Probability Risk of Heat- and Cold-Related Mortality to Temperature, Gender, and Age Using GAM Regression Analysis

Round 1

Reviewer 1 Report

This is an interesting paper and the analysis appears to be sound. 

There is no information about ethics approval, which is important when reporting about health data. In case you only got aggregated data from the data provider, please state that.

Also, report the source of the health data, please

The paper could use a rewriting for English language and clarity. e.g line 37, "Mortality risk has been..." missing some information about mortality risk in relation to what?

There is some uncertainty about the structuring of the text, e.g. lines 145-151 I think would be more appropriate in a discussion. 

Author Response

We want to thank the reviewer for the nice comments and indications for improvement.

The source of the health data has been included in the methodology. The data provided by the Health department of Cyprus were specifically divided into groups regarding the mortality cause but the age of the deceased was sub-grouped for every 5 years.

Reviewer 2 Report

Sir,

first at all thank you for the opportunity to review this interesting paper focusing on the Probability risk of heat- and cold-related mortality to temperature, gender and age in an eastern Mediterranean region (Nicosia, Cyprus).

The paper is properly written, and the authors have contributed to the ongoing debate on which analytical models are more appropriate for similar analyses by using a GAM model, that is both interesting and quite uncommon.

Still, the paper requires some minor improvements before being accepted and published on "Climate". Briefly:

1) Authors have gathered mortality data about a series of possible ICD codes (i.e. ischemic heart diseases (I20–I25), cerebrovascular diseases (I60-I69), other heart diseases (I30–I51), other circulatory diseases (I00–I15, I26–I28, I70–I99), influenza (J00–J99), pneumonia (J12–J18), chronic lower respiratory diseases (J40–J47), and other respiratory causes (J00–J06, J20–J39, J60–J99) but mortality date are presented only as cumulative ones. I warmly suggest to present, even without discerning between Males/Females/age groups, mortality data for diagnostic codes. Adding such analysis will allow the Authors to more properly discuss about the possible association between cold/heat effects and mortality.

2) please explain more extensively how you calculated DTR (I mean: have you simply calculated DTR=Tmax-Tmin ? or have you opted for another calculation?)

3) In general, the article seemly overlook the ability of the reader to understand the results of quite uncommon statistical analyses: for this reason, I warmly recommend the authors to reformulate caption for Figure 1 in order to better explain what they mean for "cross correlation" (for instance, I have assumed that with "cross correlation" authors meant a measure of similarity of two series as a function of the displacement of one relative to the other; was it correct? please explain), but also to discuss their results more extensively, reporting across the "Results" section at least the most important estimates for the analyses they performed, not forcing the reader to rely on the figures;

4) please, report the statistical software you employed: I think that you have employed a gam function from a R package similar to mgcv, but it should be more clearly reported

5) finally, while it is rather clear how you have collected meteorological data, the source for your clinical data remains unclear, and should be more clearly reported.

After such improvements, I think that the paper may be proper for a full publication on Climate.

Author Response

The authors thank the reviewer for the nice comments and the specific points indicated for improvement. We hope the changes on the corrected manuscript are as anticipated. Specifically:

• Even though the provided mortality data were divided according to the ICD codes, the small number of deceased of each group was not sufficient to be statistically valid to observe any significant relationships. Therefore, we aggregated the cardiovascular and respiratory causes of death and only used the Males/Females and age groups divisions to discuss about the association of cold/heat effects and mortality. Nevertheless, we discuss more thoroughly about the gender and age groups effects in the discussion section.
• The DTR was calculated as considered by the reviewer (DTR=Tmax-Tmin of the day). The calculation was included in the methodology section.
• The authors agree that the caption of figure 1 may not be fully understood by the readers, but the cross correlation explanation was added in the text. We also discussed the results more extensively in the results and discussion sections.
• Yes, the gam function was obtained by the R package mgcv. The source and reference is not added in the text.
• The source of the clinical data was only provided in the acknowledgements section and now the authors have also written the provider (Health Monitoring Unit of the Ministry of Health of Cyprus) in the methodology section as well.

Round 2

Reviewer 1 Report

The paper is barely understandable. It most be proofread for language.

Author Response

The authors have requested a proofread of the manuscript as recommended. we hope the revised version is as expected.