Spatial Distribution of COVID-19 Infected Cases in Kelantan, Malaysia

Round 1

Reviewer 1 Report

Based on the article received, I feel that the manuscript could be reconsidered for publication after considering all major revisions attached below: 

1. The title should be revised. The current title is very narrow and looks awkward.

2. The abstract needs to be revised substantially. There is no indication of results and recommendations in the abstract.

3. Please revise the keywords as well.

4. In the Introduction section, please clarify the justification of this research. The problem statement is not well organized and not clear for the reader. The introduction section is too short, and there is no methodological explanation and reasons for choosing the logistic regression. Also, there is a considerable gap between research novelty and the discussion of existing studies. Please make it elaborate as much as you can. Substantial changes and revisions are required in the introduction section. Please make a story and try to find out the importance of this research. You explained a lot about the Chinese scenario. Better to explain methodological novelty to examine the impact of population density on the number of COVID-19 infected cases.  

5.  Please add research questions before the research objectives.

6.  Please add ten districts' COVID-19 scenarios over the period.

7. The justification for selecting the OLS model is not properly addressed in the Materials and Methods section.

9. I am wondering about the sample points. The number of sample points is negligible (10 points). It would be better if you could collect at least subdistrict-wise points to run the regression model.

10. Please add a validation table for your LULC map preparation using UA, PA, and kappa coefficient techniques.

11. In the data analysis part, better to write this paper model instead of the general OLS model.

12. The theoretical background of the OLS model is also essential for the validation of selecting this model for this research.

13. It is also recommended that the infected area density map should be validated by considering some ground reference points.

14.  Please add an overall discussion section based on your key findings.

15. Please revise the conclusion section. In the Conclusion section, please remind the readers about your objective at the beginning of the conclusion section and conclude your result in light of your objective. Please add the conclusion section carefully and make a recommendation based on your research findings.

16. Please mention the limitation of this study and provide what could be done in the future in the conclusion section.

17. Please revise the reference style and format.

 

Finally, I would like to say that the research topic is important in the present context. Research ideas are also good, but you should focus on the consistency of your writing.

Author Response

Reviewer’s Response 1

Based on the article received, I feel that the manuscript could be reconsidered for publication after considering all major revisions attached below: 

1. The title should be revised. The current title is very narrow and looks awkward.

Thank you very much for the comment. The title has been changed to “Spatial distribution of Covid-19 Infected Cases in Kelantan, Malaysia”.

2. The abstract needs to be revised substantially. There is no indication of results and recommendations in the abstract.

Thank you very much for the comment. The sentence indicates the results of the study that has been added in the abstract.

“This study shows that the highest value of infected area density is in Kota Bharu (0.76) while the infected risk area was highest in Jeli (0.33). This study found that there is a strong relationship between COVID-19 infection cases in Kelantan and population density (R² which is 0.845).”

We also have added a recommendations sentence in the abstract:

“Understanding the potential drivers of the disease in a local setting is very important for better preparation and management”.

3. Please revise the keywords as well.

The keywords have been revised to:

Spatial distribution; COVID-19 model map; population density; infected cases; Kelantan, Malaysia

4. In the Introduction section, please clarify the justification of this research. The problem statement is not well organized and not clear for the reader. The introduction section is too short, and there is no methodological explanation and reasons for choosing the logistic regression. Also, there is a considerable gap between research novelty and the discussion of existing studies. Please make it elaborate as much as you can. Substantial changes and revisions are required in the introduction section. Please make a story and try to find out the importance of this research. You explained a lot about the Chinese scenario. Better to explain methodological novelty to examine the impact of population density on the number of COVID-19 infected cases.  

