|
Asad Moten, Medical Sciences Division, University of Oxford;
Vivian YeE, Harvard University, Cambridge MA, USA; Daniel Schafer and Elizabeth Collins, Healthnovations International, Houston TX, USA |
|
Published: 30 November 2021
Review process: Open Peer Review Draft First Uploaded: 30 November 2021 See reviewers' comments |
Open Access
|
Abstract
Disparities in the provision of healthcare have a negative impact on socially vulnerable populations, both nationally and internationally. During the COVID-19 pandemic, social determinants of health (SDHs) have played a significant role in COVID-19 incidence and fatality. It is necessary to understand the full impacts of social vulnerability and government intervention on COVID-19 outcomes. Here, we show the correlation between the twin pandemics of adverse SDHs and COVID-19 on the one hand, and the efficacy of public health intervention in mitigating this double hit on vulnerable communities on the other. Utilising the United States Centers for Disease Control and Prevention's Social Vulnerability Index (SVI), we categorised New York City counties into three vulnerability cohorts. We then performed an analysis and fitted a Susceptible-Infected-Recovered-Deceased (SIRD) model to daily cases, deaths, and hospitalisation data reported by the New York City Department of Health and Mental Hygiene and the SIRD model for the months of March and May 2020. Our results demonstrate that more socially vulnerable populations appear to experience greater COVID-19 cases and deaths and that specific city, state, or US federal government intervention correlated with reduced disparities in COVID-19 outcomes. Moreover, moderately vulnerable communities observed the greatest rate of COVID-19 cases and deaths and highly vulnerable communities exhibited the greatest cumulative number of COVID-19 cases and deaths. These findings may encourage public health interventions which can increase health equity in various settings.
Keywords: COVID-19, pandemic, Social Determinants of Health, health inequality, emergency preparedness
Keywords: COVID-19, pandemic, Social Determinants of Health, health inequality, emergency preparedness
1. Introduction
Socially vulnerable groups including low-income communities, minority groups, immigrants, refugees, children, and the elderly are historically prone to public health emergencies such as disease outbreaks and natural disasters – as are individuals who lose family members in those crises. The disparities in risk are heavily influenced by social determinants of health (SDHs): non-biological conditions such as socioeconomic class, political and cultural factors; and accessibility to nutrition, healthcare, education, housing, and employment that impact health – or, in simpler terms, the conditions in which people are born, grow, live, work, and age (World Health Organization, 2008). The effects of SDHs, in turn, are mediated by the allocation of power, resources, and money within society (New England Journal of Medicine Catalyst, 2017).
1.1 SDHs and Emergency Preparedness
The disparities in SDHs between social groups have been shown to correlate with disparities in health outcomes. Countries and regions with limited-capacity public health and medical infrastructure are particularly vulnerable to poor health, due to lack of access to quality healthcare – for geographic or financial inaccessibility – or a lack of medical provision; accessing quality healthcare is even deemed unacceptable by certain cultures (Figure 1) (Lee et al., 2015; Preda & Voigt, 2015; Pheage, 2017; Baah et al., 2018; Compton et al., 2018; Palmer et al., 2019). Historically, vulnerable health systems such as that of Sierra Leone and Myanmar have been susceptible to pandemics, natural disasters, and other health emergencies, and often encounter such barriers as improper equipment, inadequate numbers of basic and specialised staff, and a lack of facility capacity and equipment (Mills, 2011; Scheffler, 2018; Ma & Vervoot, 2020).
Figure 1. The framework of factors that affect healthcare accessibility (Peters et al., 2008).
1.2 COVID-19 Disparities in the United States
Figure 2. Timeline of pertinent COVID-19 public health interventions and correlating daily case counts in NYC.
Compiled from Intarasuwan et al. (Source: NBC News).
Compiled from Intarasuwan et al. (Source: NBC News).
While countries with limited public health capacity tend to be – though are not always – more vulnerable to pandemics, the COVID-19 pandemic has also demonstrated that countries with higher gross domestic products (GDPs), like the United States, do not adequately protect against severe impacts of health emergencies. There are disparities in the availability of hospital beds, treatments, critical care, medical staff, and the types of medical conditions across neighbourhoods; countries also have differing susceptibilities and ways of responding to health emergencies. (Rosenthal et al., 2020; Schellekens & Sourrouille, 2020). In some cases policy interventions have sought to increase accessibility to healthcare in the COVID-19 pandemic, yet there is a lack of understanding of how these interventions have impacted COVID-19 outcomes such as disparities in COVID-19 cases and deaths between communities (Figure 2) (Viner et al., 2020; Weible et al., 2020).
