Vaccine Hesitancy Toward COVID-19 Vaccines Among Humanitarian Healthcare Workers in Lebanon, 2021

 Comments 1: This is a well written article that makes a useful contribution to the field. The article has a good literature review and an appropriate methodology. The findings are well illustrated and the discussion follows well from the findings. The only question I have is the fact that the health belief model was designed in a different cultural context were any statistical checks made to see that the items had validity in the Lebanese context. E.g Confirmatory Factor Analysis. Otherwise this is a well written and well executed paper.

Response 1: Thank you for this comment. While the reviewer is correct that the HBM was designed in a different context, the HBM has previously been validated for use in the Lebanese and other MENA contexts including in relation to COVID-19 vaccination, as listed in the citations below. Additionally, in our previously published manuscript (Ali et al.), we describe our use of factor analysis to create scales according to HBM items. We have added this statement for clarification, “Principal components exploratory factor analysis using varimax rotation was conducted on the survey items to create scales according to the health belief model items. Rotated factor loadings of ≥ |0.4| were accepted.”

-          Ali Z, Perera SM, Garbern SC, Diwan EA, Othman A, Ali J, Awada N. Variations in COVID-19 vaccine attitudes and acceptance among refugees and Lebanese nationals pre-and post-vaccine rollout in Lebanon. Vaccines. 2022 Sep 15;10(9):1533.

-          Youssef D, Abou-Abbas L, Berry A, Youssef J, Hassan H. Determinants of acceptance of Coronavirus disease-2019 (COVID-19) vaccine among Lebanese health care workers using health belief model. PLoS One. 2022 Feb 22;17(2):e0264128.

-          Ghazy RM, Abdou MS, Awaidy S, Sallam M, Elbarazi I, Youssef N, Fiidow OA, Mehdad S, Hussein MF, Adam MF, Abdullah FS. Acceptance of COVID-19 vaccine booster doses using the health belief model: a cross-sectional study in low-middle-and high-income countries of the East Mediterranean region. International journal of environmental research and public health. 2022 Sep 25;19(19):12136

Comments 2: This is a very good paper with no perceptible errors. The only question is whether there have been any tests of validity, such as confirmatory factor analysis, conducted on the elements of the Health Belief Model to see that they are valid in the Lebanese context?

Response 2: Thank you for this comment. As mentioned above, while the HBM was designed in a different context, it has previously been validated for use in the Lebanese context, especially in relation to COVID-19 vaccination. We have added the following statement for clarification, “Principal components exploratory factor analysis using varimax rotation was conducted on the survey items to create scales according to the health belief model items. Rotated factor loadings of ≥ |0.4| were accepted.”

Comments 1: It is understood that the survey was validated. However, it would be advisable to know if a standardization was performed and if a pilot survey was applied that could evaluate the questions. Determine if there is a correlation between the set of questions and dimensions. The ideal in these cases is that there is a correlation. If there is no correlation, the question should be eliminated or modified.

Response 1: Thank you for this comment. Due to the urgent nature of the survey to inform vaccine campaign efforts during the height of COVID-19 cases in the country, the survey was not piloted beforehand, although was based on similar previously validated surveys done in the MENA region using the Health Beliefs Model. We have clarified that for the regression model, we selected a priori age, gender, and nationality as sociodemographic variables that have been closely correlated with vaccine hesitancy from the previous literature. We have also clarified that factor analysis was conducted to evaluate for correlation between the questions and dimensions, as written above.

Comments 2: Logistic regression. It is important to know if there were previous theoretical models. Indicating the dependent variable (0,1) and the independent variables. Control of confounding variables and interactions was performed. What was the method of variable selection to introduce or remove variables from the logistic regression model. 

Response 2: Thank you for this comment. As written above, we have clarified that for the regression model, we selected a priori age, gender, and nationality as sociodemographic variables that have been closely correlated with vaccine hesitancy from the previous literature.  For the other variables, these were directly included from the HBM constructs which is a well validated model, and therefore further variable selection was not performed for the model constructs. We have added clarification in the methods regarding the primary outcome and independent variables in the model.

Comments 3: Ethics. The survey was online; however, informed consent was sent. If confidentiality is declared, how were the consents collected if they had to be signed. How response bias is reduced: given the format is online, how is it guaranteed that the respondent is a health professional. If there are possibilities to individualize then it is not possible to waive the committee. In addition, consent is being signed.

