Overall: This is a solid paper that explores the dengue knowledge, attitude and preventive practices (KAP) among the population of tsunami-affected area of Aceh, Indonesia where dengue is a highly significant health threat. The authors assert that their study is the first one that assesses KAP of Aceh communities regarding dengue which may inform future interventions for effective dengue prevention campaigns. The findings main implication is that socio-economic status is the only independent predictor of good dengue related KAP. However, this paper adds little to knowledge about the determinants of preventive behaviour. For example, it does not shed light on why individuals did not practice specific preventive measures, and what are the main barriers to specific actions.

I have some specific comments about the data presented and with regard to clarity of methods and measures, as detailed below.

Title:

I am not sure it is appropriate to specify inhabitants as “healthy” in the title unless the authors provide more specific information on their inclusion criteria of being healthy.

Introduction

The background seems to have missing literature on dengue KAP from other settings inside and outside of Indonesia. It may be good to add other literature to frame the importance of this approach.

P 5: 101-103. “The aggregate annual economic burden…” over which period of time?

P 5: 115-118. It seems to be an oversimplification to infer that an up-ward trend of reported DF cases in Aceh between 2003 and 2011 reflects the increased DF risk after the earthquake and tsunami disaster of 2004. While the risk of DF increases in the aftermath of natural disasters, the current trend may be confounded by demographic changes, availability of diagnosis, and actually follows the global resurgence of epidemic dengue disease.

P 6: 122-124. In terms of the aim, you state “assessing and comparing KAP…” whereas your main findings report on the association of KAP and sociodemographic factors (and history of having DF). Adding more specific objectives might be more appropriate.

Methods

P 7: 154 Based on what criteria did you classify inhabitants as healthy?

P.8-9: 185-204 With regards to measures used in the survey for dengue KAP, the distribution of questions into knowledge and attitude domains seems controversial. For example, it is not clear why you placed the statement “Do you think that stagnant water around the houses in discarded tyres, broken pots and bottles are breeding places of Aedes mosquitoes?” under attitude domain, and the question “Do tightly covering water tanks reduce mosquitos?” under knowledge domain?

P 9: 202 “Breeding places”: “Oviposition” seems to be a more favourable and accurate term than “breeding”.

P 10: 219-220 The data analysis method based on designating someone of having “good” or “bad” knowledge, attitudes and practices implies that each subcomponent is equally important which is questionable considering controversy in selection of questions. This makes further interpretation difficult and limit usefulness of the study for informing preventive interventions. Perhaps worthwhile examining if any specific knowledge or attitudes were related to specific actions? This might be more insightful.

Results

P 16: 293-294 The values of OR (95% CI) in the final model described in the text do not match the values of adjusted OR (95% CI) in Table 3.

P 18: 312-313 and P 19: 334-335 Same with OR (95% CI) values described in the text for attitudes and practices domains, they do not match with the Table 3 data. Also, specify what percentage of the variation do your final regression models explain.

P 21. I am not sure it is appropriate to use correlation analysis as your data distribution is not normal. Nevertheless, your correlation coefficient ranges from 0.16-0.34. Please mention the degree of correlation (poor, fair) instead of just stating “positive significant”.

Discussion and conclusion

P 27: 503-505 It is a useful observation that messaging should be targeted for the audience (hospitalized patients and family members), which is key to health promo campaigns.

P 28: 534-541 It is good that examples of evidence-based interventions (HCW in Puskesmas) are mentioned in the recommendation for future work.