Intention to Receive the TAK-003 Dengue Vaccine and Associated Factors Among Adults in Rural Northern Thailand

Round 1

Reviewer 1 Report (Previous Reviewer 5)

Comments and Suggestions for Authors

The authors revised the manuscript according to suggestions

Author Response

Comment 1: The authors revised the manuscript according to suggestions

Response: Thank you very much for all suggestions to better this manuscript.

 

Reviewer 2 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

The authors have worked hard to revise the manuscript, but several points still need to be corrected.

-In the Methods section, it is important for the authors to explicitly clarify that convenience sampling was used, as this currently appears only in the Discussion.

-Regarding knowledge, the authors refer to “adequate knowledge” in the Conclusions, yet it is not clear what they mean by this. In the Methods, they state only that “higher scores indicate greater knowledge,” but this is insufficient, and clarification is essential for the manuscript. The authors need to establish clear cutoffs for what constitutes “adequate knowledge” or “greater knowledge,” unless they choose instead to present knowledge strictly as the proportion of correct and incorrect responses, as shown in Table 2. As the manuscript is currently written, it does not allow the reader to determine whether participants actually had a defined level of knowledge. In addition, Table 2 presents a descriptive analysis of categorical variables, whereas Table 5 uses binary logistic regression with “total knowledge scores.” It is therefore entirely unclear what the authors mean by this variable, and it appears that knowledge was treated numerically rather than categorically at this stage of the analysis. The same concern applies to the “total attitude scores” variable. The authors must clarify this issue, as the current presentation is conceptually inconsistent and requires correction.

-Moreover, the Discussion remains overly ambitious. The authors continue to use causal language that should be revised. Although the Limitations section has improved, the authors should also explicitly address the absence of a clearly defined health equity framework, despite the fact that this concept is invoked throughout the manuscript.

Author Response

Reviewer 2

The authors have worked hard to revise the manuscript, but several points still need to be corrected.

Comment 2.1:

In the Methods section, it is important for the authors to explicitly clarify that convenience sampling was used, as this currently appears only in the Discussion.

Response:

Thank you for pointing this out. We have revised Section 2.2 (Participants and Recruitment) as suggested.

“A convenience sampling approach was used” (Line 113-114, page 3)

 

Comment 2.2:

Regarding knowledge, the authors refer to “adequate knowledge” in the Conclusions, yet it is not clear what they mean by this. In the Methods, they state only that “higher scores indicate greater knowledge,” but this is insufficient, and clarification is essential for the manuscript. The authors need to establish clear cutoffs for what constitutes “adequate knowledge” or “greater knowledge,” unless they choose instead to present knowledge strictly as the proportion of correct and incorrect responses, as shown in Table 2. As the manuscript is currently written, it does not allow the reader to determine whether participants actually had a defined level of knowledge. In addition, Table 2 presents a descriptive analysis of categorical variables, whereas Table 5 uses binary logistic regression with “total knowledge scores.” It is therefore entirely unclear what the authors mean by this variable, and it appears that knowledge was treated numerically rather than categorically at this stage of the analysis. The same concern applies to the “total attitude scores” variable. The authors must clarify this issue, as the current presentation is conceptually inconsistent and requires correction.

Response:

Thank you for this important and insightful comment. We agree that the terms “adequate knowledge” and “favorable attitude” were not clearly defined in the original manuscript. To address this, we have removed these ambiguous terms from the Conclusions and revised the text for clarity.

 

“This study highlights that approximately two-thirds of adults in rural northern Thailand expressed an intention to receive the TAK-003 dengue vaccine. Around 61.2% were aware that a dengue vaccine is available in Thailand. Prior influenza vaccination and positive attitudes toward dengue vaccination were associated with higher intention to vaccinate with dengue vaccine, whereas a history of vaccine refusal was negatively associated” (Line 360-365, page 10)

In the Methods section, we have now clearly specified the scoring system and defined cut-off points for knowledge and attitude levels. This threshold was derived based on the mean score of the study population, which was comparable to the median, justifying the use of the mean as a cut-off for categorizing knowledge and attitude levels.

We also clarified that knowledge and attitude scores were treated as continuous variables in the regression analysis (Table 5), while Tables 2 and 3 present descriptive summaries of individual items to provide an overall picture of participants’ knowledge of dengue viral infection and attitudes toward the new dengue vaccine.

“2. Knowledge about dengue disease, including disease transmission, symptoms, treatment, and vaccine availability for ten items. Each of the 10 knowledge items was scored as 1 for a correct response and 0 for an incorrect response, resulting in a total possible score ranging from 0 to 10, with scores of greater than 7 indicating good knowledge.”

“3. Attitudes toward dengue disease and the dengue vaccine: Participants' attitudes toward dengue fever and the dengue vaccine were assessed using a five-level Likert scale, ranging from the lowest degree (1) to the highest degree (5) for ten items. Each of the 10 attitude items was measured on a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree), yielding a total score ranging from 10 to 50, with scores greater than 38 indicating positive attitudes toward dengue viral infection and vaccination.” (Line 130-140, page 3)

 

Comment 2.3:

Moreover, the Discussion remains overly ambitious. The authors continue to use causal language that should be revised. Although the Limitations section has improved, the authors should also explicitly address the absence of a clearly defined health equity framework, despite the fact that this concept is invoked throughout the manuscript.

Response:

Thank you for your valuable suggestion. We have carefully revised paragraph 2 to 4 in the Discussion to remove causal language and replace it with more appropriate terms to ensure more cautious and accurate interpretation of the findings. (Line 261-328, page 8-9) We also acknowledge that a comprehensive health equity framework (e.g., ETR’s health equity framework) was not applied in this study; instead, only selected aspects of health equity were discussed in relation to the findings. To avoid overstating the scope, we have removed the “health equity perspective” from the original title, which has been revised to “Intention to Receive the TAK-003 Dengue Vaccine and Associated Factors Among Adults in Rural Northern Thailand.” In addition, we have revised the Limitations section accordingly and incorporated the following statement on future research.

“A health equity framework may be applied to better understand the complex associations of health inequities and to inform targeted strategies and interventions that address these factors.” (Line 346-348, page 10)

 

Comment 2.4:

Reviewer Figures and tables can be improved

Response: Thank you for highlighting this point. We have revised Table 1 to include the number of participants with good knowledge and positive attitudes toward dengue viral infection and vaccination (Table1, page 4-5)

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

The paper is acceptable for publication in its present form. 

 

This paper provides useful information on vaccination for Dengue virus (fever) in a "local" geographic area.  It will be useful for other when conducting studies in other regions. 

Author Response

Comment1: The paper is acceptable for publication in its present form

Response: Thank you very much for all suggestions to better this manuscript. 

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Administering questionnaires to 482 people is not easy. After carefully reading your paper, I have some suggestions and comments.

Lines 94-95: The study is conducted in a specific context, so the results can only support local health initiatives and cannot be extrapolated to different sociodemographic/cultural contexts.

Lines 111-113: Since the questionnaire is self-administered, participants must be able to read and write. What kind of support did the research team provide to participants who could not read/write? Administering a questionnaire via interview is not the same as a self-administered questionnaire. The methodology should be the same for all participants. Was a pre-test of the questionnaire conducted to verify/evaluate difficulties in understanding/comprehension/completion of the questionnaire?

Line 117: Random sampling was not used, but it is unclear how or where the sample of study participants was obtained. Writing (lines 105-106) “distributed in collaboration with sub-district health promotion hospitals” does not clarify this aspect. As far as I have discovered in my research, the rural population of the Chiang Mai region lives in hundreds of small towns and villages, with a wide diversity.

Line 124: The study was conducted from September 2024 to July 2025 and asked participants if they had been vaccinated against influenza in the previous season. Does the data collection period encompass two different vaccination seasons? If so, could there have been different health education/promotion campaigns to promote vaccination during those different vaccination seasons? How might this aspect affect the study results?

Line 149: The term "using" is repeated.

