key: cord-0881477-1k08fug6
authors: Ghosh, Dr. Debarshi; Sarkar, Apurba; Chouhan, Dr. Pradip
title: Covid-19 second wave: district level study of concentration and fatality in India
date: 2021-07-24
journal: Environmental Challenges
DOI: 10.1016/j.envc.2021.100221
sha: 8caac70525d78db144ca1d34cd649fe2ad016cf5
doc_id: 881477
cord_uid: 1k08fug6

The paper aims to reveal the spatial pattern of the concentration of COVID-19 confirmed cases and the spread of the pandemic from the Case Fatality Ratio. The study has been accomplished with district-level data. The analysis of the spatial pattern decoding has been done considering the Global and Local Moran's I statistics comprising the linear trend of spatial autocorrelation for the whole India. The timeframe has been divided considering the surge of the second wave in March, 2021 and the peak of the wave in May 2021. The spatial clustering technique presents both the concentration of confirmeded cases using Location Quotient analysis and the pattern of spread of the infection-related fatality throughout the country. The high Location Quotient of the confirmeded cases strongly clustered around the Mumbai-Puna region, Kerala-Karnataka region, Garhwal Himachal, NCT of Delhi and Ladakh-Kashmir-Himachal Pradesh region during the period of the study. In May, the concentration has randomly clustered around the middle part of India. The Case Fatality Ratio was high in Maharashtra, Madhya Pradesh, Punjab and Haryana at the surge of the second wave. In peak (May), the two significant clusters of high case fatality ratio is pivoted around in Mumbai urban (Maharashtra) and NCT of Delhi (including Punjab-Haryana).

COVID-19 pandemic has been impacting the life and economy across the globe since December 2019 and has caused major disruptions (Walker et al. 2020). The COVID-19 pandemic has resurfaced in India in the form of a hard-hitting second wave. The COVID-19 has brought a threatening challenge to Indian society and the economy [1] . India's devastating second wave of COVID-19 has overwhelmed its health system and the country [2, 3] . The second wave of COVID-19, caused by severe acute respiratory syndrome (SARS-CoV-2), has struck India severely, with a significant case fatality rate [4] . The situation in India is more critical as it has a huge population, poor medical infrastructure and complex socio-economic structure, where selfisolation, social distancing and quality treatment is the key controlling factors to neutralise the impact of the disease [5, 6] . The growth of towns and the consequent need for more supplies have damaged the delicate environment of India, where there are high levels of smog, fine dust, and water pollution. Sulfur dioxide (SO2), nitrogen dioxide (NO2), and particulate matter (PM) contribute in part to the toxins causing environmental contamination [7] [8] [9] . Many Indian urban communities, including Mumbai, Kolkata, and Pune, are at risk of air contamination [10] .One out of eight (about 12.5%) deaths in 2017 in the country were attributable to high rates of respiratory disease, stroke, heart disease, diabetes, and lung cancer, all conditions for which a certain percentage of cases result from severe air pollution [11] . Some relevant scientific literature highlights that exposure to air pollution may be relevant to virus infection spread, and more recent literature focuses on COVID-19 diffusion [12] [13] [14] . On January 30 2020, Director-General WHO declared that the outbreak of novel coronavirus (2019-nCoV) constitutes a Public Health Emergency of International Concern (PHEIC) as per the advice of the International Health Regulations (IHR) Emergency Committee [15] . In the first surge in 2020, COVID-19 has infected nearly 20 million people across the globe, with 90 countries in the community transmission stage [16] . The daily reported confirmed cases started to rise from February 2021 in India [17] . The mid of April 2021, a registered sudden hike in over a thousand daily deaths in the country [18] . Multiple factors are involved in driving the second wave of COVID-19 in India, such as the complex interplay of mutant strains, violation of COVID appropriate behaviour, and government and public complacency on initiation of the vaccination drive [19] . The situation turned into a bleak one when the country witnessed the daily deaths of over three thousand at the end of April 2021. On April 26, 2021, India saw the highest daily tally of new SARS-CoV-2 infections ever recorded globally, 360 960, taking its pandemic total to 16 million cases, second only to the USA with more than 200000 deaths [20] . As of May 18, India had reported more than 26.4 million confirmeded cases and over 274000 deaths from COVID-19 [21] . 29.27 million cases have been reported in India during the pandemic, with a case fatality rate of 1.24% (363,079 deaths) up to June 11, 2021 [19] . At the beginning of the second wave, the country's Case Fatality Ratio (CFR) has hovered around 1.35% to 1.40%. At the onset of the COVID-19 pandemic, India imposed the world's strictest nationwide lockdown beginning from March 25, 2020 [22] . But, the situation during the onset of the second wave was aweary in the whole country. WHO confirmeded 15,510 new cases in India on March 1, 2021, and the peak had been formed with 414188 confirmed cases on May 7 [23] . The curve of new cases indicated a sharp rise from April onwards in India. Surprisingly, the present peak pandemic situation addresses a difference of 3,14,692 new confirmeded cases than the previous peak (September 18, 2020) [24] . It indicates the severity of the second outbreak in India. On April 17, 2021, the country peaked, considering confirmeded cases that accounted for 10.6% of the country's total population. On May 11, 2021 , the WHOeported about 16,411 confirmeded cases and 178 deaths per million [25] . On the same day, the country's CFR was 1.09%. The spread of the infection is more belligerent in the states of North-East India. From the beginning of the second wave, the outbreak was centred around the megacities of India, especially in NCT of Delhi, Mumbai and Bengaluru urban [22] . On January 4, 2021, a total of 7.57% people of in the country were infected by the coronavirus [26] . But, by May 11, the figure reached up to 16.67% [27] . At this stage, the main challenge is to organise the basic lifesaving treatment facilities at the grass-root level and especially assure the supply of oxygen and vaccines [28] . In addition to this, the legislative assembly election in Assam, West Bengal, Tamil Nadu and Kerala with mass political gatherings and rallies has made the situation out of control. India is the largest democratic country globally; however, looking at the current COVID-19 situation, the state elections (Assam, Puducherry, Kerala, Tamil Nadu, and West Bengal) could have been postponed till normality [29] . Many states did not go for the full lockdown this time and relied on night curfew to keep the local economy alive. As well the largest vaccine drive in the world was started in India in 1 st week of April. At the same time, 6.06% of the total population was vaccinated (1st and 2nd). On May 11, roughly 13.3% of people have got vaccinated [25] . The increase is not so promising for such a country with a billion-plus population. The changing nature of the virus and countrywide high oxygen demand worsened [30] . A rapid increase in the daily incidence of serious cases creates a shortage of medical instruments, oxygen, hospital beds and lifesaving drugs [31] . In a knee jerk reaction, the central government and its expert team started constructing the badly needed medical facilities to overcome the shortages. However, this frenetic activity of augmenting facilities comes in the middle of an ongoing and exponential rise in cases. In contrast, it should have started way ahead of these doldrums, say experts [15] . The country's average CFR falls significantly down from 1.45 % (in January 2021) to 1.09% (in the first week of May) [32] . Considering the second outbreak, the study incorporates a new angle to find out the district wise spatial pattern of the concentration of COVID-19 confirmeded cases and Case Fatality Ratio (CFR) in India [33] .

