Safety Monitoring of COVID‐19 Vaccines: Perspective from the European Medicines Agency

Abstract

Prior to deployment of COVID‐19 vaccines in the European Union (EU) in 2021, a high vaccine uptake leading to an unprecedented volume of safety data from spontaneous reports and real‐world evidence, was anticipated. The European Medicines Agency (EMA) implemented specific activities to ensure enhanced monitoring of emerging vaccine safety information, including intensive monitoring of reports of adverse events of special interest (AESIs) and the use of observed‐to‐expected (O/E) analyses. EMA also commissioned several independent observational studies using a large network of electronic healthcare databases and primary data collection via mobile and web‐based applications. This preparedness was key for two high‐profile safety signals: thrombosis with thrombocytopenia syndrome (TTS), a new clinical entity associated with adenovirus‐vectored vaccines, and myocarditis/pericarditis with messenger RNA (mRNA) vaccines. With no existing case definition nor background rates, the signal of TTS posed particular challenges. Nevertheless, it was rapidly identified, evaluated, contextualised and the risk minimised thanks to close surveillance and an efficient use of available evidence, clinical expertise and flexible regulatory tools. The two signals illustrated the complementarity between spontaneous and real‐world data, the former enabling rapid risk identification and communication, the latter enabling further characterisation. The COVID‐19 pandemic has tremendously enhanced the development of tools and methods to harness the unprecedented volume of safety data generated for the vaccines. Areas for further improvement include the need for better and harmonised data collection across Member States (e.g., stratified vaccine exposure) to support signal evaluation in all population groups, risk contextualisation and safety communication.

Risk of Repeated Adverse Effects following Booster Dose of mRNA COVID-19 Vaccine: Results from the MOSAICO Study

“…The safety profile of COVID-19 vaccines is under continuous evaluation, with the vast majority of known adverse effects following immunization (AEFI) being minor and self-limiting [ 9 , 10 , 11 ]. Indeed, the investigation of adverse effects following immunization (AEFI) is key to monitoring the safety of vaccination series and acceptance of repeated doses in general and in specific populations [ 12 , 13 , 14 ]. Understanding AEFI predictors after COVID-19 vaccines gives the opportunity to better design prevention strategies and encourage trust building in vaccination campaigns against COVID-19 [ 2 , 14 , 15 ].…”

Interpretation of Pharmacovigilance Disproportionality Analyses

“…Charles Khouri 1,2, * , Michele Fusaroli 3 , Francesco Salvo 4,5 and Emanuel Raschi 3 We read with interest the study by Prontskus and colleagues, 1 which analyzed the French and the World Health Organization (WHO) spontaneous reporting systems to investigate the potential link between coronavirus disease 2019 (COVID-19) vaccination and rheumatic arthritis onset and flares. The research question and its implications are important.…”

“…In the field of vaccine safety, pharmacoepidemiological studies using healthcare databases, active intense monitoring, and observed-to-expected analyses are warranted to support signal assessment, relevant contextualization, and communication. 3 Previous meta-epidemiological studies have shown common misinterpretation (also called spin) of results from disproportionality analyses in published studies. 2 Many disproportionality analyses intentionally or unintentionally overstate the strength of causal links, lack proper handling and discussion of biases, or over-extrapolate results to provide clinical recommendations or compare drug safety profiles.…”

Challenges in Assessing COVID-19 Vaccines Safety Signals—The Case of ChAdOx1 nCoV-19 Vaccine and Corneal Graft Rejection

“…Regular updates of the AESI list, including used case definitions (e.g., Brighton collaboration), remain necessary. Unfortunately, for safety signals dealing with unexpected events, the assessments had to deal with more uncertainties and a lack of accurate background incidence rates adequately stratified by age and geographical regions [ 17 ].…”