Caption for Table 1, Table 2, Table 3: These are not characteristics/knowledge/attitudes of adults residing in rural areas of Chiang Mai, but rather sociodemographic characteristics/knowledge/attitudes of the study participants (the study did not use a random sample representative of the adult population in rural areas of Chiang Mai).

Table 1: How were the knowledge and attitude scores calculated? I don't think this is indicated in the methodology section.

Table 1 and Table 3: How do you relate vaccine acceptance to attitudes towards the vaccine? Is the variable "vaccine acceptance" a composite variable, or did the question about acceptance have a dichotomous "yes" or "no" response?

Table 3: From my point of view, the phrases used refer more to the area of ​​knowledge and do not allow for the characterization of attitudes.

Line 228: When the authors write "with higher vaccine acceptance," they are referring to a degree of acceptance that has not been measured/characterized, since the answer to the question is dichotomous: "acceptance" or "non-acceptance."

Line 244: The sample is not representative, so you cannot extrapolate to the adult population living in rural Chiang Mai.

Lines 258-259: the authors cannot write "particularly among rural and older populations" because they did not compare rural populations with urban populations, nor the elderly with young people or adults.

Line 280: The authors asked about the sources of information, but did not ask if the study participants trusted the healthcare professionals. Indicating a source of information is not synonymous with trusting that source.

Line 336: I believe that instead of saying they are adults from northern Thailand, you should say they are adults from rural northern Thailand.

Author Response

Reviewer 1

Administering questionnaires to 482 people is not easy. After carefully reading your paper, I have some suggestions and comments.

Comment 1.1:

Lines 94-95: The study is conducted in a specific context, so the results can only support local health initiatives and cannot be extrapolated to different sociodemographic/cultural contexts.

Response:

Thank you for pointing this out. We have revised the manuscript per suggestion.

“Moreover, the findings can support local health initiatives to improve vaccine confidence, promote equitable access to immunization, and strengthen dengue prevention interventions, particularly for populations in resource-limited setting.” (Line 95-97, page2)

Comment 1.2:

Lines 111-113: Since the questionnaire is self-administered, participants must be able to read and write. What kind of support did the research team provide to participants who could not read/write? Administering a questionnaire via interview is not the same as a self-administered questionnaire. The methodology should be the same for all participants. Was a pre-test of the questionnaire conducted to verify/evaluate difficulties in understanding/comprehension/ completion of the questionnaire?

Response:

Thank you for pointing this out. The questionnaire was developed based on a review of relevant literature and existing instruments. It was reviewed by field experts to ensure content validity and clarity, and included items assessing knowledge of disease transmission, severity, management, and prevention. During data collection, trained research staff were available to provide clarification when needed to ensure participants’ understanding of the question. However, research staff mainly provided clarification when needed, particularly for terms such as family structure (e.g., single or extended family) and immunocompromised conditions (e.g., HIV infection or use of immunomodulatory drugs), to help participants better understand the questions. Assistance was also available to address any questions after participants completed the questionnaire. We have revised the Participants and recruitment in Methods section for better clarification as suggested.  

“Data were completed either on paper or on tablets provided by the research team, with staff assistance available when needed, for example to clarify terms or questions.” (Line 113-115, page 3)

Comment 1.3:

Line 117: Random sampling was not used, but it is unclear how or where the sample of study participants was obtained. Writing (lines 105-106) “distributed in collaboration with sub-district health promotion hospitals” does not clarify this aspect. As far as I have discovered in my research, the rural population of the Chiang Mai region lives in hundreds of small towns and villages, with a wide diversity.

Response:

Thank you for pointing this out. We have revised setting and study design and participants and recruitment in Methods section to clarify this aspect per suggestion.

“Data were collected using a structured questionnaire distributed in collaboration with subdistrict health promotion hospitals in Chiang Mai and the village health volunteers, who facilitated community outreach and household visits.” (Line 106-108, page 3)

The research team promoted the study through partner subdistrict health promotion hospitals in Chiang Mai. Trained research staff, with support from village health volunteers, conducted community visits in rural areas and personally invited eligible   residents who were at home during weekdays to participate. Data were completed either on paper or on tablets provided by the research team, with staff assistance available when needed, for example to clarify terms or questions. (Line 110-115, page 3)

Comment 1.4:

Line 124: The study was conducted from September 2024 to July 2025 and asked participants if they had been vaccinated against influenza in the previous season. Does the data collection period encompass two different vaccination seasons? If so, could there have been different health education/promotion campaigns to promote vaccination during those different vaccination seasons? How might this aspect affect the study results?

Response:
Thank you for this insightful comment. We agree that participants may have been surveyed across two different influenza vaccination campaign periods. However, this is unlikely to substantially affect the study results because influenza vaccination campaigns in Thailand are typically conducted before the rainy season (May–September). If influenza and dengue vaccination campaigns were to be promoted in Thailand, both vaccines could potentially be administered during the same period, as the two diseases share a similar transmission season in the country.

We also revised 1^st paragraph of Results section and Table1 to clarify about influenza vaccination from “History of flu vaccination” to “History of flu vaccination in the past year” (Line 187 and Table1)

 

Comment 1.5:

Line 149: The term "using" is repeated.

Response:

Thank you for pointing this out. We have removed the repeated term.

 

Comment 1.6:

Caption for Table 1, Table 2, Table 3: These are not characteristics/knowledge/attitudes of adults residing in rural areas of Chiang Mai, but rather sociodemographic characteristics/knowledge/attitudes of the study participants (the study did not use a random sample representative of the adult population in rural areas of Chiang Mai).

Response:

Thank you for this important comment. We appreciate your suggestion and  have revised the caption for Table1 to Table3 accordingly.  

“Table 1. Sociodemographic Characteristics of the Study Participants Living in Rural Areas of Chiang Mai, Thailand.” (Line 190-191, page 4)

“Table 2. Knowledge About Dengue and Vaccination Among the Study Participants Living in Rural Areas of Chiang Mai, Thailand.” (Line 205-206, page6)

“Table 3. Perceptions and Attitudes Toward Dengue and Vaccination Among the Study Participants Living in Rural Areas of Chiang Mai, Thailand.” (Line 220-221, page 6)

 

Comment 1.7:

Table 1: How were the knowledge and attitude scores calculated? I don't think this is indicated in the methodology section.

Response:

 Thank you for highlighting this important point. We have now clarified the scoring procedures in the Methods section (Questionnaire subsection) to ensure methodological clarity and reproducibility per suggestion.

“Knowledge about dengue disease, including disease transmission, symptoms, treatment, and vaccine availability for ten items. Each of the 10 knowledge items was scored as 1 for a correct response and 0 for an incorrect response, resulting in a total possible score ranging from 0 to 10, with higher scores indicating greater knowledge” (Line 129-132, page 3)

“Attitudes toward dengue disease and the dengue vaccine: Participants' attitudes toward dengue fever and the dengue vaccine were assessed using a five-level Likert scale, ranging from the lowest degree (1) to the highest degree (5) for ten items. Each of the 10 attitude items was measured on a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree), yielding a total score ranging from 10 to 50, with higher scores reflecting more positive attitudes toward dengue vaccination.” (Line 133-138, page 3)

 

Comment 1.8:

Table 1 and Table 3: How do you relate vaccine acceptance to attitudes towards the vaccine? Is the variable "vaccine acceptance" a composite variable, or did the question about acceptance have a dichotomous "yes" or "no" response?

Response:

Thank you for this important comment. Vaccine acceptance was measured using a single dichotomous question asking whether participants would accept the new dengue vaccine (yes/no); it was not a composite variable.

In this study, attitudes toward vaccination were measured using a set of Likert-scale items (5-point scale), and the total attitude score (range: 10–50) was calculated and presented in Table 1. In contrast, Table 3 provides a descriptive presentation of participants’ individual perceptions and attitudes toward dengue and vaccination across all respondents.

Relationship between vaccine acceptance and attitudes toward the vaccine was explored by multivariable analysis reported in Table5

 

Comment 1.9

 Table 3: From my point of view, the phrases used refer more to the area of ​​knowledge and do not allow for the characterization of attitudes.

Response:

Thank you for this important comment. We appreciated your suggestion. To address this, we have revised the statement in Table3 as well as Table3 caption to better reflect the perception and attitude.

