key: cord-1048136-zym0kctc
authors: Asem, Noha; Ramadan, Ahmed; Hassany, Mohamed; Ghazy, Ramy Mohamed; Abdallah, Mohamed; Gamal, Eman M.; Hassan, Shaimaa; Kamal, Nehal; Ibrahim, Mohamed; Zaid, Hala
title: Pattern of COVID-19 infection and death across countries: A pilot study
date: 2021-07-08
journal: Heliyon
DOI: 10.1016/j.heliyon.2021.e07504
sha: 5e0918f0ae44d055d77429da24459cf6f99926eb
doc_id: 1048136
cord_uid: zym0kctc

BACKGROUND: This work aimed to identify the mathematical model and ecological pattern of COVID-19 infection and mortality across different countries during the first six months of the pandemic. METHODOLOGY: In this pilot study, authors used the online available data sources of randomly selected 18 countries to collect ecological predictors of COVID-19 transmissibility and mortality. The studied determinants were environmental factors (daily average temperature, daily humidity) and socioeconomic attributes (population age structure, count and density, human development index, per capita income (PCI), gross domestic product, internet coverage, mobility trends, chronic diseases). Researchers used the linear and exponential time series analysis, and further utilize multivariate techniques to explain the variance in the monthly exponential growth rates of new cases and deaths. RESULT: In the first two months, the R(2) of linear models for the cases and deaths were higher than that of the corresponding R(2) of the exponential model. Later one, R(2) of the exponential model was occasionally relatively higher than that of the linear models. The exponential growth rate of new cases was significantly associated with mobility trends (β=0.00398, P=0.002), temperature (β=0.000679, P=0.011), humidity (β=0.000249, P=0.000), and the proportion of patients aged above 65 years (β=-0.000959, P=0.012). Similarly, the exponential growth rate of deaths was significantly associated with mobility trends (β=0.0027, P=0.049), temperature (β=0.0014, P<0.001), humidity (β=-0.0026, P=0.000), and PCI of countries. During this period, COVID-19 transmissibility was evident to be controlled as soon as social mobility is decreased by about 40% of the baseline over 3 months controlling for the other predictors. CONCLUSION: Controlling of COVID-19 pandemic is based mainly on controlling social mobility. Role of environmental determinants like temperature and humidity was well noticed on disease fatality and transmissibility. COVID-19 contagiousness and fatality were additionally affected by human modifiable risk factors like income and non-modifiable risk factors (ageing) affected.

disease fatality and transmissibility. COVID-19 contagiousness and fatality were additionally affected by human modifiable risk factors like income and non-modifiable risk factors (ageing) affected.

The World Health Organization (WHO) estimates that approximately one-third (e.g., 20 million) of the annual deaths worldwide were attributed to infectious diseases. Furthermore, three of the 10-topped causes of deaths are lower respiratory tract infection, tuberculosis, and diarrheal disease, and many of these diseases can be prevented or treated for as little as one dollar for a head. (Word Health Organization, 2020) The morbidity from infectious disease has increased during the past few decades and represents at least 70% of emerging infectious diseases (EID), which are a significant burden on global economic and public health. (Dye, 2014; Morens, Folkers, & Fauci, 2004) Emerging infectious diseases (EIDs) cause a substantial economic and public health burden in the world. (Binder, Levitt, Sacks, & Hughes, 1999; Halliday et al., 2017) The most likely causes of the emergence of EIDs are socioeconomic, environmental, and ecological factors. (Binder et al., 1999; Dye, 2014; Halliday et al., 2017; Woolhouse & Gowtage-Sequeria, 2005) It is postulated that the origins of EIDs are significantly correlated with socioeconomic, environmental, and ecological factors that provide a clue for identifying regions where new EIDs are most likely to originate. (Woolhouse & Gowtage-Sequeria, 2005) These factors also present a basis of risk for wildlife zoonotic and vector-borne EIDs originating at lower latitudes, where reporting effort is low (Jones et al., 2008; Woolhouse & Gowtage-Sequeria, 2005) .

