Introduction

One decade has passed since the vaccines Gardasil and Cervarix against human papillomavirus (HPV) were introduced. These vaccines have primarily been developed to prevent infection with HPV types 16 and 18, the types responsible for roughly 70% of all cervical cancer, affecting more than 500 000 women annually [1]. They also have a preventive effect against HPV infections causing cancer in the oropharynx and the anus. Furthermore, the quadrivalent HPV vaccine Gardasil can potentially prevent a considerable number of condylomas and papillomas [2, 3].

Despite the successful impact of these vaccines worldwide, their introduction has sparked debate not only about possible side effects of the actual vaccines, but even more on assumed effects on the behaviour of the young people being vaccinated [4–6]. A fear has been voiced that the vaccines may seduce young people into becoming more sexually adventurous and lose their moral values, or even that young people with excellent morals would not need such vaccines [5, 7]. Therefore, their introduction has been challenged not only by the economic costs, but also by compromised social acceptance of the vaccines in various geographical regions or among different communities [8–10].

Previous research indicates that several factors are important for HPV vaccination status, i.e. demographic and socioeconomic factors, individual beliefs and knowledge as well as behavioural aspects. Lower socioeconomic resources correlate with lower vaccine uptake. A Danish register study showed lower HPV vaccine initiation and completion among girls with an immigrant background and girls with mothers with lower educational level [11]. A similar study conducted in Norway found lower vaccine uptake among girls with mothers in the lowest income bracket, mothers with lower educational level and mothers with an immigrant background [12]. Moreover, ethnicity and cultural norms have an impact on vaccine uptake [13–15]. Ethnicity predicted HPV vaccine uptake among adolescent girls in the catch-up cohort in the UK and among young women in Australia [13, 15]. Findings also indicate lower acceptability of HPV vaccine among ethnic minorities even after controlling for socioeconomic status [14].

Adolescents have low knowledge about HPV, especially about the link between sexual behaviour, HPV and HPV-related cancer [16–18]. According to a recent systematic review there are also gender differences, i.e. boys know less than girls [19]. The findings among adolescents in the UK indicate low knowledge and awareness about HPV vaccination [20]. The boys commonly believed that HPV only affected girls and were ill informed about the association between HPV and cancer [20]. HPV infections and HPV-related diseases have increased in the recent decades due to increased sexual risk-taking [21, 22]. Risk-taking in other domains such as alcohol and smoking has been found to correlate with risky sexual behaviour [23]. Furthermore, there are gender differences regarding beliefs about condom use, as well as actual sexual behaviour, boys being less concerned than girls [24, 25].

In 2010, HPV vaccination was offered free of charge to all girls born 1999 or later, and in 2012 the vaccine was introduced into the school-vaccination programme for all girls aged 10–12 years, and as catch-up vaccination for young women. Boys were not included in the vaccination programme at the time. These measures have had considerable impact: it is estimated that presently almost 80% of all girls receive the vaccination [26]. School nurses are responsible for all aspects of the vaccinations, including information to parents and pupils, obtaining informed consent from the former, as well as the actual administration of the vaccine. The written information given by the nurses is standardised by the government. There is an ongoing discussion about including boys in the national vaccination programme, a strategy to which the Public Health Agency of Sweden is in favour [27]. Few studies have been conducted about adolescent boys’ attitudes to HPV vaccinations: so far, most studies on males have addressed male college students or men who have sex with men [28, 29]. Consequently, it is important to examine boys’ beliefs and knowledge about HPV and HPV vaccination.

Two years ago, we conducted a cluster randomized controlled study [30], with the overall aim to improve primary prevention of HPV infection by promoting vaccination and increased condom use among upper secondary school students. This educational intervention was delivered by the school nurse in conjunction with the general health interview that was offered to all adolescents aged 16 years. At this time roughly 50–60% of all the girls had received the vaccine [20]. We found that the intervention had favourable effects on the adolescents’ beliefs towards prevention of HPV and at follow-up after three months more girls in the intervention group were vaccinated than the controls [30].

As already mentioned above, HPV vaccination status was associated with socioeconomic factors, individual beliefs, and behavioural aspects such as risk-taking and sexual experience. This study has therefore focused on the baseline data of the participants of the intervention study.

Our aims were dual:

1. To describe sexual experiences in 16-year-old adolescent girls and boys in general and for the former group also in relation to catch-up HPV vaccination status, (see item 2).
2. To investigate HPV vaccination status in female adolescents in relation to a) sociodemographic factors, b) individual beliefs and knowledge about HPV prevention, and c) sexual experiences.

