Single-Dose Vaccination Among Infants and Toddlers Provides Modest Protection Against Influenza Illness, Which Wanes After 5 Months

Abstract

In their first season of vaccination, young children are recommended 2 doses of influenza vaccine, but a 2-dose schedule might be difficult to implement in many countries. Within a cohort study of 742 children aged 6 to <24 months in Managua, Nicaragua, this study estimated effectiveness of partial vaccination from 3 to 9 months postvaccination. Vaccine effectiveness was 74% (95% confidence interval [CI], 24%–91%) within 3 months and 55% (95% CI, 10%–77%) within 4 months. There was not significant protection beyond 5 months. Partial vaccination might confer some benefits but should be followed by a second dose.

Children aged 6 to <24 months had modest protection against influenza illness after only 1 dose of influenza vaccine, but this waned after 5 months.

Globally each year, there are 90 million new cases and 28 000 to 111 500 deaths from influenza among children aged <5 years [1]. Vaccines are approved for individuals aged ≥6 months. The World Health Organization (WHO) recommends the influenza vaccine for immunization programs with certain conditions, stating that children aged 6–23 months “should be considered a target group for influenza immunization when sufficient resources are available and with due consideration for competing health priorities and operational feasibility” [2].

For optimal protection, children 6 months through 8 years need 2 doses separated by at least 4 weeks, in their first year vaccinated [3]. The 2-dose regimen comes from studies showing reduced vaccine immunogenicity and effectiveness of vaccines with only 1 administered vaccine dose [4]. Effectiveness can also vary by age of the vaccine recipient and the degree of match between the vaccine strain and the circulating strain [2]. Indeed, many low- and middle-income countries may choose not to include influenza vaccination as part of the government-subsidized Expanded Program on Immunization because of costs, operational challenges, or competing priorities. Countries also may only stock influenza vaccine for a limited time, hampering opportunities for young children to come back to a clinic for a second shot [5].

Nicaragua, for example, recommends influenza vaccination for children aged 6 to 23 months with preexisting conditions [5] and offers this vaccine for free. Despite this policy, Nicaragua faces challenges securing sufficient influenza vaccines to fully cover its target groups, and so many children only receive 1 dose of vaccine. In this study, we use findings from a prospective cohort of children followed from birth to 2 years of age to estimate single-dose influenza vaccine effectiveness (VE) against illness.

METHODS

Study Population

A description of the Nicaraguan Influenza Birth Cohort Study is available elsewhere [6]. Briefly, infants were enrolled before 4 weeks of age and followed up to 2 years of age. Enrollment occurred from September 2011 through September 2014, with data collected through September 2016. The study was conducted out of the Health Center Sócrates Flores Vivas (HCSFV), which in 2014 had a catchment area of approximately 60 000 people. This analysis is limited to children aged ≥6 months (ie, eligible for vaccination) and excludes children fully vaccinated against influenza. Children with 2 doses in 1 season (n = 4) were excluded in all analyses. The WHO position paper on influenza states that generally “1 dose is given to previously vaccinated [children]” [7]. The US Advisory Committee on Immunization Practice further specifies that infants should still receive ≥2 doses if they only received 1 dose the prior season [3]. In our analyses, we exclude individuals had previously received 1 dose, but findings including them are listed in Supplementary Material 1.

Parents or guardians of the children completed daily symptom questionnaires, were visited at their homes annually to complete questionnaires, and were seen by study team physicians at the HCSFV for primary care visits. Data from those visits were systematically recorded into study databases. Families were encouraged to bring their child to HCSFV at the first sign of subjective fever or respiratory illness, with primary health care and laboratory testing offered for free at HCSFV.

Laboratory-Confirmed Influenza

Respiratory samples were collected from infants who had (1) fever (≥37.8°C) or reported of fever/feverishness and either rhinorrhea or cough; (2) fever or reported fever/feverishness without a defined focus; (3) severe respiratory symptoms, including apnea, stridor, nasal flaring, wheezing, chest indrawing, or central cyanosis; or (4) hospitalization for apnea, stridor, nasal flaring, wheezing, chest indrawing, central cyanosis, or sepsis. Respiratory samples were collected with nasal and oropharyngeal swabs. RNA was extracted from swabs with a QIAamp Viral RNA Mini Kit (Qiagen Corporation). Protocols from the United States Centers for Disease Control and Prevention were followed to amplify, type, and subtype/genotype influenza A and B viruses. Sequential respiratory samples found to be positive for the same influenza subtype within 28 days were considered the same illness. We defined influenza-associated acute lower respiratory infection (ALRI) as a laboratory diagnosis of influenza in the 30 days preceding a clinical diagnosis of pneumonia, bronchiolitis, or bronchitis, or a measured high respiratory rate (≥50 breaths per minute for those aged 6–11 months, or ≥40 for those aged 12–23 months).