 

Table 3. Perceptions and Attitudes Toward Dengue and Vaccination Among the Study Participants Living in Rural Areas of Chiang Mai, Thailand. (Line 220-221, page 6-7)

Statement

1. I believe dengue fever is a serious disease for children

2. I believe dengue fever is a serious disease for adults

3. I think I am at risk of contracting dengue fever in the next 5 years

4. I think my child/grandchild is at risk of contracting dengue fever in the next 5 years

5. I believe the dengue vaccine has been thoroughly tested for safety

6. I believe the dengue vaccine can helps boost the immune system in both children and adults

7.The benefits of the vaccine outweigh the side effects

8. I believe the vaccine helps prevent dengue infection

9. I believe the vaccine can reduces the likelihood of hospitalization due to dengue

10. I believe the vaccine can helps reduce deaths from dengue fever

 

Comment 1.10:

Line 228: When the authors write "with higher vaccine acceptance," they are referring to a degree of acceptance that has not been measured/characterized, since the answer to the question is dichotomous: "acceptance" or "non-acceptance."

Response:

Thank you for pointing this out. We appreciate your suggestion. We have revised 5^th paragraph in results section to correct this point per suggestion

“Factors significantly associated with intention to vaccinate with the dengue vaccine included having a secondary education, which increased the likelihood of receiving by 2.1 times compared with those with primary education or lower (aOR = 2.104, 95% CI: 1.205–3.674, p < 0.05). Participants who had received an influenza vaccine in the past year were also 1.6 times more likely to intent to receive the dengue vaccine (aOR = 1.571, 95% CI: 1.023–2.411, p < 0.05). Additionally, higher overall attitude scores were associated with an intention to vaccinate (aOR = 1.062, 95% CI: 1.024–1.101, p < 0.01). Participants who perceived the vaccine as safe were more likely to intend to vaccinate compared with those who did not perceive it as safe (aOR 1.279; 95% CI: 1.024–1.596, p <0.05) (Supplementary1). In contrast, those with a history of vaccine refusal were 62% less likely to intent to receive the dengue vaccine (aOR = 0.379, 95% CI: 0.216–0.663, p < 0.01) (Table5)” (Line 238-248, page 7)

Comment 1.11

 Line 244: The sample is not representative, so you cannot extrapolate to the adult population living in rural Chiang Mai.

Response:

Thank you for pointing this out. We have revised per suggestion.  

“This study provides early evidence on the intention to receive the TAK-003 dengue vaccine among the study participants living in rural Chiang Mai, Thailand” (Line 255-256, page 8)

 

Comment 1.12

 Lines 258-259: the authors cannot write "particularly among rural and older populations" because they did not compare rural populations with urban populations, nor the elderly with young people or adults.

Response:

Thank you for this important comment. We have revised this sentence in 2nd paragraph (Discussion section) to avoid misleading per suggestion

“These results highlight the need to address gaps in awareness and perceptions to support         dengue prevention in the study population.” (Line 270-271, page 8)

 

Comment 1.13

Line 280: The authors asked about the sources of information, but did not ask if the study participants trusted the healthcare professionals. Indicating a source of information is not synonymous with trusting that source.

Response:

Thank you for this valuable comment. We appreciated your suggestion and agree that a source of information is not equivalent to trust in that source. In Table 4, this variable referred to the trusted sources of information used by participants in making decisions about dengue vaccination. We have revised the text in Table4 to avoid misunderstanding.

“Trusted Sources of Information Influencing Dengue Vaccination Decisions” (Table4, line 234, page 7)

“Interestingly, this study found that 78% of participants reported local healthcare            providers such as community nurses, public health officers, and village health volunteers who work in sub-district health promoting hospital, as a trusted sources of information informing their decisions about dengue vaccination” (Line 295-298, page 9)

 

Comment 1.14

Line 336: I believe that instead of saying they are adults from northern Thailand, you should say they are adults from rural northern Thailand.

Response:

Thank you for pointing this out. We agree that the original wording was too broad. We have revised the text to specify that the findings refer to adults from rural northern Thailand per suggestion.

“This study highlights that approximately two-thirds of adults in rural northern  Thailand expressed an intention to receive the TAK-003 dengue vaccine” (Line 359-360, page 10)

Reviewer 2 Report

Comments and Suggestions for Authors

The authors present the manuscript “Drivers of TAK-003 Dengue Vaccine Acceptance Among Adults in Rural Thailand: A Health Equity Perspective,” which offers an interesting analysis. However, in its current form, it has several limitations and concerns that should be addressed and resolved.

1. Title
The title of the manuscript may be considered somewhat presumptuous, as it promises more than the study actually delivers. Referring to “Drivers” suggests determinants supported by a stronger, almost causal analytical basis, whereas the cross-sectional design only allows for the identification of associations. In addition, the label “A Health Equity Perspective” is not truly operationalized: the study is conducted in a rural population and describes education and income, but it does not develop an explicit equity framework, nor does it analyze inequities in access, costs, availability, structural barriers, or stratified differences that would justify such a conceptual framing. “Rural” does not automatically equate to a “health equity analysis.”

2. Sampling and external validity
The sample does not appear to be probabilistic. The authors recruited participants through collaborating hospitals and household visits, inviting residents who were at home during working days. This introduces an important selection bias, as it likely over-represents older adults, women, retirees, or individuals who spend more time at home. Indeed, the final sample had a mean age of 61.3 years and was 73.2% female. Although the authors mention this as a limitation, its impact is greater than they acknowledge, as it affects not only generalizability but also the estimate of vaccine acceptance itself, since willingness to be vaccinated may vary substantially according to age, employment status, and contact with health services.

3. Definition of the primary outcome
The definition of the primary outcome is too weak to support policy-oriented conclusions. Vaccine acceptance is essentially measured as a dichotomous “yes/no” response regarding acceptance of the new vaccine. However, the manuscript does not make clear how this question was exactly formulated, whether the vaccine was assumed to be free or paid, whether immediate availability was assumed, whether participants received standardized prior information, or whether this referred to hypothetical intention versus probable future behavior. This is crucial, because “acceptance” without a scenario involving cost, access, and medical recommendation may substantially overestimate or underestimate actual uptake. The manuscript then uses this variable to make inferences about rollout and health planning, which is more ambitious than what the instrument वास्तवically permits. Please clarify this in the manuscript.

4. Measurement of knowledge
The measurement of knowledge is psychometrically very weak, and this compromises part of the analysis. The knowledge section has a KR-20 of 0.387, which is low and suggests poor internal consistency. Nevertheless, the authors use it as a total score in the analysis and also state in the conclusion that participants had “adequate knowledge.” This is not sufficiently supported: not only were there important errors across several items, but the scale itself showed poor psychometric performance. With a KR-20 of this level, the “knowledge” construct requires a much more cautious interpretation.

5. Analytical model
The analytical model has a problem of conceptual circularity. The predictor “total attitude score” appears associated with vaccine acceptance. However, when the attitude items are examined, several are almost equivalent to the outcome or excessively close to it; for example, believing that the vaccine has been well tested, that its benefits outweigh adverse effects, and that it prevents infection, hospitalization, and death. Therefore, using this score to “explain” vaccine acceptance is almost tautological: someone who already believes the vaccine is safe, effective, and beneficial will inevitably be more likely to accept it. This weakens the analytical originality of the model, because it is not identifying distal independent predictors, but rather measuring almost the same cognitive-affective core as the outcome itself. Please correct this.

6. Statistical analyses
There are several signs of methodological fragility. First, in the sample size calculation, the authors state that they used an alpha of 0.01, but they report a Z value of 1.959964, which corresponds to 0.975 and is not consistent with a two-sided alpha of 0.01. Although this may appear technical, this inconsistency undermines confidence in the methodological reporting. Second, although the authors state that there was no cluster sampling, recruitment was conducted through facilities and community visits; therefore, independence of observations is not entirely guaranteed, and the potential clustering effect is not discussed. Third, the model includes very small categories, such as master’s degree or higher with n=5, which makes some estimates unstable. Fourth, the Methods section announces models for determinants of knowledge and attitudes, but these results do not appear in the Results section.

