The Psychometric Properties of a Romanian Version of the Vaccination Attitudes Examination (VAX) Scale

Objective: To validate a Romanian version of the Vaccination Attitudes Examination (VAX) Scale and determine its psychometric properties in Romanian people. Methods: A cross-sectional survey was conducted in different social categories. The sample consists of 124 (30.6%) males, 259 (64.0%) females, 22 (5.4%) not specified gender with ages ranging from 16 to 62 years. Data were collected on the Romanian version of the Vaccination Attitudes Examination (VAX) Scale, the Attitudes Toward Vaccination Questionnaire, the General Health Perception Scale and the Perceived Sensitivity to Medicines Scale. Results: Scale's reliability analysis revealed an overall Cronbach’s alpha of a 0.82 showing good internal consistency. The Spearman correlation between the VAX scale's total score and the Attitudes Toward Vaccination Questionnaire, indicate good levels of convergent validity. Confirmatory Analysis revealed a good fit. Conclusion: The validated Romanian version of the Vaccination Attitudes Examination (VAX) Scale is a valid and reliable measure in detecting the vaccine-hesitant Romanian people.


Introduction
Attitude towards immunization vaccines is a widely debated issue among the population. Medical, philosophical and religious considerations polarize human attitudes and behaviours, generating debates and adverse stances. Knowledge of attitudes toward vaccines can contribute to specific health and safety interventions on the population. In this study we propose a measure of attitudes to vaccines appropriate to the all social categories of general population. Many instruments address the attitudes toward different types of vaccines. In 2006 Brabin et al. [2] address the parental attitudes toward future acceptance of adolescent human papillomavirus vacination. In 2009 Kang et al. [8] developed the HPV Attitude Questionnaire that address the attitudes toward and intention to receive the human papilloma virus vaccination of female South Korean college students. Developed by McRee et al. [11] in 2010, the Carolina HPV Immunization Attitudes and Beliefs Scale (CHIAS) assesed parental attitudes and beliefs toward daughters' HPV vaccination. Opel et al. (2011) have developed the Parent Attitudes about Childhood Vaccines, instrument that identifies vaccine-hesitant parents [13]. In 2012, Chow et al. have developed a questionnaire that describes Australian parent's, attitudes, behaviours and concerns about vaccination. The respondents were parents of children age < 18. A questionnaire titled HPV and HPV vaccine-related knowledge, attitudes, and behaviours was adapted to Thai. The study was conducted in 2015 in Thailand with a sample of young women between 18 and 26 years [15]. An evaluation of attitude toward vaccines among healthcare workers was made in Italy in 2016-2017 [4]. Many of these measures are specific to the certain age groups and address parental attitudes. Other scales address only certain social categories. Developed by Martin et al. in 2017 [10], the Vaccination Attitudes Examination (VAX) Scale is a tool that asses general attitudes toward vaccines. The scale construction strategy was based on the identification of the anti-vaccination attitudes that predict vaccination behavior. Four distinct vaccine attitudes were identified: (1) mistrust of vaccine benefit, (2) worries over unforeseen future effects, (3) concerns about commercial profiteering and (4) preference for natural immunity. All four factors cover an extensive area of anti-vaccination attitudes and give the VAX Attitudes Examination (VAX) Scale scale a high and efficient degree for identifying those with vaccination resistance. The VAX scale is appropriate for identifying the attitudes towards vaccines in general and we consider that VAX scale may be a necessary and efficient instrument on Romanian people. The aim of the study was to adapt the Vaccination Attitudes Examination (VAX) Scale, determine its reliability and validity, and to verify the adequacy of the adapted version of the four-dimensional scale.

Sample and Design
A total of 405 participants (124 males, 259 females, 22 gender not specified) with ages ranging from 16 to 62 years (mean=24.99, SD=11.80), were recruted from general people of various social categories of different regions of Romania. Sampling was based on convenience. Sociodemographic data included age, gender, residential environment, educational level, marital status, and religious confession.

Measures
The instruments were translated into Romanian using the forward-backward translation design and following the guidelines given by the literature [1,6].

Attitudes Toward Vaccination Questionnaire (AVC)
Attitudes Toward Vaccination Questionnaire developed in 2002 by Busse et al. consists of 11 items. [3] The measurement is done by three levels of Likert scale (0disagree; 1-unsure; 2-agree). The total score ranging from 0 (most negative attitude toward vaccination) to 22 (most positive attitude toward vaccination). For give a generality of the scale, I replaced the item: ˮThe risk of pertussis vaccine outweighs its usefulness in preventing the disease (whooping cough)ˮ with the item: ˮThe risks of vaccines outweighs their usefulness in preventing infectious diseasesˮ. For the present sample, the Cronbach alpha for the Attitudes Toward Vaccination Questionnaire was 0.79 (mean=13.90; SD=4.46).

