key: cord-0076861-o22wwzx4
authors: Mahmood, Shakeel
title: Exploring COVID-19 incidence hotspot in Metropolitan area of Pakistan using geo-statistical approach: a study of Lahore city
date: 2022-04-10
journal: Spat
DOI: 10.1007/s41324-021-00423-4
sha: 7cc53c86bd337a4f9c8f30cafcc9c88aedcbd734
doc_id: 76861
cord_uid: o22wwzx4

Globally, COVID-19 is a top level public health concern. This paper is an attempt to identify and assess COVID-19 incidence hotspots in Metropolitan area of Pakistan using geo-statistical approach. The study is based on secondary data. The COVID-19 confirmed cases record (15/03/2020 to 15/04/2020) of entire Metropolitan area is obtained from hospitals and National Institute of Health website. Point-level geo-coding technique was applied on patient’s record and the relative location was converted into absolute location. Getis-Ord Gi* statistical model was applied in ArcGIS 10.3 to calculate Z-score and P values for each point location representing the COVID-19 incidence intensity. Then inverse distance weighted technique of spatial interpolation was applied on Z-score and spatial clusters of crime were geo-visualized in the form hotspot and cold spot. Spatially, more than 50% of land area of Allama Iqbal, Samanabad, Gulburg and Cantonment is covered by very high incidence zone which is surrounded by high incidence zone whereas Ravi, Shalimar, and north of Wagha and Aziz Bhatti towns are located in very low incidence zone. This study provides a suitable methodological framework for identification and analysis infectious disease hotspots. The study can also facilitate health and related authorities to fight war against COVID-19. Similarly, it can help policy makers to manage the movement of travelers and restrict social interaction.

The COVID-19 is a humanitarian emergency, which is started in the Wuhan province of China in early December 2019 and was declared as an emergency in the third week of January 2020 [1] [2] [3] . The World Health Organization (WHO) declared COVID-19 as Public Health Emergency of International Concern (PHEIC) on 31th of January 2020, and finally a pandemic on 11th March 2020. As of 28th April, 2020, there are 3.06million confirmed Coronavirus cases, with 0.213 million deaths while 0.907million has been recovered. In the total Coronavirus patients died, very interestingly the highest number belongs to Italy i.e. 4,032 deaths. The death toll is followed by China (3248), Iran (1433) and Spain (1044). Worldwide, the impact of this pandemic has affected countries and communities socially, economically and psychologically as well as international relations [4] . At present, COVID-19 pandemic is hot topic globally [3] .

Globally, COVID-19 pandemic became a biological disaster and tens of thousands of people have lost their lives; millions are infected and quarantined. Man has experienced various pandemics throughout the history with different intensity and impact [5] . Today we are observing and experiencing a very tough time by fighting once again with an invisible enemy; the COVID-19 [6] . The Corona virus infection spread started in early January 2020 in China [7] . The impact of this public health emergency has affected countries and communities in terms of economic, socio-psychological issues, as well as international relations [4] .

Currently, a variety of human diseases is prevailing with unknown etiology [1, 8, 9] . Viruses have been considered as cause of these diseases which has increased the continuous search for new viruses [10] [11] [12] . Corona viruses, a genus of the Coronaviridae family, are enveloped viruses with a large plus-strand RNA genome. The genomic RNA is 27-32 kb in size, capped and polyadenylated [13] . There were four human coronaviruses known to exist: human coronavirus 229E (HCoV-229E), HCoV-OC43, severe acute respiratory syndrome (SARS)-associated coronavirus (SARS-CoV) and Coronavirus. The coronaviruses have been identified in mice, rats, chickens, turkeys, swine, dogs, cats, rabbits, horses, cattle and humans, and can cause a variety of severe diseases including gastroenteritis and respiratory tract diseases [14] [15] [16] . Identification of a fourth human coronavirus, HCoVNL63 is reported by using a new method of virus discovery which was isolated from a 7-month-old child suffering from bronchiolitis and conjunctivitis [17] . Coronavirus has caused an ongoing outbreak of viral pneumonia in China [18, 19] . It spreads by the mucous containing liquid present in the human body mainly droplets of saliva or discharge of nose. After the first three months of this pandemic multiple epidemiological assessments in different countries from Asia, Europe, and North America have been published [6] . Nevertheless, there are countries, with a rapid increase and a high number of cases, with a lack of studies [20, 21] .

Worldwide, numerous studies have been conducted to assess the current and predict the future trends of COVID-19 using different approaches. Variety of models have been used in some of the recent studies like an effective model for diagnosis and classification of COVID-19 is designed [2] , Deep Sense approach is developed and applied for COVID-19 modelling [22] , Support Vector Regression (SVR) and deep neural network technique were clubbed and forecasting model is developed [23] , mortality rate is predicted by applying competence of AI-mediated methods [3] and Spatial prediction of hotspot using geo-statistical technique [24] . The current study is focusing on COVID-19 incidence hotspot modeling using Getis and Ord Geo-Statistical model in integration with Geographic Information System (GIS).

