Before the start of the 2016–2017 influenza season, the Advisory Committee on Immunization Practices withdrew its recommendation promoting the use of live attenuated influenza vaccines (LAIVs). There was concern that this might lessen the likelihood that those with a previous LAIV would return for an injectable influenza vaccine (IIV) and that child influenza immunization rates would decrease overall.
Using Oregon's statewide immunization registry, the ALERT Immunization Information System, child influenza immunization rates were compared across the 2012–2013 through 2016–2017 seasons. Additionally, matched cohorts of children were selected based on receipt of either an LAIV or an IIV during the 2015–2016 season. Differences between the IIV and LAIV cohorts in returning for the IIV in the 2016–2017 season were assessed.
Overall, influenza immunization rates for children aged 2 to 17 years were unchanged between the 2015–2016 and 2016–2017 seasons. Children aged 3 to 10 with a previous IIV were 1.03 (95% confidence interval, 1.02 to 1.04) times more likely to return for an IIV in 2016-2017 than those with a previous LAIV, whereas children aged 11 to 17 years with a previous IIV were 1.08 (95% confidence interval, 1.05 to −1.09) times more likely to return.
Withdrawal of the LAIV recommendation was not associated with an overall change in child influenza immunization rates across seasons. Children with a previous (2015–2016) IIV were slightly more likely to return during the 2016–2017 season for influenza immunization than those with a previous LAIV.
In previous seasons, the live attenuated influenza vaccine (LAIV) was used for a substantial number of childhood influenza immunizations; the impact of the US withdrawal of the recommendation to use LAIVs, starting with the 2016–2017 influenza season, was unknown.
No overall change in child influenza immunization rates was observed after the LAIV recommendation withdrawal. Children with a previous-season LAIV were slightly less likely to return for influenza immunizations when compared with children with previous injectable influenza immunization.
On June 22, 2016, the Advisory Committee on Immunization Practices (ACIP) voted that the live attenuated influenza vaccine (LAIV) not be used in the 2016–2017 influenza season.1 This decision was not without controversy because other nations continued to recommend LAIVs in the 2016–2017 season.2 In previous seasons, LAIVs were also given a preferential recommendation by the ACIP on the basis of initial evidence of superior performance compared with injectable influenza vaccines (IIVs) in children. Because LAIV administration does not involve using a needle, some researchers suggested that the intranasal administration of vaccine could potentially lead to increased influenza immunization of children by vaccine-hesitant or needle-averse parents.3,–5 Nationally, the Centers for Disease Control and Prevention’s sentinel surveillance found that 38% of children used an LAIV in the 2013–2014 season.6 In Oregon, which is known for having a substantial vaccine-hesitant population, LAIVs accounted for approximately one-third of influenza immunizations for children aged 2 to 9 years in the 2015–2016 influenza season.
A motivating concern for this study was that the withdrawal of the LAIV recommendation would lower childhood influenza immunization rates and that further interventions might be needed to promote the use of the IIV among parents who in previous seasons had opted to use an LAIV. As a first step to assess this issue, we used the Oregon ALERT Immunization Information System (ALERT IIS) to produce influenza immunization rates by vaccine type (LAIV versus IIV) for Oregon children aged 2 to 17 years from the 2011–2012 through the 2016–2017 influenza seasons. Because other, unrelated factors could result in different patterns of influenza immunization seeking from 1 year to the next, we also applied a second, more rigorous test for the assessment of LAIV withdrawal. For this second test, we assessed whether Oregon children aged 2 to 16 years at the start of the 2015–2016 season who received an LAIV during that season were less likely to return for subsequent influenza immunization in the 2016–2017 season than those who had received an IIV in the 2015–2016 season.
The data for this study are from the ALERT IIS, Oregon’s population-based immunization registry, which contains >6 million client records and >55 million immunization records and is a participant in the Centers for Disease Control and Prevention’s Sentinel Site Project for immunization surveillance. Immunization information systems and immunization registries are recognized as effective tools for tracking and improving immunization coverage.7 In comparison with Oregon clinic records, ALERT IIS child immunization histories are over 97% complete, which is similar to other published Sentinel IIS completeness evaluations.8 The ALERT IIS received 1.2 million influenza immunization reports for Oregon residents in both of the 2015–2016 and 2016–2017 seasons, accounting for ∼80% of total estimated influenza immunizations in Oregon during those seasons. In Oregon, the peak for influenza immunization receipt typically occurs by the third week in October, with over 90% of total season influenza immunizations given by December 31.
