image reference : https://www.cdc.gov
Dr. Murtaza Haider – Ryerson University, Toronto
New York Governor, Andrew Cuomo, opened a Pandora’s box on March 22, 2020 when he identified New York City’s dense built environment as an enabling factor in the spread of COVID-19. “There is a density level in NYC [New York City] that is destructive,” he wrote in a tweet. The Governor further implored the City to “develop an immediate plan to reduce density” (Cuomo, 2020).
The following day, the New York Times ran a story with the headline, “Density Is New York City’s Big ‘Enemy’ in the Coronavirus Fight” (Rosenthal, 2020). The Times noted that NYC is the densest city in the US, with an average population density of 28,000 residents per square mile followed by San Francisco at 17,000. The comparative statistics the Times quoted showed that NYC had far more confirmed cases of coronavirus than other populous yet less dense urban centres.
Many urbanists responded to the Governor’s criticism of density. Emily Badger, responding in the New York Times, warned that “It will be a shame if we come away from this moment skeptical of density itself, or if some of the benefits of density, like mass transit and bustling commercial corridors, suffer lasting damage” (Badger, 2020).
The debate about the association between high density and the spread of infectious diseases might be new to the popular press. However, this has been a subject of inquiry in the academic and professional press for decades. One of the earlier investigations, relying on an abstract mathematical model, observed that “in general a threshold density of population is found to exist, which depends upon the infectivity, recovery, and death rates, peculiar to the epidemic. No epidemic can occur if the population density is below this threshold value” (Kermack, McKendrick, and Walker, 1927).
This brief report will not settle the density-disease debate. Instead, it aims to explore the statistical correlation between various measures of density and the spread of coronavirus in the United States. Relying on measures of aspatial and spatial correlation, this report attempts to determine whether a statistical relationship exists between population density and the prevalence of coronavirus.
COVID-19 is an infectious disease caused by the severe acute respiratory syndrome associated with Novel Coronavirus that was first reported in Wuhan, China, in late 2019. By March 2020, the World Health Organization (WHO) classified COVID-19 as a pandemic. Though pandemics occur infrequently, advanced anticipatory planning plays an essential role in preventing the spread of disease while trying to minimize associated morbidity and mortality.
Earlier evaluation of the clinical characteristics of hospitalized patients with Novel Coronavirus–Infected Pneumonia in Wuhan, China, indicated that older adults and people of any age who have severe underlying medical conditions were likely to be at higher risk for severe illness from COVID-19 (Wang et al. 2020; WTO 2020; Zhou et al., 2020).
Meanwhile, a secondary conversation is questioning whether the built form contributed to the spread of this pandemic. A rich body of empirical research emphasizes the importance of built environment characteristics influencing public health outcomes (Sallis et al. 2012; Saelens and Handy 2008; Brownson et al. 2009). The literature, though, is vague about any relationship. Understanding the association between the built form (especially in urban settings) and disease transmission may prove useful in devising responses and interventions to limit the spread of the disease. Urbanization can affect disease transmission through increased contact rates and altered socioeconomic conditions (Zhang and Atkinson, 2008). The spread of a communicable disease in an urban environment can be the result of individuals’ interactions in a geospatial context. Dynamic and intense spatial interaction leads to large concentrations of people at risk in spatial clusters along the urban landscape. For example, infected individuals, who may be asymptomatic, can be responsible for disease transmission at schools, colleges, universities, and malls (Perez and Dragicevic, 2009).
Urban population density, a frequently recognized proxy of built form, has been the subject of earlier scholarship for its role in the spread of infectious diseases. Places with relatively higher densities are deemed more vulnerable to infectious diseases (Moore et al., 2016). Human interaction in high density (shared) environments, such as malls, and the use of collective resources, such as public transit, are considered channels through which viral infections may spread faster.
This report investigates the relationship between urban density and the number of patients confirmed with COVID-19 across 3,075 counties in the United States. County-level data for confirmed cases was collected from John Hopkins University’s coronavirus resource center database and includes a total of 300,254 cases in 3,075 counties that were reported between January 21 and April 04, 2020.
We hypothesize that counties with higher population density are more vulnerable to the spread of infectious diseases due to clustering, crowding, and increased social interactions. Furthermore, the same high-density characteristics of the built environment do not facilitate adherence to preventive measures, such as “social distancing.” We experimented with numerous proxies for urban density and tested their association with the spread of disease at the county level. The density measures included population density, activity density — places and facilities where people usually concentrate and interact with others such as “arts and entertainment,” “education facilities,” “healthcare facilities,” “hotels and restaurants” and “retail stores”. We also considered public transit stop density as a network-related density that is highly related to other urban density measures.
Since the county-based data is inherently spatial, we have relied on spatial statistical analysis to determine whether population density recorded at a county is spatially correlated with confirmed cases of coronavirus in neighbouring counties. The Bivariate Local Moran’s I (Figure 1) is, therefore, a generalization of measures of spatial autocorrelation (Anselin, 2003).
Descriptive statistics of the collected data from 3,075 counties indicate a total of 300,254 confirmed COVID-19 cases as of April 04, 2020. The maximum number of confirmed cases in a single county is 57,159, and the mean number of cases per county is 87.
Relying on correlation analysis, we explore the associations between urban density and confirmed COVID-19 cases in US counties. To normalize the effect of population size for each county, we use the “number of cases per 1,000 persons” in each county as our dependent variable.
