VALIDATION OF THE FILIPINO TRANSLATED QUESTIONNAIRE ON PARENT ATTITUDES ABOUT CHILDHOOD VACCINES

,


INTRODUCTION
Childhood vaccination is regarded as one of public health's groundbreaking accomplishments.
Development of immunization policies have significantly decreased child morbidities and deaths related to certain diseases globally. The success of these programs relies heavily on vaccine compliance in lessening vaccinepreventable diseases (VPD). Immunized children benefit directly from vaccinations and significant community vaccination coverage rates have added protection via herd immunity. 1 Despite the proven efficacy of vaccination and its acceptance worldwide, a growing proportion of parents have refused vaccinating their children for different reasons. 2 Diminished trust in vaccination has led to outbreaks in diseases. This has put hindrances towards global elimination of diseases such as polio which have sparked political discussions in different nations worldwide. 3 The World Health Organization (WHO) defined 'vaccine hesitancy' as the delay in acceptance or refusal of vaccination despite availability of vaccination services. 4 Given the rising global issue of vaccine hesitancy, the WHO launched the Strategic Advisory Group of Experts (SAGE) Working Group on Vaccine Hesitancy. They were tasked to approach this problem and deliver evidence-based analyses and solutions. 4 The SAGE Working Group saw the necessity to outline the reasons for vaccine hesitancy. Development of this matrix of determinants involved extensive review of literature. 5  incorporating results from previous studies in order to add to the item pool. Screening and content validation were conducted by a group of experts in the field of immunization, and pre-testing the validated tool on a group of parents. 6 A prospective cohort study on 437 parents of children under an integrated health care system based in Seattle showed that scores on the PACV predict childhood immunization status and have high reliability. It was recommended that results be validated in different geographic and demographic samples of parents. 7 The Department of Health (DOH) of the Republic of the Philippines was alarmed that a significant number of parents refused to avail of the government's various vaccination programs following the issue regarding the newly introduced dengue vaccine. 8 Several studies related to vaccine hesitancy, its determinants, and its impact in the different regions of the world are already available but there is a dearth of investigations done in the Philippines. With this in mind, efforts have been made to address the growing hesitancy and refusal of parents for vaccine administration to their child. To increase the possibility of success in this endeavor, the development of a means of measuring vaccine hesitancy in certain population groups is necessary to determine reasons contributing to it. This study was undertaken to determine the content validity, face validity and testretest reliability of a Filipino Translated Questionnaire on Parent Attitudes About Childhood Vaccines.

MATERIALS AND METHODS Description of the Study Setting
This is a tool validation study. Convenience sampling was done. Parents/legally authorized representatives of patients 15 months to 6 years seen at the Outpatient Department or admitted at the Pediatric Wards and Emergency Room were enrolled in this study. The hospital has a 1,500-bed capacity with 200 patients admitted at the UP-PGH Department of Pediatrics. Each month, there are at least 600 new pediatric admissions and 1000 outpatient consults.

Participants
Tagalog-speaking parents of Filipino children aged 15 months to 6 years who consulted at the Philippine General Hospital Pediatrics Outpatient Department Section, Pediatrics Emergency Room and Pediatrics Wards 9 and 11 from July 2019 to August 2019 were eligible to participate in the study. The minimum age was set at 15 months as it is expected that a child would have been given all the vaccine doses prescribed in the National Immunization Program by that age if the parents are fully compliant. A maximum age of 6 years was set to account for catchup immunization. Parents of patients requiring resuscitation, in cardiorespiratory distress, or were immunocompromised, were excluded from this study.
Potential respondents were identified through the census in the pediatric wards, emergency room and outpatient department. Available respondents were recruited into the study by the principal investigator. The age and clinical status of the child were identified during the recruitment of respondents. Among eligible subjects, only those who gave consent were included in this study.

Tool Validation and Data Collection
Approval from the main author of the PACV survey was sought before study initiation. The English questionnaire was translated into Filipino by a linguist from Sentro ng Wikang Filipino-Manila. This office located in the University of the Philippines in Manila, undertakes translation, writing and publication of materials into Filipino. The translated questionnaire underwent assessment for content validity by 4 experts in the field of vaccination. Each of the non-demographic questions (15 items) were assessed using a 4-point ordinal scale: 1 = not relevant, 2 = somewhat relevant, 3 = quite relevant, 4 = highly relevant. Items with low content validity index (CVI) were revised based on comments or suggestions of the 4 experts. CVI was computed again for the revised items in the translated PACV survey.
The revised translated PACV survey were then administered to ten parents/ legally authorized representatives to assess the face validity of each question based on clarity and simplicity. Agreement was assessed using yes or no responses and the level of agreement was computed.
The final translated PACV survey was administered using a test-retest design to a sample of parents/ legally authorized representatives who consented for the study. The initial assessment was done right after the informed consent process. The second assessment was done during the next patient follow-up which was at least 2 weeks from initial assessment. Privacy was ensured for all respondents while answering the questionnaire and they were provided 10 to 15 minutes to answer all items. The responses of participants during the re-test assessment were used to determine the acceptability and internal consistency reliability of the translated PACV survey.