It was hypothesised that vaccination against HPV was associated with a more favourable socioeconomic situation, more favourable beliefs towards HPV prevention, and less risk-taking. It was also hypothesised that there would be differences between boys and girls regarding sexual experiences and behaviour.

Methods

Design

This cross-sectional study is a part of the project Prevention of HPV in a school-based setting [30], with trial registration—ClinicalTrials.gov Identifier: NCT02280967. The theoretical framework was the Health Belief Model (HBM), described below. We collected the baseline data between October 2014 and February 2015. The focus of this paper is mainly on girls since only girls are offered HPV vaccination in the national vaccination programme. The study follows reports of cross-sectional studies according to STROBE (S1 File. STROBE Checklist) [31].

Setting.

Sweden is a multicultural country of 10 million inhabitants; about 23% of the population at large, and almost 30% of those below age 18 years of age, have an immigrant background with at least one parent born abroad. The most common countries of origin outside the Nordic countries are Afghanistan, Ethiopia, Iraq, Iran, Somalia and Syria–thus, many of these families have experienced war [32, 33]. Almost all 16–19 year-olds in Sweden attend upper secondary school, which comprises theoretical and vocational programmes. During their first year all students are offered a health interview with the school nurse, focused on the individual adolescent’s health and well-being. This interview is non-compulsory but the vast majority of the students do participate [34].

Procedure.

In 2014, school nurses working in upper secondary schools (n = 59) in Sweden were invited to participate in the project via direct contact at the national conference for school nurses. We also sent information about the study by email to the heads of the school health in the respective municipalities (n = 9). The north and the south extremities of the country were excluded for logistic reasons. In the end 20 school nurses, working in 18 schools situated in five counties (nine municipalities) in mid and western Sweden, were included. In a Supplemental file (S1 Fig. Flow chart.) as well as our earlier publication [30], more details and a flow chart regarding the recruitment process are provided. The participating schools comprised those both publicly and privately managed, offered vocationally as well as theoretically oriented education, and had a varying number of students.

In the second stage classes were randomised to be involved and thereafter adolescents in the included classes were invited to participate. The school nurse informed adolescents (n = 832) (both girls and boys) about the project orally and with a leaflet before the health interview started. Adolescents who agreed to participate gave informed written consent and completed a questionnaire individually in a separate room at the school nurse’s office. It took about 10–15 minutes to complete the questionnaire.

Eligible for inclusion were first-year upper secondary school students, aged 15 years or more, attending the general health interview with the school nurse. Students with severe cognitive learning disabilities and development disorders (e.g. students attending special schools) and students who were not able to read and write in Swedish (e.g. recently arrived immigrants) were excluded. For details of the procedure and randomisation see Grandahl et al [30]. Notably, none of the included young girls had received HPV vaccination as 10–12 year olds through the school based vaccination programme, thus all vaccinated girls had received catch-up vaccination.

Theoretical framework.

HBM was used as a theoretical framework for this project. According to HBM, a person’s health behaviour can be explained by individual beliefs regarding health actions. HBM includes the following central constructs: perceived susceptibility, perceived severity, perceived benefit, perceived barriers, perceived self-efficacy and cues to action. In HBM, socio-demographic factors such as age, sex, ethnicity and education level as well as knowledge are recognised as modifying factors for the individual’s behaviour [35]. The theory has previously been used in behavioural research about HPV [36–40] and sexual health [41].

The questionnaire.

The questionnaire was based on previous research [21, 30, 42, 43]. Questions about beliefs (see Table 1), knowledge and awareness about HPV and HPV vaccination (n = 24) were based on previous studies among parents [43] and school nurses [42], and comprised multiple choice alternatives and six-point verbal rating scales from”Totally agree” to”Totally disagree”, including “Do not know”. Questions about sexual health (n = 17) were based on an earlier project among university students [21] and the demographic questions (n = 14) were taken from a national questionnaire for adolescents [44]. The questionnaire has been tested for validity and reliability as described previously by Grandahl et al [30].

Download:

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

Table 1. Beliefs about HPV^* and HPV vaccine–questions based on HBM^**.

https://doi.org/10.1371/journal.pone.0187193.t001

Results

A total of 751/832 (90.3%), 391 girls and 360 boys, agreed to participate. Table 2 presents the characteristics of the participants and their parents. As indicated in Table 2, about one quarter of the participants had an immigrant background. The average age of both girls and boys was 16 years, and there were no differences in parental characteristics. Girls tended to smoke more often than boys, whereas boys used more snuff than girls did. There were no differences between the sexes with regard to alcohol consumption.