Results from correlation analysis indicate a significant positive association between all measures of urban density and COVID-19 confirmed cases. The highest correlations are reported for the Density of Arts/Entertainment business in counties. Moreover, it seems that counties with a higher density of hotels and restaurants, and retail are associated with higher confirmed cases. As expected, counties with higher concentrations of health care facilities are more likely to report patients infected with COVID-19. This is possibly an outcome of patients from other locations being transported to counties with better health facilities.
The output from the Bivariate Local Moran’s I is presented in Figure 1. The map displays the following five distinct trends between population density in a county and the spatially weighted average of confirmed cases of coronavirus per 1,000 residents in neighbouring counties:
1. High-High (Red): High population density correlated with a high weighted average of confirmed cases in neighbouring counties.
2. Low-Low (Blue): Low population density correlated with a low weighted average of confirmed cases in neighbouring counties.
3. Low-High (Light Blue): Low population density correlated with a high weighted average of confirmed cases in neighbouring counties.
4. High-Low (Pink): High population density correlated with a low weighted average of confirmed cases in neighbouring counties.
5. Not Significant (Light Gray): The correlation is not statistically significant.
The primary mode of transmission of COVID-19 is through direct physical contact between individuals. Since physical contact is likely to occur more frequently in places with high population density and more intense (diverse) land uses, it is argued that higher population densities are correlated with a higher occurrence of infections and vice versa. For this reason, developing an improved understanding of the complex relationships between built environment proxies and the transmission rates of infectious diseases is needed. This study shows that the spread of COVID-19 in the US is more pronounced in places with relatively high urban densities and less so in places with low population density.
Anselin, L. (2003). An introduction to spatial autocorrelation analysis with GeoDa. Spatial Analysis Laboratory, University of Illinois. http://www.dpi.inpe.br/gilberto/tutorials/software/geoda/tutorials/spauto.pdf
Badger, E. (2020, March 24). Density Is Normally Good for Us. That Will Be True After Coronavirus, Too. The New York Times. https://www.nytimes.com/2020/03/24/upshot/coronavirus-urban-density-risks.html.
Brownson, Ross C., Christine M. Hoehner, Kristen Day, Ann Forsyth, and James F. Sallis. 2009. Measuring the Built Environment for Physical Activity. American Journal of Preventive Medicine 36 (4): S99-S123.e12. https://doi.org/10.1016/j.amepre.2009.01.005.
Cuomo, Andrew. Twitter post. March 22, 2020, 11:36 a.m. https://twitter.com/NYGovCuomo/status/1241750717939007490
Kermack, W. O., McKendrick, A. G., & Walker, G. T. (1927). A contribution to the mathematical theory of epidemics. Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and Physical Character, 115(772), 700–721. https://doi.org/10.1098/rspa.1927.0118.
Moore, Melinda, Bill Gelfeld, Adeyemi Okunogbe, Christopher Paul, and Bill Gelfeld. 2016. Identifying Future Disease Hot Spots: Infectious Disease Vulnerability Index. Santa Monica, Calif: RAND.
Perez, Liliana, and Suzana Dragicevic. 2009. An Agent-Based Approach for Modeling Dynamics of Contagious Disease Spread. International Journal of Health Geographics 8 (1): 50. https://doi.org/10.1186/1476-072X-8-50.
Rosenthal, B. M. (2020, March 23). Density Is New York City’s Big Enemy in the Coronavirus Fight. The New York Times. https://www.nytimes.com/2020/03/23/nyregion/coronavirus-nyc-crowds-density.html.
Saelens, Brian E., and Susan L. Handy. 2008. Built Environment Correlates of Walking: A Review. Medicine & Science in Sports & Exercise 40 (Supplement): S550–66. https://doi.org/10.1249/MSS.0b013e31817c67a4.
Sallis, James F., Myron F. Floyd, Daniel A. Rodríguez, and Brian E. Saelens. 2012. Role of Built Environments in Physical Activity, Obesity, and Cardiovascular Disease. Circulation 125 (5): 729–37. https://doi.org/10.1161/CIRCULATIONAHA.110.969022.
Wang, Dawei, Bo Hu, Chang Hu, Fangfang Zhu, Xing Liu, Jing Zhang, Binbin Wang, et al. 2020. Clinical Characteristics of 138 Hospitalized Patients With 2019 Novel Coronavirus–Infected Pneumonia in Wuhan, China. JAMA 323 (11): 1061. https://doi.org/10.1001/jama.2020.1585.
WHO. 2020. Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19). World Health Organization.
Zhang, Ping, and Peter M. Atkinson. 2008. Modelling the Effect of Urbanization on the Transmission of an Infectious Disease. Mathematical Biosciences 211 (1): 166–85. https://doi.org/10.1016/j.mbs.2007.10.007.
Zhou, Fei, Ting Yu, Ronghui Du, Guohui Fan, Ying Liu, Zhibo Liu, Jie Xiang, et al. 2020. Clinical Course and Risk Factors for Mortality of Adult Inpatients with COVID-19 in Wuhan, China: A Retrospective Cohort Study. The Lancet, March, S0140673620305663. https://doi.org/10.1016/S0140-6736(20)30566-3.
If you want to know more about the project please download the report from Here