1917. Variable Influenza Vaccine Effectiveness by Type and Subtype: Meta-analysis of Studies Using the Test-negative Design
Session: Poster Abstract Session: Vaccines: Influenza
Saturday, October 10, 2015
Room: Poster Hall
  • Belongia.Meta-analysis poster.pdf (496.1 kB)
  • Background: Over the past decade, the test negative design has emerged as a valid and widely used method for estimating influenza vaccine effectiveness (VE). Patients with acute respiratory illness are systematically enrolled and tested for influenza. VE can be calculated separately for each component as 100% x (1- odds ratio) for vaccine receipt in influenza cases vs. test negative controls with adjustment for confounders. We conducted a meta-analysis to estimate component-specific VE.

    Methods: We searched Ovid MEDLINE and EMBASE for VE studies. Two authors independently screened each title and abstract, and potential VE studies were selected for in-depth review. Studies were eligible if they used the test-negative design, included systematic recruitment of outpatients, confirmed influenza by RT-PCR, and reported VE with age adjustment. We abstracted VE estimates for A/H3N2, A/H1N1, pandemic A/H1N1 (pH1N1), and B. Selected articles were independently abstracted by two reviewers and discrepancies adjudicated. We excluded duplicate reports and studies restricted to hospitalized patients or special populations. We assessed study heterogeneity and calculated pooled VE using a random effects model that accounted for within- and between-study variance.

    Results: We screened 3,629 published articles and selected 142 for full review; 60 articles and 126 VE estimates (without age restriction) met eligibility criteria and were included in the primary meta-analysis. There were 10 VE estimates for monovalent pH1N1 vaccine, 33 for seasonal pH1N1 vaccine, 39 for H3N2, 37 for type B, and 7 for pre-pandemic H1N1. There was a high level of heterogeneity (p<0.0001).

    Vaccine Type

    Antigenic Component

    Pooled VE

    95% CI

    Pooled Standard Error

    Tri or Quadrivalent



    (55, 68)


    Tri or Quadrivalent



    (28, 48)


    Tri or Quadrivalent



    (58, 68)





    (61, 86)



    Pre-pandemic H1N1


    (-23, 94)


     For age groups ≥60 years, there were 5 VE estimates for H3N2 and 3 each for B and pH1N1. Pooled VE was 34% (-11, 60) for H3N2, 63% (53, 71) for B, and 64% (51, 74) for pH1N1.

    Conclusion: Influenza vaccines provided moderate protection against medically attended pH1N1 and type B illness, including older adults. Vaccine effectiveness against H3N2 was substantially lower.

    Edward Belongia, MD1, Melissa Simpson, PhD, DVM1, Jennifer King, MPH1, Maria Sundaram, MSPH2, Nicholas Kelley, PhD3, Michael Osterholm, PhD, MPH, FIDSA4 and Huong Mclean, PhD1, (1)Center for Clinical Epidemiology and Population Health, Marshfield Clinic Research Foundation, Marshfield, WI, (2)University of Minnesota School of Public Health, Minneapolis, MN, (3)University of Minnesota School of Public Health, Minn, MN, (4)University of Minnesota, Minneapolis, MN


    E. Belongia, Medimmune: Investigator , Research support

    M. Simpson, None

    J. King, None

    M. Sundaram, None

    N. Kelley, None

    M. Osterholm, None

    H. Mclean, Medimmune: Investigator , Research support

    See more of: Vaccines: Influenza
    See more of: Poster Abstract Session

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 7th with the exception of research findings presented at the IDWeek press conferences.