Association of Prior Vaccination With Influenza Vaccine Effectiveness in Children Receiving Live Attenuated or Inactivated Vaccine.

Importance: Some studies have reported negative effects of prior-season influenza vaccination. Prior-season influenza vaccination effects on vaccine effectiveness (VE) in children are not well understood. Objective: To assess the association of prior-season influenza vaccination with subsequent VE in children aged 2 to 17 years. Design, Setting, and Participants: This multiseason, test-negative case-control study was conducted in outpatient clinics at 4 US sites among children aged 2 to 17 years with a medically attended febrile acute respiratory illness. Participants were recruited during the 2013-2014, 2014-2015, and 2015-2016 seasons when influenza circulated locally. Cases were children with influenza confirmed by reverse-transcription polymerase chain reaction. Test-negative control individuals were children with negative test results for influenza. Exposures: Vaccination history, including influenza vaccine type received in the enrollment season (live attenuated influenza vaccine [LAIV], inactivated influenza vaccine [IIV], or no vaccine) and season before enrollment (LAIV, IIV, or no vaccine), determined from medical records and immunization registries. Main Outcomes and Measures: LAIV and IIV effectiveness by influenza type and subtype (influenza A[H1N1]pdm09, influenza A[H3N2], or influenza B), estimated as 100 × (1 - odds ratio) in a logistic regression model with adjustment for potential confounders. Prior season vaccination associations were assessed with an interaction term. Results: Of 3369 children (1749 [52%] male; median age, 6.6 years [range, 2-17 years]) included in the analysis, 772 (23%) had a positive test result for influenza and 1674 (50%) were vaccinated in the enrollment season. Among LAIV recipients, VE against influenza A(H3N2) was higher among children vaccinated in both the enrollment and 1 prior season (50.3% [95% CI, 17.0% to 70.2%]) than among those without 1 prior season vaccination (-82.4% [95% CI, -267.5% to 9.5%], interaction P < .001). The effectiveness of LAIV against influenza A(H1N1)pdm09 was not associated with prior season vaccination among those with prior season vaccination (47.5% [95% CI, 11.4% to 68.9%]) and among those without prior season vaccination (7.8% [95% CI, -101.9% to 57.9%]) (interaction P = .37). Prior season vaccination was not associated with effectiveness of IIV against influenza A(H3N2) (38.7% [95% CI, 6.8% to 59.6%] among those with prior-season vaccination and 23.2% [95% CI, -38.3% to 57.4%] among those without prior-season vaccination, interaction P = .16) or with effectiveness of IIV against influenza A[H1N1]pdm09 (72.4% [95% CI, 56.0% to 82.7%] among those with prior season vaccination and 67.5% [95% CI, 32.1% to 84.4%] among those without prior season vaccination, interaction P = .93). Residual protection from prior season vaccination only (no vaccination in the enrollment season) was observed for influenza B (LAIV: 60.0% [95% CI, 36.8% to 74.7%]; IIV: 60.0% [36.9% to 74.6%]). Similar results were observed in analyses that included repeated vaccination in 2 and 3 prior seasons. Conclusions and Relevance: Influenza VE varied by influenza type and subtype and vaccine type, but prior-season vaccination was not associated with reduced VE. These findings support current recommendations for annual influenza vaccination of children.