Correlates of 2009 H1N1 influenza vaccine acceptability among parents and their adolescent children

Abstract

School-aged children were a priority group for receipt of the pandemic (2009) H1N1 influenza vaccine. Both parental and adolescent attitudes likely influence vaccination behaviors. Data were collected from surveys distributed to middle- and high-school students and their parents in two counties in rural Georgia. Multivariable logistic regression analyses were conducted to assess correlates of parental acceptance of H1N1 influenza vaccination for their children and adolescents’ acceptance of vaccination for themselves. Concordance analyses were conducted to assess agreement between parent–adolescent dyads regarding H1N1 influenza vaccine acceptance. Parental acceptance of H1N1 influenza vaccination for their children was associated with acceptance of the vaccine for themselves and feeling motivated by the H1N1 influenza pandemic to get a seasonal influenza vaccine for their child. Adolescents’ acceptance was associated with receipt of a seasonal influenza vaccine in the past year, fear of getting H1N1 influenza, feeling comfortable getting the vaccine and parental acceptance of H1N1 influenza vaccine. Half (50%) of parent–adolescent pairs included both a parent and child who expressed H1N1 influenza vaccine acceptance, and 19% of pairs would not accept the vaccine. This research highlights the need for interventions that target factors associated with H1N1 influenza vaccine acceptance among both parents and adolescents.

Introduction

In April 2009, a new strain of influenza A/H1N1 was identified in many countries around the world. In June 2009, the World Health Organization raised the influenza pandemic alert to Phase 6 [1]. In the United States to date, H1N1-related mortality has disproportionately affected children (0–17 years), who have accounted for approximately 1270 deaths and approximately 86 000 hospitalizations [2]. The proportion of school-aged children impacted by 2009 pandemic influenza A (H1N1) virus and subsequent risk of serious illness, hospitalization, and death is much greater than that typically associated with seasonal influenza [2].

Similar to seasonal influenza, school-aged children played a key role in 2009 H1N1 influenza virus transmission. A recent report from England showed that the proportion of children aged 5–14 years with hemagglutination inhibition titers of 1:32 or greater in serum samples obtained in September 2009 was higher than that in those aged 25–44 years, consistent with higher transmission rates in school-age children than in older age groups [3].

In July 2009, the Advisory Committee on Immunization Practices recommended that all school-aged children aged 5–18 years receive the 2009 H1N1 influenza vaccine [4]. Although immunization is the most effective measure to reduce the number of infections, hospitalizations and deaths from influenza, there was concern about the acceptability of the 2009 H1N1 influenza vaccine. Recent surveys of adults found that their intention to receive the 2009 H1N1 influenza vaccine was estimated at approximately 50%, with some studies showing acceptance rates as high as 65% [5–7]. No study has assessed the predictors of acceptance of 2009 H1N1 influenza vaccine among adolescents.

Recent estimates of 2009 H1N1 influenza vaccination coverage in the United States among 5–18 year olds was only 28.1% (16 million persons) [8]. Given that, parents likely played a role in the vaccination behaviors of their children, their beliefs about 2009 H1N1 influenza vaccination may have been important in determining vaccination rates among school-age populations. Several studies have assessed factors associated with parental acceptance of seasonal influenza vaccination for younger children, including 6- to 21- month- olds [9] and children with chronic medical conditions [10]. However, to date, little is known about attitudes toward influenza vaccination among parents of adolescents (ages 11–18 years). Studies assessing parental attitudes and predictors of vaccine acceptance for adolescents have mainly focused on sexually transmitted infections (STIs), including human papillomavirus (HPV) vaccine and herpes simplex virus type 2 (HSV-2) vaccine [11–17]. The results of studies conducted with parents of young children, in addition to the studies on STI vaccines, indicate that parental views on a child’s susceptibility to infection, severity of infection, child’s risk of infection and safety of vaccine may all influence parental vaccine acceptance [9, 11, 12, 14, 16, 17].