The paper aims to know the spatial clustering pattern of both the concentration of confirmeded cases and CFR. Spatial Cluster analysis (SCA) has two general purposes. The first is data preprocessing, where clustering is used as an automated, unsupervised step for organising data, increasing the efficiency of search or query algorithms, and improving the system's overall performance. The second is exploratory, where clustering is used for knowledge discovery and for motivating new hypotheses based on discovered patterns [34] . In this study, the SCA technique well presented the pattern of spread of the infection in the second wave. The work will surely help the policymakers of the country-firstly to understand the spread of the second wave from the concentration of positive cases and its spatial clustering, secondly to identify the hotspots of the concentration and thirdly to implement COVID-19 pandemic combat strategies like demarcation of containment zones, lockdown strategies, vaccine drive with paramount importance, blocking intra-state movements and most importantly thorough preparation at the medical level (district, block and village level). This study incorporates a completely new angle to reveal the hidden geospatial pattern of the second wave at the district level with spatial statistics.

The purpose of the Location Quotient (LQ) technique is to yield a coefficient to understand how well industry can concentrate in a region [35, 36] . The LQ measures the ratio between the local and national share of productive activities of a particular industry in a region, usually using employment to represent productive activities [37] . It is measured on a simple numerical scale. The value of LQ can be presented in several ways: a) LQ less than one indicates that industry is under-presented or the output is not sufficient to meet the local demand etc., b) LQ 1 means the study region's share of the industry is identical to the reference region's share or the employment is equal in the sector for the national and regional economy and c) LQ greater than 1 suggests that the study region has more than its share of an industry or the output is more than sufficient [38] . The LQ can be stated as a ratio of ratios. The technique is frequently used specially in studying industrial concentration, agglomeration and in regional economics. It can also be considered as agglomeration index like Gini Index, Maurel-Sedillot (MS) index etc [39] . The theoretical foundation of LQ was constructed by Ellison and Glaeser (1997) and Guimaraes et al. (2009) [40, 41] . Further, Billings and Johnson (2012) examined the statistical accuracy of LQ [42] . In this study, the district-wise concentration of the confirmeded cases of covid-19 has been conceptualised with LQ technique. We took the objective to know at which part of the country, the second wave of the Covid-19 pandemic has concentrated intensely. In addition to this, the district-wise concentration of the confirmeded cases can also be checked compared to the whole country. The formula of LQ has been reshaped for this study:

