A Retrospective Test-Negative Case-Control Study to Evaluate Influenza Vaccine Effectiveness in Preventing Hospitalizations in Children

Background. Vaccination is the primary strategy to reduce influenza burden. Influenza vaccine effectiveness (VE) can vary annually depending on circulating strains. Methods. We used a test-negative case-control study design to estimate influenza VE against laboratory-confirmed influenza-related hospitalizations among children (aged 6 months-17 years) across 5 influenza seasons in Atlanta, Georgia, from 2012-2013 to 2016-2017. Influenza-positive cases were randomly matched to test-negative controls based on age and influenza season in a 1:1 ratio. We used logistic regression models to compare odds ratios (ORs) of vaccination in cases to controls. We calculated VE as [100% x (1 - adjusted OR)] and computed 95% confidence intervals (CIs) around the estimates. Results. We identified 14 596 hospitalizations of children who were tested for influenza using the multiplex respiratory molecular panel; influenza infection was detected in 1017 (7.0%). After exclusions, we included 512 influenza-positive cases and 512 influenza-negative controls. The median age was 5.9 years (interquartile range, 2.7-10.3), 497 (48.5%) were female, 567 (55.4%) were non-Hispanic Black, and 654 (63.9%) children were unvaccinated. Influenza A accounted for 370 (72.3%) of 512 cases and predominated during all 5 seasons. The adjusted VE against influenza-related hospitalizations during 2012-2013 to 2016-2017 was 51.3% (95% CI, 34.8% to 63.6%) and varied by season. Influenza VE was 54.7% (95% CI, 37.4% to 67.3%) for influenza A and 37.1% (95% CI, 2.3% to 59.5%) for influenza B. Conclusions. Influenza vaccination decreased the risk of influenza-related pediatric hospitalizations by>50% across 5 influenza seasons.