Download:

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

Table 2. Characteristics of the participants and their parents.

https://doi.org/10.1371/journal.pone.0187193.t002

Sexual experiences and HPV vaccination status

Around half of the respondents had had sexual intercourse as shown in Table 3. Boys and girls were almost equally experienced in this respect. The reported ranges in sexual debut were mainly between 12 and 17 years of age, while two students reported 11 years of age and one student reported 19 years of age. Girls reported slightly older partners than boys did, had given oral sex to a higher extent, were more likely to have been tested for a sexually transmitted infection (STI), and were more often vaccinated against HPV.

Download:

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

Table 3. Sexual experiences, HPV^* vaccination status and contraceptive use among sexually active participants.

https://doi.org/10.1371/journal.pone.0187193.t003

Beliefs and knowledge about HPV among vaccinated and non-vaccinated girls

Table 4 shows differences between vaccinated and non-vaccinated girls with regard to the HBM-related beliefs, although no differences were found in knowledge about HPV. Vaccinated girls perceived HPV infection as more severe, had more insight into women’s susceptibility of the infection, perceived more benefits about the vaccines protective effect against cervical cancer and had a higher intention to engage in HPV-preventive behaviour. Non-vaccinated girls were more ignorant about how to make an appointment for HPV-vaccination.

Download:

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

Table 4. Beliefs and knowledge about HPV^* and HPV vaccination with respect to vaccination status.

https://doi.org/10.1371/journal.pone.0187193.t004

There were no differences in separate knowledge items about HPV and HPV vaccination among vaccinated or non-vaccinated girls. However, when computed together to form a knowledge score, vaccinated girls scored higher (Table 4).

Differences between HPV vaccinated and non-vaccinated girls with respect to background variables, sexual experience and other HPV related variables were also examined. The results of the bivariate analyses are shown in Table 5. There were several differences between the groups with respect to their background variables, but no differences in sexual experience. All HPV related variables differed between vaccinated and non-vaccinated girls (Table 5).

Download:

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

Table 5. Bivariate analysis of differences between HPV^* vaccinated and non-vaccinated girls.

https://doi.org/10.1371/journal.pone.0187193.t005

A multivariate logistic regression model including selected variables differing significantly in the bivariate analysis was performed as described in the Methods section. The result is displayed in Table 6, showing that non-European background and maternal education level remained associated with HPV vaccination.

Download:

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

Table 6. Multivariate logistic regression models of variables associated with HPV^* vaccination among girls.

https://doi.org/10.1371/journal.pone.0187193.t006

Finally, we examined sexual risk-taking in relation to HPV vaccination status and found that vaccinated girls had less unprotected first intercourse with their latest partner (12.7% vs. 31.8%, p<0.01), and less unprotected latest intercourse (13.9% vs. 30.6%, p = 0.01) compared to their non-vaccinated peers. No difference was found in relation to first intercourse (16.5% vs. 22.2%). The girls’ intention to prevent HPV infection in the future did not differ with respect to demographic variables, such as country of birth or parental factors.

Discussion

This study aimed to examine catch-up HPV vaccination status in youth in relation to socioeconomic factors, beliefs and knowledge about prevention of HPV, and sexual behaviour among boys and girls. The study showed that there were differences in beliefs and in health behaviour regarding sexual risk-taking. Differences in beliefs were that vaccinated girls perceived HPV infection as more severe, had more insight into women’s susceptibility of the infection, perceived more benefits about the vaccines protective effect against cervical cancer and had a higher intention to engage in HPV-preventive behaviour. Differences in sexual risk-taking were that HPV vaccinated girls had less unprotected first intercourse with their latest partner, and less unprotected latest intercourse compared to their non-vaccinated peers. However, the girls’ intention to prevent HPV infection in the future did not differ significantly with respect to demographic variables, such as country of birth or parental factors.

Our hypothesis, that vaccination against HPV was associated with a more favourable socioeconomic status, more favourable beliefs towards HPV prevention, and less risk-taking was partly confirmed. In the bivariate analysis differences were found between vaccinated and non-vaccinated girls depending on parental educational level and parental working status, discussions of HPV vaccine with parents and others and if a friend was vaccinated against HPV. However, in the final multivariate logistic regression model the only variables associated with HPV vaccination were country of birth and mother’s level of education.

