|Year : 2020 | Volume
| Issue : 3 | Page : 115-119
Thinking beyond lockdowns and social distancing: Identifying the predictors of spread for COVID-19 pandemic using ecological correlation study
Sunil Kumar Raina1, Mitasha Singh2
1 Department of Community Medicine, Dr. RP Government Medical College, Tanda, Himachal Pradesh, India
2 Department of Community Medicine, ESIC Medical College and Hospital, Faridabad, Haryana, India
|Date of Submission||19-May-2020|
|Date of Acceptance||20-Jun-2020|
|Date of Web Publication||09-Oct-2020|
Dr. Sunil Kumar Raina
Department of Community Medicine, Dr. RP Government Medical College, Tanda - 176 002, Himachal Pradesh
Source of Support: None, Conflict of Interest: None
Background: COVID-19 (SARSCoV2 or 2019 novel coronavirus [nCov]) pandemic has spread to every continent except Antarctica, and cases have been rising daily globally. However, COVID-19 is not just a health crisis. Aim: The present study was aimed to identify the predictors defining spread of the pandemic. Methodology: An ecological correlation study was conducted to identify the factors predicting the spread of COVID-19 (SARSCoV2 or 2019 nCov) pandemic worldwide. For this purpose, countries affected with Covid-19 worldwide from both the northern and southern hemispheres were included using a predefined inclusion criteria. Data from the selected countries were retrieved for the duration extending from January 1, 2020, to March 31, 2020. Results: A significant moderate positive correlation between cumulative Covid-19 cases and number of motor vehicles registered per 100 persons was observed in January 2020 (r = 0.623, P = 0.01) and March 2020 (r = 0.620, P = 0.01). Population density remained positively correlated with the cases of Covid-19. A strong significant correlation was observed in February (r = 0.746, P = 0.001). Increase in the length of highway road in countries was statistically significantly correlated with the increase of cases in the month of March 2020 (r = 0.644, P = 0.007). However, after accounting for all the variables, none of the variables could be identified as an independent predictor of cumulative cases except for time (in form of months). The R2 of the model was 41.4%. Conclusions: Urbanization, urban clustering, and population density may be the important contributors in spread; however, their role as independent predictor may be questionable. Therefore, it is time to think beyond social distancing and lockdowns and strengthen primary care and public health.
Keywords: COVID–19 pandemic, lockdowns, predictors, social distancing
|How to cite this article:|
Raina SK, Singh M. Thinking beyond lockdowns and social distancing: Identifying the predictors of spread for COVID-19 pandemic using ecological correlation study. Amrita J Med 2020;16:115-9
|How to cite this URL:|
Raina SK, Singh M. Thinking beyond lockdowns and social distancing: Identifying the predictors of spread for COVID-19 pandemic using ecological correlation study. Amrita J Med [serial online] 2020 [cited 2020 Oct 22];16:115-9. Available from: https://www.ajmonline.org.in/text.asp?2020/16/3/115/297557
| Introduction|| |
COVID-19 (SARSCoV-2 or 2019 novel coronavirus [nCov]) pandemic is defining global health crisis of our time since its emergence in Asia late last year. The virus has spread to every continent except Antarctica, and cases have been rising daily globally. However, COVID-19 is not just a health crisis, but by stretching every one of the countries, it touches, it has begun to create devastating social, economic, and political crises. People are losing jobs and income, with the International Labor Organization estimating that 195 million jobs could be lost.
Governments worldwide have been making all out efforts to slow the spread of the virus by testing and treating patients, carrying out contact tracing, limiting travel, quarantining citizens, and canceling large gatherings such as sporting events, concerts, and schools. However, it looks like that it may still take some time before we have some concrete evidence to test any of the drugs and vaccines for large scale use in preventing the spread of the disease. Therefore, researchers worldwide are looking at other predictors of this pandemic. One such predictor is temperature and humidity. Change in climate has been known to influence the emergence and re-emergence of many infectious disease as the disease agents, and their vectors each have particular environments that are optimal for growth, survival, transport, and dissemination.,,,, Furthermore, environmental factors such as precipitation, temperature, humidity, and ultraviolet radiation intensity have been known to be of importance in the transmission of diseases. Therefore, a deduction, though less well documented, but apparently plausible, some respiratory viruses prefer growing in the upper airways where the temperature is slightly lower than the core body temperatures. The current study was aimed to identify the predictors defining spread of the pandemic, and therefore, help nations prepare better. For this purpose, factors were not restricted to environmental predictors, but were empirically extended to demography and accessibility using the data currently available on the public domain.
| Methodology|| |
An ecological correlation study was conducted to identify the factors predicting the spread of COVID-19 (SARSCoV2 or 2019 nCov) pandemic worldwide. For this purpose countries affected with Covid 19 around the world from both northern and southern hemispheres were included using a predefined inclusion criteria.
- Countries with population more than 10 million
- Countries performing more than or equal to 147 Covid-19 tests daily.
Once the countries were identified using the defined criteria, 10 countries with the most number of cases from both the hemispheres were selected. However, only six countries from the southern hemisphere complied with the inclusion criteria of testing. The selected countries with their date of entry into Covid-19 pandemic is listed in [Figure 1].
|Figure 1: Countries including in the study along with the time of entry in the pandemic|
Click here to view
Data from the selected countries were retrieved for the duration extending from January 1, 2020, to March 31, 2020.
The cumulative cases from January 1, 2020, to March 31, 2020, for each country are taken as the dependent variable. The independent variables were: time in form of the month of entry in pandemic, climate (temperature, humidity, and precipitation) over 3 months, population, density, length of road highways, and number of vehicles registered per 1000 persons.
For each variable, public domain was accessed. The data on country-wise population, land area, population density, and proportion of the urban population search were obtained from the world population meter. The data on the kilometers of national road highways were obtained from the Google search for each country. The WHO's website was searched to look for the nCov (COVID-19) situation dashboard, and data on each country fulfilling the inclusion criteria were extracted. The selected countries' cumulative number of laboratory-conformed COVID cases from January 1, 2020, to March 31, 2020, was retrieved.
The words “Climate” was entered in the Google search engine for each of the selected countries and searched for the results. Out of the obtained Google searches, we selectively looked for the government authorized data and narrowed down our search to understand the trends in maximum and minimum temperature, humidity, and precipitation for the months of January, February, and March 2020.
Data and statistical analysis
The collected data were analyzed using the SPSS software version 21.0 (International Business Machines Corporation (IBM), Armonk, New York, USA). The independent variables for each country were divided into two hemispheres and presented as median and interquartile range (IQR) for each hemisphere. The climate data were also presented as the median for each hemisphere and month. The median cumulative cases of Covid-19 were presented for countries in both hemispheres at the end of the months. A bivariate correlation using the Pearson's correlation between dependent and independent variables was calculated. All the variables were subjected to the multiple linear regression models with cumulative cases for each country as the dependent variable. There was no outlier and autocorrelation in our regression data and the assumption for normality and homoscedasticity were met. The level of significance was set at 5%.
| Results|| |
Ten countries from the northern hemisphere and six from the southern hemisphere were selected based on the inclusion criteria. The median population and population density of countries in the northern hemisphere was higher as compared to southern. Southern hemisphere catered a higher median proportion of the urban population (85%). The median number of vehicles registered per 1000 persons was higher among the northern hemisphere countries (550 (IQR, 169.5)) as compared to the southern hemisphere (273 (IQR, 295)). The median number of cumulative Covid-19 cases was higher in the northern hemisphere and continued to increase from January to March [Figure 1].
A comparison of cumulative cases of Covid-19 from January to March 2020 in both the hemispheres with climate changes in 3 months is depicted in [Table 1]. The cases in both the hemispheres increased significantly over 3 months. The median cumulative cases on March 31, 2020, in the northern hemisphere (10,842.50 (IQR, 83,842.8)) was around ten times higher as compared to the southern hemisphere countries (1887.50 9(IQR, 3437.80)). The median maximum and minimum temperature increased in both hemispheres. However, the temperature in north continued to be lower in these 3 months as compared to south. The median relative humidity in the north hemisphere decreased significantly from 88.50% to 41.50% (P = 0.003), whereas the same increased in the southern hemisphere [Table 2]. However, temperature and humidity were negatively correlated with the cases in all 3 months [Table 3].
|Table 2: Climatic factors and coronavirus disease-19 cases in the north and south hemispheres from January 2020 to March 2020|
Click here to view
|Table 3: Bivariate correlation of cumulative cases with other independent variables|
Click here to view
A significant moderate positive correlation between cumulative Covid-19 cases and number of motor vehicles registered per 100 persons was observed in January 2020 (r = 0.623, P = 0.01) and March 2020 (r = 0.620, P = 0.01). Population density remained positively correlated with the cases of Covid 19. A strong significant correlation was observed in February (r = 0.746, P = 0.001). Increase in length of highway road in countries was significantly correlated with the increase of cases in the month of March 2020 (r = 0.644, P = 0.007).
