Adverse events of interest vary by influenza vaccine type and brand: Sentinel network study of eight seasons (2010–2018)
Cross JW., Joy M., McGee C., Akinyemi O., Gatenby P., de Lusignan S.
Background: Influenza contributes significantly to the burden of disease worldwide; the United Kingdom has a policy of vaccination across all ages. Influenza vaccinations are known to be associated with common minor adverse events of interest (AEIs). The European Medicines Agency (EMA) recommends ongoing surveillance of AEIs following influenza vaccination to monitor common and detect infrequent but important AEIs. Methods: A retrospective cohort study using computerised medical record data from the Royal College of General Practitioners (RCGP) Research and Surveillance Centre (RSC) sentinel network database 2010–2018 (N = 848,375). We extracted data about vaccine exposure (n = 3,121,334) and consultations for AEIs within seven days of receiving vaccinations specified by the EMA (1,488,870 consultations by 430,029 unique individuals). We used a self-case series design which employs a likelihood estimation method using conditioning of observed adverse events. Such a model assumes non-homogenous Poisson intensity processes for each exposure period and age interval. We compared AEI between QIV and TIV reporting relative incidence (RI) of AEIs. A RI < 1 signified lower AEI rate compared to TIV. Results: QIV was associated with a RI of AEIs of 1.14 (95%CI, 1.10–1.18, p < 0.01), though the number of years exposure was limited. By way of contrast, LAIV had a lower rate 0.60 (95%CI 0.63–0.68, p < 0.001). Cellular manufacture was also associated with a lower rate 0.78 (95%CI 0.61–0.99, p = 0.04). AEIs varied by season: Rash and musculoskeletal conditions are particularly pronounced in the 2014/15 season and respiratory conditions in 2016/17. In an analysis of all seasons, we found an elevated relative incidence of AEIs of 1.78 (95%CI, 1.62–1.95) in pregnant women and 1.76 (95%CI, 1.56 – 1.99) in children under 5 years. Conclusion: Routine sentinel network data can be used to contrast AEIs between vaccine types and may provide a consistent method of observation of vaccine benefit-risk over time.