Young children are especially vulnerable to influenza virus infections and along with older adults are the most likely to suffer secondary complications and be hospitalized due to influenza illness.1–5 However, relatively few studies have examined the effectiveness of inactivated trivalent influenza vaccines (IIV3) in this age group.6,7 Over the past decade, the Advisory Committee on Immunization Practices (ACIP) of the CDC in the US has progressively expanded who should receive annual influenza vaccination. Groups targeted for vaccination expanded from high-risk children only to including all children aged 6–23 months (starting as “encouraged when feasible” in 2002 and recommended since 2004),8–10 children aged 6–59 months (since 2006),11 and children aged 6 months to 18 years (since 2008).12
ACIP recommends that children aged 6 months to 8 years receive 2 doses of influenza vaccine if they are previously unvaccinated; in 2010, ACIP revised their recommendations to ensure 2 prior doses containing the A(H1N1)pdm09 virus component (see Figure, Supplemental Digital Content 1, http://links.lww.com/INF/C321). Admittedly, evidence that can inform these recommendations is limited.6,7,13 Improved immunogenicity following 2 doses of IIV3 has been noted, especially among vaccine naïve infants and toddlers, but the findings have not been consistent for children aged ≥2 years or for all vaccine components.14–20
Findings from observational studies regarding vaccine effectiveness (VE) in preventing laboratory-confirmed influenza outcomes have also been mixed. Some studies have reported significant VE for fully vaccinated children only,21–24 while others have observed similar VE for fully and partially vaccinated children in some age groups.23,25–28 Most of these studies have examined VE against all circulating influenza strains21,23,25,26; few have been able to consider the importance of full vaccination against a specific influenza type or subtype.19,24,27,29,30
Interpreting these findings is difficult because the criteria for full and partial vaccination consist of a variety of potential vaccine exposures. Children who received 2 doses within the same season (4 weeks apart) have consistently been described as fully vaccinated. However, a single dose can also result in full vaccination depending on vaccination history.7,11,31–33 Until 2007, a child with a single current season dose and at least 1 previous seasonal influenza vaccine dose was considered fully vaccinated. After 2007, a single current season dose after 2 or more doses in previous seasons (in specific combinations that have been updated each season by ACIP) constituted full vaccination.7,11,31–33 The definition of “unvaccinated” also varies across studies, with some categorizing partially vaccinated children as unvaccinated.34 Thus, it is difficult to determine the preventive benefit of particular vaccination schedules from VE studies as typically reported.
The US Influenza Vaccine Effectiveness (Flu VE) Network conducts annual evaluations of VE of seasonal influenza vaccines in preventing outpatient medical care for acute respiratory illness (ARI) with laboratory-confirmed influenza virus infections.21,35 In this study, we examine the VE of IIV3 against medically attended influenza ARI among children aged 6 months to 8 years during the relatively mild 2011–2012 influenza season36 and the moderately severe 2012–2013 season in the US.37
Specifically, our study has 3 aims. First, we estimate VE of full and partial IIV3 vaccination using the ACIP criteria for the 2 seasons. We improve upon our previous VE reports for this age group by limiting our analysis to IIV3 and excluding trivalent live-attenuated influenza vaccine (LAIV3),21,35 by considering prior doses received during multiple previous seasons,21 and by reporting VE for both fully and partially vaccinated children in both seasons.35 Second, we estimate VE of IIV3 by the number of current season doses and by alternative definitions of priming that consider doses received in the previous season, doses received across multiple prior seasons and receipt of 2 doses in at least 1 prior season. Third, we estimate VE against A(H3N2) and B influenza virus illnesses separately.
MATERIALS AND METHODS
Details of the US Flu VE Network’s methodology and enrollment during the 2011–201221 and 2012–2013 season35 have been previously published. Patients seeking outpatient medical care for an ARI (onset ≤7 days) with cough were recruited by study staff in 5 states (Michigan, Pennsylvania, Texas, Washington and Wisconsin). Consenting parents completed an enrollment interview to provide patient and household information. Patient’s history of high risk medical conditions was confirmed by extraction from medical records. Combined nasal and throat specimens (or nasal specimens only for children aged <2 years) were tested for influenza at network laboratories, using real-time reverse transcription polymerase chain reaction (RT-PCR). The study was reviewed and approved by the institutional review boards at each Network site.
