Methods: Children aged 18 months – 8 years with medically attended acute respiratory illness were recruited during 2007-08, 2010-11, 2011-12, and 2012-13 influenza seasons. Influenza status was determined by RT-PCR. Influenza vaccination dates since birth were obtained from a validated vaccine registry. Logistic regression models adjusted for age, season, and calendar time were used to assess the association between priming history and risk of A/H3N2 influenza. We used separate models for priming history based on total doses received (0, 1, ≥2) and antigenically similar H3 doses received (0, 1, ≥2). Children who received zero or two doses during the current season or live attenuated vaccine in any season were excluded.
Results: A total of 609 vaccinated children were included; 91 (15%) had influenza H3N2. Children received a median of 1 antigenically similar dose [interquartile range (IQR): 0, 2] and 3 total doses (IQR: 2, 4) in the past. 480 (24%) received antigenically similar H3 only, 207 (34%) received antigenically distinct H3 only, 219 (36%) received both antigenically similar and distinct H3, and 35 (6%) had not received any influenza vaccine in the past. There were no differences in risk of H3 influenza by total doses (any antigenic group) received [odd ratio, OR (95% confidence interval, CI): 1.9 (0.4, 8.7) and 1.4 (0.4, 4.3) for 1 and ≥2 doses compared to 0 doses, respectively]. The number of antigenically similar doses received also did not influence risk of vaccine failure [OR (95% CI): 1.8 (0.8, 4.1) and 1.3 (0.5, 3.4) for 1 and ≥2 doses compared to 0 doses, respectively].
Conclusion: Risk of A/H3N2 vaccine failure among young children vaccinated with inactivated influenza vaccine did not differ with regard to priming history. Further studies should examine the importance of priming history for H1 and B influenza to inform vaccine policy.
E. Belongia, Medimmune: Investigator , Research support