The General Health Perceptions Scale (PGS)
The general health perception (PGS) was measured with a single item by the General Health Perceptions Scale developed by Ware et al. in 1992 [16]: "In general, would you say your health is." Respondents indicated their agreement of the item, on a 5-point scale (1-poor, 2-fair, 3good, 4-very good, 5-excellent). For the present sample, respondents indicated low score (mean=3.92; SD=0.68).

The Perceived Sensitivity to Medicines Scale (SPM)
The perceived sensitivity to medicines was measured with a single item of the Perceived Sensitivity to Medicines Scale developed by Horne et al. in 2013 [7]: ˮMy body is very sensitive to medicinesˮ on a six levels of Likert scale (1strongly disagree; 6-strongly agree). Martin et al. [10] assume that higher scores of perceived sensitivity to medicines correlate with higher symptom reporting. In our study the respondents indicated low score (mean=2.79; SD=1.27)

Procedure
For some scales only single items have been filled. The scales were completed by participants after they were ensured that their participation in the study was anonymous and confidential. Every participant was explained the purpose of the research. The data were completed on a voluntary basis and informed consent. The respondents completed the scales individually without receiving any compensation.

Statistical Analysis
We conduct preliminary analyses to examine the descriptive statistics and the association of all analyzed variables in the study. We used exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) with SPSS v. 20 in order to examine the matrix structure and the fit of the VAX scale. We chose (a) absolute match measures (CMIN/DF) that determine the degree to which the model predicts the observed correlation matrix and whose value is recommended to be below 5 and the RMSEA that indicates approximate fits of the pattern in population and (b) incremental measures (TLI, CFI) that compare the proposed model to a baseline model that all other models should overtake and which indicates the discrepancy between the two models [12]. A scale has a good reliability if on different occasions, under different conditions and administered by different people the measurements are repeatable [5]. Ranging between 0 and 1, an internal consistency index of over 0.7 ensures a good reliability to the scale. Validity refers to the quality of an instrument to measure what it has intended to measure [9]. Convergent validity was examined with Spearman correlation calculations between the scores of the VAX and the scores of Attitudes Toward Vaccination Questionnaire.

Sample Characteristics
In Table 1, we present the socio-demographic characteristics of the participants in the present study.

Preliminary Analysis
Kruskal-Wallis test significantly indicated statistical differences by the age categories and by the religious confessions (see Table 2). People with age 16-30 year old indicate the higher score to VAX scale toward people with age 31-45 and 46-62 year old. We did not find significant statistical differences between group 16-30 and 46-62 year old and between group 31-45 and 46-62 year old. We found significant statistical differences between Orthodox and Pentecostals and between Orthodox and Seventh-day Adventists. We did not find significant statistical differences between Pentecostals and Seventh-day Adventists (see Table 3). Note: VAX_T-total score to VAX scale; ** p<.01.

Reliability and Validity
The scale's reliability analysis revealed an overall Cronbach's alpha of 0.82. The corrected item to total correlation, was greater than 0.3 except item 4 (Although most vaccines appear to be safe, there may be problems that we have not yet discovered). For theoretical relevance, we have decided to keep item 4 on a scale, because the removal did not lead to a significant increase in reliability (see Table 4). VAX12 Fiind expus (ă) la boli în mod natural este mai sigur pentru sistemul imunitar decât expunerea prin vaccinare (Being exposed to diseases naturally is safer for the immune system than being exposed through vaccination) 0.406 0.807 Total Alpha=0.82. Note: VAX_T-total score to VAX scale; MVB-total score to Mistrust of vaccine benefit; WFE-total score to Worries over unforeseen future effects; CCP-total score to Concerns about commercial profiteering; PNI-total score to Preference for natural immunity; AVC_T-total score to Attitudes Toward Vaccination Questionnaire; SPM-item: "My body is very sensitive to medicines"; PGS-item: "In general, would you say your health is"; ** p <.01; * p <.05.
Convergent validity was supported by a significant correlation (-.62**) between the VAX total score and the Attitudes Toward Vaccination Questionnaire total score (see Table 5). The high score of VAX scale strongly correlates with the low score of the Attitudes Toward Vaccination Questionnaire (AVC_T). In our study the respondents indicated a higher score of VAX scale (mean=3.64; median=3.58) that reflects the anti-vaccination attitudes [10] and a lower score of AVC (mean=13.90; median=14.00) that reflects the negative attitude toward vaccination [3]. The VAX total score poorly correlates with SPM and PGS.
Discriminative validity of VAX scale was tested by the method of known groups. Based on the answer of the two items of Attitudes Toward Vaccination Questionnaire: "Vaccines should never be given to infants under 1 year of age" and "You are in favor of vaccination in general?" we divided the respondents into two groups. Those that were in favor of the vaccination in general and those that agreed the vaccination to infants under 1 year of age, have indicated a significant low score to VAX scale toward those who indicated the opposite answers. The results are depicted in Table 6.