Where, X ij represents confirmeded cases i in a region j, X i is the total population of an area j, X * j is total confirmeded cases of the whole country, and X k -is the total population of the entire country.

This approach will also lead to identify the spatial pattern of the concentration of confirmeded cases in the country. Thus, the mobility of the spatial concentration (confirmeded cases) from one region to another has been presented with maps considering a time lag.

The Case specific Fatality Ratio (CFR) is a measure of the severity of the condition as it corresponds to the proportion of the population with a certain condition who die from that condition during the reference period [24, 43] . But the measurement is crude in sense that it does not account for the changes in the demography of positive cases and deaths during the different stages of an epidemic [44, 45] . Considering the lacuna, the CFR can be used as a basis for estimating and monitoring the number of infected individuals in a population, which may be subsequently used to inform policy decisions relating to public health interventions and lockdown strategies [46, 47] . In epidemiology, the ratio stands for finding the proportion of the people who die from specific diseases among all the individuals diagnosed with the diseases over a while. As many cases of previous covid-19 were asymptomatic, generalised data on the true number of persons infected lack in India. Mortality rates, therefore, are calculated from confirmed cases, which overestimate the Case Fatality Ratio (CFR). The ratio can also be used to evaluate the effect of new treatments [37] . It is conventionally expressed as a percentage and represents a measure of disease's severity. The calculation comprises the proportion of deaths among all infected individuals. The CFR can be measured as the proportion of individuals diagnosed with a disease who die from that disease and is therefore a measure of severity among detected cases. In this study, we found out the fatality risk among the confirmeded covid-19 cases. The calculation technique is:

The CFR clustering pattern has been cross checked with the dot density mapping considering the deaths on the selected dates. It tells us about the spread of the severity of the infection in the country.

The spatial pattern of the CFR has been studied using Global Moran-I statistics (spatial statistics) considering Moran's I index, Z-score, P-value and Expected Index [48, 49] . The Global Moran's I determine the probability that the observed value could be found in the random distribution or not [50] . The spatial distribution pattern of the Fatality Ratio has been scaled considering contiguity scale of distribution which varies from dispersed to a clustered pattern. The statistics conceptualises the spatial pattern based on contiguity edge only function. It only finds neighboring polygon features that share a common boundary. The spatial distance among the neighbour polygons has been measured using the Euclidean distance method. The Global Moran I summary indicates z-score value in the standardised normal probability distribution and determines the likelihood of spatial pattern (dispersed, random and clustered) [51] .

Associations (LISA) cluster maps were prepared to assess the univariate local and global patterns of spatial dependence for the neighbourhood states considering the Fatality Ratio [52, 53] . It is useful in determining spatial outliers. The statistics produce Local Moran's I Index, z-score, pseudo-value and cluster-outlier type (COType) ( Table 1 ). The analysis wants to check whether the apparent similarity (spatial clustering of high or low values) or dissimilarity (an outlier) is more pronounced than one would expect in a random distribution. The COType indicates statistically significant clusters and outliers for a 0.05 level of significance. The authors used the spatial weight matrix that yielded the highest Moran's I values for understanding the spatial association. The matrix weights are a key component in measuring spatial dependence.

Here, the author's created contiguity weights following the Queen's contiguity method. The method considers the common boundary as well as the vertices of the neighbour polygons. The Queen's criteria consider eight neighbours. The matrix weights have been generated carefully considering the topology of the spatial structure produced as 'Connectivity Graph' in the GIS environment. Finally, to identify the clusters of the districts (Fatality Ratio), the Local Moran cluster map or Local Indicator of Spatial Association (LISA) clustering has been adopted. This clustering approach is designed to reject the null hypothesis of spatial randomness favouring an alternative clustering. LISA approach functions in two ways: a) it brings out an assessment of significance for each location attributes and b) it reveals a proportional relationship between the local and global statistics. The LISA clustering has been done with permutations of 999 where the smallest possible pseudo-p-value is 0.001 and all other pseudo values will be even multiples of this value. The LISA clustering on FR in India produces a good and rationale fit to compare the India's district-level data with the world's reality. Finally, Moran's scatter plot helps to determine the linear fit of the regression slope exactly following the Moran's I value [54] . The plot classified the nature of spatial auto-correlation into four categories that correspond with the LISA CoTypes. The four quadrants in the scatter plot connect between global and local spatial autocorrelation. 

