Theoretical framework

The theoretical framework used is the Health Belief Model (HBM). The HBM has previously been used in studies about parents’ beliefs about HPV and the HPV vaccination [20, 29, 44, 45]. According to the HBM, a person’s health behavior can be explained by the individual’s beliefs regarding the health action. The HBM includes these central constructs: perceived benefit, perceived barriers, perceived susceptibility, and perceived severity and also cues to action. Knowledge and socio-demographic variables such as age, sex, religion, and ethnicity as well as the education level and income are modifying factors for the individual behavior. Moreover, it is important for a person with risk behaviors to recognize the risk in order to be able to change his or her behavior. The benefits have to outweigh the barriers for a person to act upon the health promotion, for example, participating in a screening (cervical cancer) or vaccination program. The main limitation of the model is that it does not consider emotional aspects involved in an individual’s decision regarding their health behavior [46].

Our theoretical framework is based on previous research and clinical experiences. Socio-economic, cultural and environmental conditions, the community and social network, including work and health services, and individual lifestyle factors have an impact on health status. Health inequalities are related to socio-economic status and life conditions overall. Lower socio-economic (i.e., lower educational level and lower income) status is related to a poorer standard of health. Alcohol consumption, tobacco use, risky sexual behavior, and lack of adherence in attending recommended medical screening programs, such as cervical cancer screening, are associated with unhealthy behavior and risk for cancer. Addressing access to free and adequate healthcare services and changing health behaviors are important factors for reducing unhealthy behaviors as well as reducing the burden of cancer worldwide [47, 48].

Based on the HBM and literatures we reviewed, it was hypothesized that there would be differences among parents due to the demographic variables; parents with higher socio-economic status 1) had greater knowledge, 2) perceived more benefits, 3) perceived less barriers for the HPV vaccination, and 4) had higher HPV vaccine acceptance compared with parents with lower socio-economic status. It was also hypothesized that the importance of religion was associated with knowledge, beliefs, and acceptance of the HPV vaccination.

Study design, sample, and setting

We used a cross-sectional approach; thus, this design cannot provide information about causality regarding the relation between parental attitudes, beliefs, behavior, knowledge, or acceptance of the HPV vaccination. Consequently, no experiments were conducted, and we cannot provide the causal relationship. Measurements of the parents’ attitudes, beliefs, knowledge, and acceptance of HPV and the HPV vaccination were undertaken on one occasion. Thus, when referring to relation throughout the paper, we do not point out the causality but an association between the variables.

The data were collected using a self-reporting questionnaire. A purposive sample of parents of daughters aged 9–12, with the ability to read and write in Thai, were invited to participate. Data were collected between September and November 2015 in three schools. Two schools in Bangkok as representative of the city (urban area) and one school in Nakhon Pathom Province as representative of the suburbs of Bangkok were selected. A combination of city and suburban schools was included to ensure the various enrolments of pupils living in low- and high- income areas. All classes from grades 3–6 in each school were recruited, targeting the parents of all female pupils. Inclusion criteria were being a self-identified parent or a primary caregiver of female pupil(s) aged 9–12-years-old in the schools and classes, as described above. The Health Belief Model [46] was used as the theoretical framework. The study follows reports of cross-sectional studies according to STROBE (S1 File.) [49].

Procedure

One of the researchers (PS) contacted the principals at the respective schools and asked for permission to conduct the study. The principals received information about the study’s aim and procedure. After getting the ethical approval and permission from the schools, a research assistant who had been trained in collecting the data contacted the teachers, who were responsible for all the classes in grades 3–6 in each school, to schedule an appointment to give information about the study and deliver the questionnaires. The teachers delivered the informational letter about the study, the rights for participation, and the questionnaires to all female students in the class with the instruction to give to her mother or father (i.e., legal guardian) to complete. Parents of female students in the three schools were provided with written information, including that participation was voluntary and confidential. Those who agreed to participate completed the questionnaire at home. After three weeks, the research assistant contacted the teachers to collect the questionnaires. In total, 359 questionnaires were distributed: 186 in the two schools in Bangkok and 173 questionnaires in the school in the Nakhon Pathom Province.

Questionnaire

The validated questionnaire (S2 File.) was based on previous studies [20, 50, 51] and comprised 18 principle questions (60 items) grouped into four areas: 1) Socio-demographic variables and health behavior (13 items); 2) knowledge related to HPV, the HPV vaccine, and cervical cancer (22 items); 3) beliefs about the HPV vaccine (16 items); and 4) acceptance of the HPV vaccination (9 items). In addition, one open-ended question for additional comments was included.

A five-point Likert scale was used for questions regarding beliefs about HPV and the HPV vaccine in relation to the HBM, with possible responses ranging from “strongly disagree” to “strongly agree.” A five-point scale was also used for questions about acceptance of the HPV vaccine. The questions regarding knowledge about HPV and the HPV vaccine had the response alternatives dichotomized: yes/no or true, false, or do not know. The remaining questions had multiple-choice alternatives. It took about 15–20 minutes to complete the questionnaire.