In Pakistan, first case was reported on February 25, 2020. The total confirmed cases are 14,612, active cases 11,067, deaths 312 and recovered cases 3233 till 28th April, 2020. The main cause of COVID-19 pandemic in Pakistan is the entry of infected pilgrims from Iran, pilgrims and migrants from Saudi Arab. Similarly, infected people from UAE, Malaysia and China have also transmitted and spread the virus. These infected migrants travelled and meet different people over different locations and transmit corona virus to healthy people. This transmission continued till the complete lockdown. Currently, lockdown has minimized the interaction of infected and healthy persons. Similarly, the quarantine centers across the country have reduced the exposure of healthy citizens [25] . The COVID-19 is still a biological hazard and the entire population is at high risk in Pakistan. Therefore, the aim of this study is to carryout spatio-temporal analysis of COVID-19 incidence hotspot in metropolitan area of Pakistan using Geo-statistical Approach.

Lahore is selected for COVID-19 incidence hotspot mapping which is the second largest city in Pakistan with population more than 6.5 million. Administratively, Lahore is divided into nine towns; Ravi, Shalimar, Wagah, Aziz Bhatti, Data Ganj Bakhsh (DGB), Samanabad, Allama Iqbal, Nishtar and Johar Town (Fig. 1) . This study is based on secondary data. The data regarding registered COVID-19 confirmed cases were acquired from National Institute of Health (NIH) of Pakistan website. The data were classified in MS Excel. Point-level geo-coding technique is applied to geo-code 492 confirmed COVID-19 cases of past 30 days i.e. 15/03/2020 to 15/04/2020. The hotspots is determined by applying Getis-OrdGi* (G-i-star) statistical model in GIS environment to quantify Z-score and P values. Finally, Inverse Distance Weighting (IDW) technique of spatial interpolation is implemented to generate a spatial layer depicting hotspots and coldspot zones by utilizing Z-score as input data. In the current study, Getis-OrdGi* geo-statistical model has utilized for spatial clustering after. The Gi* statistic is more interested in assessing incident intensity than in analyzing the spatial clustering of any particular value associated with COVID-19 incidence. Similarly, Gi* statistic examines the individual locations, enabling hotspots to be identified based on a comparison with the neighboring samples and has been successfully applied in hotspot identification and mapping.

The Getis-Ord local statistic is given in Eqs. 1, 2 and 3 [26] [27] [28] The Getis-Ord local statistic is given as:

where x j is the attribute value for feature j, w i,j is the spatial weight between feature i and j, n is equal to the total number of features and:

The G i * statistic is a z-score so no further calculations are required.

The geo-coded COVID-19 cases records were imported into ArcGIS10.2 and geo-visualized in the form of point feature class as spatial layer (Fig. 2) . The layer was added as input data to perform hotspot analysis. The hotspot analyst, identifies spatial clusters of high values (hotspots) and low values (cold spots) by implementing Getis-OrdGi* statistical model. It converts the geo-coded COVID-19 cases record into a new output feature having attributes of the z-score, P value, and Gi-Bin confidence level ranging from -3 to ? 3. High counts of cases close together have ? ve z-score indicating hotspot and vice versa. The z-score range ± 3 reflect statistical significance with 99-percent confidence level, ± 2 bins reflect 95-percent confidence level, and ± 1 bins reflect 90-percent confidence level, whereas 0 bin is not significant (Fig. 3) . Then the IDW technique of spatial interpolation is implemented to generate the raster spatial layer depicting COVID-19 cases intensity as hotspot or coldspot. The z-score values are utilized as input data in spatial interpolation. The COVID-19 incidence intensity is categorized into five classes namely very low (-2.1 to -3), low (-1.1 to -2), moderate (? 1 to -1), high (1.1 to 2) and very high (2.1 to 3). Linear regression analysis is applied on daily COVID-19 confirmed cases.

The results obtained from processed data are explained in the following sections. The confirmed cases were 150 in Allama Iqbal town followed by Cantonment and DGB with less than 100 cases (Fig. 4) . In Gulberg and Samanabad the number of confirmed cases was 60 and 64 respectively. Aziz Bhatti town is lowest number of confirm cases (Fig. 4) . The spatial pattern of COVID-19 confirmed cases is variable across the district Lahore. 

In the high to very high COVID-19 incidence zone has high to very high intensity of confirmed cases. Spatially, more than 50% of land area of Allama Iqbal, Samanabad, Gulberg and Cantonment is covered by hotspot (Fig. 5) . Some areas in the north and west of Nishter Town and south of the DGB town are located in very high incidence zone. Whereas the remaining parts of the same towns are covered by high incidence zones. As a whole 63% land area of Lahore is covered by high to very high intensity zone.

The moderate COVID-19 incidence zone has moderate level of confirmed cases. Spatially, west of Wagha and south of Nishter Town has moderate incidence zone.

