Improved Epidemic Dynamics Model and Its Prediction for COVID-19 in Italy
Round 1
Reviewer 1 Report
Summary
I read the manuscript with much interest. The research question addressed, it goes without saying, is extremely important at this point in time. I should mention at the outset that events have clearly overtaken us between the time this manuscript has been submitted and when I write this review, and I do not intend to judge the authors based on whether their predictions came true. Rather, I wish to focus on the methodology they have used, and the scientific contribution that they have made by presenting their methodology. Nevertheless, I urge the authors to revise their paper in such as a way that it matches the presently known information about the COVID 19 pandemic in Italy, and make their predictions from this point in time when they revise.
Overall, the manuscript gives me a sense that it was prepared and submitted in a rush, without due diligence. This is understandable but more care should be taken when revising. See below for examples.
I give my detailed comments below. Comments are offered in good faith. I recommend a major revision where the below mentioned issues are addressed with care.
Scope, Novelty and Focus
I am not going to thrash the paper with the aid of hindsight. I would rather like to focus on their methodology. Nevertheless, the fact is that Italy has reached 208,000 cases by early May, and it looks like it will get another 25,000 cases at least. So the authors have to think about re-writing their paper in light of this information, also keeping in mind that by the time the paper gets published, we will have even more information. So if the authors like to highlight the predictions in their paper, it has limited utility after the end of epidemic. It makes for awkward reading if someone reads it after say three months and the predictions are inaccurate. I would suggest therefore the authors put more emphasis on their modelling effort and highlight that as their contribution.
Technical issues and methodology
The derivation of equations seem technically sound, though I didn’t check rigorously.
Discussion: when you talk about inflection points, inflection point in which curve? incidence, prevalence, or cumulative incidence? Paper does not say. In fact, the words “incidence, prevalence, or cumulative incidence” hardly figure in the paper at all. Use these technical terms and explain clearly which curve you are talking about.
Appendix: why use casual terms such as “Newly confirmed cases, Cumulatively confirmed cases, Existing confirmed cases”? It is very awkward to read. This is a scientific paper: say incidence, cumulative incidence, and prevalence! (If that is what you mean)
Anyways, check the data for April 9th and 10th. How come your cumulative incidence is less than your prevalence in the last two rows? Also, “a cumulative number” can never drop from previous day to next day, yet it drops from 9th to 10th in your data!!! if your predictions were made by fitting this clearly innaccurate data, they were clearly compromised.
The proposed compartmental model seem reasonable otherwise.
Readability
The manuscript has several grammatical mistakes, unconventional uses of English, and careless typos. See some examples below:
University of New South Wakes University of New South Wales !!!
Abstract: SEIR_QJD Do not use abbreviations in abstract. Do not use abbreviations before you introduce them in text.
Do not use citations in abstract.
Currently confirmed cases How can ‘currently’ confirmed cases reach a peak in a future date? ‘current’ means ‘at the time of writing’. Anyway what do you mean by ‘currently confirmed cases’? Number of active cases or number of total cases?
The authors should keep in mind that computational / mathematical epidemiology is a well established field with very specific terminology. Throughout the manuscript the authors use terms loosely, as if this is the first paper in the field. The above is just one example of that. Please use established terms and phrases after understanding the convention of using them.
about 190,742 by the end of May This is only a prediction (which seems wrong in hindsight anyway), and the authors say “about”, so why give such a very specific number? Say “about 190,000”. It makes no sense to give a very specific number when it is only a prediction and you say “about”.
Keywords: “Novel Coronavirus Pneumonia”: Again, what is this? The virus is “SARS-CoV-2”. The Disease is COVID-19. The condition the disease causes, which has been responsible for the most deaths in Italy, is “Bilateral Interstitial Pneumonia”. This is an established medical term. “Novel Coronavirus Pneumonia” mixes up the virus and the condition. I have indeed seen researchers use this term, but I am not sure if it is a medical term. It may have been just made up casually and repeated. Please double check. I would recommend you use the professional keywords SARS-CoV-2, COVID-19, Bilateral Interstitial Pneumonia.