A pre-test was performed among 20 parents of female students aged 9–12-years-old before the actual data collection commenced. Some questions in the questionnaire were rewritten, i.e., clarified in order to make it clearer for the participants and increase the face- and construct validity.

Variables and measurement

A total of six dependent variables were measured as continuous variables and used for the analysis, including (1) knowledge of HPV and cervical cancer, (2) susceptibility, (3) severity, (4) benefit, (5) barriers of HPV and the HPV vaccine, and (6) acceptance of the HPV vaccination. For knowledge of HPV and cervical cancer, each individual respondent was asked to answer a total of fourteen true or false questions, and the total number of correct answers was represented as the level of knowledge of HPV and cervical cancer. The value of the knowledge ranged from zero to fourteen, and a higher value indicated a higher level of knowledge.

The other five dependent variables were measured by the following steps: (1) For each dependent variable, the numeric values chosen by each individual respondent for each question, which is five-point Likert scale question, were added and (2) the sum-score for each question was then divided by the total number of questions to obtain an average score. For instance, for acceptance of the HPV vaccination, each individual was asked to answer a total of seven questions; thereafter, the total score of each question was divided by seven. The value of the acceptance score ranged from one to five, and a higher value indicated a higher level of acceptance.

For independent variables, school location was measured as a binary variable (urban vs. suburban). The survey respondent was measured as a nominal variable with three levels, which are mother, father, and others including aunt, grandfather, grandmother, and uncle. Age was used as an ordinal variable with three levels, which are the age under 30, 31–35, and over 35 years. Education was measured as an ordinal variable with four levels, namely primary school or below, secondary school, vocational school, and college or above.

Employment was used as a binary variable (yes vs. no) in which the respondents classified into the group ‘yes’ were employee, owner of business/merchant, government officer, or farmer. Income, which represented the monthly household income in THB, was measured as an ordinal variable. The number of daughters was also used as an ordinal variable. Any childhood vaccines were measured as a binary variable (yes vs. others). Each individual respondent was asked to answer a question ‘my daughter has received other recommended childhood vaccines.’ If respondents answered ‘yes all’ or ‘yes some,’ they were classified into the group ‘yes.’ If respondents answered ‘no’ or ‘not sure,’ they were classified into the group ‘others’.

Religious importance was measured as a binary variable (very/rather important vs. rather little/very little importance). Both alcohol use and smoking were measured as binary variables (regular/irregular vs. non-drinker/smoker). Respondents were classified into the group ‘regular/irregular’ based on if they were weekly, monthly, or non-regular drinkers or smokers. Both health check-up and Pap Smear test were measured as ordinal variables with four levels, which were never, 1- or 2-year interval, 2- or 5-year interval, and more than 5-year interval. Mother’s abnormal Pap Smear and mother’s cervical cancer history were measured as nominal variables with three levels, which were yes, no, and not sure.

Background knowledge of HPV was measured as a binary variable (yes vs. no). Each individual respondent was asked to answer three yes or no questions such as ‘have you ever heard about HPV?’ If the respondents answered ‘no’ to all three questions, they were classified into the group ‘no.’ If not, they were classified into the group ‘yes.’ The variable background knowledge of the HPV vaccine was also addressed in the same manner.

Statistical analysis

Since all dependent variables were continuous, ordinary least squares (OLS) regression analyses were performed. A total of six OLS analyses were conducted separately. First, we performed an OLS analysis to examine the relationship between the independent variables and knowledge of HPV and cervical cancer (Model 1). Second, four separate OLS analyses were conducted to examine how the knowledge from Model 1 is associated with beliefs about HPV and the HPV vaccine, which includes susceptibility, severity, benefit, and barriers of HPV and the HPV vaccine (Models 2–5). Finally, an OLS analysis was conducted to examine how the knowledge and the beliefs were related to acceptance of the HPV vaccination (Model 6).

Assumptions of the OLS models, particularly multicollinearity and residual normality, were tested. For multicollinearity, both Pearson and Spearman correlation analyses were conducted for continuous variables, and Cramer’s V was calculated for the categorical variables. Modest correlation coefficients obtained from the analyses suggested that multicollinearity did not appear to be problematic. Additionally, tolerance tests in each OLS model did not show multicollinearity to be an issue. For residual normality, we inspected the residual distributions of each OLS model and found that there was no strong violation of the normality assumption. All data management and statistical analyses in this study were performed by using IBM SPSS Statistics version 20 software (S3 File. Data set).

Ethical considerations

The study was conducted according to the ethical requirements stated in the Declaration of Helsinki [52]. All participants received information before giving their written consent. The participants were informed that participation was voluntary and that they could withdraw participation at any time. They were also informed that only the researchers would have access to the data and that all data would be presented on a group level. The Committee for Research Ethics, Faculty of Social Sciences Humanities, Mahidol University, Nahkon Pathon, Thailand, approved the study. Approval no: 2015/331.2010; MU SSIRB no: 2015/389 (B2).

