Vaccine effectiveness against influenza-associated hospitalization among children aged < 13 years using a hospital-based surveillance system in Minnesota, 2013 – 2016

BACKGROUND: Due to marked variability in circulating influenza viruses each year, annual evaluation of the vaccine’s effectiveness against severe outcomes is essential. We used the Minnesota Department of Health’s (MDH) Severe Acute Respiratory Illness (SARI) surveillance to evaluate vaccine effectiveness (VE) against influenza-associated hospitalization over three influenza seasons. METHODS: Residual respiratory specimens from patients admitted with SARI were sent to the MDH laboratory for influenza RT-PCR testing. Medical records were reviewed to collect patient data. Vaccination history was verified using the state immunization registry. We included patients aged ≥6 months to < 13 years, after which immunization reporting is not required, hospitalized from the earliest influenza detection after July through April each year. We defined vaccinated patients as those ≥1 dose of influenza vaccine in the current season. Children aged < 9 years with no history of vaccination were considered vaccinated if 2 were doses given a month apart. Partially vaccinated children were excluded. We estimated VE as 1 minus the adjusted odds ratio (x100%) of influenza vaccination among influenza cases vs. negative controls, controlling for age, race, days from onset to admission, comorbidities, and admission month. RESULTS: Among 2198 SARI patients, 763 (35%) were vaccinated for influenza, 180 (8.2%) were partially vaccinated, and 1255 (57%) were unvaccinated. Influenza was detected among 202 (9.2%) children, and significantly more frequently among children aged ≥5 years (17%) compared with younger children (7.4%). The adjusted VE in 2013–14 was 68% (95% Confidence Interval: 34, 85), but was non-significant during the 2014–15 and 2015–16 seasons (Figure). Estimates of VE by influenza A subtypes varied substantially by year; VE against influenza B viruses was significant, but could not be stratified by year. VE was impacted when live attenuated influenza vaccine recipients were excluded. CONCLUSION: We report moderately high influenza VE in 2013–14 and a point estimate higher than other published estimates from outpatient data in 2014–15. These results, underscore the importance of influenza vaccination to prevent severe outcomes such as hospitalization. DISCLOSURES: All authors: No reported disclosures.