7. Discussion
The Discussion section overinterprets several findings. The manuscript moves from fairly basic observational associations to broad recommendations regarding primary prevention, equitable rollout strategies, and public policy. The problem is not that these recommendations are unreasonable, but rather that the study lacks sufficient causal depth and sufficient richness of variables to support them firmly. The authors did not measure willingness to pay, real barriers to access, distance to health services, insurance coverage, previous adverse experiences, campaign exposure, or institutional trust in a structured manner. Therefore, the study may generate useful hypotheses, but it should not be presented as a solid basis for more complex rollout decisions. Please revise this section.

8. Limitations
The study limitations are acknowledged, but they are underestimated. The authors are encouraged to explicitly mention important limitations, including the non-probabilistic nature of the sampling, the poor reliability of the knowledge score, the possible endogeneity between attitudes and acceptance, the lack of an explicit equity framework, and the ambiguity between hypothetical acceptance and actual future uptake.

Author Response

Reviewer 2

The authors present the manuscript “Drivers of TAK-003 Dengue Vaccine Acceptance Among Adults in Rural Thailand: A Health Equity Perspective,” which offers an interesting analysis. However, in its current form, it has several limitations and concerns that should be addressed and resolved.

Comment2.1:Title
The title of the manuscript may be considered somewhat presumptuous, as it promises more than the study actually delivers. Referring to “Drivers” suggests determinants supported by a stronger, almost causal analytical basis, whereas the cross-sectional design only allows for the identification of associations. In addition, the label “A Health Equity Perspective” is not truly operationalized: the study is conducted in a rural population and describes education and income, but it does not develop an explicit equity framework, nor does it analyze inequities in access, costs, availability, structural barriers, or stratified differences that would justify such a conceptual framing. “Rural” does not automatically equate to a “health equity analysis.”

Response:

Thank you for suggestion this important point. We appredicate your suggestion and have revised the resesarch title to avoid misleading as suggested.

“Intention to Receive the TAK-003 Dengue Vaccine and Associated Factors Among Adults in Rural Northern Thailand” (Line 2-3, page 1)

 

Comment 2.2: Sampling and external validity

The sample does not appear to be probabilistic. The authors recruited participants through collaborating hospitals and household visits, inviting residents who were at home during working days. This introduces an important selection bias, as it likely over-represents older adults, women, retirees, or individuals who spend more time at home. Indeed, the final sample had a mean age of 61.3 years and was 73.2% female. Although the authors mention this as a limitation, its impact is greater than they acknowledge, as it affects not only generalizability but also the estimate of vaccine acceptance itself, since willingness to be vaccinated may vary substantially according to age, employment status, and contact with health services.

Response:

Thank you for this important comment. We agree that the recruitment approach may have introduced selection bias and also limited generalizability.  We have revised the Limitations section to explicitly acknowledge this issue and have tempered the interpretation of our findings per suggestion

“Due to convenience sampling in the    community, study sample was predominantly composed of adults aged 50 years and older, as recruitment targeted individuals who were at home during weekdays. This may have introduced selection and social desirability biases, potentially underestimating or overestimating the intention to vaccinate. Additionally, the study was conducted in a single center, which may limit the generalizability of the findings to other regions of Thailand or other countries with differing sociodemographic contexts.” (Line 332-338, page 9)

 

Comment 2.3: Definition of the primary outcome

The definition of the primary outcome is too weak to support policy-oriented conclusions. Vaccine acceptance is essentially measured as a dichotomous “yes/no” response regarding acceptance of the new vaccine. However, the manuscript does not make clear how this question was exactly formulated, whether the vaccine was assumed to be free or paid, whether immediate availability was assumed, whether participants received standardized prior information, or whether this referred to hypothetical intention versus probable future behavior. This is crucial, because “acceptance” without a scenario involving cost, access, and medical recommendation may substantially overestimate or underestimate actual uptake. The manuscript then uses this variable to make inferences about rollout and health planning, which is more ambitious than what the instrument वास्तवically permits. Please clarify this in the manuscript.

Response:

Thank you for this important comment. We agree that the term “vaccine acceptance” may imply a stronger behavioral outcome than what our measurement instrument can support. To address this concern, we have revised the primary outcome throughout the manuscript from “vaccine acceptance” to “intention to vaccinate”, which more accurately reflects the hypothetical nature of the question and the cross-sectional design of the study.

We acknowledge that participants’ responses may vary depending on contextual factors such as vaccine cost, access, availability, and medical recommendation, as suggested by the reviewer. In the local context of Thailand, the TAK-003 dengue vaccine is paid out-of-pocket, which may further influence individuals’ vaccination decisions. Therefore, our study aimed to explore behavioral intention as a proxy indicator to understand community perspectives toward the vaccine (when they need to pay out of pocket to protect themselves), particularly among populations in rural settings who may have lower educational attainment, lower income and limited access to healthcare services.

We have revised the last paragraph of discussion section by adding a note suggesting future study to acknowledge this point.  

“The future studies should examine intention to vaccinate under different scenarios e.g., varying cost, accessibility, previous experiences, campaign exposure, and physician recommendation to better understand and reflect potential uptake in real-world settings.” (Line 344-347, page 10)  

We also revised the conclusion to make the implication from this result more realistic per suggestion

“Despite these limitations, the findings offer important implications. They can inform the development of tailored strategies to address misconceptions, foster more positive attitudes, and improve communication approaches to enhance intention to vaccinate and reduce health disparities among rural adults. Physicians can play a leading role by collaborating with local healthcare providers to generate context-specific evidence and address concerns regarding the safety and efficacy of the new TAK quadrivalent dengue vaccine. Furthermore, the results support the implementation of community-based vaccination program to strengthen vaccine confidence, promote informed decision making, and advance dengue prevention intervention efforts to increase vaccine uptake, particularly among populations in resource-limited settings.” (Line 347-357, page 10)

 

Comment 2.4: Measurement of knowledge

The measurement of knowledge is psychometrically very weak, and this compromises part of the analysis. The knowledge section has a KR-20 of 0.387, which is low and suggests poor internal consistency. Nevertheless, the authors use it as a total score in the analysis and also state in the conclusion that participants had “adequate knowledge.” This is not sufficiently supported: not only were there important errors across several items, but the scale itself showed poor psychometric performance. With a KR-20 of this level, the “knowledge” construct requires a much more cautious interpretation.

Response:

Thank you for pointing this out. We appreciate your suggestion and revised the limitation of this KR-20 per suggestion

“The internal consistency of the 10-item knowledge scale was relatively low, indicated limited reliability. the findings related to knowledge should be interpreted with caution as the scale may not fully capture participants’ knowledge levels” (Line 341-344, page 9)

 

Comment 2.5: Analytical model

The analytical model has a problem of conceptual circularity. The predictor “total attitude score” appears associated with vaccine acceptance. However, when the attitude items are examined, several are almost equivalent to the outcome or excessively close to it; for example, believing that the vaccine has been well tested, that its benefits outweigh adverse effects, and that it prevents infection, hospitalization, and death. Therefore, using this score to “explain” vaccine acceptance is almost tautological: someone who already believes the vaccine is safe, effective, and beneficial will inevitably be more likely to accept it. This weakens the analytical originality of the model, because it is not identifying distal independent predictors, but rather measuring almost the same cognitive-affective core as the outcome itself. Please correct this.

Response:

Thank you for this valuable comment. We agree that individuals who believe a vaccine is safe, effective, and beneficial are naturally more likely to accept it. To further explore this relationship, we conducted additional analyses to examine which perceptions about dengue and the dengue vaccine were associated with intention to vaccinate (Supplementary Table 1). The results showed that participants who perceived the vaccine as safe were more likely to intend to vaccinate compared with those who did not perceive it as safe (aOR 1.28; 95% CI: 1.02–1.60). Understanding this relationship is important from a public health perspective, as it highlights the potential value of interventions aimed at strengthening positive perceptions of vaccine safety, which may influence future vaccination decisions.