Results of the OLS regression analysis

Table 3 presents the results of the OLS regression analyses. In model 1 (socio-demographic variables vs. knowledge of HPV and cervical cancer), a significant relationship was found between four variables, including income, importance of religion, background knowledge of HPV, and background knowledge of the HPV vaccine.

Download:

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

Table 3. Results from the OLS regression analysis with participants’ characteristics and beliefs and knowledge about HPV and the HPV vaccine in relation to the HBM.

https://doi.org/10.1371/journal.pone.0193054.t003

For income, participants with more than 30,000 THB were positively related to knowledge, compared to those with less than 5,000 THB. This indicated that knowledge of HPV and cervical cancer was significantly higher in participants with a high-income level than those with a low-income level. However, participants with a middle-income level (i.e., participants with 5,001–9,999 THB and with 10,000–29,999 THB) were not significantly related to knowledge compared to those with a low-income level (i.e., participants with less than 5,000 THB).

Importance of religion was positively associated with knowledge, with the coefficient equal to 1.67. Participants who reported that religion was ‘very/rather’ important had greater knowledge of HPV and cervical cancer than those who reported that religion was of ‘rather little/very little’ importance. Lastly, both background knowledge of HPV and the HPV vaccine were positively related to knowledge of HPV and cervical cancer, with the coefficients equal to 1.12 and 1.88, respectively. This indicated that knowledge about HPV and cervical cancer was significantly higher in participants who had heard of or been informed about HPV or the HPV vaccine than those who had not.

In models 2–5 (knowledge of HPV and cervical cancer vs. belief of HPV and the HPV vaccine), knowledge was found to be positively associated with susceptibility, severity, and benefit, with the coefficient equal to 0.05, 0.06, and 0.03, respectively. It meant that participants with greater knowledge of HPV and cervical cancer had higher susceptibility, severity, or benefit of HPV and the HPV vaccine. However, the relationship between knowledge and barriers was positive but not statistically significant.

For socio-demographic variables, school location was negatively related to benefit. Specifically, participants who had daughters attending the schools in Bangkok had lower benefit than the school in the suburb. Age was negatively associated with severity. Older participants, those who were over 31, had lower severity than younger participants, those who were under 30. For income, participants with 5,001–9,999 THB had lower benefit than those with less than 5,000 THB.

Importance of religion was positively related to both susceptibility and benefit. Specifically, participants who reported that religion was ‘very/rather’ important had higher susceptibility and benefit than those who reported that religion was of ‘rather little/very little’ importance. For alcohol use, participants who were ‘regular/irregular’ drinkers reported lower susceptibility than those who were non-drinkers. Lastly, for background knowledge of HPV, participants who had heard of or been informed about HPV had higher benefit than those who had not.

Finally, model 6 presents the results of the OLS analysis to investigate how knowledge of HPV and cervical cancer and belief of HPV and the HPV vaccine are associated with acceptance of the HPV vaccination. Knowledge was positively related to acceptance, with the coefficient equal to 0.03. Participants with greater knowledge had higher acceptance of the HPV vaccination. For belief, participants with higher susceptibility were found to have higher acceptance of the HPV vaccination.

Additionally, for socio-demographic variables, participants who were ‘regular/irregular’ smokers had lower acceptance of the HPV vaccination than those who were non-smokers. For background knowledge of the HPV vaccine, participants who had heard of or been informed about the HPV vaccine had higher acceptance than those who had not.

Strengths and limitations

The study was conducted in Thailand, a country with the highest incidence of cervical cancer worldwide. The study is undertaken in different socio-demographic and geographical areas, both city and rural areas, and includes a diverse range of parents of different beliefs and socio-economic status. Another strength was the theoretical framework of the HBM, which is a systematic way to explain a person’s health beliefs [46]. The design was considered appropriate since there is no previous knowledge about Thai parents’ beliefs and acceptance of the HPV vaccination. However, the cross-sectional design cannot provide information about causality regarding parental acceptance of the HPV vaccination.

The majority of the participants were mothers. One reason for this might be that in Thailand mothers are usually the decision-makers about the child’s health. Research also shows that mothers are the decision-makers regarding the HPV vaccination [15, 16]. The parents completed the self-reported questionnaire in their homes, and self-reported data must always be interpleaded with caution. Furthermore, we purposively selected three schools, and such purposive sampling method with a small sample size may include overgeneralization. Thus, this issue should be taken into consideration in a future study. In addition, we recommend longitudinal analysis in order to decrease the issue of simultaneity that our OLS analyses include.

Conclusions

We found associations between parents’ knowledge, beliefs, and acceptance of the HPV vaccination for their young daughter, in relation to their socio-demographic variables and religious beliefs. Parents who reported religion as being important were more in favor of the HPV vaccination compared to parents who reported religion as being of less importance. Four out of ten of the included mothers had never undergone a cervical cancer screening but had accepted previous childhood vaccinations for their daughters. Acceptance of the vaccine was high, and we believe our results are promising for future implementation of the HPV vaccination in the national childhood vaccination program in Thailand.