COVID-19 and Systematic Risk in New York City

The emergence of COVID-19 has manifested to be a major global problem in 2020. In the U.S. especially, COVID-19 infections have grown exponentially, affecting many aspects of daily life. New York began as the epicenter of the outbreak in the U.S., with more than 230,000 citizens testing positive [1].

As New York City begins to reopen on a phase by phase basis, socioeconomic parameters were used to track how the infection is spreading, if at all, through selected systematic factors. Our parameters: age, sex, race, poverty and income cohesively determine a region’s social vulnerability - a given community’s ability to endure the negative effects of external burdens such as catastrophes and epidemics [2]. Once we detect any spatial cluster(s) of COVID-19 cases, we will look to explain these clusters.

Spatial-temporal analysis was used to observe the development of the virus throughout a duration of time. Confirmed cases, hospitalizations, and deaths were tracked since the first laboratory-confirmed case in the New York City area (February 29, 2020). Geographical Information Systems technology (GIS) was used to model disparities of COVID-19 geographical distribution with each individual parameter [3]. Mapping such data assists in identifying COVID-19 hotspots - areas in New York City that are of a high vulnerability to the virus. 

The parameters that were observed included: age, race, sex, poverty and income.

• Poverty : "Levels" were determined by if one made more or less than the Federal Poverty Threshold.
• Sex : Divided by the sex that one officially identified as however, there was not enough information for people who identified as transgender or gender non-conforming.
• Age : Divided into these age categories: 0-17, 18-44, 45-64, 65-74, and 75+. These age groups were divided as such because COVID-19 trends were similar within these groups.
• Race : Information on confirmed cases, hospitalizations, and probable and/or confirmed deaths were recorded.
• Income : Information was gathered on both the median and average salary of each zip code in New York City [4].
• Hotspot Analysis - Used to find where features with high or low values cluster spatially.
• Exploratory Regression - Used to find the best possible model with the case count as its dependent variable [3].
• Ordinary Least Squares Regression - Used to see how our explanatory variables relate to the COVID case incidence [3].
• Geographically Weighted Regression - Used to find out if our dependent variable is affected geographically by our dependent variables [3].

Our hot spot analysis demonstrates that there is in fact a geographical significance for the case distribution in New York City. The Ordinary Least Squares regression tool was applied on the best model that the exploratory regression tool calculated, returning a strong R^2 value of 0.88. The Geographically Weighted Regression analysis on that same model returned an adjusted R^2 value of 0.92. The explanatory variables that were part of the model were: population between 25 and 34 years, male population, black/African American population, and Hispanic/Latino population. The Hispanic/Latino population and black/African American population were the variables that appear to have the most relevant geographical impact.

Analysis of our data revealed information that can prove to be useful in combating COVID-19. It was observed that individuals of certain ethnic or socio-economic backgrounds, sex, and age were more susceptible to the virus. Ordinary Least Squares and Geographically Weighted Regression analysis results show a strong correlation between case incidence and Hispanic/Latino and black/African American people; this is consistent with the Center for Diseases Control's (CDC) Social Vulnerability Index (SVI). Hotspot analysis demonstrates the geographical significance of case distributions in NYC. Our analysis shows the most vulnerable locations in the city. This could be of value to decision makers when thinking of handling the pandemic. Future modeling can potentially improve by incorporating other variables such premedical conditions.

WHY?

Race : Hispanic/Latino & Black/African American

• Educational and wealth gaps, which also ties into an inability to access proper healthcare services [5]
• Tend to live in densely populated areas [6]

Age : 75+ years

• Underlying medical conditions [7]
• Inexperienced/weakened immune systems [7]
• White blood cell reduction [7]
• Inability to recognize viruses quickly [7]
• Ineffective antibodies [7]

Sex : Male vs. Female

• Stronger immune systems
• Less concentrations of harmful enzymes (ex. ACE2 )
• Men being less compliant with pandemic-related restrictions and not taking symptoms too seriously [8]

Income : $20K-175 K

• Working in high-risk areas
• Unable to access proper health services

1. N. D. o. Health, "COVID-19: Data," 21 July 2020. [Online]. Available:https://www1.nyc.gov/site/doh/covid/covid-19-data.page. [Accessed 21 July 2020]
2. A. f. T. S. a. D. Registration, "CDC's Social Vulnerability Index (SVI)," 12 September 2018. [Online]. Available: https://svi.cdc.gov/. [Accessed 21 July 2020].
3. esri. (2020). esri. Retrieved 2020, from ArcGis Pro:  https://pro.arcgis.com/en/pro-app/tool-reference/spatial-statistics/how-geographicallyweightedregression-works.ht ml
4. Morales, A., & Carney, K. (2020). Retrieved 2020, from Income by Zip Code: https://www.incomebyzipcode.com/
5. CDC. (2020, July 24). Retrieved 2020, from cdc.gov: https://www.cdc.gov/coronavirus/2019-ncov/community/health-equity/race-ethnicity.html
6. HOROWITZ, J. M. (2020, June 26). Pew Research Center. Retrieved from https://www.pewresearch.org/fact-tank/2020/06/26/views-on-why-black-americans-face-higher-covid-19-hospitalization-rates-vary-by-party-race-and-ethnicity/
7. Khan, A. (2020, March 29). Al Jazeera. Retrieved from https://www.aljazeera.com/indepth/features/doctors-note-older-people-vulnerable-covid-19-200329060131663.html
8. Curley, B. (2020, May 12). Retrieved 2020, from health line: https://www.healthline.com/health-news/men-more-susceptible-to-serious-covid-19-illnesses
9. NYC Health Data Repository  https://github.com/nychealth/coronavirus-data/ 
10. Health, N. D. (2020, August 7). NYC Health. Retrieved from https://www1.nyc.gov/site/doh/index.page
11. (Baruch College The William and Anita Newman Library, 2020)