The authors used the Health Belief Model as theoretical framework to guide this study and used the Pillar Integration Process for integration for qualitative and quantitative data. This mixed-method approach is suitable, novel and promising for this topic. However, the results are not surprising and provide rarely any new knowledge. One reason is probably that the Health Belief Model seems not to guide the whole analysis and integration of qualitative and quantitative remains on a rather superficial level. Also the discussion of results is rather repetitive and results are not discussed in the light of the Health Belief Model. As the data is interesting, I recommend that the authors move the integration on a more in-depth level. Furthermore, some issues need to be specified:
Page(s) Line(s) Comment
Page 2 Abstract Correlation between constructs are not presented in the paper. Please adapt this, also in the discussion.
Pages 3-6 Introduction The paper has two focuses: international and Ireland - I suggest focusing on international level and present the background accordingly. The Ireland-specific issues could then be elaborated in the discussion section.
Perhaps the literature could be presented for countries where vaccination is mandatory versus those where vaccination is non-mandatory.
Page 3 Line 24 Vaccine reduces the risk for getting influenza, but does not fully prevent it. Please specify this.
Page 3 Line 29 ‘Subgroups’ could refer to patient groups as well, perhaps it would be clearer to mention ‘health-care workers’ directly (and not in the parenthesis)?
Page 3 Lines 31-38 It would be interesting to know how much the risk decreases. Could the authors mention some quantitative data, e.g. in odds ratio (if available) here?
Page 3 Lines 40-47 I suggest shortening the list of the organizations - as the introduction is rather long.
Page 4 Lines 46-50 A reference from 2010 does not seem suitable to argue for limited research - as much has been published in the past 10 years.
Page 6 Line 10 It would be interesting to know a little bit more why the authors chose this model, given that there are various models.
Page 6 Line 37 I suggest deleting ‘in two acute hospital settings’ in the aim.
Page 6 Line 3 The concepts of the Health Belief Model (HBM) are mentioned, but the relationships between the concepts are missing. A picture of the HBM would be helpful here.
Page 7 Method The authors should specify how far the Health Belief Model guided the study:
 the quantitative part: were all concepts / variables of the HBM assessed?
 the qualitative part: was the interview guide developed on basis of the HBM)
 the integration of data: how far was the integration of data guided by the HBM?
Page 7 Lines 24-54 The selected variables (if possible in connection with your theoretical framework, the HBM) and the operationalisation of each variable is not fully clear. It would be helpful to see them in a table.
Page 7 Lines 24-54 Please provide range and direction (e.g. high numbers = high or low knowledge) for each instrument.
Page 7 Lines 24-54 Please provide information regarding validity and reliability for each instrument.
Page 8 Lines 18-23 Which were the two groups? Vaccinated versus non vaccinated?
Page 8 Lines 23-27 The authors should provide more detail about the multivariable analysis: What kind of analysis? (logistic regression model / endpoint e.g. 0=vaccinated, 1=not vaccinated or vice-versa). A reference and / or reason for selecting variables with p<0.2 should be provided. As p-values are dependent on sample size, it does not seem to be a very robust criteria for selecting variables.
Page 8 Line 37 What were the criteria for purposive sampling? I suggest presenting the sample characteristics in the result section accordingly.
Page 10 Lines 47-56 Does this section refer to the questionnaire from the quantitative study? If yes, suggest moving this passage to the quantitative part.
Page 13 Lines 6-13 Means and SDs per group for all p-values should be provided. Table 2 could be deleted.
Page 13 Line 13 Did the authors analyse per site - not treated as one group? Why?
Page 13 Line 25 For the logistic regression model: Provide goodness-of-fit.
Pages 13-14 Lines 52-6 Are the qualitative findings described in another paper - if yes, the reference should be cited.
Page 14 Lines 10-20 The titles of the integrated findings are more or less the same titles as those of the qualitative findings -> this gives the impression that the qualitative findings have driven the integration and that the integration has been stayed close to the qualitative findings. There are also no gaps which are considered as important by Johnson et al. 2017 (step 3 of integration). It would be important to know how the authors handled this issue.
Pages 16-20 Results qualitative The text in the result section repeats Table 3 to a great extend (e.g. p 16 l 10-14, 21-30, …). I would expect much more integration of data in the result section. As far, it repeats mainly qualitative and quantitative results.
The author mentioned the HBM in the introduction as their theoretical framework and I expected data are integrated under this theoretical framework. It is not clear how the titles of the integration (pillar) arose. The authors should expand on this.
Page 24 Discussion Strength and limitation of methodology and method, as well as trustworthiness should be discussed.
Pages 38-39 Table Under ‘protection, risk and vulnerability’, the nurses own vulnerability to influenza and then, his or her role (professional identity) referring to the patient’s vulnerability are summarized. This seems to me to be two different themes.