We have added this further analysis in the last paragraph of result section

“Participants who perceived the vaccine as safe were more likely to intend to vaccinate compared with those who did not perceive it as safe (aOR 1.28; 95% CI: 1.02–1.60, p <0.05) (Supplementary1)”. (Line 244-246, page 8)

 

Comment 2.6: Statistical analyses:There are several signs of methodological fragility.

2.6.1, in the sample size calculation, the authors state that they used an alpha of 0.01, but they report a Z value of 1.959964, which corresponds to 0.975 and is not consistent with a two-sided alpha of 0.01. Although this may appear technical, this inconsistency undermines confidence in the methodological reporting.

Response:

Thank you for pointing this out. This was a typo-eror. We have revised the sample size calculation as suggested.

“The sample size was calculated for estimating an infinite population proportion. The proportion of individuals was assumed to be 0.883 [12], with an error margin (d) of 0.05, an alpha level of 0.01, and a Z (0.995) value of 2.58. Cluster sampling was not applied. As a result, the required sample size was 276 participants, and accounting for a 20% dropout rate, a total of 332 participants were needed.” (Line 118-122, page 3)

 

2.6.2, although the authors state that there was no cluster sampling, recruitment was conducted through facilities and community visits; therefore, independence of observations is not entirely guaranteed, and the potential clustering effect is not discussed.

Response:

 Thank you for this important comment. We agree that the original wording did not sufficiently clarify how the study participants were identified and recruited. Participants were not selected through cluster sampling. Instead, the questionnaire was distributed in collaboration with subdistrict health promotion hospitals in Chiang Mai and the village health volunteers, who facilitated community outreach and household visits. Through this established network, eligible rural residents in the areas were randomly invited to participate in the study. We have revised setting and study design as well as participants and recruitmet subsection in Methods section to clarify this recruitment process.

“Data were collected using a structured questionnaire distributed in collaboration with subdistrict health promotion hospitals in Chiang Mai and the village health volunteers, who facilitated community outreach and household visits” (Line 106-108, page 3)

 “The research team promoted the study through partner subdistrict health promotion hospitals in Chiang Mai. Trained research staff, with support from village health volunteers, conducted community visits in rural areas and personally invited eligible residents who were at home during weekdays to participate” (Line 110-113, page 3)

 

2.6.3, the model includes very small categories, such as master’s degree or higher with n=5, which makes some estimates unstable.

Response:

Thank you for pointing this out. We agree that the small sample size in some education categories (e.g., master’s degree or higher, n = 5) could lead to unstable estimates. To address this issue, we combined the bachelor’s degree category with the master’s degree or higher category and re-ran the analysis using the revised grouping to obtain more stable estimates. The updated results are presented in Table 1, page 5. And Table 5, page 7-8

“Table 1. Sociodemographic Characteristics of the Study Participants Living in Rural Areas of Chiang Mai, Thailand.” (Line 190-191, page 4-5)

Parameters	Total	Intention to receive vaccine		P value
		Yes (n=331)	No (n=151)
Mean age (SD)	61.3(11.5)	60.1(11.5)	64.1(10.9)	<0.001**a
Sex
Male	129 (26.8)	91 (27.5)	38 (25.2)	0.658b
Female	353 (73.2)	240 (72.5)	113 (74.8)
Education
Primary or lower	310 (64.3)	197 (59.5)	113 (74.8)	0.002**b
Secondary	138 (28.6)	111 (33.5)	27 (17.9)
Bachelor or above	34 (7.1)	23 (6.9)	11 (7.3)

 

Table 5. Binary Logistic Regression for investigating Factors Associated with Intention to Vaccination Among Adults Living in Rural Areas of Chiang Mai, Thailand.

Factors	Adjusted OR	95% CI	P-value
Age	0.978	0.956, 1.001	0.065
Education levels (Primary or lower as ref.)
Secondary	2.104	1.205, 3.674	0.009*
Bachelor	1.142	0.490, 2.663	0.759
History of flu vaccination	1.571	1.023, 2.411	0.039*
History of refuse vaccine	0.379	0.216, 0.663	<0.001**
Total scores of knowledges	1.034	0.909, 1.176	0.610
Total scores of attitudes	1.062	1.024, 1.101	<0.001**

 

2.6.4, the Methods section announces models for determinants of knowledge and attitudes, but these results do not appear in the Results section.

Response:

Thank you for pointing this out. We have revised the Methods section to clarify how the knowledge and attitude scores were measured and analyzed. Specifically, the knowledge and attitude scores are now clearly described, and their summaries (total knowledges and attitudes score) are presented in Table 1. We also clarified the statistical analysis used to examine their association with intention to vaccinate, which is reported in Table 5. In addition, descriptive results for knowledge and attitudes are presented in Tables 2 and Table3, respectively.

 

Comment 2.7: Discussion

The Discussion section overinterprets several findings. The manuscript moves from fairly basic observational associations to broad recommendations regarding primary prevention, equitable rollout strategies, and public policy. The problem is not that these recommendations are unreasonable, but rather that the study lacks sufficient causal depth and sufficient richness of variables to support them firmly. The authors did not measure willingness to pay, real barriers to access, distance to health services, insurance coverage, previous adverse experiences, campaign exposure, or institutional trust in a structured manner. Therefore, the study may generate useful hypotheses, but it should not be presented as a solid basis for more complex rollout decisions. Please revise this section.

Response:

Thank you for this important comment. We acknowledge that the study did not assess several important factors relevant to vaccine rollout decisions as mentioned by reviewer. We have revised the limitation and last paragraph in discussion section to adopt a more cautious interpretation, to avoid overstating policy and rollout implications and add future research plan to comprehensively understand intention to vaccinate under different scenario. We also added the future research plan to study on other scenarios that could effect the vaccine uptake in the real world.   

“The future studies should examine intention to vaccinate under different scenarios e.g., varying cost, accessibility, previous experiences, campaign exposure, and physician recommendation to better understand and reflect potential uptake in real-world settings.” (Line 344-347, page 9)

“Despite these limitations, the findings offer important implications. They can inform the development of tailored strategies to address misconceptions, foster more positive attitudes, and improve communication approaches to enhance intention to vaccinate and reduce health              disparities among rural adults. Physicians can play a leading role by collaborating with local healthcare providers to generate context-specific evidence and address concerns regarding the safety and efficacy of the new TAK quadrivalent dengue vaccine. Furthermore, the results support the implementation of community-based vaccination program to strengthen vaccine confidence, promote informed decision making, and advance dengue prevention intervention efforts to increase vaccine uptake, particularly among populations in resource-limited settings..” (Line 347-357, page 9-10)

 

Comment 2.8: Limitations

The study limitations are acknowledged, but they are underestimated. The authors are encouraged to explicitly mention important limitations, including the non-probabilistic nature of the sampling, the poor reliability of the knowledge score, the possible endogeneity between attitudes and acceptance, the lack of an explicit equity framework, and the ambiguity between hypothetical acceptance and actual future uptake.

