Abstract The COVID Infection Survey monitored daily positivity through the COVID-19 pandemic from April 26, 2020 to March 13, 2023. In total 451 079 participants in private residential households were enrolled in England and tested at regular intervals for SARS-CoV-2. Here, we estimated the cumulative incidence of PCR-positive infections using a multilevel regression and post-stratification model to obtain estimates of daily positivity, combined with a distribution of the duration of positivity from regular testing data. We estimated cumulative incidence by epoch (approximated by the dominance of successive SARS-CoV-2 variants) and calculated the corresponding infection-hospitalisation ratios. We found cumulative incidence was relatively low during pre-Alpha and Alpha-dominant epochs, rose steadily during the Delta-dominant epoch, and was highest during successive Omicron-dominant epochs. High cumulative incidences in successive Omicron-dominant epochs are consistent with lack of protection from previous infections. However, infection-hospitalisation ratios, whilst higher at the start of the pandemic, remained low after the Delta-dominant epoch and vaccine introduction. Stratified estimates show hospitalisation risk was consistently very low for younger age groups, increasing with age. Surveys with random sampling and longitudinal designs facilitate direct estimation of prevalence and incidence, however should be complemented by dense sampling to estimate duration of infection to maximise their value.
Gaughan et al. (Wed,) studied this question.