The low to very low COVID-19 incidence zone has low incidences. Spatially, Ravi, Shalimar, and north of Wagha and Aziz Bhatti towns are located in this zone. Spatially, the COVID-19 incidence is very high (Hotspot) in central and south west of district Lahore whereas the surrounding of hotspots are covered by high incidence areas. The incidence is moderate in the south and west. Towards north incidence is decreasing forming coldspot with low incidence of COVID-19. The most affected (34%) age group is 31 to 40 years old followed (26%) by 21 to 30 years old (Fig. 6) . The most vulnerable age group above 50 years old is also severely affected because 12% of confirmed cases were of this group. No confirmed case is reported in age group less than 10 years old. The number of confirmed cases is growing very rapidly and still on rise (Fig. 7) . This trend is depicting that the infected cases is likely to increase which will further grow the spatial extent of hotspot.

In Pakistan, first case was reported on February 25, 2020. Initially, District Sukker and Karachi were most affected with 273 and 154 confirmed cases respectively till March, 26th, 2020. The district Sukker has highest number of cases because of returning pilgrims from Iran to Taftan (a town located in district Chaghai, Baluchistan) and then Sukker. This has resulted in high incidence in Sindh province but later on the travelling of unaware infected persons has transferred to Punjab province and transmitted in healthy people. Currently, Punjab has the highest number of confirmed cases followed by Sindh. The hotspot Prediction of infected persons can be made on the bases of existing data. The number of infected persons is likely to be increased but most importantly, it will require intensive health care units. This prediction is of crucial importance to plan for new facilities in hospitals. The very high incidence zones (Hotspots) are spatially located in the center of the city where population density is very high ([ 1000 persons/Km 2 ), crossed by number of major and minor roads, dense economic activities and high level of people's interaction. These conditions have further increased the probability of new cases which will grow the spatial extent of hotspot. In such predictions, travelling restrictions towards hotspots need to be implemented and extraordinary community measures must be taken within and outside the hotspots. Currently, lock down has minimized the interaction of infected and healthy persons. Similarly, the quarantine facility across the country has decrease the exposure of healthy citizens.

It is concluded from the study that Pakistan is still in a good position than many other countries to react to the current outbreak. The aggressive and systematic approach taken by the Government of Pakistan has played a crucial role and resulted in containing the spread. Similarly, good health facilities have recovered more than 2500 infected persons. The quarantine centers are supporting the existing health system. It is predicted in the study that if the exponential trend continues for the next few weeks then the health system will be not able to respond. The measures taken by the government are in the right direction, but prediction tells us that they need to be implemented without delay. Otherwise, a substantial number will become inevitable.

Spatially, the COVID-19 incidence is very high (Hotspot) in central and south west of district Lahore. The incidence is moderate in the south and west. Towards north incidence is decreasing forming cold spot with low incidence of COVID-19. The spatial extent of hotspot seems to extend towards the borders of Lahore. In this regard, research on emergency bases is highly needed to minimize the spread of COVID-19 pandemic.

The study further concludes that the maximum number of patients will require intensive health care units. The spatio-temporal pattern and prediction of COVID-19 is of crucial importance to plan for new facilities in Pakistani hospitals and to calculate the time period in which they need to be available. Finally, lock down has minimized the interaction of infected and healthy persons. Similarly, the quarantine facility has decreased the exposure of healthy citizens to COVID-19. But still, the entire population in main cities particularly in Lahore is at high risk due to the rapidly growing COVID-19 confirmed cases.

Conflict of interest There is no conflict of interest.

Ethical approval There are no ethical issue related to the data and manuscript. 

Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia

Automated COVID-19 diagnosis and classification using convolutional neural network with fusion based feature extraction model

Predicting mortality rate and associated risks in COVID-19 patients

Corona Virus (COVID-19) ''Infodemic'' and emerging issues through a data lens: The case of China

COVID-19 and Italy: What next? The Lancet

The first 2019 novel coronavirus case in Nepal. The Lancet Infectious Diseases

Tourism flows and death rates suggest covid-19 is being under-reported

History is repeating itself, a probable zoonotic spillover as a cause of an epidemic: The case of 2019 novel Coronavirus

Time to use the p-word? Coronavirus enter dangerous new phase

Viral induced demyelination

Does viral disease underlie ALS? Lessons from the AIDS pandemic

Could a herpesvirus be the cause of Kawasaki disease?

Characterization of a novel coronavirus associated with severe acute respiratory syndrome

Virology: SARS virus infection of cats and ferrets

COVID-19): A literature review

Characteristics of health care personnel with COVID-19-United States

Identification of a new human coronavirus

Presumed asymptomatic carrier transmission of COVID-19

Real-time nowcast and forecast on the extent of the Wuhan CoV outbreak, domestic and international spread

Going global -travel and the 2019 novel coronavirus

Mapping the incidence of the COVID-19 hotspot in Iran-Implications for Travellers

Analysis of COVID-19 infections on a CT image using deepsense model. Frontiers in Public Health

A deep learning method to forecast COVID-19 outbreak

Spatial prediction and mapping of the COVID-19 hotspot in India using geostatistical technique

National Institute of Health, Islamabad. Ministry of National Health Services, Regulation and Coordination

The analysis of spatial association by use of distance statistics

Local spatial autocorrelation statistics: Distributional issues and an application

The analysis of spatial association by use of distance statistics

Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations