The data presented are observational and have a range of possible biases that have previously been described.(2, 6) There are some notable biases that could impact on waning in particular: firstly, within the all age analysis, and even within age groups, there will be different targeted cohorts with different lengths of follow-up and potentially different levels of waning and this may vary by vaccine. For example, healthcare workers are more likely to have received the Pfizer vaccine, whereas persons in clinical risk groups and care home residents were more likely to have received the AstraZeneca vaccine. Adjustments and stratified analyses are used to account for or investigate these factors, however, this is unlikely to account for all confounding. Secondly, as the pandemic has progressed, there will be an increasing number of individuals who have been previously infected, both in the vaccinated and the unvaccinated group. Protection due to previous infection in the comparator group will attenuate VE over time. Lastly, there have been changes in testing policy over time, for example, the widespread rollout of lateral flow testing means that PCR testing is increasingly being used as confirmatory testing, therefore there is an increasing risk of false negative PCRs – the resulting misclassification could attenuate VE.