Fig. 2

Illustration of different approaches for SARS-CoV-2 variant exposure status ascertainment: (1) exposure defined based on WGS, (2) exposure defined based on WGS or presumptive genotyping and (3) exposure defined based on time periods with known variant circulation. The different approaches are compared in terms of the obtained sample sizes, the inference estimates and their 95% confidence intervals (expressed as standardized risk and risk ratio) for severe COVID-19 (experiencing an acute respiratory distress syndrome (ARDS) event, being admitted to the intensive care unit (ICU), or in-hospital mortality), and the distribution over time of the number of hospitalized COVID-19 patients included in the analysis by week of diagnosis (7-day rolling average). Please note the scale differences in the y-axis in the distribution plots