Influenza vaccination status for the study season and for prior seasons was documented from medical records and immunization registries. Participants with indeterminate current season vaccination status (with illness onset <14 days from vaccination or receipt of a second dose of vaccine <28 days from the first dose) and those who received LAIV3 during the study season were excluded. For both seasons, we extracted dates of IIV3 vaccination for the current study season, up to 3 doses of any influenza vaccine from previous seasons, and the earliest recorded influenza vaccination. Thus, up to 4 prior season doses were documented, but this likely underestimated the number of children who received 2 prior doses within the same prior season, especially among repeatedly vaccinated older children.
Because we could not distinguish between seasonal and monovalent A(H1N1)pdm09 vaccines in 2009, we considered doses after October 1, 2009 as monovalent A(H1N1)pdm09 vaccines; monovalent vaccine was introduced and seasonal vaccination declined in early October.38,39 Receipt of the monovalent vaccine was considered as it applied to ACIP criteria, but it was not included in the count of prior seasonal vaccinations. There was little change in the percentage of primed children or in our VE results when we used alternative November and December cut-offs for counting monovalent A(H1N1)pdm09 vaccinations.
For the 2012–2013 season only, we were able to confirm whether the child had consistently been enrolled in the site’s health system since 6 months of age. For this season, we excluded children aged <2 years who lacked medical records going back at least 1 year and excluded children 2–8 years without records going back at least 2 years.
Vaccine components for the study years and prior seasons are listed in Table, Supplemental Digital Content 2, http://links.lww.com/INF/C322; the A(H3N2) and B influenza virus components changed from 2011–2012 to 2012–2013.36 The vast majority of tested circulating viruses in the US were antigenically similar to vaccine prototype components during both seasons.36,37 However, antigenic differences were noted between circulating viruses and the egg-adapted A(H3N2) strain in the 2012–2013 vaccine.40
Influenza VE is assessed using a test-negative design,41,42 whereby VE equals 100% × (1 − odds ratio [ratio of odds of vaccination among influenza-positive cases to the odds of vaccination among influenza-negative controls]) using logistic regression. Similar to previous VE estimates for this Network,21,25,35 adjusted models included values for age (months), race/ethnicity, high-risk medical conditions, days from illness onset to specimen collection, study site and calendar time (month of enrollment). Statistically significant or nonzero VE is indicated by a 95% confidence interval that does not overlap with 0.
We define full and partial vaccination using the ACIP criteria for each study season33,43 (see Figure, Supplemental Digital Content 1, http://links.lww.com/INF/C321). We then examine VE using 4 definitions of vaccine priming: ≥1 doses of any influenza vaccine during the prior season, ≥1 doses recorded across prior seasons, ≥2 doses recorded across prior seasons and receipt of 2 doses within the same prior season. To test for priming, we examine the conditional hypothesis that the effect of current season vaccine varies depending on prior vaccination history; we examine an interaction term for receipt of current season IIV3 by receipt of prior doses after adjusting for both contributing terms.21,44,45 We estimate VE for each combination of current season and prior doses with those unvaccinated in the current season and not meeting the priming criteria as the reference group.21,44–46 We also test for differences between VE estimates (eg, VE of fully vs. partially vaccinated; VE for primed vs. unprimed children) by calculating the odds ratio with one of the 2 vaccine exposures as the referent.
We estimate that 89 cases are needed to achieve 80% power (α = 0.05) to detect a VE of 50% with an influenza positivity rate of 33% and a vaccination rate of 50% among influenza negative controls. We meet or exceed these criteria in most estimates of VE for fully vaccinated children and those with common prior vaccination exposures (eg, current season dose plus 1 or more prior doses). However, we have lower power to estimate VE for partially vaccinated children and for those with less common prior vaccination exposures. We consider as noteworthy consistent patterns of results across models, but there is no adjustment for family wise error. Statistical analyses were conducted using SAS statistical software (version 9.3; SAS Institute, Cary, NC).
