After the JCVI recommended against offering vaccines to children aged 12 to 15 on health grounds, the government asked the four chief medical officers to consider the broader case, including the impact on schooling.
As we know, the government has now accepted the chief medical officers’ recommendation: that all 12 to 15 year olds should be offered one dose of Pfizer on the grounds that doing so will reduce disruption to education. The government has released details of the modelling that underpins that rationale. The approach was first to estimate the number of infections with and without vaccination under different scenarios of infection spread. Next, they used this to model the number of days of lost education that could be prevented by vaccination.
Under a central scenario, the modelling paper estimates that vaccinating 60 per cent of 12 to 15-year-olds would prevent the loss of about 110,000 days of school in the six months between October and March 2022. At first glance, that sounds like a big number. But when you take account of the number of pupils, it works out to 41 days per thousand pupils or, put another way, just 15 minutes per child. And yes, that is the total over the full six months.
To put that into context, if the government had ended the schools’ lockdown just a day earlier back in March, children would have got a benefit 25 times more than that estimated from vaccinating over one million 12 to 15-year-olds.
But there are even more worrying questions that need to be asked about the modelling. First, the paper uses a vaccine effectiveness figure of 55 per cent taken from recent research including an ONS-Oxford study. That paper found that one dose of Pfizer was 57 per cent effective in preventing infection.
The trouble is that figure of 57 per cent comes from comparing those vaccinated against unvaccinated who have no evidence of having been previously infected. The same study reports that previous infection alone achieved 72 per cent protection against the Indian variant, much higher than one dose of vaccination. Vaccination may provide some further protection but relative effectiveness of vaccinating children previously infected will be drastically lower than 55 per cent.
The ONS antibody studies suggest that nearly half of 16 and 17 year olds have been previously infected. We don’t know the equivalent figure for 12 to fives but it is likely to be similar. That means the vaccine effect relative to all unvaccinated (previously infected and not) will be drastically lower than the figure used in the modelling paper. In turn, even the 15 minutes of prevented lost schooling will be a significant overestimate.
The point about previous infection is so important and obvious that it seems scarcely believable that the modelling underpinning the chief medical officers’ recommendation has ignored it. But the issue does not even get a mention in the modelling document, even to note it as a limitation of the study.
But that is not the end of the story. There are two further very significant omissions in the paper, though at least these are noted as limitations. The first is that the modelling takes no account of lost schooling due to the time taken to administer the vaccine. If, as seems likely, the programme is administered mainly in schools, we need to account for the time out of the classroom by pupils waiting for and receiving the vaccine, as well as the 15 minutes' rest time that is recommended after the Pfizer vaccine in case of immediate adverse reactions. That alone would more than wipe out the estimated 15 minutes lost education prevented. Professor Robert Dingwall, formerly a JCVI member, makes the further point that another 15 to 20 minutes per child would be necessary to establish that their consent is free and informed.
The other omission is that the modelling ignores any lost education due to vaccine side effects. Research in Lancet Infectious Diseases found that 13 per cent suffered systemic (whole body) side effects such as diarrhoea, fever and nausea and 72 per cent suffered localised side effects. That study was for adults but found that side effects were more common among younger people. If the numbers for children were similar – and even if only those suffering systemic reactions took just a single day off school – the net effect would again be more than enough to reverse the (overestimated) 15 minutes of saved education.
The modelling paper does point to other assumptions that might suggest their 15 minutes is an under-estimate. But even if that figure was doubled or tripled, once we allow for previous infection and time lost for vaccine administration, observation and side effects, the only possibly conclusion from the modelling is that far from preventing school time being lost, vaccinating 12 to 15-year-olds will lead to a significant net loss of education.
Clearly the four chief medical officers have some serious questions to answer. But although it is their job to advise their respective governments, in the end policy decisions remain the responsibility of ministers. You would hope that MPs and journalists would be asking some very urgent questions of the Prime Minster and his cabinet about the evidence on which they have based their decision to vaccinate children.