Introduction

Influenza causes high risk of complications such as pneumonia and bronchitis, which may lead to hospitalizations or even deaths among young children [1,2]. The World Health Organization recommends annual influenza vaccination (IV) to children aged 6–59 months [3]; the Advisory Committee for Immunization Practices in the U.S. also recommended it to children aged 6–23 (2004) and 24–59 months (2006) [4]. Age is important as prevalence of IV concerning <2 or ≥2 years old children differ. Previous studies in the U.S. reported IV prevalence specifically for <2 years old children (56% in 2003 and 31.9% in 2006) [5,6]. Other studies conducted in the U.S. reported IV prevalence both for children aged 6–23 (41.0%-74.4%) and 24–59 months (range = 36%-53.3%) [7–9]. In Spain, there was no noticeable difference between the two age groups (2.2% versus 5.2%) [10]; the younger group showed lower prevalence than the older group of children in Poland (4.1% in 0–1 year, 26.4% in 1–2 years, and 42.3% in 2–5 years) [11]; reverse trends were observed in the U.S. (56% versus 38%) [12] and France (21% versus 15%) [13].

Both sociocultural factors such as policies (e.g., IV providers’ visits to schools) and experience related to IV are important in explaining age differences [8,9]. The factors of IV may differ between <2 and >2 years old children, although no study has directly compared such between-group differences. Different strategies might hence be required to promote IV for the two age groups. While there are a number of studies conducted among children >2 years old and reported factors such as perceived vaccine effectiveness, perceived influenza severity, and subjective norm [14–16]. There are only two related studies conducted among children < 2 years old, which found that health care workers’ recommendation, parental concerns about side-effects, and parental positive attitudes toward IV were significant factors [17,18]. Hence, there is a knowledge gap about the understanding of the factors of IV for children under two years old. Such research is potentially useful for design of programs.

Furthermore, IV is free for young children in many developed countries including Canada, Slovakia, Australia, and the U.K. [19–22]. In 2008/09, the Hong Kong government recommended 6–71 months old children to take up IV annually and reimbursed HKD 80 (about USD 10) to private practitioners for IV. The market price was about HKD200 in 2011/12. The subsidy was increased to HKD 130 and HKD 160 in 2012/13 and 2014/15, respectively. Children of families receiving the Comprehensive Social Security Assistance (CSSA) scheme can take up IV free of charge at specific clinics. A study reported low IV prevalence of 9.0% among children aged 6–23 months in Hong Kong in 2006 [17]. The previous study in children IV in Hong Kong was conducted in 2006. As the government launched the subsidy policy on IV in 2008/2009, it is meaningful to conduct a study in 2011 to see if the prevalence of IV has increased over a two-year period since the implementation of the new policy. In addition, another local study, which was conducted from March to June 2011 [16], reported higher IV prevalence of 58.9% among children aged 24–59 months [17].

Behavioral health theories are commonly used to understand factors of IV. The Health Belief Model (HBM) prescribes parental perceived susceptibility and perceived severity of influenza, and perceived benefit, perceived barrier, cue to actions and self-efficacy related to IV as determinants of young children’s IV [23]. It has been used to guide variable selection in a local study among 24–59 months old children [16] and a Taiwan study investigating IV among children aged 6–36 months old [24], and other studies targeting older people, health care workers, pregnant women, and chronic diseases patients [25–29]. Given its wide applicability, the present study used it as the theoretical framework. In addition, subjective norm, defined as perceived support from significant others for performing a health-related behavior, a construct of the Theory of Planned Behaviors (TPB) [30], was relevant as family members and friends influence parental decision on children’s IV. In the local study [16], factors of IV for 24–59 months old children included family members’ IV experience, subjective norm and those related to the HBM. Other potential factors of IV included family members’ IV experience [8] and exposure to influenza-related media messages [8,31]. Fear experienced during the H1N1 pandemic was also potentially important, as fear induced from SARS/H1N1 was associated with intention to take up IV [32,33]. H1N1 has become a major source of seasonal influenza. The present study considered such factors.

