In our trilogy of posts, we looked at the evidence base provided for the UK’s JCVI (Joint Committee for Vaccination and Immunisation) recommendation to suspend COVID vaccination of pregnant women from spring 2025. We also reviewed the critique of the recommendation in the online outlet “The Conversation”. The motto of “The Conversation” is academic rigour with journalistic flair.
According to the JCVI, the recommendation is based on the lack of cost-effectiveness of vaccines for pregnant women. The subject is very delicate, and oddly, the JCVI has suddenly discovered this lack of cost-effectiveness. So, we looked at the relevant economic study, which appeared recently in preprint form (part 2), before looking at the basis for the criticisms (part 3). However, first, we wrote a very basic introduction to economic logic (part 1) before going into the economic evaluation.
We pointed out that health economics is about choices based on the idea of scarcity of resources. All health systems have finite resources, so you need to justify investing in A and not B. Scarcity was indeed the justification provided by the cited preprint (Part 2). Reading it was heavy going.
So, let’s look at the economic evaluation. All good economic evaluations should have a clear list of assumptions, their sources, and the uncertainty surrounding them. This is vital in pregnant women, as there are no published trials. Therefore, we do not have reliable information; we only have observational studies, which are well-known sources of bias. Several other items struck me as odd.
The lack of clear explanation of the methods and models (supposedly respectively 3 and 5), the lack of clarity about the assumptions of costs, effects, i.e. outcomes: no mention of harms at all and no definitions of hospitalisation, deaths or cases. The mention of the use of “bespoke” economic variables is even less explicit. The killer is probably the sentence at the end of the study (page 22). This makes it clear we can all bug off if we want to see the data: “The raw study data are protected and are not available due to data privacy laws”. This sounds like a rerun of the UKHSA’s Dr Jenny Harries's secret squirrel refusal to release death data to Parliament based on the number of vaccine doses, citing "commercial confidentiality."
How that can be possible for a global intervention in which society has invested so much is beyond our understanding. Our Comirnaty series sheds some light on this issue.
The whole economic evaluation is very unclear until you read the Intro (which we usually read last): 100s of thousands of doses of the vaccines had been bought and prepaid by Hancock and friends on the “let’s do something, anything” rationale. Now that stocks are running out and we are resorting to freezing pensioners and expropriating farms to keep the ship afloat, we have to ration the vaccines. It’s now Mum’s turn.
Now, we come to the third part of our trilogy.
The Conversation critics (part 3) picked up on the lack of economic clarity, and it’s the only convincing part of the critique. The rest is based on highly selected and shaky evidence.
The points made are:
Pregnant women are at higher risk of Covid infection.
The meat of this statement is a US CDC surveillance study in 325,335 women aged 15-44 who tested positive for SARS-CoV-2, 28% of whom had data on pregnancy status, and 9% were pregnant.
The pearl in this study (apart from lack of definitions) is the following statement on the severity of the disease: “Hospitalization was reported by a substantially higher percentage of pregnant women (31.5%) than nonpregnant women (5.8%) (Table 2). Data were not available to distinguish hospitalization for COVID-19–related circumstances (e.g., worsening respiratory status) from hospital admission for pregnancy-related treatment or procedures (e.g., delivery)” (our emphasis). So these pregnant women could have been admitted for, say, a breech presentation or signs of foetal hypoxia or pre-eclampsia. With such a potentially highly biased sub-sample, it is impossible to draw conclusions.
When we were in medical school, we were taught that pregnancy is not a deadly disease; it is a physiological state in a woman's lifespan if doctors can stop interfering.
So we agree with the JCVI, which thinks pregnancy is not a state of heightened risk for either mum or baby.
Now we come to the second Conversation criticism: The risk of Covid infection can be reduced by vaccination.
This statement is supported by a UK cohort study of 3,699 pregnant women admitted to hospital with SARS-CoV-2 by symptom group in the UK between 15 December 2021 and 14 March 2022. Again, no details of the likelihood of these women being infectious are given. The authors compare the pregnant women by the presence and severity of symptoms with a list of sociodemographic variables (age, BMI, etc), including one or more exposures to the vaccines (Table 1).
Immunisation status was “not known or not documented” in 635/1886 pregnant women (34%). Any inference from such a faulty dataset is meaningless.
The third and last Conversation criticism focuses on the Indirect benefit to the newborn from the mother’s vaccination.
This one is supported by a CDC presentation that does not mention the function of maternal antibodies. Antibodies, in general, are an unclear correlate of protection.
The CDC presentation does mention absolute vaccine effectiveness but warns about the likely bias in the estimates. We consider it bad academic practice to present knowingly biased estimates. The basic rule is that benefits can only be reliably quantified with trials.
We are asked to believe either side of the argument: do or do we not vaccinate in pregnancy?
Quite frankly, we cannot give you a bottom line other than reminding everyone of the precautionary principle, especially in pregnant women and unborn babies: if you do not know what you are doing, sit on your hands until you do. No document that we consulted while writing mentioned potential harm.
At medical school, we were also taught that drugs should only be prescribed during pregnancy if the evidence shows that the expected benefit (to the mother) is greater than the potential risk (to the foetus). A risk-benefit assessment should also be conducted on an individual patient basis, and the consenting clinician should ensure that the most up-to-date evidence is used to ensure the safe and effective use of medicines and vaccines.
This is Trust the Evidence. Thank you all for your support. We will continue to monitor this story and its consequences.
This summary was written by two old geezers who do not play politics or games with mothers' and babies’ lives.
Share this post