Vaccination Status

Vaccination status was abstracted from the child’s vaccine cards or clinic records, or verified from parental report during annual surveys. The primary independent variable in this analysis was a single dose of influenza vaccine at least 14 days prior to the laboratory-confirmed influenza diagnosis. We modeled the risk of influenza infection following vaccination across a range of plausible durations of protection, specifically 3, 6, and 9 months, versus a comparison of no vaccination. During the study period, HCSFV distributed the Green Cross Southern formulation of influenza vaccine (0.25 mL trivalent vaccine with 7.5 μg per antigen).

Covariates

Other covariates included socioeconomic status quintiles (based on ownership of household appliances, housing density, and housing material; see a previous analysis for more detail [8]), child’s age (in 6-month groups), and child’s sex. As children aged, they moved across different age groups. Although comprehensive data from primary care visits were available, we were underpowered to detect any confounding from preexisting or chronic conditions in children.

Statistical Analysis

To determine if vaccination coverage varied by child’s sex or family socioeconomic status, we computed Pearson χ² tests for these relationships. The rate ratio (RR) of influenza was modeled through Poisson regressions, which specified a population offset and which included multiple person-months' worth of data per child. We estimated partial influenza VE through subtracting RR estimates from 1. The outcome was a laboratory-confirmed case of influenza within that month, and the main exposure was vaccination in the prior 3, 6, or 9 months, adjusted for age, sex, and socioeconomic status. We excluded the time period January through May from analysis because influenza did not circulate in the study population during those months [8]. We used general estimating equations with an autoregressive correlation matrix to account for subject repeated measures. Analyses were conducted in SAS version 9.4 (SAS Institute). We present RRs with 95% confidence intervals (CI) and assess significance at an α level of .05. Additional subgroup analyses include stratifying the age of children into those 6 to <12 months and those 12 to <24 months, and using an outcome of influenza-associated ALRI instead of influenza diagnosis of any illness.

Ethical Approval

This study was approved by the University of Michigan Health Sciences and Behavioral Sciences Institutional Review Board (No. HUM00091607) and the Nicaraguan Ministry of Health Institutional Review Board (No. NIC-MINSA/CNDR CIRE-05/05/11-029). Parents or guardians of children provided written informed consent.

RESULTS

The original cohort included 833 children (Figure 1A); 4 were excluded because they were fully vaccinated against influenza, 86 children exited the study before turning 6 months, and 1 child had incomplete survey data. The remaining 742 children contributed 7483 person-months of data (or on average 10 months per child).

Figure 1.

Flowchart of individuals into the vaccine effectiveness analysis (A), timing of influenza vaccination (B), and duration (mean and 95% confidence interval) of 1-dose influenza vaccine protection (C) in Managua, Nicaragua.

Overall, 9% of the children were vaccinated between ages 6 months and 1 year, and 20% received a vaccine between ages 1 and 2 years (Table 1). Vaccinations were typically administered in May (Figure 1B), prior to the anticipated influenza season. Vaccination coverage did not differ by child sex (P = .97 in first year, P = .71 in second year) or socioeconomic status (P = .59 in first year, P = .98 in second year). There was a total of 292 polymerase chain reaction (PCR)-confirmed influenza illnesses: 36 (12%) occurred prior to 6 months of age, 92 (32%) between 6 and 12 months of age, and 164 (56%) in the second year of life. Of those cases that occurred between 6 and 24 months of age, 142 (56%) were H3N2, 43 (17%) were H1N1pdm, and 70 (27%) were influenza B. In this age group, 32% of influenza illnesses (92) were associated with ALRI.

Table 1.

Characteristics of Study Population, Managua, Nicaragua

Characteristic	Total No. (Column %)	Vaccinated at Age 6 to <12 mo, No. (Row %)	Vaccinated at Age 1 to <2 y, No. (Row %)
Overall	833	76 (9)	169 (20)
Child's sexa
Female	418 (50)	38 (9)	87 (21)
Male	415 (50)	38 (9)	82 (20)
Socioeconomic status quintileb,c
Lowest	208 (25)	18 (9)	40 (19)
Lower	174 (21)	11 (6)	34 (20)
Middle	169 (20)	17 (10)	35 (21)
Higher	148 (18)	15 (10)	31 (21)
Highest	133 (16)	15 (11)	29 (22)
Vaccinated at <1 y of age
No	757 (91)	…	160 (21)
Yes	76 (9)	…	9 (12)

^aχ² test for child’s sex and vaccination: P = .97 in first year, P = .71 in second year.