One of these EIDs is the novel coronavirus (COVID-19) that was firstly identified in December 2019 in Wuhan city, in China. It resulted in unusual pneumonia after visiting an animal market that sells poultry, fish ,and other animals to the public (Xu et al., 2020) . This outbreak was J o u r n a l P r e -p r o o f soon reported to the WHO. Due to this pandemic, the entire life has been changed. Millions of people have been infected (179, 598, 446) , and hundreds of thousands have been deceased (3, 890, 035) by the novel coronavirus.(Worldmeter, 2021)Countries adopted different strategies to combat the spread of this pandemic. Many countries implemented the national lockdown policy that extended for different duration during the daytime. Another containment strategy was social distancing and panning going out of homes without wearing facemasks. International travel bans from and to infected countries was another effective strategy used to face the spread of this pandemic. (Cheng et al., 2020; Movsisyan et al., 2021; World Health Organization, 2020) It is worthy to note that COVID-19 does not affect everyone in the same way, furthermore, it is not easy to understand the consequences or to predict how this pandemic affects differently various countries. There are several reasons could explain why different population are affected by this pandemic in different ways. These conditions can include socioeconomic factors (income, population density, distribution of human population, urban and rural settings, education level, and lifestyle, the size of household, and homeowners & tenants), behavior factors (direct contact with domestic and wild animals, migration, social interactions), and environment factors (humidity, temperature, wind spread, climate change, deforestation, agricultural growth) (Barratt, Shaban, & Gilbert, 2019; Dehghani & Kassiri, 2020) Prior studies reported several epidemiological factors that are associated with coronavirus transmission and or the test positivity using a regression statistical model of the online available data to predict the country-level conditions related to the COVID-19 pandemic. (Karmakar, Lantz, & Tipirneni, 2021; Kathe & Wani, 2020) In March 2020, using data retrieved from John Hopkin's Hospital database for the United states of America (USA) and all reported infected countries, it was hypothesized that temperature and humidity have a certain effect on the number of COVID-J o u r n a l P r e -p r o o f

Considering that a total of 193 countries are members of the United Nation and 15 countries reported 0 incidences of COVID-19 by the end of June 2019. A pilot sample of 18 countries (10%) was selected randomly using a simple random sampling technique. We applied the random generation function of Microsoft Excel to select randomly the 18 studied countries. The selected countries were Australia, Brazil, China, Canada, Egypt, France, Germany, Indonesia, Iran, Italy, Japan, Lebanon, Norway, Oman, Spain, Sweden, United Kingdom, and United States of America (USA). Of note, the selected countries represented the 6 major continents, had different climates, socioeconomic status, population densities, and numbers of confirmed COVID-19 cases and deaths.

Through the period from 1 st of January to 30 th of June 2020, the daily counts of COVID-19 confirmed cases and deaths, and socioeconomic attributes for the 18 selected countries were collected from Our Word in Data website (Our World in Data, 2021) . Sociodemographic attributes included the population density per Km 2 , human development index (HDI), per capita income (PCI), gross domestic product (GDP) in (USD), internet coverage, population count, and proportion of the population aged 65 years or above, and proportion of the population suffering from chronic diseases i.e diabetes mellitus (DM). The daily average temperature and relative humidity of the capital cities of the selected countries were collected from the Weather and Iran), google daily documentation on mobility trends across different categories of places such as retail and recreation, groceries and pharmacies, parks, transit stations, workplaces, and residential was retrieved. Google documentation was presented by percent change in mobility across the mentioned places compared to the baseline, which is the median value of mobility trends for the corresponding day of the week, during 5 weeks from Jan 3 -Feb 6, 2020 (Community Mobility Reports, 2021).

Data analysis was conducted using Statistical Package for Social Science (SPSS) software (version 25, Chicago, USA). The Daily counts of confirmed cases and deaths for each country were split by month and were further linearly and exponentially modeled with time. The linear equation was based on the equation (y1 = a +βx1), where y is the confirmed cases or deaths for a given day of this month, the constant (a) is the number of cases at day 0 (the count before the start of that month), β is the linear slope coefficient which is presenting the average daily increase in the count of confirmed cases or deaths within this month, and x is the day number of that month.

On the other hand, the exponential equation is based on the equation (y1= a e βx1 ), where e β is the average daily multiplying factor for increase or in other words, the exponential growth rate in each month. The daily average temperatures of the capital cities were aggregated to the monthly temperature for each country. Google documentation was handled using big data analytics with R software (R Foundation for Statistical Computing, 2021) to aggregate the data into the monthly mean percent change from the baseline across six categories of places for each of the selected countries.