As a first assessment, rates of influenza immunization for Oregon children aged 2 to 17 years were calculated from the ALERT IIS for the 5 influenza immunization seasons of 2012 to 2013 through 2016 to 2017, with each influenza immunization season defined as running from August 1 through May 1. Counts of influenza immunizations from the ALERT IIS were used as numerators for rate determination, with yearly population forecasts from the Oregon demographer’s office as denominators.9
As a second and potentially more rigorous approach, we also conducted a matched cohort assessment of LAIV withdrawal effects. For this second assessment, a study population consisting of 2 cohorts of children age 2 to 16 at the start of the 2015–2016 influenza season was selected. The first cohort included children who received an LAIV during the first part of the 2015–2016 season, between August 1 and December 31. The second cohort included children who received an IIV during the same period and who were matched on a one-to-one basis to the LAIV cohort. Matches between the LAIV and IIV cohorts were made by age, sex, zip code or county of residence, vaccine funding source (such as Vaccines for Children or Medicaid), and week of influenza immunization during the 2015–2016 season Across the 2 influenza seasons, we anticipated that <2% of children would move out of the study area and that no difference in mobility would occur between the LAIV and IIV cohorts; any children with known mobility out of area or without a last known address in Oregon were excluded. This study design was chosen to mitigate factors such as intermittent (rather than annual) influenza immunization seeking and yearly externalities that may affect immunization seeking, such as seasonal disease severity, news reporting, and immunization promotion campaigns.
The primary study question was whether those children who had previously received an LAIV would return for an IIV at a lower rate than those who had previously received an injectable vaccine. To test this, we assessed the difference in rates of return for the IIV in the current 2016–2017 season between those who had previously received an IIV versus those who had previously received an LAIV by the assessment date of December 31. Analyses of the study data were done for the matched IIV and LAIV cohorts by using WinPepi software, with Wald 95% confidence intervals (CIs) for the ratio of proportions.10
Overall, the rates of childhood immunization for influenza in Oregon have shown minimal changes across the last 5 seasons regardless of changes in the proportion of LAIV recipients across seasons (Fig 1). Influenza immunization rates among children aged 2 to 17 years in the 2015–2016 and 2016–2017 seasons were statistically equivalent at 32.6% (95% CI of the difference between the 2 seasons: −0.2% to 0.1%), indicating no difference between the seasons immediately before and after the withdrawal of the recommendation for use of the LAIV.
For the matched cohort assessment, the study data set consisted of 60 151 matched pairs of children aged 2 to 16 years in the 2015–2016 influenza season, in which 1 child in each pair received an LAIV and the other received an IIV. Although the LAIV was still available in the Oregon market during the 2016–2017 season, the withdrawal of the ACIP recommendation for its use had a marked effect: only 19 doses of an LAIV for children in the study age range were reported to ALERT IIS through December 31. Among the matched pairs, 50.1% were girls and 49.9% were boys, 21.6% lived in a rural county, 92.3% received their baseline 2015–2016 influenza immunization in a private clinic setting, 4.7% used a public setting such as a county health department, 0.7% used an Indian Health clinic, 0.4% used a pharmacy, and 1.9% used another type or unspecified setting. Baseline influenza immunizations for 63.4% of matched pairs were paid for through private insurance, 33.1% received influenza immunization through Medicaid and Medicaid expansion programs, and 3.5% received publicly funded, non-Medicaid vaccines under the Vaccines for Children program or other public sources.
Overall, 53.1% of children in the study with previous LAIV and 56.4% with a previous IIV returned for an IIV during the 2016–2017 season. Those with a previous IIV were 1.05 times (95% CI, 1.04 to 1.06) more likely to return than those with previous LAIV. Among study children aged 3 to 10 during the 2016–2017 season, 56.7% with a previous LAIV and 58.3% with a previous IIV returned for an IIV during the 2016–2017 season. Those with a previous IIV were 1.03 times more likely to return than those with a previous LAIV (95% CI, 1.02 to 1.04). Among study adolescents aged 11 to 17 years during the 2016–2017 season, 49.1% of those with a previous LAIV and 53.1% of those with a previous IIV returned during the 2016–2017 for an IIV, so those with a previous IIV were 1.08 times more likely to return (95% CI, 1.06 to 1.10). Individual year of age comparisons are presented in Fig 2. Of 15 individual year-cohort age bands, 6 reveal no significant difference between a previous IIV and LAIV.
