The Covid-19 vaccines don’t have ‘negative efficacy’

Several Facebook posts have shared data from the UK Health Security Agency (UKHSA) with the claim that the Covid-19 vaccines have negative effectiveness against the disease, meaning they would make vaccinated people more likely to catch Covid.

One post, which appeared on 24 December 2021, linked directly to a Twitter thread about the UKHSA data. The Facebook post claims: “The figures below show you're more likely to have Covid and pass it on, if you're vaccinated.”

Other posts from January link to a blog, which takes its data from another thread by the same Twitter user, with newer data. The blog claims “this data [is] showing negative efficacy of the shots” and “the U.K. data has been showing negative efficacy in all but the youngest age group”.

These claims are not true. The vaccines make people less likely to catch Covid.

The UKHSA report containing this data has been misunderstood and misused in this way many times before. It shows a rate of Covid infections among unvaccinated people that is probably much lower than the true rate in the real world.

The reports used in both the blog and the Facebook post also contain estimates showing that the Covid vaccines are effective.  

What is “vaccine effectiveness”?

A vaccine is considered effective if it lowers the risk of an unwanted outcome. It would be “negatively effective”, as the blog claims, if it somehow raised the risk.

The UKHSA publishes a weekly ‘Vaccine Surveillance Report’, which includes the most recent data from England on a variety of outcomes, including the number of people who test positive for Covid.

On one page, it includes what it calls the “unadjusted rates” at which people test positive, are admitted to hospital, or die following a positive test. The table shows separate rates for different age groups, and depending on whether people were vaccinated or not.

Both the December Facebook post and the blog make the mistake of assuming that the data on this page can be used to estimate vaccine effectiveness. It can’t, and the UKHSA is clear about this, saying in the first footnote below the table: “Comparing case rates among vaccinated and unvaccinated populations should not be used to estimate vaccine effectiveness against COVID-19 infection.”

What have they got wrong?

As we have said before, the data on these tables is very likely to underestimate the rate of Covid cases among unvaccinated people, because it probably overestimates the number of unvaccinated people in England.

It is also likely to be biased by other factors, which may affect the chance of a vaccinated or unvaccinated person testing positive. In a footnote, the UKHSA lists several of these possible factors, which is why the figures are labelled “unadjusted”.

The unknown extent of these potential biases does mean that it is conceivable that vaccinated people are more likely to test positive, for example if they are more likely to get tested, or feel safer, take more risks and so are exposed more often than unvaccinated people. However, we don’t know this, and the UKHSA doesn’t suggest it.

Estimating vaccine effectiveness correctly is a more complex process, which has been conducted by many studies mentioned in the same report.

Photo by Roger Starnes Sr on Unsplash