After request from a science reporter in Italy, we have produced a similar graphic for Italy to those we have produced with UK data to illustrate the potential benefits and harms of the AstraZeneca COVID-19 vaccine.

There are two elements to the graphic and the methodology is broadly based on the methodology used in the assessment of the UK harm / benefits.

Potential harms: This was taken directly from the European Medicines Agency Assessment from 23 April 2021 and represent exact rates for each age bracket (i.e. no modelling as we did for the UK data).

Potential benefits: This was calculated in the following way:

• All cases aged ≥ 20 years tested positive for SARS-CoV-2 infection between 21 October 2020 and 20 November 2020 (symptomatic or asymptomatic) in Italy were collected. These are the cases notified to the COVID-19 Italian Surveillance System (information extracted on 21 April 2021). Hospitalisations and admissions to ICU were counted if occurred within 28 days since the date of testing positive.
• A constant (not age dependent) underreporting rate of 25% was assumed. This is in line with previous Italian estimates. This is because, unlike with the UK Office of National Statistics' Infection Survey, in Italy there is no randomised population testing to give a more direct assessment of the number of cases of COVID-19 (including asymptomatic cases).
• The direct positive case to ICU percentages were used for each age group to determine risk of entering ICU.
• A vaccine efficacy of 80% at reducing ICU admissions was used.
• A high rate of 200 new infections per 100,000 per day was used; a medium rate of 100 new infections per 100,000 per day was used; a low rate of 20 new infections per 100,000 per day was used.
• Benefits were calculated over 16 weeks.