Sample Size
The minimum sample size needed was computed using an online software by Arifin (2018). 9

STATISTICAL ANALYSIS
The respondents' characteristics were described using frequencies and proportion.
Content validity was determined using the item content validity index (I-CVI), based on the expert panel's rating on item relevance. A CVI greater than 0.78 per item was considered as acceptable. 11 Face validity was determined by computing for the level of agreement among the ten respondents based on clarity and simplicity. The level of agreement was computed by the proportion of "yes" response per item. Items with at least 75% level of agreement were considered acceptable. 12 Test-retest reliability was determined by computing for the Intraclass correlation coefficient. A cut-off value of 0.70 was considered evidence of acceptable reliability. 13 Acceptability was determined by obtaining the percentage of no response per item. Finally, Cronbach's alpha was computed for the 10 items using 5-point likert scales to determine internal consistency. A value of at least 0.70 was considered acceptable. 14 All analyses was done using Microsoft Excel and SPSS v23.

Population
A focus group of four experts in the field of vaccination rated the content of each item based on its relevance. Two of the experts are physicians who work in the field of public health while the other two are pediatricians, one of which is an Infectious Disease Specialist. Seventyseven participants from the Pediatric OPD, Pediatric Emergency Room and Pediatric Wards who fulfilled the inclusion criteria were recruited. Ten of the participants rated each item on the level of agreement based on its simplicity and clarity. The sixty-seven participants then answered the questionnaire for its test-retest reliability. Majority (70.12%) of participants were females. The mean age of participants was 34.5 years with 63% of participants aged 30 years and above. Thirty five percent of participants were married, while 66% were in a common law relationship. The highest educational status attained by majority of participants is secondary education at 66%. (See Table 1). Table 2 shows that majority (3-4) gave a high content validity rating on all questions. The content validity index for each question is high at 0.75 to 1.0 and the overall CVI is 0.95 which supports the content validity of the questionnaire.

Content and Face Validity
All questions had 100% rating in terms of clarity and simplicity which supports the face validity of the tool.

Test-retest reliability
The questionnaire had a high testretest Pearson's correlation (0.970) which supports the reliability of the tool. Cronbach's alpha is low at 0.687. Item number 5 can be removed to increase alpha to an acceptable rating of 0.711.
The fifteen non-demographic items with the corresponding responses are listed in Table 3. These 15 items were translated from the original PACV questionnaire which were developed from three domains.

DISCUSSION
Vaccine hesitancy has context specific determinants which span different regions in the world. These sentiments have been extensively studied in other countries as mentioned by Larson et al. 3 Tools have been created as a result of decreasing vaccination rates to identify such determinants. Tool development has since then helped understand the generally positive response of some regions toward vaccination.
In the Philippines there is a looming need to address vaccine hesitancy from a public health perspective given the recent drops in vaccination rates. 15 Despite the vast majority of studies done in this field, there are no tools developed to measure vaccine hesitancy in the Philippines. Translated from an accepted and validated PACV survey, this tool was validated with the goal of developing a means of assessing vaccine hesitancy in the Philippine setting.
In this study, the Filipino Translated Questionnaire on Parent Attitudes About Childhood Vaccines showed high content validity among the 4 experts with overall agreement in 12 of the 15 nondemographic items in the tool. The remaining three items were reworded to improve readability among laypeople. Upon administration to the first ten participants, there was a unanimous agreement among parents that the survey items were clear and simple to understand (N=10). This was then administered to the 67 participants which showed that the translated PACV has a high test-retest reliability as supported by a high intraclass correlation coefficient of 0.970.
The internal consistency however had a low Cronbach's α coefficient with an overall coefficient of 0.687. On review of responses and feedback of patients, some of the respondents' answers would have incongruences under similar domains. The original validation study by Opel et. al. had three domains identified in its questionnaire. 2 These were Immunization Behavior, Vaccine Safety and Efficacy, and Trust. In this study, the internal consistency for individual domains were low ranging from 0.468-0.632. On data analysis, removal of item number 5 "Naniniwala ako na ang mga sakit na naaagapan ng bakuna ay malulubhang sakit." would increase the overall coefficient to 0.711 (See Table 3).

LIMITATIONS
The team recognizes that this study has limitations. The test-retest phase of tool development was completed after the measles epidemic. The timing of the survey may have affected parents' responses to the questionnaire. In addition, all participants were derived from one institution in the National Capital Region limiting the generalizability of results. The sample population were obtained from convenience sampling. This questionnaire was translated to Filipino and will still bring about different contextual meanings. Although this test has been administered globally, the team understands the need for tools to be available in the native tongue. The study team attempted to preserve each item's readability and inherent meaning by having it officially translated by the Sentro ng Wikang Filipino as well as having it screened by both an expert panel and the target study population.

CONCLUSION
The translated Filipino PACV is a useful 21-item tool to identify possible reasons for vaccine hesitancy among Filipino parents. The remaining 14 nondemographic items on the Filipino Translated Questionnaire on PACV have face and content validity with an acceptable internal consistency.

RECOMMENDATIONS
Future studies can be done geared towards improving internal consistency in the questionnaire by adjusting item phrasing to improve readability. Expanding the study sites to include communities outside a tertiary hospital setting will also bring about a broader study population. The next phase of this study should test the tool's predictive and content validity. Focus should be placed on further psychometric evaluation to measure association of sociodemographic features and vaccine hesitancy. This is already being done globally in the original PACV survey. This study can serve as a framework for future studies in correlating behaviors affecting vaccine hesitancy. Original tool development should be contemplated by future researchers to uncover domains and behaviors towards vaccine hesitancy which are more appropriate to the Filipino context.