^bχ² test for family socioeconomic status and vaccination: P = .59 in first year, P = .98 in second year.

^cOne individual had missing data.

Among all children aged 6 to <24 months, VE against laboratory-confirmed influenza illness following 1 dose was 74% (95% CI, 24%–91%) within 3 months and 55% (95% CI, 10%–77%) within 4 months (Figure 1C) of vaccination. We did not observe significant protection after 5 months post vaccination. In the subgroup analysis (Supplementary Material 1), there was 69% VE for children aged 12 to <24 months within 3 months of vaccine administration (95% CI, 3%–90%). We were underpowered to estimate VE against influenza-associated ALRI at 4 months (VE, 67%; 95% CI, −88% to 94%) and 5 months (VE, 32%; 95% CI, −99% to 77%). Including or excluding individuals based on having 1 dose of influenza vaccine in a previous season did not substantially impact results.

DISCUSSION

This study found partial, single-dose influenza vaccination provided modest protection against influenza illness for 3–5 months after administration among infants aged 6–24 months. Protection waned 6 months after vaccination. Previously, we had also documented that duration of illness was shorter with partial versus no vaccination [8]. Full, 2-dose, vaccination would likely provide greater protection. Between the 2012–2013 and 2015–2016 influenza seasons, influenza VE in children aged 6 months to 8 years in the United States ranged between 25% and 57% over the entire influenza season [9]. The available evidence from our studies and others suggest a 2-dose vaccine schedule would be recommended over a 1-dose schedule in a young child’s first year of vaccination.

Waning immunity has been extensively documented. A review of previous studies found noticeable declines in VE after 90 days [10]. Similarly, pooled data from 2011–2012 through 2014–2015 influenza seasons among individuals aged ≥9 years demonstrated that VE against H3N2 peaked at 35% at 14 days after vaccination, and declined to 0 after 158 days; for H1N1pdm, VE declined from 80% at 14 days postvaccination to 46% at 180 days, and for influenza B, VE declined from 59% to 23% over the same time period [11]. Several hypotheses might explain waning immunity, including reductions in antibody titers [12] or antigenic drift in circulating viruses, which become increasingly dissimilar from those in vaccine formulations.

Single-dose influenza vaccination campaigns could be useful in low- and middle-income countries where 2-dose vaccination is not imminently practicable because of cost or logistics. Single-dose vaccination might help protect infants against influenza illnesses during the 3 to 6 months when their maternally conferred anti-influenza antibodies are at their nadir and their risk of hospitalization are greatest [13]. It is possible that single-dose short-term VE against severe hospitalized illness might be better than single-dose short-term VE against mild outpatient illnesses. Future studies could assess the severity of illness in vaccinated versus unvaccinated children. One study in Ontario, for example, found partial influenza vaccination moderately effective (39%) against hospitalization [14].

Although a 2-dose influenza vaccination strategy is preferable for children in their first year of vaccination, additional studies might be warranted to replicate single-dose VE against outpatient illness, better quantify the magnitude and duration of single-dose protection against severe illness, and the marginal cost-benefit of single-dose versus full-dose vaccination. Our study also highlights the need to carefully consider timing of vaccination if only 1 dose is available. Given the substantial waning of protection after 3 months postvaccination, administration of 1 dose would be most useful when influenza epidemic activity seems imminent, as determined by local respiratory virus surveillance [15]. Studies could also examine if it would be effective to push back the second dose from the currently recommended 1 month to 3 months after the first vaccine dose.

This was a longitudinal study, where we obtained vaccination status from health cards and our outcome from laboratory-confirmed influenza diagnoses from active surveillance. However, we acknowledge certain limitations. Our study does not have the statistical power to examine type or subtype-specific VE, and to contrast VE between those with and without prior infection.

CONCLUSIONS

Many low- and middle-income countries have low rates of full pediatric influenza vaccination. In a longitudinal study in Nicaragua, we found single-dose vaccination among infants aged 6–24 months provided modest protection against influenza illness for 3–5 months after administration. This study provides evidence for some benefits of receiving only 1 dose of influenza vaccine relative to zero doses in the first season but, overall, highlights the need for more research on how to implement pediatric vaccine programs, which would ideally include a 2-dose vaccine series.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.