J o u r n a l P r e -p r o o f 8 The linear and exponential slope coefficients hereinafter referred to as linear growth rate and exponential growth rate, respectively. Non-parametric correlation between variables was assessed using Spearman rho's test. The monthly exponential growth rates of confirmed cases and deaths were regressed versus the collected predictors for each country using mixed-effects model with unstructured covariance. The highly correlated variables were subjected to principal component analysis (PCA) with varimax rotation, and the extracted PCs were utilized for the regression models.

This study was approved by the Ethical Committee of the Ministry of Health and Population, Egypt.

J o u r n a l P r e -p r o o f

Time series analysis was conducted for the monthly records of COVID-19 confirmed cases and deaths, both linear and exponential growth models were studied. The R 2 , constant, and β (the slope coefficient) of the linear and exponential models are presented in (Supplementary material 1). In January 2020, China was the only country that reported confirmed deaths with COVID-19.

China, Australia, Canada, France, Germany, and the USA have reported confirmed cases with COVID-19 that are valid for statistical modeling. During January, the R 2 of linear models for confirmed cases by time in most of the selected countries was higher than that of the corresponding R 2 of the exponential model for the same data, in other words, the growing data for confirmed cases were better described with linear equations.

Starting from February 2020, and except for Egypt, Indonesia, and Brazil, all the selected countries started to report an increasing number of confirmed COVID-19 cases. Until June 2020, the R 2 of the exponential models were occasionally relatively higher than that of the linear models for confirmed COVID-19 cases and deaths, while in most models they were nearly equal. The monthly exponential growth rates of confirmed cases and deaths were perfectly correlated (Spearman correlation coefficient = 0.945, P <0.01). For the 18 countries, the median monthly growth rates for confirmed cases and deaths are presented in (Figures 1 and 2) , respectively. We remarked that the increase in exponential growth in COVID-19 cases and deaths in a given month implies an increase in linear growth in the next month, so the highest exponential growth in March J o u r n a l P r e -p r o o f implied the highest linear growth in April 2020. This is because the exponential growth multiplies the counts over time, which in turn increases the absolute counts in the next month.

The monthly change in mobility trends from baseline across different place categories is presented in (Figure 3) . In February 2020, mobility trends almost did not change from the baseline, then the median value started to decrease in March and April for all place categories except for mobility trends towards residential places. Afterward, an inflection has occurred and the median values for mobility trends increased towards all place categories until the final time point in June 2020, in which most of the median values were returned to the baseline. Interestingly, the median value of mobility trends toward parks, beaches, and public gardens has even exceeded the baseline in June 2020 (+25%).

It is worth noting that changes in mobility trends affected the exponential growth of COVID-19 new cases and deaths in the next month. So, we can notice a decrease in mobility trends from baseline in March, whose median exponential growth is the highest for cases and deaths, but we can further notice a decrease in exponential growth in the next month of April. This is because the confirmed cases usually get infected in a given month and been reported in the next month after a lag period of incubation and confirmation of infection. This remark was tested by assessing the correlation between the exponential growth rates and mobility trends in the same month. The results are presented in (Table 1) , most of the correlations were week and insignificant, while when mobility trends were tested for correlation with the exponential growth rates of the next month, we got significant strong positive correlations for new cases and deaths ( Table 2) .

J o u r n a l P r e -p r o o f

The monthly lagged exponential growth rates of cases showed an intermediate positive correlation with mobility trends to glossaries and pharmacies (r = 0.452, P <0.01), and a strong positive correlation with MPC in mobility trends to transit stations (r = 0.712, P<0.001), mobility trends to workplaces (r =0.77, P< 0.01), and mobility trends to retail and recreation (r=0.688, P<0.01), while it showed strong negative correlation with mobility trends to residential places (r= -0.727, P<0.001).

The monthly lagged exponential growth rates of deaths showed an intermediate significant positive correlation with changes in mobility trends to groceries and pharmacies (r= 0.495, P<0.01), while it had a strong positive correlation with changes in mobility trends to retail and recreation places (r =0.7, P<0.01), changes in mobility trends to workplaces (r=0.775, P<0.01), and changes in mobility trends to transit stations (r=0.748, P<0.01), and it showed a strong negative correlation with changes in mobility trends to residential places (r= -0.744, P<0.01).

