Real-world effectiveness of the Covid jabs | David Spiegelhalter and Anthony Masters

Behind the numbers: the policy of giving as many first jabs as possible is already cutting hospital admission rates

We should by now be familiar with the idea of a vaccine's efficacy, but last Monday three analyses of vaccine effectiveness were published. Despite the confusing similarity, these are different concepts: efficacy is measured in tightly controlled clinical trials, effectiveness is how well a vaccine works in the messy real world.

In trials, healthy volunteers are put in vaccinated and control groups at random - this ensures the groups are comparable and differences in outcomes must be due to the vaccine. If we simply compare people who have been jabbed with those who have not, they will differ in all sorts of ways: older and other higher-risk people will be first in the queue, while communities that are hesitant to be vaccinated may also be at higher risk. These confounders can lead to systemic bias in estimating effectiveness. So, studies use elaborate statistical analysis to make fair comparisons.