During 2011–2012, 1562 children aged 6 months to 8 years were enrolled during weeks with local influenza virus circulation at the 5 Network sites; 30 (2%) with indeterminate vaccination status and 91 (6%) who received LAIV3 were excluded, resulting in an analysis sample of 1441 children (see Figure, Supplemental Digital Content 3, http://links.lww.com/INF/C323). During 2012–2013, 1732 children were enrolled; 227 (13%) with incomplete medical records, 30 (2%) with indeterminate vaccination status and 148 (9%) who received LAIV3 were excluded, resulting in an analysis sample of 1327 children (see Figure, Supplemental Digital Content 4, http://links.lww.com/INF/C324).
In the 2011–2012 season, influenza circulation began late and peaked in March and April; only 13% (190/1441) of enrolled children were influenza positive (Network site range = 6%–17%). In contrast, during the 2012–2013 season, circulation peaked in December and January, and influenza positives made up 38% (499/1327) of enrollees (site range = 8%–55%). In both seasons, A(H3N2) viruses were prominent at most sites; influenza positivity increased with age and was higher among Black children (Table 1).
The proportion receiving ≥1 doses of IIV3 was similar in 2011–2012 and 2012–2013 (50% and 45%, respectively). However, the percentage of enrolled children who were vaccinated increased per month in 2012–2013, climbing from 34% in December to 54% in March (Table 1). In both seasons, vaccine coverage declined with age and was lower among Black children.
Potential confounders (associated with both vaccination and influenza positivity) identified in the descriptive analyses were age, race, month of enrollment and Network site. All of these potential confounders were already included a priori in the adjusted models.
VE for Full and Partial Vaccination Using ACIP Criteria
The majority of children who received at least 1 dose of IIV3 in 2011–2012 and 2012–13 had sufficient prior doses to meet the definition of fully vaccinated according to ACIP criteria (81%, 591/726 and 79%, 473/596, respectively). Therefore, in analyses that combined children from both seasons, the vast majority of fully vaccinated children (79%, 845/1064) had received a single dose of IIV3; the other 21% (219/1064) received 2 recommended doses, and the vast majority of these were less than 2 years old (88%, 193/219).
We observed significant unadjusted VE for both partial and full vaccination associated with all circulating influenza viruses (both seasons), A(H3N2) viruses (both seasons) and B influenza viruses (2012–2013 season; Table 2). Adjusted VE point estimates were 11–31 percentage points lower than the crude estimates. VE for full vaccination remained significant in all models, but VE for partial vaccination was only significant in the 2012–2013 models.
VE estimates for fully vaccinated children were not significantly higher than VE for partially vaccinated children in any of the models. In fact, the only significant difference noted was opposite of the expected direction; in 2012–2013, adjusted VE against A(H3N2) illness was significantly lower among fully vaccinated children (Table 2).
VE with Different Prior Vaccination Exposures
Among children aged 2–8 years, we examined 4 definitions of influenza vaccine priming (Table 3). Because VE estimates associated with 1 versus 2 current season doses of IIV3 were similar (with wide and overlapping CIs; see Table, Supplemental Digital Content 5, http://links.lww.com/INF/C325), we defined current season vaccination as receipt of 1 or more doses for the following priming analyses. Without priming, we observed adjusted VE for ≥1 current season dose(s) of IIV3 that ranged from 47% to 53% across the 4 influenza outcomes (Table 3).
In the first analysis, priming consisted of 1 or more doses of any influenza vaccine in the previous season only. We observed significant effect modification of current season VE by prior season priming in all models (ie, interaction terms P values < 0.1). This cross-season effect is driven by significant residual VE for children who received no current season IIV3 but were vaccinated the previous season; this was noted in all models (VE range = 36%–40%).
In the second and third analyses, we expanded the definition of priming to include ≥1 and then ≥2 doses recorded in any prior season (Table 3). However, we did not observe a significant interaction effect for current season IIV3 by prior vaccination history using either of these definitions. Nonetheless, for most of these models, the only significant VE was observed for children with current and prior seasonal doses.