We investigated prevalence and associated factors of IV among Chinese children aged 12–23 months in Hong Kong. Besides socio-demographics, potential factors included: 1) family members’ IV status, 2) parental perceptions of the HBM and subjective norm of the TPB, 3) parental fear experienced during the H1N1 pandemic, and 4) exposure to related mass media messages.

Methods

Results

Background characteristics

About half of the index children were male; 47.4%, 51.1% and 1.4% were 12–15, 16–19, and 20–23 months old, respectively (Table 1). Majority were born in Hong Kong (99.2%), with both parents being Hong Kong residents (89.6%); 61.9 were the first child. Majority of the participants was mother of the child (80.6%) and had attained senior high or above education (90.0%); 67.7% were 30–39 years old; 55.6% were employed full-time; 3.1% of the participants’ families were covered by the CSSA. About one-fourth (26.4%) reported that their family members’ had taken up IV in the last year.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of the index children and their parent participants (N = 489).

https://doi.org/10.1371/journal.pone.0205561.t001

Perceptions and knowledge

Of the participants, 7.5% and 45.2% perceived high susceptibility (score ≥6) for their child to contract H1N1 and seasonal influenza, respectively; 28.4% and 19.1% perceived high severity of H1N1 and seasonal influenza, respectively. Over half (57.1% to 68.1%) perceived some benefits of IV; 43.1% believed that the currently available influenza vaccines were more effective than the previously available ones. Concerning perceived barriers, prevalence of items reflecting safety concerns ranged from 35.0% (“This vaccine is not safer than the previous one”) to 58.5% (“My child is too young to take up IV”); concern for barriers of cost and inconvenience was less common (23.7% and 16.4%). Regarding cue to action, only 12.9%-19.2% had received an advice concerning the child’s IV from healthcare providers, peer parents, and relatives/friends; 17.2% believed that it was common for children of similar age to take up IV; 48.3% believed in the governmental recommendation on IV for young children. Perceived self-efficacy was very high (90.6% and 92.8%). Regarding subjective norm, less than half found their spouse (42.9%) and other family members (40.7%) supportive of the child’s IV (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Parental perceptions and knowledge toward influenza and IV.

https://doi.org/10.1371/journal.pone.0205561.t002

Majority of the participants knew that the local government advised children of similar age to take up IV (75.3%); 40.1% and 28.0% knew that the local government and WHO recommended current influenza vaccines including H1N1 to children of similar ages; about one-fourth knew that the vaccines currently provided by the local government for children included the H1N1 vaccine (23.9%). Only about two-third knew that children aged 6–71 months should take up IV annually (59.3%) (Table 2).

Factors of IV

Child’s age was the only significant background factor (16–19 months: ORu = 2.88, 95%CI = 1.52, 5.46; 20–23 months: ORu = 11.68, 95%CI = 2.38, 57.36; Table 3); it was adjusted for in subsequent analyses. A number of variables were significantly associated with IV status and in both the univariate and adjusted analyses (Table 3). Their AOR were: (1) family members’ IV experience in the last 12 months (AOR = 6.60, 95% CI = 3.60, 12.11), (2) knowledge about governmental policy and recommendations on IV for young children (3–4 correct responses: AOR = 2.64, 95% CI = 1.09, 6.40; reference: no correct response), (3) knowledge about the annual vaccination requirement (AOR = 2.30, 95% CI = 1.21, 4.38), (4) the HBM-related variables [perceived barrier related to safety (1–2 item responses: AOR = 0.24, 95% CI = 0.12, 0.47; 3–4 item responses: AOR = 0.14, 95% CI = 0.06, 0.33; reference: 0 item response), and cue to action (3–5 item responses: AOR = 7.79, 95%CI = 3.45,17.58; reference: 0 item response)], and (5) supportive subjective norm (AOR = 4.59, 95% CI = 2.34, 9.00). Two other variables, influenza experience (ORu = 1.79, 95% CI = 1.02, 3.12) and perceived severity of seasonal influenza (ORu = 0.38, 95% CI = 0.15, 0.98) were significant in the univariate analysis but not in the adjusted analysis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Factors associated with IV (ever received) of the index child (among all participants).