 

Response:

Thank you for this valuable suggestion. We have revised the limitation per suggestion

“This study has several limitations. Its cross-sectional design prevents causal inference, and the use of self-reported data introduces the possibility of recall biases especially for dengue and vaccination history. Due to convenience sampling in the    community, study sample was predominantly composed of adults aged 50 years and older, as recruitment targeted individuals who were at home during weekdays. This may have introduced selection and social desirability biases, potentially underestimating or overestimating the intention to vaccinate. Additionally, the study was conducted in a   single center, which may limit the generalizability of the findings to other regions of   Thailand or other countries with differing sociodemographic contexts. The absence of information on personal prevention practices, such as mosquito-bite prevention or         mosquito-breeding site elimination, also limits the ability to assess how these behaviors may influence knowledge, attitudes, or intention to receive the new dengue vaccine. The internal consistency of the 10-item knowledge scale was relatively low, indicated limited reliability. the findings related to knowledge should be interpreted with caution as the scale may not fully capture participants’ knowledge levels. The future studies should examine intention to vaccinate under different scenarios e.g., varying cost, accessibility, previous experiences, campaign exposure, and physician recommendation to better understand and reflect potential uptake in real-world settings. Despite these limitations, the findings offer important implications. They can inform the development of tailored strategies to address misconceptions, foster more positive attitudes, and improve communication approaches to enhance intention to vaccinate and reduce health disparities among rural adults. Physicians can play a leading role by collaborating with local healthcare providers to generate context-specific evidence and address concerns regarding the safety and efficacy of the new TAK quadrivalent dengue vaccine. Furthermore, the results support the implementation of community-based vaccination program to strengthen vaccine confidence, promote informed decision making, and advance dengue prevention intervention efforts to increase vaccine uptake, particularly among populations in resource-limited settings.” (Line 330-357, page 9-10)

 

Reviewer 3 Report

Comments and Suggestions for Authors

The authors present an interesting and timely study exploring the acceptance of the newly introduced TAK-003 dengue vaccine among adults in rural Thailand. Given the burden of dengue in endemic settings, gathering early evidence from underserved populations through a health equity lens is highly relevant. The manuscript is generally well-structured, and the insights regarding trust in local healthcare providers and prior influenza vaccination histories are valuable for future immunization program planning.

However, there are several critical methodological and reporting issues—particularly concerning the statistical analysis, potential selection bias, and the exact definition of the primary outcome—that must be addressed before the manuscript can be considered for publication.

Major Comments

Questionnaire reliability and internal consistency: The reported KR-20 for the 10-item knowledge scale is 0.387, which indicates poor internal consistency. However, later in the text, it is stated that values ≥0.70 indicate acceptable reliability, implying the overall tool is reliable. The authors must explicitly acknowledge the low reliability of the knowledge scale and critically discuss its implications, especially since this specific score is used as a predictor in the regression models.

Sample size calculation discrepancies: There is an inconsistency in the reported statistical parameters in the Methods section. The text states α=0.01, but the calculation uses Z=1.959964. This Z-value corresponds to a two-sided α=0.05, not 0.01. Please correct this discrepancy and clearly state the exact formula and assumptions used.

Clarification of the primary outcome: The manuscript frequently uses the term "vaccine acceptance." Since previously vaccinated individuals were excluded and TAK-003 is newly introduced, it appears the study is measuring hypothetical acceptability or intention to vaccinate rather than actual vaccine uptake. Please define this outcome precisely in the Methods and ensure the terminology remains consistent throughout the text.

Selection bias and generalizability: Recruitment was conducted through weekday home visits, resulting in a predominantly older (mean age 61.3 years) and female (73.2%) sample. This demographic skew introduces significant selection bias. This must be addressed explicitly in the limitations section, and the authors should refrain from generalizing these findings to the broader adult population of rural Thailand.

Statistical methods and reporting: The statistical methodology requires more detail. Please clarify how missing data was handled. Furthermore, provide model diagnostics and goodness-of-fit assessments (e.g., calibration), multicollinearity checks, and confirm whether linearity assumptions were met for continuous predictors (age, knowledge, and attitude scores). Finally, the Methods section mentions using logistic regression to assess determinants of knowledge and attitudes, but these results are missing. Please either present these findings or remove the reference from the Methods.

Epidemiological metrics: In the Introduction, it is stated that there were 181 dengue deaths in Thailand in 2023, translating to a mortality rate of 0.82 per 100,000. Please verify this metric and its denominator to ensure this is an accurate population mortality rate rather than a case fatality rate or another case-related measure.

Ethics and consent terminology: The informed consent section refers to "patients." Given this is a community-based survey of adults, the term "participants" or "respondents" is more accurate. Please revise accordingly.

Minor Comments

Health equity framing: The title heavily emphasizes a "health equity perspective," yet the analysis remains largely descriptive. To justify the title, I encourage the authors to strengthen the equity discussion—for example, by exploring educational and income gradients and what they imply for vaccine access in these rural settings.

Score construction: Please explicitly state the full possible score ranges for both the knowledge and attitude scales, and clarify if any items were reverse-coded.

Table formatting: Verify the percentages in Table 1 (there appears to be a mathematical inconsistency in the "Family type" category). Additionally, Table 5 would greatly benefit from the inclusion of exact p-values or clearer footnotes explaining the significance markers.

Typographical errors: * Correct the repetition in "using using STATA".

Fix the truncated affiliation line ("Biomedical Informati...").

Review the manuscript for awkward line-break hyphenations before the final submission.

Author Response

Reviewer 3

The authors present an interesting and timely study exploring the acceptance of the newly introduced TAK-003 dengue vaccine among adults in rural Thailand. Given the burden of dengue in endemic settings, gathering early evidence from underserved populations through a health equity lens is highly relevant. The manuscript is generally well-structured, and the insights regarding trust in local healthcare providers and prior influenza vaccination histories are valuable for future immunization program planning.

However, there are several critical methodological and reporting issues—particularly concerning the statistical analysis, potential selection bias, and the exact definition of the primary outcome—that must be addressed before the manuscript can be considered for publication.

Major Comments

Major Comment 3.1

Questionnaire reliability and internal consistency: The reported KR-20 for the 10-item knowledge scale is 0.387, which indicates poor internal consistency. However, later in the text, it is stated that values ≥0.70 indicate acceptable reliability, implying the overall tool is reliable. The authors must explicitly acknowledge the low reliability of the knowledge scale and critically discuss its implications, especially since this specific score is used as a predictor in the regression models.

Response:

Thank you for pointing this out. We agree that 10-item knowledge scale reported poor internal consistency. We have revised the limitation section to acknowledge low reliability of the knowledge and its implications. We also removed potential misunderstanding statement regarding the reliability of overall tools.  

“The internal consistency of the 10-item knowledge scale was relatively low, indicated limited reliability. the findings related to knowledge should be interpreted with caution as the scale may not fully capture participants’ knowledge levels” (Line 341-344, page 9)

We also removed the statement “values of at least 0.70 indicated acceptable reliability of the questionnaire” to avoid potential misunderstanding regarding the reliability of the knowledge scale

Major Comment 3.2

Sample size calculation discrepancies: There is an inconsistency in the reported statistical parameters in the Methods section. The text states α=0.01, but the calculation uses Z=1.959964. This Z-value corresponds to a two-sided α=0.05, not 0.01. Please correct this discrepancy and clearly state the exact formula and assumptions used.

 

 

Response:

Thank you for pointing this out. This was a typo-error. We have revised the sample size calculation per suggestion.

“The sample size was calculated for estimating an infinite population proportion. The proportion of individuals was assumed to be 0.883 [12], with an error margin (d) of 0.05, an alpha level of 0.01, and a Z (0.995) value of 2.58. Cluster sampling was not applied. As a result, the required sample size was 276 participants, and accounting for a 20% dropout rate, a total of 332 participants were needed.” (Line 118-122, page 3)

Major Comment 3.3

Clarification of the primary outcome: The manuscript frequently uses the term "vaccine acceptance." Since previously vaccinated individuals were excluded and TAK-003 is newly introduced, it appears the study is measuring hypothetical acceptability or intention to vaccinate rather than actual vaccine uptake. Please define this outcome precisely in the Methods and ensure the terminology remains consistent throughout the text.

Response:

Thank you for this valuable comment. We agree with the reviewer that the outcome measured in this study reflects intention to vaccinate rather than actual vaccine uptake. Although the terms “vaccine acceptance” and “intention to vaccinate” are sometimes used interchangeably in the literature, to avoid ambiguity we have revised the terminology throughout the manuscript. Specifically, “vaccine acceptance” has been replaced with “intention to vaccinate or intention to receive vaccine”. We have revised this point and ensured consistency throughout the text

Major Comment 3.4:

Selection bias and generalizability: Recruitment was conducted through weekday home visits, resulting in a predominantly older (mean age 61.3 years) and female (73.2%) sample. This demographic skew introduces significant selection bias. This must be addressed explicitly in the limitations section, and the authors should refrain from generalizing these findings to the broader adult population of rural Thailand.

Response:

Thank you for pointing this out. We agree that our recruitment may have introduced selection bias and limited generalizability. We have revised the limitations section per suggestion.