In the fourth analysis, only children who had received 2 doses in the same prior season were considered primed. We observed significant interaction terms for current season IIV3 by 2-dose priming history against influenza outcomes in all four models (P values < 0.05). Although we observed significant VE for current season IIV3 with or without priming, the VE point estimates were consistently higher for children who had received 2 doses in a prior season compared with those without (VE range of 58%–80% vs. 33%–42%, respectively; Table 3). This was statistically significant for VE against A(H3N2) illness; the odds of being A(H3N2) positive were 2.4-fold (95% confidence interval: 1.4–4.3) higher among children who received IIV3 but were unprimed compared with children who received IIV3 after receiving 2 dose priming in a previous season. Although the referent category in this model included unvaccinated children with no record of prior vaccination and those with prior vaccinations but without 2 doses within the same season, we observed similar VE estimates when the referent group was limited to only unvaccinated children with no known prior vaccination (see Table, Supplemental Digital Content 6, http://links.lww.com/INF/C326).
In our study of children aged 6 months to 8 years receiving outpatient medical care for ARI during 2011–2012 and 2012–2013, we found that children aged 2–8 years who received IIV3 after receiving 2 priming doses of any influenza vaccine during the same previous season had significantly higher VE against A(H3N2) illness than children without 2 priming doses. Children with 2-dose priming also had high VE point estimates against ARI associated with all influenza viruses (in both seasons) and B viruses (2012–2013). The effect was most evident when we examined vaccine failure against A(H3N2) illness; among children aged 2–8 years who received IIV3, the odds of A(H3N2) illness despite vaccination were 2.4-fold higher among unprimed children compared with those who received 2 doses of vaccine in a previous season. Our findings are consistent with studies that observed better IIV3 immunogenicity following 2 doses, although this has been demonstrated most consistently among vaccine naïve infants and children younger than those in our analyses.14–20 Nonetheless, our findings add support to recommendations by multiple international organizations to offer 2 doses of influenza vaccine during a child’s first season33,43 and suggests the preventive benefits of this initial investment may extend to future seasons.
We did not observe higher VE for children fully versus partially vaccinated with IIV3 as defined by ACIP in any of our models. The relatively small number of partially vaccinated children combined with the need to adjust for potential confounding (by age, race/ethnicity, geography and simultaneous changes in vaccine coverage and influenza circulation) made VE estimates for partially vaccinated children imprecise, especially for 2011–2012. Admittedly, the relevance of these observations is limited because during the study period ACIP recommendations for children aged 6 months to 8 years considered doses of vaccine containing the A(H1N1)pdm09 component specifically; yet, we observed few A(H1N1)pdm09 virus infections in our study.
Nonetheless, our finding of significant VE for both full and partial vaccination against ARI associated with all influenza viruses, A(H3N2) viruses and B influenza viruses in 2012–2013 is consistent with several prior studies.25–27 In contrast, our observation that partially vaccinated children experienced higher VE against A(H3N2) illness than fully vaccinated children was unexpected and is difficult to interpret. Without a much larger sample, it is impossible to disentangle and examine separately the diverse influenza vaccine exposures that constitute partial versus full vaccination. In the 2012–2013 sample, for example, partially vaccinated children, as defined by US ACIP, include those with a single current season dose and at least 4 kinds of vaccination histories (eg, no prior vaccination, only 1 prior dose, ≥2 prior doses but without at least 1 dose before and 1 dose after July 2010). These differences in vaccine exposure are also likely interconnected with differences in children’s history of natural infections with influenza viruses. Nonetheless, the observed higher VE against A(H3N2) for partially vaccinated children warrants further investigation, especially given recent studies of children aged ≥9 years and adults that reported higher point estimates for VE against influenza A(H3N2) illness among those not vaccinated during the previous season21,40,44,45,47,48 or over the previous 5 seasons.47 These future investigations should include measures of immune markers, such as preseason A(H3N2) antibodies. Indeed, Ohmit et al.’s48 household cohort study observed lower preseason A(H3N2) antibody titers which corresponded to lower VE against A(H3N2) illness among those vaccinated in both the current and prior year compared with those vaccinated in the current season only.