The monthly average temperatures of countries had a weak significant negative correlation with exponential growth rates of deaths in the next month (r = -0.25, P=0.038), while it did not significantly correlate with exponential growth rates of new cases (-0.08, P=0.49). Relative humidity was significantly correlated with the exponential growth rates of confirmed cases (r = 0.28, P=.017) and deaths (r=0.31, P =0.008) of the next month (Table 3) . On the other hand, J o u r n a l P r e -p r o o f researchers did not get any significant correlations upon testing the overall exponential growth rates for cases and deaths with the socioeconomic factors of the selected countries.

We modeled the exponential growth rates for cases and deaths with time (months), the principal component of mobility trends of the previous month, the temperature of the previous month, the humidity of the previous month, the proportion of population above 65 years, the prevalence of DM, and PCI of each country as predictors.

The exponential growth rate of confirmed COVID-19 cases was significantly different across However, this is not the worst scenario, scientist warn the word from more waves of COVID-19 especially worldwide herd immunity is still away(Wise, 2020).

In this research, we tried to address different ecological factors that could affect the transmissibility and fatality of COVID-19. To the best of our knowledge, this study is the first to highlight the effect of different ecological factors like temperature, and humidity concurrently with studying other humanitarian factors like social mobility across different countries simultaneously.

In this work we researchers reported that for the first two months, the R 2 of linear models for new cases and deaths were higher than that of the corresponding R 2 of the exponential model. Later one, R 2 of the exponential models were occasionally relatively higher than that of the linear models, while in most models they were nearly equal. New cases of COVID-19 were significantly associated with mobility trends, temperature, humidity, and the proportion of patients aged above 65 years. Similarly, COVID-19 deaths were significantly associated with mobility trends, temperature, humidity, and PCI of countries. During this period, COVID-19 incidence was evident to be controlled as soon as social mobility is decreased by about 40% of the baseline over 3 months controlling for the other predictors.

During the first wave of the pandemic, it was so hard to predict the exact determents related to the severity of COVID-19 disease due to the large discrepancy of testing regulations and the pandemic precautionary measures between the countries. Compliance to social distancing and J o u r n a l P r e -p r o o f travel ban are important factors to be considered, which appeared in the study by Pana TA et al. on 38 countries which have more than 25 COVID-19 related deaths till 8th of June 2020. They studied the association between demographic, social, environmental, and economic parameters and the mean mortality rate. The multi-variate analysis revealed that international arrivals condition is the major determinant associated with the death rate, while BCG vaccination, the prevalence of hypertension and testing capacity were slightly associated. The other studied parameters such as temperature, population capacity, GDP and other ones were not statistically significant. (Pana et al., 2021) Similarly, Hassan et al, conducted geospatial study to identify the ecological determinant of COVID-19 related incidence and mortality in Africa.

They concluded that COVID-19 incidence rate was positively associated with overcrowding, health expenditure, human immunodeficiency virus (HIV) infection and air pollution and negatively associated with BCG vaccine while, COVID-19 fatality was positively related to asthma prevalence and tobacco use.

Here, we randomly piloted 18 countries representing 6 continents with variable COVID-19 incidence and mortality and different environmental and social determinants. We remarked that the increase in Exp (β)of a given month implies an increase in linear (β)in the next month, so the highest calculated median Exp (β)in March implied for the highest median linear (β)in April 2020. This is because the exponential growth multiplies the counts over time, which in turn increases the absolute counts in the next months. We further plotted the median monthly Exp (β)of confirmed cases and deaths of different continents.

In this research, we studied both exponential growth and the linear pattern of COVID-19 spread within the earliest 6 months of the COVID-19 pandemic. Similarly, in a study conducted by Comarova et al, (Komarova, Schang, & Wodarz, 2020 ) the pattern of growth was either exponential or power-law growth. They further cleared that the pattern of spread of each country depended on the time when the pandemic took place; if the pandemic started early the pattern was exponential, while later pandemic following Italy was power law.

Many countries adopt different strategies of national lockdown despite its damaging effect on the economy and education. Mobility can be used as a proxy measure of contact frequency, so social mobility is drastically affecting the disease transmission. It is considered as one of the main control measures until an effective vaccine is discovered. In the work presented the exponential growth of infection was significantly correlated with the social mobility pattern of the preceded month. This is explained by the duration of the incubation period of COVID-19 and reporting was recorded in the month following. Moreover, Carteni et al, (Cartenì, Di Francesco, & Martino, 2020) proclaimed that both the number of tested cases, proximity to the outbreak area, and mobility of the citizen are the main predictors of COVID-transmission. One of the significant types of social mobility was trips; within 3 weeks trips which were significantly associated with transmission risk.

Interestingly, this duration exceeded the containment duration which is 14 days. In this research, the effect of social mobility may be diluted due to the inclusion of countries with different wealthy indexes. Weil et al, reported that wealthier communities had lower social mobility than poorer communities, furthermore, the direction of mobility was more toward the least crowded areas before the pandemic than the more crowded ones and vice versa in poorer communities. (Weill, Stigler, Deschenes, & Springborn, 2020) In the United Kingdom, reduction of social mobility to J o u r n a l P r e -p r o o f 57.3% to 65.9% of the pre-lockdown situation would decrease the reproduction rate below 1 which means pandemic control. Reports noted that social mobility contributes to 80% of disease transmissibility. (Sussex, 2020) Similarly, Badr et al, (Badr et al., 2020) reported the Pearson correlation between COVID-19 transmission and social mobility exceeded 0.7 in 25 states in the USA. In this research, the effect of social mobility may be diluted due to the inclusion of countries with different wealthy indexes and studying other important environmental factors. Weil et al, reported that wealthier communities had lower social mobility than poorer communities, furthermore, the direction of mobility was more toward the least crowded areas before the pandemic than the more crowded ones and vice versa in poorer communities. (Weill et al., 2020) In the current study, social mobility was revealed to be the most important determinant of the incidence and virulence of COVID-19 infections. This was simultaneously controlled by the presence of the other different determinants in the same model. The mobility patterns across the studied countries were a reflection for the nationwide decisions taken for controlling the pandemic.

Most of these decisions were effective by March 2020, responding to the dramatic increase in confirmed cases and deaths. Accordingly, social mobility was decreased which in turn decreased the incidence of the disease. However, due to several causalities, most countries decided to lessen the precautionary measures and made a gradual re-opening. Most of the re-opening decisions were effective by May and June 2020. Interestingly, we remarked an inflection in the social mobility graphs in April 2020, thus we are worried about a potential correlation with new waves of the pandemic. Interestingly, the model suggests that a 40% decrease in social mobility would control the incidence of the infections in 3 months, taking into considerations that the model was utilizing data from March to June 2020, during which the toughest decisions were taken by the governments. On the other hand, the lockdown policies don't alone prevent the transmission of the J o u r n a l P r e -p r o o f disease, but combined with other regulations like wearing face masks, using sanitizers and hand washing could be more effective in decreasing the incidence of the disease. (Savaris, Pumi, Dalzochio, & Kunst, 2021) 

A questionable issue arose during this pandemic, whether climatic or demographic characteristics enable a more significant transmission of the virus. The Environmental temperature was thought to affect the virus survival on surfaces consequently it affects viral transmissibility.

In this research, we also remarked that temperature and humidity had a significant determinant on the incidence of COVID-19, however, the effect was trivial compared to that of social mobility

The results support the first reported statistically significant relationship of negative correlation between the average environmental temperature and exponential growth rates of the infected cases. (Livadiotis, 2020) Mecenas et al, (Mecenas, Bastos, Vallinoto, & Normando, 2020) reviewed the results of 517 published research articles that evaluated the effect of humidity and temperature on the viral spread. Only 17 articles fulfilled the inclusion criteria, they concluded that hot and wet weather significantly affect the viral transmission. On the same line, Prata et al (Prata, Rodrigues, & Bermejo, 2020) demonstrated that there is more than 4.8% reduction in the cumulative incidence of newly diagnosed cases with each rise in temperature by 1 °C. Similarly, (Basray et al., 2021) showed that for every unit increase in humidity, there was a 3.345 reduction in the daily reported cases of COVID-19 cases, while for every unit increase in humidity, there was a noticed increase in the number of daily reported cases by 10 times. Interestingly, Anis (Anis, 2020) demonstrated that the optimal temperature for virus transmission was between 13-24 ° C based on analysis of data from March to November 2020.