The difference in HPV vaccination status between ethnic groups was not surprising since the most common origins of immigrants in Sweden today are from non-European countries such as Afghanistan, Syria, Ethiopia, and Somalia. Girls with an immigrant background are vulnerable groups that often need catch-up vaccinations [45]. Cultural norms and religious beliefs can have an impact on health behaviour and HPV vaccination status [13, 14]. Bowyer et al found ethnicity independently associated with vaccine uptake, and adolescent girls with an immigrant background had lower vaccine uptake [13]. Similar findings have been found in registry studies undertaken in Norway and Denmark [11, 12]. More specifically, differences were found in vaccine initiation and completion, depending on socioeconomic factors such as parental education, employment and income level, with lower vaccination status among girls with immigrant or poorly educated mothers [11, 12]. However, national vaccination programmes can reduce disparities in vaccine uptake due to socioeconomic factors, a fact which emphasises the need for school-based vaccinations.

Generally, the responses of the HPV vaccinated girls revealed more informed beliefs and sound intentions than those of their non-vaccinated peers. This is not surprising, since individual beliefs about HPV vaccination surely influence the decision to take the vaccine. It should also be noted that parents have to consent to vaccination for children and parental beliefs are thus important as well [4, 43, 46–49]. This fits well with our finding in the bivariate analyses that HPV vaccination was associated both with discussions with parents or others and the presence of vaccinated friends. Peers evidently have a substantial influence on the adolescents’ decisions, an observation which is supported by HBM. The discussions with friends might have strengthened the girls’ self-efficacy and influenced the girls to vaccinate, i.e. they might have functioned as cues to action.

According to HBM, knowledge is a modifying factor for the individual health behaviour, yet, in the final regression model the knowledge index did not remain associated with HPV vaccination. However, the knowledge about HPV and HPV vaccine was low among the participants overall. This lack of knowledge was disheartening, but has been described previously [17, 30, 43, 50]. There is surely a need for more public education in this important field, especially about the link between HPV infection and cancer [17, 30]. As found in previous studies educational intervention can have favourable effects on adolescents’ awareness and beliefs [18, 30, 50] and also on vaccination status [30].

For girls whose mothers had a low education level it was less common to be vaccinated. For girls born outside Europe, however, the lower odds ratios for being vaccinated persisted also after taking the level of knowledge among the girls into account. A previous study has shown positive attitudes towards preventive measures for cervical cancer among migrant women in Sweden [51]. However, the health literacy, i.e. the ability to access, understand value and use health information, is often low among migrant groups in Sweden. In order to not only receive information about vaccination, but also to value and use the information and take action to vaccinate, more advanced health literacy is needed. The lower vaccination rates among non-European girls thus may be due to lower levels of complex health literacy as well as to poor knowledge about the Swedish healthcare system. Preventive strategies may need to take this into account by not only informing about vaccinations, but also give room for an active dialogue about how to value and use this information [52].

Boys and girls were almost equally sexually experienced, although girls reported slightly older partners than boys did, had given oral sex to a higher extent, and were more likely to have been tested for an STI. This is not surprising, since the vast majority attending youth clinics in Sweden are girls. Previous studies also indicate that boys have lower knowledge about HPV and the HPV vaccine and are less worried about sexual health risks [19, 20, 23–25]. Thus, it is important to provide information to both boys and girls, especially about the link between sexual behaviour, HPV infection and HPV-related cancer. Sadly, the low condom use is a common finding in different contexts among youth, which might indicate the need for gender-neural vaccinations, i.e., HPV vaccination to both boys and girls.

No differences were found in experience of intercourse and oral sex among HPV vaccinated and non-vaccinated girls. HPV vaccination was not associated with earlier sexual debut or more sexual risk-taking (unprotected sexual intercourse). This very important finding needs to be emphasized, since the fear that HPV vaccination could promote “promiscuity” is a recognized obstacle against vaccination [4, 5, 7]. Our data show that this fear is ungrounded. Our findings were also in line with previous research undertaken in the USA among girls and young women aged 13–24 [7]. If authorities and the community at large acted in accordance with this knowledge more adolescents would become vaccinated and, in the end, lives would be saved.

Supporting information

Acknowledgments

We thank all the participants; especially the adolescents and the school nurses. We also thank the heads of the school health in each municipality and the principals of the schools. A special thanks to the Swedish Association of School nurses for support, and for allowing us to present the project at the national conference.