Exploratory Analysis
For testing the factorial structure of VAX scale, we conducted an exploratory factor analysis (EFA) with SPSS 20.0. Table 7 and Figure 1 depict the results of the analysis. (Natural exposure to viruses and germs gives the safest protection). .844 VAX12 Fiind expus (ă) la boli în mod natural este mai sigur pentru sistemul imunitar decât expunerea prin vaccinare. (Being exposed to diseases naturally is safer for the immune system than being exposed through vaccination).
.871 The percentage of cumulative variance explained by the factors extracted is 71.22%.

Discussion
The aim of the present study was to test the psychometric properties of the Romanian version of the Vaccination Attitudes Examination (VAX) Scale in a sample of different social categories on Romanian people. This scale has not been used in Romania before. In their study, Martin et al. [10] found a good reliability for the VAX scale 0.84 (0.86-0.92 for subscales). Our study showed good Cronbach's alpha of 0.82 (0.71-0.86 for subscales). Except item 4 (Although most vaccines appear to be safe, there may be problems that we have not syet discovered) all items had correlation greater than 0.3 with the sum of the other items together. We consider that the small correlation is due to the induced suggestion of the item that assumes that there may be problems although most vaccines appear to be safe, which led to incorrect answers from many respondents. This assumption is in agreement with [5] that misinterpretation of scale instructions and the existence of a small number of items, can generate answers that lead to a greater degree of error.
The Romanian version of the VAX scale maintains the four factors of the original version. The Romanian version of the VAX scale is related to age categories and religious confessions. The age group 16-30 year old, Pentecostals and Seventh-day Adventists reported a higher score of the VAX scale, but also a higher level of general perception of health.
The Romanian version of the VAX scale is unrelated with gender, marital status, residential environment and level of education. These findings suggest intervention for the increase of acceptability of the young people, the certain religious confessions, and the groups with any level of education. The score of the VAX scale poorly correlates with perceived sensitivity to medicine and the general perception of health. The Romanian version of VAX scale is related with vaccination behavior and the intentions to receive vaccines for one'children. These findings are in agreement with Martin's results [10]. A significant correlation between the Romanian version of VAX total score and the Attitudes Toward Vaccination Questionnaire has indicated a high convergent validity. The method of known groups revealed a good discriminative validity. The confirmatory analysis showed a good fit of the Romanian version of VAX scale. The present study has several limitations. A first limitation refers to the convenience sample that does not use the random sampling. This study was conducted with a relative sample (n=405). Some social categories were poorly represented and the respondents were recruited only from four areas of Romania. Further research with the Romanian version of VAX scale is needed to identify and other factors that are associated with vaccination behavior (eg. information on vaccine efficacy, type of source of information).

Conclusions
The Romanian version of VAX scale showed a high reliability and a good discriminative and convergent validity. The good fit of the scale recommend its use in detecting the vaccine-hesitant Romanian people. The positive correlation between VAX score and the perceived sensitivity to medicines (SPM) is in agreement with Martin'results and it represent an indicator of personal concern about how one body might react to substances recommended by health care providers [10]. Also the positive correlation between VAX score and the general health perception (PGS) is a subjective indicator that shows that people who perceive themselves as healthier have a stronger anti-vaccination attitude. These indicators allows decision makers to develop programs to increase the vaccines acceptability. The high psychometric qualities of the Romanian version of VAX Scale recommend it for both screening and research, becoming a useful tool for both health care providers and researchers. To confirm the high psychometric qualities of the Romanian version of VAX scale, future studies should include and homogenious populations (eg. only adolescents, only adults).

Funding
This research received no external funding.

Conflicts of Interests
The author declares no conflict of interest.