In addition to parental attitudes, attitudes among school-age children, particularly adolescents, may also play a key role in vaccination uptake. Several studies have focused on adolescents’ attitudes toward vaccinations, and the role that parental attitudes may play in influencing adolescents’ beliefs. These studies have shown that young people are more likely to accept hepatitis B vaccination if they perceive it is important to their parents [18], to accept a hypothetical HSV-2 vaccination if they thought their parents would encourage it [19] and to accept the HPV vaccine if their mothers approved [20] or if their parents demonstrated an intention to get them vaccinated [17]. One study demonstrated that social norms, a composite measure including perceived parent/guardian approval, may impact adolescents’ intention to receive influenza vaccine [21]. However, to our knowledge, no studies have assessed parental and adolescent attitudes toward vaccine acceptance for 2009 H1N1 influenza.

The purpose of this study is to examine: (i) correlates of parental acceptance of 2009 H1N1 influenza vaccine for their adolescent children, (ii) correlates of adolescents’ acceptance of 2009 H1N1 influenza vaccine for themselves and (iii) level of agreement (concordance) between parents and their adolescent children regarding 2009 H1N1 influenza vaccine acceptance. The current study is of particular importance because it assesses these factors during an influenza pandemic and can be informative for mass vaccination campaigns and acceptance of new vaccines in the future.

Methods

Participants

Parent–adolescent dyads were recruited from two rural counties participating in an ongoing school-based seasonal influenza vaccination intervention in rural Georgia. The participating counties were relatively small, with a single middle and high school per county. Dyads (n = 122) included 122 adolescents and 102 parents as 18 parents were paired with 2 children and 1 parent was paired with 3 children. Due to the small average cluster size (average number of siblings in our study) and small intraclass correlation coefficient, analyses were not adjusted for non-independence. Data were collected from surveys distributed to parents and their adolescent children prior to implementing the interventions to increase immunization against seasonal influenza virus among students in those schools [22]. This survey was administered shortly before the Food and Drug Administration’s licensure of four 2009 H1N1 influenza vaccines on 15 September 2009 and the first release of the vaccine on 30 September 2009. All parents and students in participating counties were invited to complete the survey. Both counties comprised rural, high minority, low-income populations. Data from the 2008–2009 academic year indicated that 95% of students in County A were African-American. Additionally, 88% of students were eligible to receive free or reduced cost meals [23]. In County B, 38% of students were African–American and 61% were eligible to receive free or reduced cost meals [23]. Eligibility criteria for participation in this study included: (i) being a student enrolled in a participating middle or high school or (a) being a parent of a student enrolled in a participating middle or high school and (ii) providing written informed consent to participate in the study (parents) or (a) providing parental informed consent and written assent to participate in the study (adolescents). All procedures were approved by the Institutional Review Board of Emory University.

Instrument

Parents and adolescents completed similar, but distinct survey instruments. The instruments were designed to assess demographic, psychosocial and behavioral factors associated with the 2009 H1N1 influenza pandemic. Based on previous research and recent theoretical development [24–26], psychosocial survey items were guided by the Health Belief Model [27] and the Integrated Behavioral Model [26]. Questions were adapted from existing surveys assessing related concepts because previous questionnaires did not assess attitudes and beliefs toward 2009 H1N1 influenza [24, 28–30].

Main outcome variables

The main outcome variable for parents was acceptance of 2009 H1N1 influenza vaccination for their adolescent children: ‘I would allow my children to get a swine flu vaccine’, while the main outcome variable for adolescents was acceptance of 2009 H1N1 influenza vaccination for themselves: ‘I plan to get a swine flu vaccine this fall or winter, if it is available’. Both variables were initially asked on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Due to high levels of skewness, variables were dichotomized into ‘strongly agree/agree’ versus ‘strongly disagree/disagree/neutral’ for analyses.

Demographic variables

All participants reported their age, race, gender and seasonal influenza immunization history. Parents also reported their marital status and number of children and adolescents reported their grade level.

Knowledge

Five questions were developed to assess the participants’ knowledge about 2009 H1N1 influenza. Knowledge questions were the same for both parents and adolescents. The questions were based on five-point Likert scales ranging from 1 (strongly disagree) to 5 (strongly agree). For logistic regression analyses, items were summed to form a scale ranging from 5 to 25 (α = 0.65 for parents, α = 0.54 for adolescents). Sample items included ‘Washing your hands can help prevent getting the swine flu’ and ‘There has been a human infected by swine flu in Georgia’.

Attitudes

All psychosocial attitude questions were asked on five-point Likert scales ranging from 1 (strongly disagree) to 5 (strongly agree). Although many questions were similar for parents and adolescents, there were some differences.

