Research in context
Evidence before this study
In March, 2014, we did an informal review of the literature by searching PubMed for original reports of influenza vaccine effectiveness (VE) published in English from 1990 to 2013. We restricted our review to studies that reported single-season VE against H3N2, H1N1, H1N1pdm09, or type B. To minimise potential bias, we further restricted our preliminary review to studies using the test-negative design with outpatient recruitment based on predefined criteria, those that had confirmation of influenza with RT-PCR or culture, and those that had age adjustment. We identified 43 publications that met these criteria, leading to a decision to do a formal meta-analysis. Our preliminary review indicated that the earliest test-negative design study of influenza VE was conducted in 2004–05, and the formal meta-analysis was therefore restricted to the period from Jan 1, 2004, to March 31, 2015.
Added value of this study
Findings from this study show substantial variation in VE across influenza types and subtypes. Influenza vaccine provided moderate to high protection against H1N1pdm09, H1N1 (pre-2009), and type B, and substantially lower protection against H3N2. Differences across age groups were minimal for H1N1pdm09 and type B. VE against H3N2 was highest in paediatric age groups and lowest in older adults. VE against H3N2 was low regardless of reported antigenic match, but this comparison was limited by the absence of standardised antigenic characterisation and information about antigenic distance. In this systematic review and meta-analysis, we found that relevant information about patient recruitment, symptom eligibility, and vaccine ascertainment was inconsistently reported, and we have made recommendations to optimise VE methods in the outpatient setting. These recommendations are consistent with draft recommendations being developed by WHO.
Implications of all the available evidence
H3N2 is associated with higher morbidity and mortality than are other subtypes, and vaccine improvements are needed to generate greater protection against H3N2 than against other subtypes. Alternatives to egg-based manufacturing should be pursued since egg-induced mutations in H3N2 vaccine strains contribute to antigenic mismatch.