https://doi.org/10.1371/journal.pone.0205561.t003

Using all the significant variables found in the adjusted analysis as initial variables, three independent variables were kept by the stepwise backward multivariate logistic regression model, while the rest were removed from the final multivariate model. The variables of the model included family members’ IV experience (ORm = 6.24, 95% CI = 3.20, 12.18), the scale on perceived barrier related to safety of IV (1–2 item responses: ORm = 0.29, 95% CI = 0.13, 0.62; 3–4 item responses: ORm = 0.23, 95% CI = 0.09, 0.60; reference group: 0 item response), and supportive subjective norm (ORm = 2.37, 95% CI = 1.02,5 .53). Such results are not shown in the tables.

Discussion

Based on the data of 489 Chinese parents of children aged 12–23 months, the results showed that only 11.5% of the children had ever received IV. The significant factors of the children’s IV included knowledge about the governmental policy/recommendation, knowledge about annual IV requirement, perceived safety-related barrier, cue to action and subjective norm.

More countries/regions are providing full financial support than providing partial financial support (e.g., Latvia and Hong Kong) for IV among young children [19,36]. Despite introduction of the local subsidization policy in 2008/09, we found that IV prevalence among very young children (aged 12–23 months) remained very low in 2011 (11.5%). As a comparison, IV prevalence was 9.0% among 6–23 months old children (and 9.3% among 12–23 months old children) in the 2006 study [17] that used same sampling and data collection methods as those of this study. Thus, improvement in coverage of IV in the 12–23 months group over the five years was minimal. The prevalence was much lower than that of children aged 6–23 months in the U.S. (56%) and France (21%) [13,37]. Thus, the subsidization policy in Hong Kong and related health promotion did not seem to have increased coverage of IV in this age group substantially. Cost was not a barrier in this age group, as only 23.7% perceived that cost for IV was relatively high and perceived cost was not a significant factor of IV. The non-significance of cost suggests that the sampled parents were not responsive to the subsidization policy.

An important insight comes from the very strong positive association between the child’s age and IV uptake observed in this study. It seems that many parents did not object to IV for their children, but instead, postponed their action, as 58.9% of the children aged 24–59 months had taken up IV in 2011(during approximately the same study period as this one) [16] versus 11.5% of the 12–23 months old children involved in this study. Also, 58.5% of the parents perceived that their children were too young to take up IV. The concern of the child’s age might be due to worry about side-effects of IV. In both this study and the 2006 survey targeting children aged <2 years [17], perceived safety of IV was strongly associated with the child’s IV status; the findings also corroborate those of other studies [16,17,24]. Young children’s parents need to know that side-effect of IV among young children is not age dependent and that children aged 12–23 months need as much protection as older children [3], and as early as possible.

This study, the 2006 survey (6–23 months old children) [17], and other studies targeting young children [16,24] consistently found strong associations between health care workers’ advice and the child’s IV status. The adjusted analysis of both this and the independent concurrently conducted survey in a 24–59 months group [16] found very strong associations between cue to action (including that given by health professionals) and IV status. About 20% of the participants had been advised about their children’s IV, which was almost doubled when compared to the data (10.6%) obtained from a comparable survey conducted in 2006 [17]. The prevalence of this study was still low, despite improvement. It is important as about 90% of the local newborns would visit MCHC [38]. Thus, health care workers need to play a stronger role in information dissemination and promotion of IV for children under two years old. Furthermore, about 24.7% did not know that IV has been recommended for children aged 6–71 months, while that knowledge was a significant factor of IV. The government should also promote the knowledge that IV is required annually for young children, which was another significant factor of IV in this study.