 

We have added the paragraph in the introduction section stating on justification of the research, problem statement, methods, and Kelantan scenario. Here are the sentences:

 

 

According to recent research (Hatta et al., 2021), Kelantan has 166 cases of COVID-19 detected locally as of the end of July 2020. The capital of Kota Bharu (56.6%) had the largest population density among the other 10 districts in Kelantan, where most of the cases were located. Eighty percent of the cases in Kelantan were related to the mass gathering at Masjid Seri Petaling (MSP), which occurred from February 27 to 29, 2020. The confluence produced the Seri Petaling cluster, the biggest cluster in Malaysia to date, which affected 3,375 cases nationally (38.9% of all cases as of July 8, 2020) between March 11, 2020, and July 8, 2020 [6]. In this study, 54 (32.5%) cases involved gathering attendees, whereas 79 (47.6%) cases involved their close connections, whose transmission could be tracked up to the fifth generation. 10.8% of the additional cases were contracted while travelling to affected nations. By using active case detection, more than 57% of cases were found. 50.0% of people were screened based on contact tracing, and 45.2% were based on risk assessment. Household connections made up around 67.4% of confirmed patients' contacts; the remaining contacts were social. The case fatality rate for COVID-19 in Kelantan is comparable to China (2.3%) but lower than Italy (7.7%), Europe (4.2%), and Asia in general (3.8%) (Boehmer et al., 2020). The incidence of COVID-19 in Kelantan is lower than the United States (316/100,000 in July) (Boehmer et al., 2020), and its case fatality rate is comparable to China (2.3%) (Wu & McGoogan, 2020; Bhagavathula et al., 2020).

Various protocols were  then  developed  to  facilitate   systematic   and   rapid   data   collections   as   well   as analyses   across   the   globe   to   provide   a   better comprehension of the disease and subsequently guide the   implementation   of   various   measures   (WHO, 2020). 

 

However, epidemiological studies should be cautiously interpreted as not only the disease is very dynamic but its characteristic is heavily influenced by different factors of local population and implementation of  public  health  measures  that  vary from country to country (Lai et al., 2020). Hence, it is important to  understand the impact of population density as the  potential driver of epidemiological features  of  COVID-19  at  the  local  level  towards  a better understanding of the disease to be prepared for impending outbreaks with specific countermeasures.

 

 

5. Please add research questions before the research objectives.

The research questions have been added:

Is the population density impacting the total COVID-19 cases? and what is the relationship between population density and COVID-19 cases?

6. Please add ten districts' COVID-19 scenarios over the period.

 

In Malaysia, the first case of COVID-19 was reported on January 25, 2020 [5]. According to a recent research (Hatta et al., 2021), Kelantan has 166 cases of COVID-19 detected locally as of the end of July 2020. The capital of Kota Bharu (56.6%) had the largest population density among the other 10 districts in Kelantan, where the majority of the cases were located. Eighty percent of the cases in Kelantan were related to the mass gathering at Masjid Seri Petaling (MSP), which occurred from February 27 to 29, 2020. The confluence produced the Seri Petaling cluster, the biggest cluster in Malaysia to date, which affected 3,375 cases nationally (38.9% of all cases as of July 8, 2020) between March 11, 2020, and July 8, 2020 [6].

                                                                                                    

7. The justification for selecting the OLS model is not properly addressed in the Materials and Methods section..

Linear regression model was used in this study to determine the relationship between population density and COVID-19 cases. We have added a few justifications for selecting OLS model in the Materials and Method section: Data analysis: Linear regression.

“Ordinary Least Squares (OLS) method is a linear regression technique that is used to find the best fitted line for a dataset in which the dependent and independent variables have a linear relationship. The method relies on minimizing the sum of squared error between the actual and predicted values. Since it is simple to train and understand, it is one of the most widely used techniques for estimating an unknown parameter in a linear regression model (Dismuke and Lindrooth, 2006).”

8. I am wondering about the sample points. The number of sample points is negligible (10 points). It would be better if you could collect at least subdistrict-wise points to run the regression model.

Although appropriate sample size is required for validity of regression model, some researchers say that there should be at least 10 observations per variable. Since we only have one independent variable, it passed the minimum requirement. However, we will consider to get the subdistrict data in future research to produce a more reliable model.

9. Please add a validation table for your LULC map preparation using UA, PA, and kappa coefficient techniques.

Thank you very much for the suggestion. Here we provide the table of our LULC map preparation.

