Variable influenza vaccine effectiveness by subtype: a systematic review and meta-analysis of test-negative design studies

Edward A. Belongia, Melissa D. Simpson, Jennifer P. King, Maria E. Sundaram, Nicholas S. Kelley, Michael T. Osterholm, Huong Q. McLean

Research output: Contribution to journalArticlepeer-review

494 Scopus citations


Background Influenza vaccine effectiveness (VE) can vary by type and subtype. Over the past decade, the test-negative design has emerged as a valid method for estimation of VE. In this design, VE is calculated as 100% × (1 – odds ratio) for vaccine receipt in influenza cases versus test-negative controls. We did a systematic review and meta-analysis to estimate VE by type and subtype. Methods In this systematic review and meta-analysis, we searched PubMed and Embase from Jan 1, 2004, to March 31, 2015. Test-negative design studies of influenza VE were eligible if they enrolled outpatients on the basis of predefined illness criteria, reported subtype-level VE by season, used PCR to confirm influenza, and adjusted for age. We excluded studies restricted to hospitalised patients or special populations, duplicate reports, interim reports superseded by a final report, studies of live-attenuated vaccine, and studies of prepandemic seasonal vaccine against H1N1pdm09. Two reviewers independently assessed titles and abstracts to identify articles for full review. Discrepancies in inclusion and exclusion criteria and VE estimates were adjudicated by consensus. Outcomes were VE against H3N2, H1N1pdm09, H1N1 (pre-2009), and type B. We calculated pooled VE using a random-effects model. Findings We identified 3368 unduplicated publications, selected 142 for full review, and included 56 in the meta-analysis. Pooled VE was 33% (95% CI 26–39; I2=44·4) for H3N2, 54% (46–61; I2=61·3) for type B, 61% (57–65; I2=0·0) for H1N1pdm09, and 67% (29–85; I2=57·6) for H1N1; VE was 73% (61–81; I2=31·4) for monovalent vaccine against H1N1pdm09. VE against H3N2 for antigenically matched viruses was 33% (22–43; I2=56·1) and for variant viruses was 23% (2–40; I2=55·6). Among older adults (aged >60 years), pooled VE was 24% (−6 to 45; I2=17·6) for H3N2, 63% (33–79; I2=0·0) for type B, and 62% (36–78; I2=0·0) for H1N1pdm09. Interpretation Influenza vaccines provided substantial protection against H1N1pdm09, H1N1 (pre-2009), and type B, and reduced protection against H3N2. Vaccine improvements are needed to generate greater protection against H3N2 than with current vaccines. Funding None.

Original languageEnglish (US)
Pages (from-to)942-951
Number of pages10
JournalThe Lancet Infectious Diseases
Issue number8
StatePublished - Aug 1 2016


Dive into the research topics of 'Variable influenza vaccine effectiveness by subtype: a systematic review and meta-analysis of test-negative design studies'. Together they form a unique fingerprint.

Cite this