Parent variables included: Social norms (two-item scale, α = 0.78): ‘We talk about swine flu at home’ and ‘I talk about swine flu with my friends’; Perceived severity: ‘The swine flu is a serious disease’; Perceived susceptibility: ‘I am fearful of my children getting swine flu’; Cues to action for seasonal influenza vaccination: ‘Swine flu has motivated me to get the seasonal flu vaccine for my children’; School preventive action: ‘I support school closings as a precaution against swine flu’ and Personal acceptance: ‘I would get a swine flu vaccine’.

Adolescent variables included: Social norms (two-item scale, α = 0.65): ‘We talk about swine flu at home’ and ‘We talk about swine flu at school’; Perceived severity: ‘The swine flu is a serious disease’; Perceived susceptibility: ‘I am fearful of getting swine flu’; Perceived barriers: ‘The swine flu vaccine would make me sick’ and Self-efficacy: ‘I feel comfortable getting a swine flu vaccine’.

Behaviors

‘Vaccine history’ items included questions about past influenza vaccination behaviors, including: ‘Have you ever received the regular seasonal flu vaccine?’ and ‘Did you receive the regular seasonal flu vaccine last fall or winter?’ Questions were the same for parents and adolescents.

Procedure

Packets containing consent forms and survey instruments were distributed on 11 September 2009, prior to implementing the intervention to increase vaccination against seasonal influenza virus among students. At that time, the 2009 H1N1 influenza vaccine was in production but had not yet been approved for release by the Food and Drug Administration. The packets contained an informed consent form (for parents and adolescents) and a written assent form (for adolescents), which was separated upon receipt of the completed survey and stored separately from completed questionnaires. The survey instrument was a onetime self-administered, paper-and-pencil questionnaire distributed to participants via home mailing. Packets were returned before the seasonal influenza educational interventions for students in the schools, which were scheduled to occur about 2 weeks later. Parents completing the survey received a $20 gift card and adolescents received a $10 gift card as compensation for their time.

Data analysis

All data management and analyses were conducted using SPSS/PASW Statistics version 17. Descriptive analyses were performed to assess the distributions of demographic variables and variables assessing 2009 H1N1 influenza-related knowledge, attitudes and behaviors among parents and their adolescent children. Questions assessing knowledge and social norms were combined into scales, and Cronbach’s alphas were calculated for each scale to assess its internal consistency. Next, bivariate analyses were conducted to assess associations between demographic variables, attitude variables and (i) parental acceptance of 2009 H1N1 influenza vaccine for their adolescent children and (ii) adolescents’ acceptance of 2009 H1N1 influenza vaccine for themselves. Variables that demonstrated significant bivariate associations at the P = 0.10 level were included in multivariate logistic regression analyses. Multivariable logistic regression analyses were conducted to determine: (i) correlates of parental acceptance of 2009 H1N1 influenza vaccination for their adolescent children (among parents) and (ii) acceptance of 2009 H1N1 influenza vaccine (among adolescents).

A concordance analysis was conducted to assess level of agreement regarding 2009 H1N1 vaccination acceptance among parents and their adolescent children. Agreement was measured by Cohen’s Kappa statistic (κ) [31].

Results

Response rates and participant characteristics

The survey was completed and returned by 102 of 890 parents (12%) and 122 of 1036 adolescents (12%). Demographic information is displayed in Table I. A slight majority of parents identified themselves as black (55%), while a larger majority were female (91%) and married (66%). Parental mean age was 42. A slight majority of adolescents identified themselves as black (57%), female (52%) and in middle school (54%), with a mean age of 15. (Table I). Slightly over half of parents (52%) and under half of adolescents (48%) reported ever receiving a seasonal influenza vaccine, with 28% of parents and 39% of adolescents reporting receipt of a seasonal influenza vaccine in the previous year. Forty-five percent of parents and 60% of students reported planning to get a 2009 H1N1 influenza vaccine.

Table I.