Table 1: The overall accuracy and Kappa coefficient of land use types

Name	Totals	Totals	Correct	Accuracy	Accuracy
Class	Reference	Classified	Number	Producers	Users
built-up area	24	23	19	79.17%	82.61%
cloud	9	3	3	33.33%	100.00%
commercial   agriculture	49	50	45	91.84%	90.00%
forest	100	118	98	98.00%	83.05%
others agriculture	32	33	28	87.50%	84.85%
paddy	16	14	14	87.50%	100.00%
swamp forest	15	9	9	60.00%	100.00%
waterbody	11	6	6	54.55%	100.00%
Total	256	256	222
Overall Classification Accuracy =     86.72%

10. In the data analysis part, better to write this paper model instead of the general OLS model.

The spatial COVID-19 model map has been addressed in the result and discussion section.

Spatial COVID-19 integrated model map has been developed using spatial statistics tools by inserted linear regression value in attribute table. Figure 3 shows the land use in Kelantan in the year 2020. Infected area density map (total infected/population density) shows the highest value of infected area density in Kota Bharu (0.76) compared to Gua Musang (0) (Figure 4). However, infected area risk map (Infected density/population) was highest in Jeli (0.33) compared to Tumpat district (0.02) (Figure 5). Spatial analysis has shown the spread of COVID-19 in areas with high population density. It is supported by research from (Urban & Nakada, 2021), that shown that the high population density found in informal settlements in the city of São Paulo is a determining factor influencing the spread of COVID-19, possibly because people living in socioeconomic vulnerability may not be able to adhere to social distancing measures.

11. The theoretical background of the OLS model is also essential for the validation of selecting this model for this research.

We have added more details about OLS model in the Materials and Methods section.

Ordinary Least Squares (OLS) method is a linear regression technique that is used to find the best fitted line for a dataset in which the dependent and independent variables have a linear relationship. The method relies on minimizing the sum of squared error between the actual and predicted values. Since it is simple to train and understand, it is one of the most widely used techniques for estimating an unknown parameter in a linear regression model.

12. It is also recommended that the infected area density map should be validated by considering some ground reference points.

Thank you very much for the suggestion. The infected area density map is not validated by ground reference points because the data information of total infected and population density that we used in the formula (total infected/population density) is the authorized data from Ministry of Health Malaysia and Department of Statistic Malaysia respectively. We have added the reference of Ministry of Health Malaysia and Department of Statistic Malaysia in the method section.

13. Please add an overall discussion section based on your key findings.

Our results are consistent with the impact of population density on COVID-19 cases reported by Bhadra et al., 2021). Population density has been proven to be a determining variable influencing the spread of COVID-19, corroborating data presented by Ganasegera et al., 2021. However, the risk infected cases map shown that Jeli district (the lowest population density among the district) might have higher risk to infected by the COVID-19, possibly due to connection between other district and interaction. This study could have important public health implications and therefore, strategic planning should be taken by the decision makers and health care in order to decrease the COVID-19 cases and other spread diseases.

14. Please revise the conclusion section. In the Conclusion section, please remind the readers about your objective at the beginning of the conclusion section and conclude your result in light of your objective. Please add the conclusion section carefully and make a recommendation based on your research findings.

In this study, we have used spatial modelling to assess the spread of COVID-19 in 10 districts in the state of Kelantan, Malaysia. This research provides the model of COVID-19 infection cases and population density using linear regression. Linear re-gression model, using OLS, has proven to represent the correlation of population den-sity and infected cases at the district level. The results found that there is a strong rela-tionship between COVID-19 infected cases in Kelantan with population density. When population density increases, infection cases also increase. More importantly, our study has revealed a significant correlation between COVID-19 cases and population density (R² = 0.845). Our findings have shown that the high population density found in Kota Bharu is a determining factor influencing the spread of COVID-19, possibly because people living in socioeconomic vulnerability may not be able to adhere to social dis-tancing measures. Our study corroborates reports from recent literature, pointing to the need for special attention in high population density and also low population density. Therefore, decision makers and stakeholders can plan the intervention and strategic planning such as lockdown or vaccination program at the specific location of high population density to combat the COVID-19.