Based on our preliminary assessment of childhood influenza immunization in the 2016–2017 season, the withdrawal of the LAIV recommendation was not associated with an overall reduction in childhood immunization rates. Those children who had previously received an LAIV were slightly less likely to return for an IIV in the next season (2016–2017) when compared with a matched cohort that had received an IIV. For the great majority of parents, the likelihood of getting an influenza immunization for their children in the 2016–2017 season was only weakly dependent on whether they had previously chosen an LAIV or an IIV.
Our findings call into question whether having a noninjectable LAIV immunization available added substantially to yearly influenza immunization rates. Assuming timely influenza vaccine availability each season, decisions about whether or how to vaccinate a child involve an interplay between provider, parent, and child, although provider anticipation of this interplay is also a significant factor. Survey results before and after the introduction of the LAIV revealed that the majorities of providers, parents, and children believed that both the LAIV and the IIV should be available or that the LAIV should be given preference.11,–14 Among providers, as an example, in 2000 a majority (55%) of fellows of the American Academy of Pediatrics surveyed believed that access to an intranasal vaccine was an important consideration in recommending universal influenza immunization.15 However, it is likely that provider perception of the need for the LAIV to accommodate parent preference was overstated because researchers in 1 study have found that health care providers tend to overestimate the importance of route of injection as a factor in parental immunization decisions in comparison with actual parental reports.11
Among parents, survey reports of preference for the LAIV have to be considered against other factors and not taken in isolation. To date, researchers have found that parental preference is more related to the perceived effectiveness of a vaccine than to its route of administration.13,16,17 Reports of LAIV ineffectiveness may then drive rapid swings in parental preference. Survey results also reveal, regardless of preferences, that having an LAIV as an option does not increase parental acceptance of influenza immunization nor is lack of an LAIV commonly cited as a reason for not immunizing.14,18 LAIV acceptance by parents and children may also be related to the provision of influenza immunizations in school settings because the manufacturer of LAIVs supported the development of large-scale school influenza immunization programs featuring LAIVs.19 Among school programs in which both the IIV and the LAIV are available, researchers in 1 study have found that the majority of parents will choose an IIV, although with a caveat that parents preferred to use the same vaccine type across seasons.20
In this study, the withdrawal of the LAIV had a greater impact on rates of influenza vaccination among adolescents with previous LAIV use than it did among younger LAIV recipients. In a recent survey of preadolescents (aged 8–12 years), researchers found that although children most frequently selected effectiveness as the most important attribute in a vaccine, 69% reported preferring an intranasal preparation to an injection.5 In a focus group investigation of adolescents’ perspectives on school-based influenza vaccination programs, 62% of respondents said that availability of an intranasal preparation would influence their decision about whether to be vaccinated.21 It is possible that the lower rates of IIV immunization among adolescents who previously received an LAIV withdrawal is also a reflection of their ability to mount effective resistance, verbally or otherwise, to the IIV should they choose not to receive it.
In the results we present here, we intended to address the concern that sudden withdrawal of the LAIV recommendation may have led to a drop in childhood influenza immunization rates. As a caveat, some small amount of childhood influenza immunization is not captured in the ALERT IIS. Although this amount is expected to be small, it is unknown whether unreported doses were more likely to be LAIVs or IIVs. Similarly, no subanalysis was done assessing communities with high vaccine hesitancy rates; the impact of LAIV withdrawal could be substantially higher in such places. In our findings, we also do not factor in whether those who immunized in the 2015–2016 season were representative of LAIV and IIV use over a longer time period. In addition, if parents who previously used LAIV in vaccinating their children were more likely to seek influenza immunization for them only intermittently, our approach may overstate the effects of an LAIV withdrawal.
With our findings, we suggest that the ACIP withdrawal of its recommendation for use of the LAIV during the 2016–2017 influenza season had a minimal impact on childhood influenza immunization. No drop in overall influenza immunization rates for children occurred after the recommendation withdrawal, and only a small reduction in the likelihood of returning for an IIV was noted for children who previously received an LAIV.
Mr Robison conceptualized and designed the study, conducted the initial analysis, and drafted the initial manuscript; Mr Dunn assisted in conceptualizing and designing the study and revised the manuscript; Ms Richards assisted in conceptualizing and designing the study, performed data extracts and analyses, and contributed to manuscript text sections; Dr Leman conceptualized and designed the study, assisted with analyses, and contributed to manuscript sections and revisions; and all authors approved the final manuscript as submitted.
FUNDING: Funded in part by the Centers for Disease Control and Prevention grants to the state of Oregon for immunization surveillance.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.