On the contrary, among 63 areas in China and more than a hundred locations in other seven different countries, the high temperature, humidity and ultraviolet ray exposure were not significant metrological factors with the COVID-19 infectivity. (Pan et al., 2021) 

Since the beginning of the pandemic, there was no clue on the contribution of countries' socioeconomic attributes to the propagation of COVID-19 cases. To the best of our knowledge, no previous studies investigated these factors as determinants of transmission. In the current study, and despite showing a non-significant contribution to the incidence and virulence of the infection, these attributes remain of interest for the across-country studies for controlling the other investigated attributes.

To the best of our knowledge, this research is the first to address the pattern of the spread either linear or exponential, it modeled the effect of social mobility environmental temperature, and humidity simultaneously, we did include other important determinants of health like income, health care facilities, access to health systems, comorbidities and extra. An important point of limitation was that case definition was variable across different countries. Furthermore, case definition of a suspected case or confirmed case may vary across time in the same country, and this may affect the number of reported cases. The same problem can be encountered for number of deaths. These facts may affect the external validity of our models.

Data-driven suppositions can efficiently and proactively guide the governmental measures taken to lessen the social, health, and economic impacts of the COVID-19 pandemic. During the first six months of the pandemic, the multivariate analysis showed that the changes in mobility trends across countries dramatically affected the incidence and mortality rates across different countries, thus, it might be used as a proxy measure of contact frequency. Studying a set of different predictors in our models makes the estimates more accurate and precise because the contribution of each predictor is controlled for the contribution of the other predictors in the models.

Emerging literature separately investigated social mobility trends, ecological and socioeconomic factors to explain and further predict COVID-19 incidence and mortality rates across different provinces of the same country.

This study assembled social mobility trends, ecological and socioeconomic attributes in one statistical model explaining the incidence and mortality rates of COVID-19 in 18 different countries. Hence, it presents more accurate weight for each predictor, controlled for the presence of the other factors.

The model suggests that one unit decline in social mobility is equally effective in declining COVID-19's incidence to about 6 degrees Celsius decline in temp and about 16 degrees in relative humidity. J o u r n a l P r e -p r o o f J o u r n a l P r e -p r o o f J o u r n a l P r e -p r o o f 

The Effect of Temperature Upon Transmission of COVID-19: Australia And Egypt Case Study

Association between mobility patterns and COVID-19 transmission in the USA: a mathematical modelling study. The Lancet Infectious Diseases

Clinician perceptions of respiratory infection risk; a rationale for research into mask use in routine practice

Impact of environmental factors on COVID-19 cases and mortalities in major cities of Pakistan

Emerging infectious diseases: public health issues for the 21st century

How mobility habits influenced the spread of the COVID-19 pandemic: Results from the Italian case study

The role of community-wide wearing of face mask for control of coronavirus disease 2019 (COVID-19) epidemic due to SARS-CoV-2

See how your community is moving around differently due to COVID-19

A brief review on the possible role of houseflies and cockroaches in the mechanical transmission of coronavirus disease 2019 (COVID-19). Archives of Clinical Infectious Diseases

After 2015: infectious diseases in a new era of health and development

Driving improvements in emerging disease surveillance through locally relevant capacity strengthening

Temperature, humidity, and wind speed are associated with lower Covid-19 incidence

Global trends in emerging infectious diseases

Association of Social and Demographic Factors With COVID-19 Incidence and Death Rates in the US

Determinants of COVID-19 Incidence and Mortality in the US: Spatial Analysis. medRxiv

Patterns of the COVID-19 pandemic spread around the world: exponential versus power laws

Statistical analysis of the impact of environmental temperature on the exponential growth rate of cases infected by COVID-19

Effects of temperature and humidity on the spread of COVID-19: A systematic review

The challenge of emerging and re-emerging infectious diseases

Travel-related control measures to contain the COVID-19 pandemic: an evidence map

Warmer weather unlikely to reduce the COVID-19 transmission: An ecological study in 202 locations in 8 countries

Country-level determinants of the severity of the first global wave of the COVID-19 pandemic: an ecological study

Temperature significantly changes COVID-19 transmission in (sub)tropical cities of Brazil

Stay-at-home policy is a case of exception fallacy: an internet-based ecological study

Risk of return to exponential growth in Coronavirus cases if UK mobility not kept below half of pre-lockdown levels

Social distancing responses to COVID-19 emergency declarations strongly differentiated by income

Covid-19: Risk of second wave is very real, say researchers

Host range and emerging and reemerging pathogens

Corona Virus

Pathological findings of COVID-19 associated with acute respiratory distress syndrome