15. Please mention the limitation of this study and provide what could be done in the future in the conclusion section.

In Kelantan, Malaysia, the spread of COVID-19 cases is affected by population density. Other factors of different scopes can be further studied to find the difference between other factors and potential drivers that influence the spread of pandemic. Prior to using nontherapeutic interventions in the fight to control the epidemic, it would be desirable to create standard operating procedures that take population density into account as a risk factor for COVID-19 spread and assess them geographically. Thus, this study is essential to improve our understanding of the impact of population density of COVID-19 cases, and develop more focused and effective strategies to combat pandemic diseases.

16. Please revise the reference style and format.

Thank you very much for the comment. The reference style and format have been revised.

Finally, I would like to say that the research topic is important in the present context. Research ideas are also good, but you should focus on the consistency of your writing.

Based on the reviewer comments and suggestions, the manuscript has clearly been revised.

Reviewer 2 Report

In the current manuscript titled “Impact of Population Density on Total Covid-19 Infected Cases in Kelantan”, authors present data suggesting the impact of high population density as a potential driver of transmission of COVID- 19. My comments are below.

Some of the nomenclature used are not defined. For example -Line 200- Units of population density used for measurements should be mentioned. Figure 3-Meaning of Built-up-area should be clearly defined.

Figure 5 legend -Infected density/population. I think number infection count per square kilometer is more straight forward and meaningful way to represent this data. Also, what is the meaning of “Risk Map” if author is quantifying Infected density/population.

 Two maps side by side one showing infected population per square kilometers and another showing Covid positive cases per square km or merge of these two parameters will be helpful.

Household crowding parameter should be included while evaluating effect of population on Covid-19

Rational behind presenting data certain data set should be discussed in the results.

Author Response

Reviewer’s response 2

In the current manuscript titled “Impact of Population Density on Total Covid-19 Infected Cases in Kelantan”, authors present data suggesting the impact of high population density as a potential driver of transmission of COVID- 19. My comments are below.

Some of the nomenclature used are not defined. For example -Line 200- Units of population density used for measurements should be mentioned. Figure 3-Meaning of Built-up-area should be clearly defined.

Thank you very much for the comment. We now have mentioned the unit of population density and meaning of built up area. The sentences are:

Population density was calculated as the number of inhabitants living in an area per kilometer square (inhabitants/km2) for each district. Population density is the average number of people per unit of land area. In the U.S., the most common unit of population density is persons per square mile. The land area measurement excludes lakes and other water areas within (Ganasegeran et al., 2021).

We define built-up areas as the aerial units recording the full or partial presence of buildings and the space in-between buildings (Tenerelli & Ehrlich, 2011; Sabo et al., 2018)

Figure 5 legend -Infected density/population. I think number infection count per square kilometer is more straight forward and meaningful way to represent this data. Also, what is the meaning of “Risk Map” if author is quantifying Infected density/population.

Thank you very much for the question. Agree that infected area density map (total infected/population density) is the number of infections count per square kilometer. However, for infected area risk map (Infected density/population) has been developed by using spatial statistics tools and inserting linear regression value in attribute table to produce spatial COVID-19 integrated model map. The linear regression value provides the prediction data on the area that could be risked for spreading COVID-19. Therefore, the map is called as infected area risk map.

Two maps side by side one showing infected population per square kilometers and another showing Covid positive cases per square km or merge of these two parameters will be helpful.

Thank you very much for the suggestions. Two maps showing infected population per square kilometers and another showing COVID-19 positive cases per square km have been placed side by side.

Household crowding parameter should be included while evaluating effect of population on Covid-19

Rational behind presenting data certain data set should be discussed in the results.

Thank you very much for the suggestion. We have added the sentence that supported our results in the discussion part.

Rader et al 2020 found that population aggregation and heterogeneity have a significant impact on the COVID-19 epidemic's peaking degree. As a result, epidemics in densely populated areas spread more slowly over time and have higher overall attack rates than those in less densely populated areas. The observed variations in epidemic peaking are consistent with a COVID-19 meta-population model that explicitly takes into consideration geographical hierarchies. The finding is consistent with our result that epidemics may last longer in congested cities throughout the world (Rader et al., 2020; Sahin, 2020).