Children who were vaccinated during the previous season but who missed IIV3 vaccination in the study season appeared to benefit from residual protection. VE point estimates of 36%–40% were noted for previously vaccinated children aged 2–8 years in all models; this was noted in both the first study season (2011–2012) when the prior year vaccine components had not changed, as well as in the second study year (2012–2013) when the A(H3N2) and B influenza components had changed.21,35 Our findings are consistent with observations of persistent influenza A virus antibodies in children49 and recent observations of cross-season VE among children27 and adults.21,35,45,47,50
From a methodological standpoint, our findings suggest that reports of VE for children should estimate VE for specific combinations of current and prior season(s)’ vaccine exposure, similar to recent recommendations for the presentation of VE estimates for adults.21,35,45,47 This would also aid in VE comparisons across countries, where definitions of full and partial vaccination vary. In regards to vaccine policy, our findings combined with earlier observations of persistent protection offered by Spring (postseason) doses of IIV317,51 suggest there may be value in considering alternative vaccine schedules that take advantage of opportunities to vaccinate young children months before influenza season, especially when vaccine components are unchanged. This may be especially relevant in countries and regions with year-round or twice-yearly influenza circulation.52
Among the strengths of our study is our use of RT-PCR-confirmed influenza outcomes with a test-negative design41,42,53 in a multisite study using a common protocol and procedures that directly recruited children receiving outpatient care for ARI without relying on clinician referral. Our ability to stratify results for A(H3N2) and B influenza virus illnesses improved the specificity of our results. Other study strengths include our ability to examine multiple years of vaccination records from health plan and state registries and the fact that ACIP recommendations for annual seasonal vaccination had consistently applied to all the children in our study.
Our study is limited by several potential sources of information bias. We only documented up to 4 prior vaccination records. While it is reasonable to assume that most children receiving 2 doses in the same season were vaccine naïve before that, we cannot confirm this. We also could not examine the cumulative total of all influenza vaccinations. Also, we may have falsely classified some of the children as unprimed due to incomplete vaccination records. Although our Network now documents all influenza vaccination records of enrollees, in the current study we could not verify the completeness of medical records for children during the 2011–2012 season. In an attempt to remove monovalent A(H1N1)pdm09 vaccinations, we likely excluded some seasonal influenza vaccinations administered after October 2009. Taken together, these potential sources of information bias would result in underestimation of VE associated with full vaccination or multidose priming, especially among older children for whom more records need to be captured. Thus, although we are confident in the significant effects we observed, these information biases combined with small sample sizes for some vaccine exposures reduced the precision of our estimates and likely limited our ability to detect true differences in VE.
In addition, our study has at least 4 other limitations. First, although we excluded children who received LAIV3 during the study season, we could not determine the vaccine types of all previous vaccinations. However, LAIV3 has accounted for a relatively small proportion of vaccinations in our Network to date.21,25,35 Second, unmeasured or residual confounding may have biased our results in unknown ways, since children who consistently receive recommended influenza vaccination doses may differ from those who do not in susceptibility to influenza, access to medical care and other characteristics. Third, we lack information on past infections and possible immune mechanisms that could explain the differences in clinical protection we observed. Finally, given limitations in our sample size and the number of prior vaccination records we retrieved, we were unable to examine VE associated with other potentially valuable priming histories. For example, future studies should examine whether 2 doses of vaccine in consecutive seasons when A(H3N2) vaccine components were unchanged offer similar priming value in reducing the risk of A(H3N2) illness as the receipt of 2 homologous doses in the same year.
In conclusion, during 2011–2012 and 2012–2013, vaccination with IIV3 reduced the risk of outpatient medical visits for ARI associated with influenza virus infection by about half. Vaccinations received in previous seasons had preventive value in subsequent seasons in the form of residual protection for those who missed vaccination in and priming which boosted the benefit of IIV. Most notably, children primed with 2 doses of influenza vaccine in the previous same season appeared to enjoy even greater preventive benefit from IIV3 and may have reduced their risk of A(H3N2) illness by two-thirds. Taken together, our findings reinforce the importance of examining influenza VE among children within the context of prior vaccination history and priming. Observational studies (and possibly randomized controlled trials in countries without universal vaccination recommendations for children) are needed to confirm the importance of priming doses to immunogenicity and VE and to inform international vaccine policy.