Response 3: Thank you for your comment. No informed consents were physically collected as this was an online survey, but the identity of the participants was anonymous. The informed consent was included at the start of the survey indicating that participation was voluntary and would be confidential. Data collected were only from those individuals that agreed to participate in the survey. The survey was sent by the MOPH via a secure online link to health care providers working in the national primary health care facilities network. No one outside of the primary health care centers’ health care providers had access to the link. No confidential/ identifying information was collected from participants, therefore their responses remained anonymous. The MOPH of Lebanon reviewed the study details and waived the need for a full ethical review. These details have been included in lines 97-99 and 116-120. We hope this has sufficiently addressed your questions.

Comments 1: A report of a convenience sample to analyze the attitude of Lebanese healthcare workers (HCW) to COVID-19 vaccines conducted online between February 8-21, 2021, before vaccines were available in that country. The authors found significant vaccine hesitancy and a predominant belief that the Lebanese healthcare system could not administer the vaccines safely. The strengths of this study are that it is country-wide and includes a variety of respondents identifying as HCWs. Furthermore, the data are well analyzed with clear tables presenting the results. The limitations are that there is insufficient information provided to replicate this study, as there are no reasons for selecting the materials and methods. A more significant concern is that, as a result of the wording of the questionnaire, some of the results may be the opposite of what the respondent intended. This possibility must be part of the Limitations section.  

Response 1: Thank you for these comments. We do not completely understand the reviewer’s meaning regarding the wording of the questionnaire, however one of the concerns is addressed below in response to Comment #15 that the response “Yes” is understood as “Agree” in Arabic, although we understand the English translation may have caused confusion – we have added “Yes (Agree”) and also a footnote to clarify meaning “Agree with the statement” to the responses for clarity in the tables. We also clarify that the survey was developed by native speakers, and given the online nature of the survey there was no time limitation to completion and therefore we believe unintentionally incorrect responses on the part of the respondent are unlikely to have significantly impacted our findings.

Comments 2: Line 33-34 Please provide a citation to current research for the claim of the first sentence.

Response 2: Thank you for your suggestion, we have added the following citation: Shah A, Coiado OC. COVID-19 vaccine and booster hesitation around the world: A literature review. Front Med (Lausanne). 2023 Jan 12;9:1054557. doi: 10.3389/fmed.2022.1054557. PMID: 36714110; PMCID: PMC9878297.

 Comments 3: Line 37 As both citations 3 and 4 are outdated and not COVID-19 related, please provide a supporting citation to 3 and 4 of COVID-19-related research regarding the meaning of vaccine hesitancy.

Response 3: Thank you for your suggestion, we have added the following citation: Larson HJ, Gakidou E, Murray CJL. The Vaccine-Hesitant Moment. N Engl J Med. 2022 Jul 7;387(1):58-65. doi: 10.1056/NEJMra2106441. Epub 2022 Jun 29. PMID: 35767527; PMCID: PMC9258752.

Comments 4: Line 43-44 Please cite current research for the claim of the first sentence of this paragraph.

Response 4: Thank you for your suggestion, we have added citations to this sentence.

Comments 5: Line 84 Please explain in the text the decision to use the IMC to conduct the survey, citing a current peer-reviewed reference to support this decision..

Response 5: Thank you for your comment; to clarify, IMC conducted this study. IMC has programming in Lebanon and has been involved with the COVID-19 vaccination rollout in the country. Additionally, at the time of the study IMC supported a network of PHCCs and dispensaries across the country, more than other partners which is the reason for the MOPH’s approval for IMC to conduct this survey. As requested, we have added a citation to our previous publication:

-          Ali Z, Perera SM, Garbern SC, Diwan EA, Othman A, Ali J, Awada N. Variations in COVID-19 vaccine attitudes and acceptance among refugees and Lebanese nationals pre-and post-vaccine rollout in Lebanon. Vaccines. 2022 Sep 15;10(9):1533.

Comments 6: Line 86 Please explain in the text the selection of a convenience sample to analyze the HCW.

Response 6: Thank you for your comment. Please note that in Line 95-96, we have indicated that this survey used a convenience sample of healthcare workers (HCW) working at 90 national primary health care centers (PHCC). We hope this adequately addresses your comment.