Numerous production sectors have grown and in addition to the steel and metallurgical sectors, the textile and oil refining industries have also expanded. This economic growth has increased the number of jobs in cities and the populations of large urban areas, such as Delhi and Mumbai [55] . These megacities have high population concentration, and both the cities are overcrowded with immigrating populations from all over India. The population density of Mumbai Urban and NCT of Delhi are 19,652 per sq. km. and 11,320 per sq. km. Respectively. Social distancing has considered a preventive approach worldwide which has ensured that Covid-19 can be contained. But, in the Dharavi slum of Mumbai, the population density is more than 277,136 per sq. km. In NCT of Delhi, Munirka (urban village) and Kirti Nagar are highly congested urban slums.

The farmer protest in November 2020 at the New Delhi border areas with approximately 150 to 300 thousand mobs made the situation highly vulnerable for the infection to spread. The prescribed 'Social distancing' was at stake considering the entire situation. The household density of the urban slum areas is one of the most vital causes of infection spread. The intra-urban mobility from slum to town central Business Districts (CBD) and other workplaces increases the chance of such infection spread. The international airports of NCT of Delhi, Mumbai, Bengaluru and Kolkata acted as the gateways to allow the dangerous mutating Covid-19 strains to spread across the main urban hubs of the country. The migrant labours borrow the new strains to the different corners of the country. In addition, the relaxation phase of domestic transportation (especially railways) from November 2020 up to almost mid of April made the situation favourable for rapid infection spread. A recent study revealed the composite analysis based on socio-economic-political parameters at the district level to construct a new Socio-Economic Vulnerability Index [1] . The index indicated the high vulnerability of North-West Delhi, Central Delhi, Kolkata, Mumbai and Ghaziabad. In our study, the spatial auto-correlation technique for LQ and CFR indices also indicates the same hot spots during the second wave. Previously, many printed discussions and analyses focused on the socio-economic impact of the Covid-19 pandemic in India [56] . We also have much literature dealing with finding the cause of infection spread from clinical epidemiological [57] [58] [59] and societal viewpoints [60, 61] . In this paper, the infection spread phenomenon and its concentration have been identified with keeping a relation between geographical association (spatial auto-correlation), logic of Geographical Information system (GIS), and statistical foundation.

The LQ analysis identified the Covid-19 infection concentration district-wise ( Fig. 1 a & b) . The concentration has been checked based on confirmed cases on a particular day. On the first day of March 2021, the spatial pattern of high concentration (greater than 10) was seen, especially in Delhi, Bengaluru, and Mumbai. These urban centres were the pivot of infection at the beginning of March. The National Capital of Delhi indicated a large concentration gap considering the confirmed cases and even outmatched with Bengaluru Urban (second highest). The spatial scaling of the LQ has been modified to new categories: a) less than 1 (poor concentration), b) 1.0 to 5.0 (moderate concentration), c) 5.0 to 12.0 (High concentration), and d) greater than 12 (very high concentration The moderate to high concentration in the middle part is quite random in nature. The LISA cluster map (Fig. 1c) ) . Finally, the pattern would give us some idea how to combat with the present situation and to get prepare for the predicted third wave.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

As the corresponding author, I give you the publisher's consent on behalf of all the authors.

On behalf of all the authors, I approve that we have gone through all the journal's ethical standards and confirm that all the ethics were taken into consideration during this research.

This article does not contain any studies with human participants or animals performed by the authors.

