Background: Influenza (flu) vaccination is routinely advised for adults aged≥65 years. Previous studies have suggested that flu vaccination can have secondary benefits in protecting against adverse cardiovascular events. However, it is not known whether lipid modifying agents might influence any potential benefits of the vaccine on cardiovascular events.
Objectives: We aimed to estimate real-world effectiveness of the influenza vaccine against myocardial infarction (MI) in older adults taking statins.
Methods: We identified annual cohorts from 2010-18 of adults aged ≥ 65 years and taking statins, without previous flu vaccination, from UK general practices, registered with the Clinical Practice Research Datalink. The primary outcome was hospitalisation for MI. We used the self-controlled case series method to remove the risk of bias from time-invariant confounding. For each flu season, the observation period ran from September 1 for 12 months or loss to follow-up or death. Only individuals who experienced an MI and were vaccinated during the study period were included. Exposed person time was categorised into risk windows: (1) 1-14 days before being vaccinated (pre-vaccination ‘risk’ period), (2) 1-14 days after vaccination, (3) 15-90 days after vaccination. All remaining time within the study period was used as the unexposed control period. The models were adjusted for the effects of age and season. Further analyses were conducted using the Prior Event Rate Ratio (PERR) method, an alternative approach to within-patient control of time-invariant confounding.
Results: There were 305 vaccinated patients taking statins who experienced an MI during the 2010/11, 2011/12 and 2012/2013 flu seasons. Before accounting for seasonality, the risk of MI was significantly reduced in the 14-90 days following flu vaccination (Incident rate ratio, IRR=0.59, 95% CI=0.37-0.93, p=0.023). After accounting for season, the effect of vaccination was attenuated (IRR=0.73, 95% CI=0.43-1.22, p=0.231). There was strong evidence of a healthy vaccinee effect with reduced risk of MI in the 14 days before vaccination (IRR=0.09, 95% CI=0.01-0.78, p=0.028).
Conclusions: After adjustment for time-varying confounding and seasonality, flu vaccination was not associated with a reduced risk of MI in patients on ongoing statin therapy across three influenza seasons. However, the IRR point estimate of 0.73 is consistent with previous studies showing a protective effect of vaccination in cardiovascular patients including a recent meta-analysis of randomised trials with pooled relative risk of 0.74. Our results show the importance of considering healthy vaccinee bias. Aggregate results for 2010-2018 including triangulation of evidence with the PERR method will be presented.