The authors thank Sarah Spencer, PhD, Tricia Nowalk, PhD, David Shay, MD, Jerome Tokars, MD and Lisa Grohskopf, MD for their feedback on early versions of this manuscript. The authors also appreciate the contributions of our research staff at our study sites and the children and parents who participated in this study.
1. Nair H, Brooks WA, Katz M, et al. Global burden of respiratory infections due to seasonal influenza in young children: a systematic review and meta-analysis. Lancet. 2011;378:1917–1930
2. Liu L, Johnson HL, Cousens S, et al. Global, regional, and national causes of child mortality: an updated systematic analysis for 2010 with time trends since 2000. Lancet. 2012;379:2151–2161
3. Jules A, Grijalva CG, Zhu Y, et al. Influenza-related hospitalization and ED visits in children less than 5 years: 2000–2011. Pediatrics. 2015;135:e66–e74
4. Poehling KA, Edwards KM, Griffin MR, et al. The burden of influenza in young children, 2004–2009. Pediatrics. 2013;131:207–216
5. Poehling KA, Edwards KM, Weinberg GA, et al.New Vaccine Surveillance Network. The underrecognized burden of influenza in young children. N Engl J Med. 2006;355:31–40
6. Osterholm MT, Kelley NS, Sommer A, et al. Efficacy and effectiveness of influenza vaccines: a systematic review and meta-analysis. Lancet Infect Dis. 2012;12:36–44
7. Centers for Disease Control. . Prevention and control of seasonal influenza with vaccines. Recommendations of the Advisory Committee on Immunization Practices–United States, 2013–2014. 2013;62:1–43
8. Bridges CB, Harper SA, Fukuda K, et al.Advisory Committee on Immunization Practices. Prevention and control of influenza. Recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep. 2003;52:1–34
9. Harper SA, Fukuda K, Uyeki TM, et al.Centers for Disease Control and Prevention (CDC) Advisory Committee on Immunization Practices (ACIP). Prevention and control of influenza: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep. 2004;53:1–40
10. Bridges CB, Fukuda K, Uyeki TM, et al.Centers for Disease Control and Prevention, Advisory Committee on Immunization Practices. Prevention and control of influenza. Recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep. 2002;51:1–31
11. Smith NM, Bresee JS, Shay DK, et al.Prevention and Control of Influenza: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR recommendations and reports: morbidity and mortality weekly report recommendations and reports/centers for disease control. MMWR Recomm Rep. 2006;55:1–42
12. Fiore AE, Shay DK, Broder K, et al.Centers for Disease Control and Prevention (CDC); Advisory Committee on Immunization Practices (ACIP). Prevention and control of influenza: recommendations of the Advisory Committee on Immunization Practices (ACIP), 2008. MMWR Recomm Rep. 2008;57:1–60
13. Neuzil KM, Englund JA. Influenza vaccine for young children: two doses are better than one. J Pediatr. 2006;149:737–738
14. Phillips CF, Phillips CA, Hodgkin WE, et al. Killed subunit influenza vaccine in children. Pediatrics. 1973;52:416–419
15. Neuzil KM, Jackson LA, Nelson J, et al. Immunogenicity and reactogenicity of 1 versus 2 doses of trivalent inactivated influenza vaccine in vaccine-naive 5-8-year-old children. J Infect Dis. 2006;194:1032–1039
16. Walter EB, Rajagopal S, Zhu Y, et al. Trivalent inactivated influenza vaccine (TIV) immunogenicity in children 6 through 23 months of age: do children of all ages respond equally? Vaccine. 2010;28:4376–4383
17. Englund JA, Walter EB, Fairchok MP, et al. A comparison of 2 influenza vaccine schedules in 6- to 23-month-old children. Pediatrics. 2005;115:1039–1047