In this study, about 40% of the indexed children had been infected by influenza but the experience was not associated with IV status. It implies that prior experience of influenza cannot motivate parents to vaccinate their children. On the other hand, both family members’ IV experience and supportive subjective norm (of spouse and family members) were significant in both the adjusted and stepwise analyses. The independent concurrently conducted study [16] also found significant associations between these two factors and IV among children aged 24–59 months. Although IV among young children is a family matter, only half of the spouse and family members were supportive of the child’s IV both in this and the 24–59 months group studies [16]. As a new approach, counseling is needed to harmonize divergent opinions about children’s IV and to enhance communication skills among parents. The significant composite variable on cue to action found in this and the 24–59 months group studies [16] included items that involved friends, relatives, and other peer parents. We should hence engage parents’ social networks, possibly through social media or other means. Pilot studies are warranted to take this new approach.

Only two HBM constructs (perceived barriers and cue to action) were found to be significant factors; perceived susceptibility and perceived benefits were significant in the concurrently conducted local study of the 24–59 months group [16] but not in this study’s 12–23 months group. Perceived severity and perceived efficacy were non-significant in both studies. Likewise, perceived severity and perceived benefit were not significantly associated with the child’s IV experience in the 2006 study (6–23 months group) [17]. Although HBM is applicable to understand IV in different high risk groups [26,27], it seems to be working less well in explaining IV among children aged 12–23 months. Making IV more convenient and less costly might not be effective in promoting IV, as perceived efficacy and cost as perceived barriers were not significant factors in this study’s age group. Related media exposure was non-significant in both this study and the concurrently conducted 24–59 months group local study [16], suggesting general media campaigns are not promising. Fear experienced during the H1N1 pandemic was a significant factor for the 24–59 months group [16] but not for this study’s age group. Explanation of the difference is beyond the scope of this report. The observation that most of the local health promotion campaigns for IV among young children have been engaging some of the aforementioned non-significant components might partially explain the low IV prevalence in this age group.

The study has the strength of having some degree of comparability with two other local studies involving young children below [17] and above two years old [16]. It also has some limitations. It did not include children who were 6–11 months old during the study period (July-September 2011), as those children were younger than six months during the 2010/2011 IV season (November 2010 to March 2011) and were not by then recommended to take up IV. We contended that IV prevalence of the 6-11-month age group was even lower than that of the 12–23 month group, as IV was positively associated with age. Given that 92% of the Hong Kong population are Chinese [39], we only investigated the IV among Chinese children. Findings of other ethnic groups might be different and require another study. Also, the study’s response rate was not measured. The nurses were engaged in clinical work and did not have time to record such data. Furthermore, the 10% of the newborns who did not visit MCHC were not included in this study. We did not interview both parents but just one parent of the index child because usually only one parent visited the MCHC with the child. The scales were constructed by the research team, although they were almost identical to those used in a published study for the 24–59 months group [16]. Data were self-reported and might involve reporting bias. Our data were collected in 2011; future studies are recommended to use this study’s data to construct trends. Moreover, the current findings cannot infer causality, as they are cross-sectional in nature.

To conclude, prevalence of IV among 12–23 months old Chinese children in Hong Kong was low and showed little improvement as compared to 2006. Parents might be postponing these children’s IV unnecessarily. Effectiveness of the governmental recommendation, subsidy, use of some HBM constructs and mass media in increasing such children’s IV is uncertain. The government should revamp related health promotion strategies and treat the young age group (12–23 months) as a distinct segment. It should also emphasize dissemination of governmental policies, increase perceived safety of IV for these children, and solicit support from primary care doctors, family and friends. Children aged 12–23 months in Hong Kong were not well protected against influenza and comprehensive strategies are needed.