The post-authorisation safety of COVID-19 vaccines: real-world evidence

Vaccines against COVID-19 were developed, approved, and distributed at an unprecedented speed during the coronavirus disease pandemic. These vaccines have shown high efficacy in preventing severe COVID-19 and acceptable safety profiles in clinical trials. However, potential rare adverse events related to these new vaccines have been reported, and continuous vaccine safety surveillance is needed as mass immunisation against COVID-19 continues. With the ability of capturing information from larger and more diverse populations, routinely collected health data, also known as 'real-world data', can provide valuable real-world insights in post-marketing surveillance, complementing the knowledge gained from clinical trials. This thesis aims to assess the post-authorisation safety of COVID-19 vaccines by applying epidemiological and statistical methods using real-world data. I began with a literature review of real-world studies examining the safety of COVID-19 vaccines. I then introduced the methods and different data sources being used in the thesis. Three analytical studies were conducted using multiple electronic health records and claims datasets. In the first study, I characterised the background incidence rates of 15 pre-specified adverse events of special interest associated with COVID-19 vaccines. I observed considerable variability in the rates with respect to age and sex, emphasising the need for standardisation or stratification of the background rates used for vaccine surveillance. This study also found substantial heterogeneity in the estimated rates across databases, suggesting that observed rates among COVID-19 vaccine recipients should be compared with background rates obtained from the same database where possible. I then assessed the association between COVID-19 vaccines, SARS-CoV-2 infection, and the risk of immune-mediated neurological events. I applied the observed-to-expected analysis and the self-controlled case series methods using primary care records from the UK and Catalonia, Spain. This study found no increased risk of the included immune-mediated neurological events after COVID-19 vaccination, but increased risk after SARS-CoV-2 infection. These findings reaffirmed the safety of the studied COVID-19 vaccines, underscoring the importance of vaccination. Finally, I examined the comparative risk of thrombosis with thrombocytopenia syndrome or thromboembolic events associated with COVID-19 vaccines using datasets from Europe and the US. I compared adenovirus-based COVID-19 vaccines with mRNA-based COVID-19 vaccines, and the secondary analysis compared two brands of mRNA vaccines. This study found an increased risk of thrombocytopenia after the first dose of ChAdOx1 (Oxford-AstraZeneca) compared with BNT162b2 (Pfizer-BioNTech). While the studied events were rare, these observed risks after adenovirus-based vaccines should be considered in planning future immunisation campaigns and vaccine development. In these analyses, I have demonstrated that real-world data can generate timely and reliable evidence on post-authorisation vaccine safety. These findings have important implications for clinical practice, health policy, and future research. Above all, in light of the well-established benefits of the COVID-19 vaccination, my findings should encourage continued confidence in vaccination.