Infection individuals infected individuals
Severe disaster area Italy is not an area, it is a country. Please don’t write casually.
include the SIR model, SEIR model and SIJR Do not use abbreviations before you introduce them in text.
more complete infectious disease model is urgently need for scientific basis in severe epidemic regions grammatically wrong and unnecessarily verbose. Just say “more complete infectious disease model is urgently needed”
Many scholars have optimized the dynamic model Many scholars have adopted the dynamic model
therefore disobey the feature of COVID-19 therefore does not reflect the dynamics of COVID-19.
The word “feature” is misused throughout the manuscript. Change them all.
In turn lead in turn leads
Italy's medical level Italy’s medical system
Later in the text, several lines of text appear to be ‘superscript’ when it should be in-line with the level of other text. For examples, Rc0 = 2.4522, and many others like this.
individuals with asymptomatic individuals who are asymptomatic
These are only examples. Manuscript is riddled with them. It is not my role to list them or correct them all, especially when there are too many. Check carefully and get professional help if needed in writing.
Literature review / Background
As a journal paper, the literature review is quite limited. There is a paragraph thrown in which speak about the existing literature, almost as an afterthought. Also there is no mention at all about alternative methodologies which are or could be used for this purpose. All of this make the paper sound like it is presented somewhat out of context in scientific terms. I urge the authors to address this extensively in the revision.
As an aid, let me refer the authors to papers like:
“Modelling transmission and control of the COVID-19 pandemic in Australia” (Chang et al)
“Artificial Intelligence in the Battle against Coronavirus (COVID-19): A Survey and Future Research Directions” T. Nguyen
Which tackle the COVID-19 modelling problem,
And
Papers like
“Game theoretic modelling of infectious disease dynamics and intervention methods: a review” Chang et al
“Vaccination and epidemics in networked populations - An introduction” Wang et al
“Investigating Spatiotemporal Dynamics and Synchrony of Influenza Epidemics in Australia: An Agent-Based Modelling Approach” Cliff et al
“Urbanization affects peak timing, prevalence, and bimodality of influenza pandemics in Australia: Results of a census-calibrated model”, Zachreson et al
“The effects of imitation dynamics on vaccination behaviors in SIR-network model” Chang et al
Which present / use /review alternative modelling techniques to model and predict epidemics.
Including these and similar papers in literature discussion will give the reader a broader idea about the field, and give the paper a scholarly outlook which is lacking at present. This is very important to make the paper feel scholarly and put the work of the authors in context.
Figures and other issues
Figure 3: name the axes. Say 'plot' rather than curve. Explicitly say "death removal ...patients against [what is in the x-axis].
Figure 6: caption appears in next page!!!!
Recommendation
On balance, the manuscript is likely publishable and of interest to the readership of the journal but needs to undergo major revision and must include more background information and literature survey. Should a revision be offered, I would be happy to review it again.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Overall the manuscript is of High interest and seems well written.
I tend to prefer a past tense written scientific writing. This article shows some future tense about results that must already have happens which makes it confusing at certain moments.
some comments:
Abstract,
line 19: (1) there is a remarkable - I propose "(1) there was a remarkable"
lines 19-20: (3) the currently confirmed cases of Italy will reach the epidemic maximum around April 14 - been this revision occurring more than two months after the referred date this is a strange statement that should be confirmed or not. It seems to mean that the data at the time of manuscript writing and the proposed methodology point that the epidemic in Italy should have reached its maximum by April 14 which was (or was not) confirmed. this must be clarified.
Introduction,
line 45 - it seems that the notations SIR, SEIR and SIJR should have should have a description of its meaning when first introduced.
line73: a space is needed before "(IA)"
line 82: "The traditional SEIR [model] only divides the Population into four categories:" - the five categories are extensively showed this seems strange, is this correct?
line 121: there is double space
lines 139-140: there is an apparent unnecessary line break
Lines 172-175: these results relate to the "future tense" problem already pointed. I would like that these would be more conditional ("would" instead of "will" in both occurrences). some confirmation or discussion is expected at current dates.
lines 191: "Figure 6" instead of "Figure6"
Discussion
line 215: at this point starts the discussion of the predicted peak of the epidemic in Italy. I propose a conditional tense phrasing ("would" instead of "will"). As proposed above, further discussion in now necessary
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
I am satisfied that the authors have addressed my reviewer comments reasonably.