Demographic information among the sample population

Variables	Parents, N = 102	Students, N = 122
	N (%) or mean (SD)	N (%) or mean (SD)
Age	41.7 (7.1)	14.6 (1.9)
Race
White	46 (45.1)	50 (41.0)
Black	56 (54.9)	69 (56.6)
Other	—	3 (2)
Gender
Male	9 (8.8)	58 (47.5)
Female	93 (91.2)	64 (52.5)
Marital status
Married	67 (65.7)	—
Single	28 (27.5)	—
Divorced/widowed	7 (6.9)	—
Number of children
1	19 (18.6)	—
2	27 (26.5)	—
3 or more	56 (54.9)	—
School type
Middle school	—	66 (54.1)
High school	—	56 (45.9)
Seasonal influenza history
Ever received influenza vaccine	53 (52.0)	58 (47.5)
Received influenza vaccine in the past year	29 (28.4)	47 (38.5)
Plan to get an H1N1 influenza vaccine this year, if it is available	46 (45.1)	73 (59.8)

Parental predictors of 2009 H1N1 influenza vaccine acceptance for their children

Correlates of parental 2009 H1N1 influenza vaccine acceptance are presented in Table II. Three quarters of parents (75%) agreed or strongly agreed that they would accept a 2009 H1N1 influenza vaccine for their children. In bivariate analyses, correlates of parental 2009 H1N1 influenza vaccination acceptance included social norms, perceived severity, perceived susceptibility (fear of children getting swine flu), school preventive action (supporting school closings as a precaution against swine flu), cues to action (motivated by the swine flu pandemic to get a seasonal influenza vaccination for their child) and personal acceptance (willingness to accept a 2009 H1N1 influenza vaccination for themselves). Demographic variables, knowledge and receipt of a seasonal influenza vaccine in the past year were not significantly associated with parental acceptance of 2009 H1N1 influenza vaccination for their children.

Table II.

Predictors of parental acceptance of 2009 H1N1 flu vaccine for their children

Variable	Unadjusted logistic regression analysis, OR (95% CI)	P-value	Adjusted logistic regression analysis, OR (95% CI)	P-value
Age	1.02 (0.96–1.10)	0.489
Gender
Male	Ref
Female	0.82 (0.16–4.23)	0.814
Race
White	Ref
Black	1.30 (0.53–3.18)	0.561
Received seasonal flu vaccine last year	1.11 (0.41–3.00)	0.843
Knowledge	1.17 (0.98–1.40)	0.091	1.28 (0.99–1.67)	0.062
Social norms	1.38 (1.03–1.86)	0.032	0.87 (0.52–1.43)	0.576
Perceived severity	2.33 (1.22–4.44)	0.010	1.96 (0.84–4.60)	0.121
Fearful of my children getting swine flu	1.50 (1.50–2.16)	0.029	1.22 (0.63–2.34)	0.557
Support school closings as a precaution against swine flu	1.50 (1.06–2.11)	0.023	0.81 (0.47–1.42)	0.465
Motivated by swine flu to get seasonal flu vaccine for children	2.59 (1.62–4.14)	<0.001	2.28 (1.14–4.53)	0.019
Would accept swine flu vaccine for self	3.35 (2.05–5.45)	<0.001	3.49 (1.87–6.50)	<0.001

CI, confidence interval. OR, odds ratio.

In multivariate analyses, only two variables remained significantly associated with parental acceptance of 2009 H1N1 influenza vaccination for their children. The odds of parental acceptance of 2009 H1N1 influenza vaccination for their children were 2.3 times greater among parents motivated by the H1N1 influenza pandemic to get a seasonal influenza vaccination for their child (P = 0.019) and 3.5 times greater among parents who would accept a 2009 H1N1 influenza vaccination for themselves (P < 0.001).

Adolescent predictors of 2009 H1N1 influenza vaccine acceptance

Correlates of 2009 H1N1 influenza vaccine acceptance among adolescents are presented in Table III. Among adolescents, 60% reported that they would accept a 2009 H1N1 influenza vaccination. In bivariate analyses, correlates of 2009 H1N1 influenza vaccination acceptance included receipt of a seasonal influenza vaccine in the past year, social norms, perceived susceptibility (fear of getting swine flu), self-efficacy (feeling comfortable getting swine flu vaccine), perceived barriers (the swine flu vaccine would make me sick) and parental acceptance of 2009 H1N1 influenza vaccination for their children. Demographic variables, knowledge and perceived severity were not significantly associated with adolescents’ acceptance of 2009 H1N1 influenza vaccination.

Table III.