18. Schmidt-Ott R, Schwarz T, Haase R, et al. Immunogenicity and reactogenicity of a trivalent influenza split vaccine in previously unvaccinated children aged 6–9 and 10–13 years. Vaccine. 2007;26:32–40
19. Ng S, Ip DK, Fang VJ, et al. The effect of age and recent influenza vaccination history on the immunogenicity and efficacy of 2009–10 seasonal trivalent inactivated influenza vaccination in children. PloS One. 2013;8:e59077
20. Bernstein DI, Zahradnik JM, DeAngelis CJ, et al. Influenza immunization in children and young adults: clinical reactions and total and IgM antibody responses after immunization with whole-virus or split-product influenza vaccines. Am J Dis Chil (1960). 1982;136:513–517
21. Ohmit SE, Thompson MG, Petrie JG, et al. Influenza vaccine effectiveness in the 2011–2012 season: protection against each circulating virus and the effect of prior vaccination on estimates. Clin Infect Dis. 2014;58:319–327
22. Eisenberg KW, Szilagyi PG, Fairbrother G, et al. Vaccine effectiveness against laboratory-confirmed influenza in children 6 to 59 months of age during the 2003–2004 and 2004–2005 influenza seasons. Pediatrics. 2008;122:911–919
23. Shuler CM, Iwamoto M, Bridges CB, et al. Vaccine effectiveness against medically attended, laboratory-confirmed influenza among children aged 6 to 59 months, 2003–2004. Pediatrics. 2007;119:e587–95
24. Ferdinands JM, Olsho LE, Agan AA, et al. Effectiveness of influenza vaccine against life-threatening RT-PCR-confirmed influenza illness in US children, 2010–2012. J Infect Dis. 2014;210:674–683
25. Treanor JJ, Talbot HK, Ohmit SE, et al.US Flu-VE Network. Effectiveness of seasonal influenza vaccines in the United States during a season with circulation of all three vaccine strains. Clin Infect Dis. 2012;55:951–959
26. Joshi AY, Iyer VN, St Sauver JL, et al. Effectiveness of inactivated influenza vaccine in children less than 5 years of age over multiple influenza seasons: a case-control study. Vaccine. 2009;27:4457–4461
27. Fu C, Xu J, Lin J, et al. Concurrent and cross-season protection of inactivated influenza vaccine against A(H1N1)pdm09 illness among young children: 2012–13 case-control evaluation of influenza vaccine effectiveness. Vaccine. 2015;33:2917–2921
28. Szilagyi PG, Fairbrother G, Griffin MR, et al.New Vaccine Surveillance Network. Influenza vaccine effectiveness among children 6 to 59 months of age during 2 influenza seasons: a case-cohort study. Arch Pediatr Adolesc Med. 2008;162:943–951
29. Katayose M, Hosoya M, Haneda T, et al. The effectiveness of trivalent inactivated influenza vaccine in children over six consecutive influenza seasons. Vaccine. 2011;29:1844–1849
30. Sugaya N, Nerome K, Ishida M, et al. Efficacy of inactivated vaccine in preventing antigenically drifted influenza type A and well-matched type B. JAMA. 1994;272:1122–1126
31. Fiore AE, Uyeki TM, Broder K, et al.Centers for Disease Control and Prevention (CDC). Prevention and control of influenza with vaccines: recommendations of the advisory committee on immunization practices (ACIP), 2010. MMWR Recomm Rep. 2010;59:1–62
32. Fiore AE, Shay DK, Haber P, et al.Advisory Committee on Immunization Practices (ACIP), Centers for Disease Control and Prevention (CDC). Prevention and control of influenza. Recommendations of the advisory committee on immunization practices (ACIP), 2007. MMWR Recomm Rep. 2007;56:1–54
33. . Prevention and control of influenza with vaccines: recommendations of the Advisory Committee on Immunization Practices (ACIP)–United States, 2012–13 influenza season. MMWR Morbidity Mortality Weekly Report. 2012;61:613–618
34. Heinonen S, Silvennoinen H, Lehtinen P, et al. Effectiveness of inactivated influenza vaccine in children aged 9 months to 3 years: an observational cohort study. Lancet Infect Dis. 2011;11:23–29