Reviewer 3 Report

The research from Amal Najihah Muhamad Nor et al. demonstrated a positive correlation between population density and COVID19 infection based on the Kelantan area in Malaysia. However, more factors can be addressed for better-addressed modeling.

1. Authors mentioned that wind speed, humidity, and temperature are also factors associated with COVID19 cases, it will be more convincing if climate information within the state is provided to exclude interference introduced by the non-population factors.

 

2. Population mobility is a factor driving population density and COVID19 transmission among the districts. Authors should involve it in the discussion.

 

3. As for the COVID19 infection data, looks like the authors only obtained a total of infected cases from two hospitals (Hospital Perempuan Zainab II Kota Bharu and Hospital Sultan Ismail Petra, Kuala Krai). How do authors exclude bias introduced by case coverage from the two regional hospitals? Will it be more accurate by including more data from local clinics?

 

 

 

Author Response

Reviewer’s response 3

The research from Amal Najihah Muhamad Nor et al. demonstrated a positive correlation between population density and COVID19 infection based on the Kelantan area in Malaysia. However, more factors can be addressed for better-addressed modeling.

 

1. Authors mentioned that wind speed, humidity, and temperature are also factors associated with COVID19 cases, it will be more convincing if climate information within the state is provided to exclude interference introduced by the non-population factors.

Thank you very much for the suggestion.

Climate information has been added in the study area, method section. The sentences are:

According to Faizalhakim et al (2017), major significant changes in climate were detected in Kelantan state i.e. annual rainfall by 41.13 mm, annual raindays (1.58 days), temperature changes (0.07°C), global radiation (0.17 MJm-2), atmospheric pressure (0.07 hPa ), cloud cover (< 0.01 oktas), annual evaporation by -0.50 mm, relative humidity (-0.25%), sunshine hour (-0.04 hour) and wind speed (-0.02 m/s year-1 ).

 

2. Population mobility is a factor driving population density and COVID19 transmission among the districts. Authors should involve it in the discussion.

Thank you very much for the suggestion. Agree that population mobility is a factor driving population density and COVID19 transmission among the districts. Therefore, we have added the sentences in the discussion. The sentences are:

We show that population density can be important factors in transmission but only in the absence of mobility-restricting policy measures, although particularly strong policy measures may be required to mitigate the highest population densities (Smith et al., 2021). Souch et al., 2021 found that, during the pandemic, cases appeared in more rural locations and involved transportation and mobility concerns. A lot of rural places turned into national hubs for prevalence and transmission. COVID-19 was transported along the route of interstates. Quantitative findings showed that COVID-19's anticipated arrival times varied across rural and urban areas. Counties that are crossed by interstates arrived before counties that are not. The biggest arrival time discrepancy was seen in the most rural counties, which suggests that road traffic is a factor in the spread of disease into rural areas. Road travel enabled human mobility, bringing COVID-19 to more rural areas. Interstate and road traffic limitations would have aided more effective mitigation efforts during the early COVID-19 epidemic stages and decreased transmission. This study supported our finding that the rural area, Jeli district could be risk of infected cased because of population mobility. These observations can provide insights for how social gatherings and communal events might enable rural communities, while having a lower population density, to become hotspots similar to their urban counterparts. Therefore, that rural areas as the infected risk area require an effective strategy and should also protected from the spread of COVID-19.

 

3. As for the COVID19 infection data, looks like the authors only obtained a total of infected cases from two hospitals (Hospital Perempuan Zainab II Kota Bharu and Hospital Sultan Ismail Petra, Kuala Krai). How do authors exclude bias introduced by case coverage from the two regional hospitals? Will it be more accurate by including more data from local clinics?

 

Thank you very much for the comment. The information of the total of infected cases from two hospitals (Hospital Perempuan Zainab II Kota Bharu and Hospital Sultan Ismail Petra, Kuala Krai) was authorized by the Ministry of Health Malaysia and have been exposed to the public. Therefore, this information are the valid information and have been mentioned accurately.

 

Round 2

Reviewer 1 Report

Thank you so much for the revision. 

Reviewer 3 Report

My concerns are well addressed by the revised version. Thank you.