“Due to convenience sampling in the    community, study sample was predominantly composed of adults aged 50 years and older, as recruitment targeted individuals who were at home during weekdays. This may have introduced selection and social desirability biases, potentially underestimating or overestimating the intention to vaccinate. Additionally, the study was conducted in a single center, which may limit the generalizability of the findings to other regions of Thailand or other countries with differing sociodemographic contexts.” (Line 332-338, page 9)

 

Major Comment 3.5

Statistical methods and reporting: The statistical methodology requires more detail. Please clarify how missing data was handled. Furthermore, provide model diagnostics and goodness-of-fit assessments (e.g., calibration), multicollinearity checks, and confirm whether linearity assumptions were met for continuous predictors (age, knowledge, and attitude scores). Finally, the Methods section mentions using logistic regression to assess determinants of knowledge and attitudes, but these results are missing. Please either present these findings or remove the reference from the Methods.

Response:

Thank you for pointing this out. We have excluded the  missing data into the analysis. This study reported the response rate for complete questionnaire about 89.1%, with 59 participants excluded due to incomplete responses (Line 115-116, page 3)

In line with our study objective, we primarily aimed to identify factors associated with the intention to receive the dengue vaccine rather than to develop a predictive model. Therefore, a goodness-of-fit assessment was not considered directly relevant to the purpose of our analysis. However, we assessed multicollinearity among the predictors, and the results are presented in Supplementary Table 2. The variance inflation factor (VIF) values for all variables were below 10, indicating no significant multicollinearity and suggesting that the regression coefficients were reliable.

Supplementary2. Multicollinearity assessment of predictors in the binary logistic regression model for factors associated with intention to vaccinate among adults in rural Chiang Mai, Thailand

Determinant	VIF	Square-root VIF
Age	1.49	1.22
Less than primary	(reference category)
Secondary	1.33	1.15
Bachelor	1.16	1.08
Flu vaccination	1.08	1.04
No vaccination	1.05	1.02
Total attitudes score	1.06	1.03
Total knowledge score	1.09	1.04

 

Major Comment 3.6:

Epidemiological metrics: In the Introduction, it is stated that there were 181 dengue deaths in Thailand in 2023, translating to a mortality rate of 0.82 per 100,000. Please verify this metric and its denominator to ensure this is an accurate population mortality rate rather than a case fatality rate or another case-related measure.

Response:

Thank you for pointing this out. We have revised the sentence to read: “In 2023 alone, Thailand reported 158,620 dengue cases and 181 deaths, with a case fatality rate of 0.11%,” instead of reporting the mortality rate per population (Line 52-53, page2)

Major Comment 3.7

Ethics and consent terminology: The informed consent section refers to "patients." Given this is a community-based survey of adults, the term "participants" or "respondents" is more accurate. Please revise accordingly.

Response:

Thank you for this important comment. We agree that the term “patients” is not appropriate in the context of this community-based survey. We have revised the informed consent statement accordingly by replacing “patients” with “participants.” (Line 383, page 10)

 

Minor Comments

Minor Comment 3.8

Health equity framing: The title heavily emphasizes a "health equity perspective," yet the analysis remains largely descriptive. To justify the title, I encourage the authors to strengthen the equity discussion—for example, by exploring educational and income gradients and what they imply for vaccine access in these rural settings.

Response:

Thank you for this insightful suggestion. We agree that our study does not explicitly use a health equity framework; therefore, we have revised the title to a more neutral wording: ‘Intention to Receive the TAK-003 Dengue Vaccine and Associated Factors Among Adults in Rural Northern Thailand.’ We also agree that health equity is an important consideration. Accordingly, we have revised the second paragraph of the Discussion section to further address this point, as suggested.

 “Socioeconomic disparities, including lower education and income, may also contribute to inequitable vaccine uptake in rural settings. Therefore, local health providers should promote equitable access by improving vaccine availability and implementing targeted educational programs to enhance knowledge and awareness of dengue prevention and vaccination” (Line 274-278, page 8)

Minor Comment 3.9

Score construction: Please explicitly state the full possible score ranges for both the knowledge and attitude scales, and clarify if any items were reverse-coded.

 

Response:

Thank you for highlighting this important point. We have now clarified the scoring procedures in the Methods section (Questionnaire subsection) to ensure methodological clarity and reproducibility per suggestion. There is also no reverse-coded in our questionnaire.  

“Knowledge about dengue disease, including disease transmission, symptoms, treatment, and vaccine availability for ten items. Each of the 10 knowledge items was scored as 1 for a correct response and 0 for an incorrect response, resulting in a total possible score ranging from 0 to 10, with higher scores indicating greater knowledge” (Line 129-132, page 3)

“Attitudes toward dengue disease and the dengue vaccine: Participants' attitudes toward dengue fever and the dengue vaccine were assessed using a five-level Likert scale, ranging from the lowest degree (1) to the highest degree (5) for ten items. Each of the 10 attitude items was measured on a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree), yielding a total score ranging from 10 to 50, with higher scores reflecting more positive attitudes toward dengue vaccination.” (Line 133-138, page 3)

Minor Comment 3.10

Table formatting: Verify the percentages in Table 1 (there appears to be a mathematical inconsistency in the "Family type" category)  Additionally, Table 5 would greatly benefit from the inclusion of exact p-values or clearer footnotes explaining the significance markers.

Response

Thank you for pointing this out. The percentage in the “Family type” category in Table 1, page5 has been corrected from 6.9% to 68.9%. We have added the exact p-value and clearer footnotes per suggestion

Table 5. Binary Logistic Regression for investigating Factors Associated with Intention to Vaccination Among Adults Living in Rural Areas of Chiang Mai, Thailand.

Factors	Adjusted OR	95% CI	P-value
Age	0.978	0.956, 1.001	0.065
Education levels (Primary or lower as ref.)
Secondary	2.104	1.205, 3.674	0.009*
Bachelor	1.142	0.490, 2.663	0.759
History of flu vaccination	1.571	1.023, 2.411	0.039*
History of refuse vaccine	0.379	0.216, 0.663	<0.001**
Total scores of knowledges	1.034	0.909, 1.176	0.610
Total scores of attitudes	1.062	1.024, 1.101	<0.001**

Analyzed with binary logistic regression analysis with the Enter method

*p-value<0.05, **p-value<0.001

 

 

Minor Comment 3.11

Typographical errors: * Correct the repetition in "using using STATA".

Response:

Thank you for pointing this out. We have removed the repeated term. (Line 156, page4)

 

Minor Comment 3.12

Fix the truncated affiliation line ("Biomedical Informati...").

Response:

Thank you for pointing this out. The truncated affiliation line has been corrected. (Line 14)

 

Minor Comment 3.13

Review the manuscript for awkward line-break hyphenations before the final submission.

Response

We have reviewed the manuscript for any awkward line-breaking hypenations before the final submission per suggestion.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

This paper reviews Thai community participants perception of Dengue Vaccine Acceptance.  Abstract are very well written.

I don’t believe the sample size calculation is correctly described.  Please review this.  

I also wonder about the assumption that 88 out of 100 people would be confident in getting the vaccine based on the background. This does not seem reasonable and you have gathered data from many more people so your power calculation can be adjusted. 

 

For the evaluation of the survey, the attitude scale showed acceptable internal consistency (alpha ≈ 0.81, 95% CI roughly 0.77–0.84). In contrast, the knowledge scale had low reliability (KR‑20 ≈ 0.39), suggesting that the knowledge items did not form a consistent scale. This limits the precision and interpretability of the knowledge scores, and should be treated as a key methodological limitation and noted in the limitation section.

 

The ‘enter method’ could be described in the analysis section for those who are not as familiar with this.  --- non-stepwise method, including all variables regardless of adding or removing variables based on p-values.

 

Thorough discussion that covers many relevant areas. 

Author Response

Reviewer 4

This paper reviews Thai community participants perception of Dengue Vaccine Acceptance.  Abstract are very well written.

 

Comment 4.1

I don’t believe the sample size calculation is correctly described.  Please review this.  