35. McLean HQ, Thompson MG, Sundaram ME, et al. Influenza vaccine effectiveness in the United States during 2012–2013: variable protection by age and virus type. J Infect Dis. 2015;211:1529–1540
36. . Update: influenza activity - United States, 2011–12 season and composition of the 2012–13 influenza vaccine. MMWR Morbidity Mortality Weekly Report. 2012;61:414–420
37. . Influenza activity–United States, 2012–13 season and composition of the 2013–14 influenza vaccine. MMWR Morbidity mortality Weekly Report. 2013;62:473–479
38. . Interim results: state-specific seasonal influenza vaccination coverage - United States, August 2009-January 2010. MMWR Morbidity Mortality Weekly Report. 2010;59:477–484
39. . Interim results: state-specific influenza A (H1N1) 2009 monovalent vaccination coverage - United States, October 2009-January 2010. MMWR Morbidity Mortality Weekly Report. 2010;59:363–368
40. Skowronski DM, Janjua NZ, De Serres G, et al. Low 2012–13 influenza vaccine effectiveness associated with mutation in the egg-adapted H3N2 vaccine strain not antigenic drift in circulating viruses. PloS One. 2014;9:e92153
41. Jackson ML, Nelson JC. The test-negative design for estimating influenza vaccine effectiveness. Vaccine. 2013;31:2165–2168
42. Foppa IM, Haber M, Ferdinands JM, et al. The case test-negative design for studies of the effectiveness of influenza vaccine. Vaccine. 2013;31:3104–3109
43. . Prevention and control of influenza with vaccines: recommendations of the Advisory Committee on Immunization Practices (ACIP), 2011. MMWR Morbidity mortality Weekly Report. 2011;60:1128–1132
44. Ohmit SE, Petrie JG, Malosh RE, et al. Influenza vaccine effectiveness in the community and the household. Clin Infect Dis. 2013;56:1363–1369
45. Thompson MG, Li DK, Shifflett P, et al. Effectiveness of seasonal trivalent influenza vaccine for preventing influenza virus illness among pregnant women: a population-based case-control study during the 2010–2011 and 2011–2012 influenza seasons. Clin Infect Dis. 2014;58:449–457
46. Sullivan SG, Kelly H. Stratified estimates of influenza vaccine effectiveness by prior vaccination: caution required. Clin Infect Dis. 2013;57:474–476
47. McLean HQ, Thompson MG, Sundaram ME, et al. Impact of repeated vaccination on vaccine effectiveness against influenza A(H3N2) and B during 8 seasons. Clin Infect Dis. 2014;59:1375–1385
48. Ohmit SE, Petrie JG, Malosh RE, et al. Influenza vaccine effectiveness in households with children during the 2012–2013 season: assessments of prior vaccination and serologic susceptibility. J Infect Dis. 2015;211:1519–1528
49. Gilca V, De Serres G, Hamelin ME, et al. Antibody persistence and response to 2010–2011 trivalent influenza vaccine one year after a single dose of 2009 AS03-adjuvanted pandemic H1N1 vaccine in children. Vaccine. 2011;30:35–41
50. Skowronski DM, Janjua NZ, Sabaiduc S, et al. Influenza A/subtype and B/lineage effectiveness estimates for the 2011–2012 trivalent vaccine: cross-season and cross-lineage protection with unchanged vaccine. J Infect Dis. 2014;210:126–137
51. Walter EB, Neuzil KM, Zhu Y, et al. Influenza vaccine immunogenicity in 6- to 23-month-old children: are identical antigens necessary for priming? Pediatrics. 2006;118:e570–e578
52. Azziz Baumgartner E, Dao CN, Nasreen S, et al. Seasonality, timing, and climate drivers of influenza activity worldwide. J Infect Dis. 2012;206:838–846
53. De Serres G, Skowronski DM, Wu XW, et al. The test-negative design: validity, accuracy and precision of vaccine efficacy estimates compared to the gold standard of randomised placebo-controlled clinical trials. Eurosurveillance. 2013;18