Predictors of adolescents’ acceptance of 2009 H1N1 flu vaccine

Variable	Unadjusted logistic regression analysis, OR (95% CI)	P-value	Adjusted logistic regression analysis, OR (95% CI)	P-value
Age	0.89 (0.73–1.08)	0.249
Gender
Male	Ref
Female	1.45 (0.70–3.00)	0.318
Race
White	Ref
Black	0.74 (0.35–1.56)	0.435
Received seasonal flu vaccine last year	3.36 (1.49–7.58)	0.003	3.96 (1.35–11.63)	0.012
Knowledge	1.02 (0.87–1.20)	0.816
Social norms	1.30 (1.03–1.65)	0.027	0.96 (0.69, 1.34)	0.811
Perceived severity	1.10 (0.79–1.53)	0.565
Fear of getting swine flu	1.54 (1.17–2.02)	0.002	1.54 (1.05–2.25)	0.027
Feel comfortable getting swine flu vaccine	2.78 (1.85–4.18)	<0.001	2.27 (1.41–3.66)	0.001
The swine flu vaccine will make me sick	0.58 (0.40–0.84)	0.004	0.52 (0.30–0.89)	0.016
Parental acceptance of swine flu vaccine for their children	2.31 (1.54–3.45)	<0.001	1.84 (1.14–2.95)	0.012

CI, confidence interval. OR, odds ratio.

In multivariate analyses, almost all of the variables that demonstrated significance in bivariate analyses remained significant. The odds of adolescents’ acceptance of 2009 H1N1 influenza vaccination were four times greater among adolescents who received a seasonal influenza vaccination in the past year (P= 0.012), 1.5 times greater among adolescents who expressed fear of getting the swine flu (P = 0.027), 2.3 times greater among adolescents who felt comfortable getting a swine flu vaccine (P = 0.001) and 1.8 times greater among adolescents whose parents indicated that they would accept a 2009 H1N1 influenza vaccination for their children (P = 0.012). Adolescents who feared that the 2009 H1N1 influenza vaccination would make them sick were significantly less likely to report intention to receive an influenza vaccination (odds ratio = 0.5, P < 0.001).

Concordance among parents and their adolescent children

Of the 122 parent–adolescent child pairs, 61 (50%) included both a parent and a child who expressed 2009 H1N1 influenza vaccine acceptance and 23 (19%) of pairs included both a parent and a child who would not accept a 2009 H1N1 influenza vaccine. Twenty-six (21%) of parents would accept a 2009 H1N1 influenza vaccine for their child when the adolescent would not accept the vaccine for themselves and 12 (10%) of adolescents would accept a 2009 H1N1 influenza vaccine for themselves even though their parents would not accept the vaccine. When a κ coefficient was used to assess percent agreement among parents and their adolescent children, fair agreement was found (κ = 0.320) [31].

Discussion

The present study is unique in its focus on factors associated with 2009 H1N1 vaccine acceptance among parents and their adolescent children, as well as level of agreement in 2009 H1N1 influenza vaccine acceptance among parent–adolescent pairs. This study is also distinctive in its focus on 2009 H1N1 influenza vaccine acceptance among a traditionally under-vaccinated population: low-income, rural and minority adolescents. Among parents, the level of intent to get the 2009 H1N1 influenza vaccine observed in the sample was comparable to that reported in other studies [5, 32].

Our results may help identify in advance those parents who may or may not be open to vaccinating their child against influenza (seasonal or pandemic) and may provide insight on possible interventions to encourage vaccination. For instance, parents who would accept a 2009 H1N1 influenza vaccine for themselves were more likely to accept the vaccine for their children. This may provide at least one point of access for vaccination efforts. Additionally, parents who felt motivated by the H1N1 influenza pandemic to get a seasonal influenza vaccine for their children were also more likely to accept a 2009 H1N1 vaccine for their children. Thus, there may be an underlying subset of parents who are highly interested in protecting their children against multiple strains of influenza and therefore, more willing to accept multiple types of influenza vaccines. Alternatively, this finding may also indicate some confusion regarding the prevention potential of the 2009 seasonal influenza vaccination (i.e. believing that the seasonal influenza vaccine would protect against 2009 H1N1 influenza), which highlights the need for clear communication of health information to lay audiences. Our study did not find any significant associations between demographic variables, knowledge or receipt of a seasonal influenza vaccine in the past year and parental acceptance of 2009 H1N1 influenza vaccination for their children. However, due to the small sample size, it is possible that we were underpowered to detect these associations.

