pandemic in the UK. medRxiv 2024.11.06.24316782;