Review Reports

Round 1

Reviewer 1 Report

 The authors of the article used the HFMD with temperature, rainfall, humidity and internet search patterns to forecast model that could inform preventive strategies to reduce the disease burden of HFMD. I have a few comments that may improve the readability of the paper.

1.    It is necessary to explain the specific form of the HFMD model and what factors are involved.

2.    The source of each data reference shall be marked such as Line 132 in page 3 and Table 1, and the data cannot be limited in Malaysia.

3.    Besides the temperature and other factors mentioned in this article, whether there are other factors that can affect HFMD.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Dear Sir/Ma'am,

This study and data provided is meticulous but diagnosing a self limiting and common disease like HMFD does not require such an algorithm 

The moment cases increase in OPD, we know an epidemic is impending. No treatment is required. So why do we need to study so extensively for a benign self limiting disease. neither it changes the disease course or prognosis. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Increased attention is being paid to HFMD because it is an infectious disease spread by a virus. Preventative measures in Sabah, Malaysia were the focus of this research, hence a forecasting tool was created using climate predictors and internet search queries. From 2010 to 2018, the authors used HFMD case data collected by the Sabah state health department, climate predictors developed by the Malaysia meteorological department, and Google search trends gleaned from the Google trends platform. A seasonal auto-regressive moving average (SARIMA) model was developed with external regressors guided by measuring model fit, and cross-correlations were computed. Then, test data was used to validate the selected variables with metrics like mean average percentage error (MAPE). Temperature showed weaker positive associations to HFMD instances than Google search trends did, and the strongest association occurred between 0 and 1 weeks. The best model was a SARIMA model that used regressors based on Google search patterns at lag 1 and average temperature at lag 0. Over the course of the four-week period, its forecasts were the most consistent, and it made the most accurate forecasts two weeks in advance. The model's trajectory forecasting oscillations are steady up to four weeks in advance, with the maximum accuracy two weeks beforehand, suggesting it may be beneficial in outbreak preparedness.

The keywords must be different from the title.

English can be improved.

Some box-whisker plots must be added for further clarification.

Future directions are missing.

The conclusion section should completely be revised. The current one has no discussion of the major findings in the manuscript.

After above changes, the manuscript can be resubmitted.

Author Response

Please see the attachment.

Author Response File: Author Response.docx