Comments 7: Line 88 Please explain in the text the reason for distributing the survey online and the time selected for conducting it.

Response 7: Thank you for your comment. Data collection occurred between February 8–21, 2021 while the COVID-19 pandemic was still ongoing and just prior to vaccine rollout. For the safety of our staff and the participants, we felt it was best to conduct the survey online instead of face-to-face. January 2021 was considered one of the deadliest months in Lebanon and the country was in total lockdown, as the number of cases in that month alone exceeded 100,000. The sharp increase in morbidity and mortality prompted the government to impose a strict nationwide lockdown on 15 January 2021, which extended until 8 February 2021. We have included details in the text as suggested in Lines 97 and 118-119.  

Comments 8: Line 96-103 The information provided in this paragraph regarding the IMC should follow the first mention of the IMC. Consequently, the first paragraph of this subsection should concern the IMC.

Response 8: Thank you for this valuable comment, we have restructured this section as suggested.

Comments 9: Line 105 Please define “healthcare worker”.

Response 9: Thank you, we have revised as suggested “…HCW (doctor/ nurse/ pharmacist /other health provider) working at one of the 90 PHCCs…”.

Comments 10: Line 106 Please state the exclusion criteria

Response 10: Thank you for your comment. There were no exclusion criteria other than having completed the survey previously.

Comments 11: Line 109-110 Please explain in the text the reason for selecting the list of questions of the SAGE to adapt for this survey.

Response 11: Thank you for your suggestion, details have been added to the text. The SAGE questions were adapted by the national RCCE taskforce for COVID-19 in January 2021. In fact, the RCCE taskforce “rumors and misconception tracker“ for COVID-19 vaccine, which was developed and updated constantly by health cluster partners, validated the relevance of these questions to the country context.  

Comments 12: Line 115 Please explain in the text the selection of Google Forms for the online survey.

Response 12: Thank you for your comment. Since the MOPH has its own google workspace account, it allowed them to easily share the survey with the 90 PHCC Directors across the country and collect answers securely during the COVID-19 lockdown. This has been added in Line 117.

Comments 13: Line 138-141 Please explain in the text the selection of the particular descriptive analysis, citing a COVID-19 publication using a similar data analysis.

Response 13: Thank you for your comment, we have added a citation of other studies with similar descriptive analysis.  

Comments 14: Line 146-147 Please explain in the text the selection of STATA version 16 for the analyses and cite COVID-19 research using this version of STATA for analyses.

Response 14: Thank you for your comment, we have added a citation to the text.

Comments 15: 175-176 Table 2: Given the statements presented, it is possible that selecting “Yes” or “No” as a response could have the opposite result from the intention of the HCW. For example, “I am not at risk of severe complications of COVID-19 so I will not take the vaccine”, could be interpreted by the respondent as “Yes, I agree with that statement” or “No, I am not at risk, so I will not take the vaccine”. Rather than “Yes” or “No” as the response, the correct answer to not produce misunderstanding is “Agree” or “Disagree”. As a result, the authors must include this poor choice of words as the answer as one of the study's limitations. Another limitation of the statements is that more than one is compound. The example above is one such statement. The responders could agree they are not at risk but also decide to take the vaccine—this option is unavailable because there is a conflation of the two points. Another example is “I think the side effects of the COVID-19 vaccine are very serious/could lead to death”. This statement suggests that a very serious side effect is only one that leads to death. The responders may have agreed that the side effects lead to morbidity but not mortality. This result is unanalyzable from the statement word choice—representing another limitation to this survey. In future studies dependent on responses to a survey, please ensure each question asked pertains to one matter alone.

Response 15: Thank you for this comment. We have clarified that as written in Arabic, the potential for misunderstanding or incorrect response is extremely low, given the “Yes” response would be understood to mean also “agree.” The survey was written and checked by multiple native Arabic speakers and we clarify that while we appreciate the reviewer’s concern, this is only a concern in the English translation but not in Arabic where the responses would be understood as “agree with statement”. For example, the statements in Arabic are straightforward and allow for a clear answer, such as:' I do not need to take the vaccine because I am in good health” so a “yes” answer means that the participant “agrees”  that there is no need for the vaccine once he is in good health. Another question correlates vaccine refusal when a person has already contracted COVID-19. These specific correlations between vaccine hesitancy and a specific cause were based on vaccine rumors that were collected first hand from the community by the national RCCE taskforce during December and January 2021 upon the launch of Lebanon’s National Deployment and Vaccination Plan (NDVP). We have added information in the Limitations section that “a few survey items had compound components and in the future should be revised to separate questions to understand the nuance in the respondents’ answers”.