The data are collected from the data repository of https://www.covid19india.org which is publicly available and could be sassed upon a request subject to non-profit and academic interest only. Informed

COVID-19: District level vulnerability assessment in India

Temporal Dynamics of COVID-19 Outbreak and Future Projections: A Data-Driven Approach

How India is dealing with COVID-19 pandemic

Prediction for the spread of COVID-19 in India and effectiveness of preventive measures

The second wave of COVID-19: time to think of strategic stockpiles

India begins COVID-19 vaccination amid trial allegations

Modeling spatial determinants of urban expansion of Siliguri a metropolitan city of India using logistic regression, Modeling Earth Systems and Environment

Population-weighted exposure to air pollution and COVID-19 incidence in Germany

Qualitative and quantitative analyses of impact of COVID-19 on sustainable development goals (SDGs) in Indian subcontinent with a focus on air quality

Stringent Emission Control Policies Can Provide Large Improvements in Air Quality and Public Health in India

Air pollution trends over Indian megacities and their localto-global implications

Escalating infection control response to the rapidly evolving epidemiology of the coronavirus disease 2019 (COVID-19) due to SARS-CoV-2 in Hong Kong

Evaluation of the potential relationship between Particulate Matter (PM) pollution and COVID-19 infection spread in Italy

Lockdown and unlock for the COVID-19 pandemic and associated residential mobility in India

COVID-19: the case for healthcare worker screening to prevent hospital transmission, The Lancet

Qualitative and quantitative analyses of impact of COVID-19 on sustainable development goals (SDGs) in Indian subcontinent with a focus on air quality

An analysis of COVID-19 clusters in India: Two case studies on Nizamuddin and Dharavi

Impact of Migration of Labour Force due to Global COVID-19 Pandemic with Reference to India

Second wave of COVID-19 pandemic in India: Barriers to effective governmental response

Why is India having a covid-19 surge?

Contextualising evidence-based recommendations for the second wave of the COVID-19 pandemic in India, The Lancet Infectious Diseases

The Lancet, India's COVID-19 emergency

Case fatality ratio of pandemic influenza, The Lancet Infectious Diseases

COVID-19 : STATUS ACROSS STATES

River bank erosion hazard study of river Ganga, upstream of Farakka barrage using remote sensing and GIS, Natural Hazards

Precision shielding for COVID-19: Metrics of assessment and feasibility of deployment

India grapples with second wave of COVID-19

Addressing Impact of COVID-19 Post Pandemic on Farming and Agricultural Deeds

Transmission dynamics of the COVID-19 epidemic in India and modeling optimal lockdown exit strategies

COVID-19: A global and continental overview of the second wave and its (relatively) attenuated case fatality ratio

Case-Fatality Ratio and Recovery Rate of COVID-19 : Scenario of Most Affected Countries and Indian States A Situational Analysis Paper for Policy Makers International Institute for Population Sciences

An analysis of COVID-19 clusters in India: Two case studies on Nizamuddin and Dharavi

Analysis of import changes through shift-share, location quotient and BCG techniques: Gwangyang Port in Asia

Core periphery analysis of the European Union: A location quotient approach

Measuring agglomeration using the standardized location quotient with a bootstrap method

Spatial Cluster Analysis for Point Data: Location Quotients verses Kernel Density

Making Location Quotients More Relevant as a Policy Aid in Regional Spatial Analysis

Geographic concentration in U.S. manufacturing industries: A dartboard approach

Dartboard tests for the location quotient

The location quotient as an estimator of industrial concentration, Regional Science and Urban Economics

Condition-specific mortality risk can explain differences in COVID-19 case fatality ratios around the globe

Temporal trends in general and age-specific fertility rates among women with schizophrenia

The COVID-19 pandemic and global environmental change: Emerging research needs

Estimating the infection and case fatality ratio for coronavirus disease (COVID-19) using age-adjusted data from the outbreak on the Diamond Princess cruise ship

Using early data to estimate the actual infection fatality ratio from covid-19 in france

Extending Moran's Index for Measuring Spatiotemporal Clustering of Geographic Events

New Approaches for Calculating Moran's Index of Spatial Autocorrelation

Location matters: Unravelling the spatial dimensions of neighbourhood level housing quality in Kolkata, India, Habitat International

Spatial cluster detection of air pollution exposure inequities across the United States

Local Indicators of Spatial Association-LISA, Geographical Analysis

Handbook of Applied Spatial Analysis

An Introduction to Spatial Autocorrelation Analysis with GeoDa

COVID-19 pandemic and challenges for socio-economic issues, healthcare and National Health Programs in India, Diabetes and Metabolic Syndrome

A brief review of socio-economic and environmental impact of Covid-19, Air Quality, Atmosphere and Health

Epidemiology and transmission dynamics of COVID-19 in two Indian states

Inferring COVID-19 spreading rates and potential change points for case number forecasts

Epidemiological data from the COVID-19 outbreak, real-time case information

The COVID-19 pandemic: Impacts on cities and major lessons for urban planning, design, and management

Indirect implications of COVID-19 towards sustainable environment: An investigation in Indian context