All the independent variables were subjected to the linear regression model with cumulative cases as the dependent variable. After accounting for all the variables, none of the variables could be identified as an independent predictor of cumulative cases except for time (in form of months). The R2 of the model was 41.4% [Table 4].
|Table 4: Linear regression model with cumulative cases as the dependent variable|
Click here to view
| Discussion|| |
The epidemiologic triad (agent, host, and environmental factors) have been seen to provide inspiration for adding these as the parameters in addition to the availability of public health resources to study health-care problems. Various aspects of the built environment and other factors have been known to influence the transmission dynamics in infectious diseases. Therefore, it is more expected than not that these factors are considered relevant for the estimation of basic reproduction number (R0) of the infection. The current pandemic of COVID–19 (SARSCoV2 or 2019 nCov) is not an exception. The current study was conducted with the aim to study the role of various environmental and demographic factors on this pandemic.
The study based on the analysis of the data obtained for the past 3 months for selected countries reveals that high temperature and humidity were correlated with the reduced number of cases. Southern hemisphere, which witnessed summers from January to March 2020, showed lesser number of cases of Covid-19. This observation has been reported by some previous studies as well. According to a study by Wang et al., a rise in temperature decreases R0 for Covid-19. Lowen et al. in their animal experiment on influenza virus demonstrated that temperature more than 30 degrees blocks aerosol transmission but not contact transmission. However, it appears it is too early to draw the conclusions from these studies. Our study was unable to establish the temperature and humidity as the independent predicator of Covid-19 pandemic. Importantly, we still have to witness the natural history of novel coronavirus when the temperature, humidity, and precipitation reverse in the north and south hemisphere.
The transmission dynamics may also be dependent on the population parameters of a country. For the purpose of this study, some population parameters population density were included. It was observed that countries with high-population density demonstrated a higher number of Covid-19 cases. However, the countries included in the current analysis had a clear cut demarcation as countries in the northern hemisphere were observed to have high population density as compared to the south. A strong correlation with population density was observed in the month of February 2020. One of the probable reasons could be increase in testing during the month of February across countries. February also saw immigration and emigration of population across countries. It was hypothesized that the majority of cases were reported from the urban areas. Hence, we included the variable proportion of the urban population in countries in our analysis. Southern hemisphere was observed to have high proportion of the urban population. Higher urban population appeared to have higher number of cases. However, when other variables were taken into account, a negligible correlation was observed.
To account for migration and the movement of population in a country, the length of national highways and number of vehicles registered per 1000 population were included. Both were on higher side in the northern hemisphere. Higher number of vehicles was significantly correlated with higher number of cases in January and March 2020. Taking other variables in account, it did not emerge as an independent predictor. These are the indirect measures of social interactions. The correlation of urbanization, urban clustering, and population density with the number of cases appears to be the driving force behind public health interventions such as social distancing and lock downs. There apparent benefits on slowing the speed of epidemic may be because of their impact on urbanization and urban clustering. However, their role as the independent predictors of the spread of the pandemic may still be questionable.
The R2 of our regression is about 41%, which means that 59% of change of cumulative cases over time cannot be explained alone by climate, population factors, and travel factors. Therefore, integrating approaches focusing on the cultural and social behaviors of population coupled with strengthening of the primary health-care system of a country may have a long-term impact.
| Conclusions|| |
Urbanization, urban clustering, and population density may be important contributors in spread; however, their role as independent predictor may be questionable. Therefore, it is time to think beyond social distancing and lockdowns and strengthen primary care and public health.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
IPCC. Climate Change 2014 Synthesis Report Summary for Policy Makers. Geneva: Inter-Governmental Panel on Climate Change; 2014.
Field CB, Barros VR, Dokken DJ, Mach KJ, Mastrandrea MD, Bilir TE, et al
. Summary for policymakers. In: Climate Change 2014: Impacts, Adaptation, and Vulnerability. Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press: IPCC; 2014. p. 1-32.
Cascio A, Bosilkovski M, Rodriguez-Morales AJ, Pappas G. The socio-ecology of zoonotic infections. Clin Microbiol 2011;17:336-42.
National Research Council (US) Committee on Climate, Ecosystems, Infectious Diseases, and Human Health, Under the Weather: Climate, Ecosystems, and Infectious Disease. Washington (DC): National Academies Press (US); 2001.
Wang J, Tang K, Feng K, Weifeng LV. High Temperature and High Humidity Reduce the Transmission of COVID-19; 09 March, 2020. Available from: https://ssrn.com/abstract=3551767
. [Last accessed on 2020 Apr 16].
Lowen AC, Steel J, Mubareka S, Palese P. High temperature (30 degrees C) blocks aerosol but not contact transmission of influenza virus. J Virol 2008;82:5650-2.
[Table 1], [Table 2], [Table 3], [Table 4]