Response:

Thank you for pointing this out. This was a typo-errors. We have revised the sample size calculation per suggestion

“The sample size was calculated for one population proportion. The proportion of individuals was assumed to be 0.883 [12], with an error margin (d) of 0.05, an alpha level of 0.01, and a Z (0.995) value of 2.58. Cluster sampling was not applied. As a result, the required sample size was 276 participants, and accounting for a 20% dropout rate, a total of 332 participants were needed” (Line 118-122, page 3)

 

Comment 4.2

I also wonder about the assumption that 88 out of 100 people would be confident in getting the vaccine based on the background. This does not seem reasonable and you have gathered data from many more people so your power calculation can be adjusted. 

Response:

Thank you for this valuable comment. The assumption of 88% vaccine acceptance used for the sample size calculation was based on the only previously published study on dengue vaccine acceptance available at the time of study design. The population characteristics in that study were considered relatively comparable to our target population in Thailand.

However, we acknowledge that this estimate may not fully represent the Thai context. In response to the reviewer’s suggestion, we have revisited and revised the sample size calculation. (Line 118-122, page3) Importantly, the final number of participants included in the analysis exceeded the minimum sample size required based on the revised assumptions, suggesting that the study maintained adequate statistical power.

 

Comment 4.3

For the evaluation of the survey, the attitude scale showed acceptable internal consistency (alpha ≈ 0.81, 95% CI roughly 0.77–0.84). In contrast, the knowledge scale had low reliability (KR‑20 ≈ 0.39), suggesting that the knowledge items did not form a consistent scale. This limits the precision and interpretability of the knowledge scores, and should be treated as a key methodological limitation and noted in the limitation section.

Response:

Thank you for pointing this out. We agreed that our knowledge scale had low reliability and we acknowledged this point in limitation per suggestion.

“The internal consistency of the 10-item knowledge scale was relatively low, indicated limited reliability. the findings related to knowledge should be interpreted with caution as the scale may not fully capture participants’ knowledge levels.” (Line 341-344, page 9) 

 

Comment 4.4

The ‘enter method’ could be described in the analysis section for those who are not as familiar with this.  --- non-stepwise method, including all variables regardless of adding or removing variables based on p-values.

Response:

Thank you for this valuable suggestion. We have revised the Statistical Analysis section to clarify the use of the enter method. Specifically, we have added a description indicating that the enter method is a non-stepwise approach in which all selected independent variables are simultaneously included in the multivariable model, without automatic selection or removal based on statistical criteria (e.g., p-values). This clarification has been incorporated to improve transparency and facilitate understanding for readers who may be less familiar with this approach.

We have revised Section 2.5, Statistical Analysis

“A univariable analysis was initially performed to identify potential factors associated with the dependent variables. Variables with a p-value <0.2 in the univariable analysis were subsequently included in the multivariable models. Binary logistic regression analysis was performed to determine factors associated with intention to vaccinate with dengue vaccine. Adjusted odds ratios (aORs) and 95% confidence intervals (95% CI) were estimated using the enter method. The enter method is a non-stepwise approach in which all selected independent variables are entered into the model simultaneously, without automatic selection or removal based on statistical criteria. Independent variables included in the binary logistic model were age, education levels, history of flu vaccination, history of refuse vaccine, total scores of knowledges, and total scores of attitudes. All tests were two-tailed, and a p-value <0.05 was considered statistically significant” (Line 165-176, page 4)

Reviewer 5 Report

Comments and Suggestions for Authors

This study aims to evaluate the acceptance of dengue vaccine in Thailand. The survey was conducted by recruiting village residents in a province of Thailand; a convenience sample of sufficient size was thus recruited.

1. The survey was conducted using a specially designed questionnaire. It showed low reliability with the KR-20 in binary questions. This is a limitation of the study that must be reported.
2. One aspect that should be clarified is whether vaccine acceptance is understood as a positive response to a question about willingness to be vaccinated, or actual acceptance of vaccination. In the former case, the authors should acknowledge and adequately discuss the possibility that the positive response was driven by social desirability.
3. The authors should indicate in the methods the statistical tests they used to ascertain the type of distribution of the variables.
4. The authors should complete the vaccination acceptability percentage, as well as other prevalences calculated by questionnaire, with confidence intervals.
5. The authors state in the methods that they conducted univariate and multivariate logistic regression analyses, but the results in Table 5 do not clearly indicate which of these methods they are referring to. They should clarify which was the dependent variable and which were the predictor variables in the presented model.

Author Response

Reviewer 5

This study aims to evaluate the acceptance of dengue vaccine in Thailand. The survey was conducted by recruiting village residents in a province of Thailand; a convenience sample of sufficient size was thus recruited.

 

Comment 5.1

The survey was conducted using a specially designed questionnaire. It showed low reliability with the KR-20 in binary questions. This is a limitation of the study that must be reported.

 

Response

Thank you for pointing this out. We have revised limitation per suggestion.

“The internal consistency of the 10-item knowledge scale was relatively low, indicated limited reliability. the findings related to knowledge should be interpreted with caution as the scale may not fully capture participants’ knowledge levels.” (Line 341-344, page 9)

 

Comment 5.2

One aspect that should be clarified is whether vaccine acceptance is understood as a positive response to a question about willingness to be vaccinated, or actual acceptance of vaccination. In the former case, the authors should acknowledge and adequately discuss the possibility that the positive response was driven by social desirability.

Response

Thank you for this important point. In our study, a positive response to a question relates to an intention to get this new dengue vaccine. We appreciate your suggestion and have revised limitation on social desirability bias.   

“Due to convenience sampling in the community, study sample was predominantly composed of adults aged 50 years and older, as recruitment targeted individuals who were at home during weekdays. This may have introduced selection and social desirability biases, potentially underestimating or overestimating the intention to vaccinate” (Line 332-336, page 9)

 

Comment 5.3

The authors should indicate in the methods the statistical tests they used to ascertain the type of distribution of the variables.

Response

Thank you for pointing this out. We have revised the statistical analysis section to clearly indicate the statistical tests used according to the distribution of the variables, as suggested.

“The normality of continuous variables was assessed using the Shapiro–Wilk test, supplemented by visual inspection of histograms. Continuous variables with normal distribution were reported as mean and standard deviation (SD) and were compared between groups using Independent t-test, while non-normally distributed variables were summarized as median and interquartile range (IQR)” (Line 159-164, page4)

 

Comment 5.4

The authors should complete the vaccination acceptability percentage, as well as other prevalences calculated by questionnaire, with confidence intervals.

 

Response:

Thank you for your valuable suggestion. We have revised the acceptability with confidence internval as suggested.

“Overall, 68.7% (95% CI: 64.3-72.8%, n = 331) reported acceptance of the dengue vaccine” (Line 178-179, page 4)

 

Comment 5.5

The authors state in the methods that they conducted univariate and multivariate logistic regression analyses, but the results in Table 5 do not clearly indicate which of these methods they are referring to. They should clarify which was the dependent variable and which were the predictor variables in the presented model.

Response:

Thank you for pointing this out. We have revised statistical analysis to clarify the methods and also revised the legend of Table5 for clarification per suggestion.

 

“A univariable analysis was initially performed to identify potential factors associated with the dependent variables. Variables with a p-value <0.2 in the univariable analysis were subsequently included in the multivariable models. Binary logistic regression analysis was performed to determine factors associated with intention to vaccinate with dengue vaccine. Adjusted odds ratios (aORs) and 95% confidence intervals (95% CI) were estimated using the enter method. The enter method is a non-stepwise approach in which all selected independent variables are entered into the model simultaneously, without automatic selection or removal based on statistical criteria. Independent variables included in the binary logistic model were age, education levels, history of flu vaccination, history of refuse vaccine, total scores of knowledges, and total scores of attitudes. All tests were two-tailed, and a p-value <0.05 was considered statistically significant” (Line 165-176, page4)

 

“Table 5. Binary Logistic Regression for investigating Factors Associated with Intention to Vaccination Among Adults Living in Rural Areas of Chiang Mai, Thailand.” (Line 250-251, page 7)