A strength of our study is that we assessed attitudes toward 2009 H1N1 influenza vaccination among adolescents, in addition to their parents. Per the Patient-Self Determination Act, adolescents are increasingly encouraged to participate in medical decision making [33]. Therefore, it is important to gain an understanding of adolescents’ attitudes toward 2009 H1N1 vaccination, and the role that adolescent and parental attitudes play in vaccine acceptance. Adolescents who believed that the 2009 H1N1 influenza vaccine could make them sick were significantly less likely to accept the vaccine. This finding indicates that vaccine information materials geared toward adolescents may benefit from emphasizing vaccine safety. Conversely, adolescents who felt comfortable getting the 2009 H1N1 influenza vaccine were more likely to accept it. This finding highlights the importance of increasing adolescents’ comfort with and self-efficacy for 2009 H1N1 influenza vaccination. Other correlates of 2009 H1N1 vaccine acceptance among adolescents were fear of getting the swine flu and receipt of a seasonal influenza vaccine in the past year. Thus, other important areas to target for interventions may be perceived susceptibility to infection and seasonal influenza vaccine acceptance. Finally, we also tested whether parents’ acceptance of 2009 H1N1 influenza vaccine for their children would predict acceptance among adolescents. In both bivariate and multivariate analysis, parental acceptance was a significant predictor of adolescent acceptance. Thus, while adolescents’ own attitudes toward 2009 H1N1 influenza vaccination influence their vaccine acceptance, their parents’ attitudes may also play a role. It may therefore be important to create dyad-level interventions, which focus on increasing vaccine acceptance among parents and their adolescent children.

Finally, we examined the concordance between parents and adolescents regarding 2009 H1N1 influenza vaccine acceptance. A total of 69% of parent–adolescent pairs demonstrated agreement on vaccine acceptance (50%) or non-acceptance (19%). This is a fair degree of agreement, and provides further support for the importance of dyad-level interventions, which focus on increasing vaccine acceptance among parents and their adolescent children.

Limitations

The findings in this report are subject to several limitations. First, the response rate was relatively low, indicating the possibility of response bias among parents and adolescents who were more interested in 2009 H1N1 influenza vaccination. Also, as a result of the low response rate, the sample size was small and reduced the power of our analysis and the precision of our effect estimates. Second, the results are representative only of the populations of two small counties in rural Georgia and may not be generalizable to populations residing in urban areas or other geographic locations. Third, this is a cross-sectional study. Thus, a causal link between attitudes toward vaccination and intention to receive an influenza vaccination could not be established. Fourth, the outcome is 2009 H1N1 influenza vaccination acceptance, not actual vaccination. It is possible that participants who reported vaccine acceptance would not follow through with actual vaccination. Additionally, the measure of ‘perceived susceptibility’ asked whether participants were ‘fearful’ of getting swine flu, which may have assessed an emotional reaction as opposed to a strict measure of perceived risk. Finally, the findings of this study are specific to vaccination against 2009 H1N1 influenza and may not be generalizable to other types of influenza such as seasonal or avian influenza. It is also possible that parents and adolescents may not have clearly understood the difference between H1N1 and seasonal influenza.

Conclusions

Identifying factors associated with influenza vaccine uptake is important to enable public health campaigns to tailor their advice to eligible patients more effectively. Our study found that key attitudes toward 2009 H1N1 influenza vaccination may influence vaccine acceptance both among parents and their adolescent children. Certain correlates of 2009 H1N1 influenza vaccine acceptance differed among parents and their children. However, our study also found a fair degree of concordance in 2009 H1N1 influenza vaccine acceptance among parents and their adolescent children and that parental acceptance toward 2009 H1N1 influenza vaccination was a significant correlate of vaccine acceptance among adolescents. Taken together, these findings highlight the need for interventions, which aim to increase influenza vaccine acceptance among both parents and adolescents in the event of future influenza pandemics. Educational interventions should be sensitive to parents’ and adolescents’ separate and distinct needs for information regarding influenza vaccination, while also addressing the importance of joint decision making. Public health education efforts should focus on developing multi-component messages for both parents and their adolescent children.

Funding

Centers for Disease Control and Prevention, CDC (R18 IP000166) and by Award Number T32AI074492 from the National Institute of Allergy and Infectious Diseases.

Conflict of interest statement

None declared.