Comments 16: 188-189 It is notable that the answers provided to the respondents were “Agree” or “Disagree” rather than “Yes” and “No”, as in Table 2—this is an improvement. However, the same problem regarding the compound statements to which respondents were to reply exists in Table 3. It is possible to know that the vaccine is available for all nationalities but not know that it is free of charge. This result is impossible to document with the survey construction—another limitation. One statement to which the HCWs are to respond has three different parts, “Given the variety of vaccines and their evolution I would prefer to wait until they are proven to be safe” A HCW could have three separate views on (1) the variety of vaccines, (2) their evolution, and (3) preferring to wait. This possibility cannot be identified with this statement to which the HCW must respond—another limitation.

Response 16: We appreciate the reviewer’s concern; however, we have clarified that in Arabic, this statement would be clearly understood to mean the question regarding preference to wait. As mentioned above, we have added information in the Limitations section about a few survey items having compound components.

Comments 17: Line 199-202 Please produce one or two tables regarding Communication and Barriers to vaccination and preference for vaccination sites.

Response 17: Thank you for your comment, however we prefer to avoid duplicative information and do not feel a table will contribute to the information presented already in the narrative format related to trusted sources and communication. We have, however, included a table related to barriers to vaccination (See Table 4, Line 212).

Comments 18: Line 292 Please provide a supporting citation for 38 of research published since 2020.

Response 18: Thank you for your valuable suggestion, we have added additional citations as requested.

Comments 1:

Major comments: The authors welcomed all the suggestions and comments made in the first installment. In addition, other suggestions pointed out by other reviewers have been incorporated. These changes are highlighted in the manuscript. I have no additional comments.

Detail comments: The authors welcomed all the suggestions and comments made in the first installment. In addition, other suggestions pointed out by other reviewers have been incorporated. These changes are highlighted in the manuscript. The summary has been improved. The methodology was improved and the variables that comprise the study are clarified. I have no further comments.

Response 1: Thank you for your review and feedback. 

Comments 1: 64-66 Citation 15 is missing. Please add citation 15.

Response 1: Thank you for pointing this out, we have addressed this omission in the manuscript.

Comments 2: 112 The matrix created by SAGE was in 2015—well before COVID-19. Here is an update: WHO interim guidance: Guiding principles for immunization activities during the COVID-19 pandemic. Geneva: World Health Organization; 2020 (https://apps.who. int/iris/bitstream/handle/10665/331590/WHO-2019-nCoV-immunization_services-2020.1-eng.pdf). Please include a citation to this update.

Response 2: Thank you for this suggestion, we have added this citation as requested.

Comments 3: 137-140 Please include “agree” and “disagree” here and inform the reader that “yes” is understood as “agree” and “no” is understood as “disagree” in Lebanese, as the authors have mentioned in their response to the reviewer.

Response 3: Thank you for your comment, we have included the following clarifying sentence in Lines 143-44: “In the survey, a response of “yes” or “no” is understood in Arabic as “agreeing” or “disagreeing” with the statement.”

Comments 4: 199 Please italicize “3.5. Communication”.

Response 4: Thank you for pointing this out, we have made the requested change.

Comments 5: 204 The authors state in their response to the reviewer that they have not added the table for Communication because “we prefer to avoid duplicative information and do not feel a table will contribute to the information presented already in the narrative format related to trusted sources and communication”. The reviewer disagrees. Please include a table for Communication, similar to the tables provided for all the other subsections of the results.

Response 5: We have included a Communications table as the reviewer has requested.

Comments 6: 351-353 Thank you to the authors for adding this statement to the Limitations. The authors must also indicate that it is possible that having compound questions might have compromised the responses for those questions that had them.

Response 6: We have added this detail as requested into the revised addition to the Limitations Section.

Comments 7: 379-476 Please redo the reference in the preferred style of the journal: https://www.mdpi.com/journal/covid/instructions.

Response 7: Thank you, we have revised to the preferred reference style using the Zotero file provided by the journal.