2. Materials & Methods
2.1 NYC as a Microcosm – and categorising counties by SVI
We chose NYC as the region of focus for this study. The once epicentre of the American COVID-19 outbreak provides some of the strongest data sets of COVID-19 in the US: the counties within NYC observe the same social distancing policy and data collection techniques, and NYC is noted for its socioeconomic diversity, which is pivotal to this study of SDHs (Furman Center for Real Estate and Urban Policy, 2011; Economic Policy Institute, 2018). The socioeconomic variation in NYC allows for consideration of a greater range of social vulnerability backgrounds.
To measure social vulnerability, the United States (US) Centers for Disease Control and Prevention (CDC) has crafted a metric called the Social Vulnerability Index (SVI), which accounts for 15 social factors observed to affect health and wellbeing (Flanagan et al., 2018). The CDC assigns each county a numerical indication of social vulnerability to health complications (Figure 3).
To measure social vulnerability, the United States (US) Centers for Disease Control and Prevention (CDC) has crafted a metric called the Social Vulnerability Index (SVI), which accounts for 15 social factors observed to affect health and wellbeing (Flanagan et al., 2018). The CDC assigns each county a numerical indication of social vulnerability to health complications (Figure 3).
Figure 3. Map of Social Vulnerability Index of US counties (Flanagan et al., 2018).
In this article, we study the implications of SDHs on COVID-19 transmission and mortality rates in reported data and SIRD model simulations, which use the SVI to account for predispositions that might affect health outcomes. Where the majority of previous studies focuses only on one single indicator, the use of the SVI here produces a greater analysis of overall social vulnerability at a county level, by accounting for 15 social factors (Flanagan et al., 2018; Hendryx & Luo, 2020; Chowkwanyun & Reed, 2020; Rosenthal et al., 2020; Schellekens & Sourrouille, 2020). Specifically, we analyse COVID-19 cases, deaths, and hospitalisation data in different counties to model the spread of COVID-19 within counties in New York City (NYC), grouped by SVI. To explore the significance of local, state, and federal responses to minimise inequality in COVID-19 outcomes in NYC, we compare COVID-19 trends in March 2020 and May 2020 to analyse the evolution of the disease.
The CDC's SVI was employed to categorise NYC counties; its 15 social factors are categorised into four tracts (Figure 4). The SVI was chosen because it is the most comprehensive and well-established metric for social vulnerability. This metric is reported on a county level, and it has been tailored for use in public health (Flanagan et al., 2018).
The CDC's SVI was employed to categorise NYC counties; its 15 social factors are categorised into four tracts (Figure 4). The SVI was chosen because it is the most comprehensive and well-established metric for social vulnerability. This metric is reported on a county level, and it has been tailored for use in public health (Flanagan et al., 2018).
Figure 4. Factors that affect Social Vulnerability Index (Flanagan, 2018).
The NYC counties were grouped into three categories – low vulnerability (LV), moderate vulnerability (MV), and high vulnerability (HV) – based on their reported SVI (Figure 5). This was accomplished by taking the range of SVIs per NYC borough and dividing the range into three equivalent groups. The population of each county is as follows: 2,736,074 (Kings County), 2,405,464 (Queens County), 1,694,251 (New York County), 1,472,654 (Bronx County), and 495,747 (Richmond County).
Figure 5. NYC counties categorized by SVI. The most vulnerable NYC boroughs are
prominently characterised by socioeconomic strife and marginalised demographics.
prominently characterised by socioeconomic strife and marginalised demographics.
2.2 Time Period Stratification and Determination of Trends in Reported Data
We evaluated the impacts of federal resources, of changes in infrastructure capabilities and in healthcare accessibility, and of social distancing measures on the course of COVID-19 in these counties, during the months of March and May 2020. NYC COVID-19 response capabilities evolved significantly between these months. In March 2020, federal resources were deployed in NYC, and the Coronavirus Aid, Relief, and Economic Security Act (CARES) was enacted (Trump, 2020; United States Department of Treasury, n.d.). By May 2020, NYC reported 28,490 COVID-19 cases, a decrease from 109,432 in the previous month (New York City Department of Health and Mental Hygiene, 2020; Siedner et al., 2020). Differences in COVID-19 transmission and mortality rates between the cohorts during this period may indicate whether the public health interventions were more helpful for certain social groups than for the socially very vulnerable.
To analyse trends in COVID-19 outcomes between cohorts, we compiled data from the NYC Department of Health (DoH) on the daily number of new cases, daily deaths at any location following a positive COVID-19 laboratory test, and daily hospitalisations during March and May 2020 (New York City Department of Health and Dental Hygiene, 2020). The number of daily deaths was subtracted from the number of daily hospitalisations on the same date to estimate daily recoveries as data on recoveries from COVID-19 infection were not reported (Martin, 2020; White, 2020; Hall, 2020). This method relies on the assumption that those who are hospitalised but do not succumb to COVID-19 have recovered. COVID-19 data collected by the NYCDoH are noted to be robust and having a three-day lag time (New York City Department of Health and Dental Hygiene, 2020). We calculated the daily cases per 10,000 people, and daily deaths and daily recoveries per 100,000 people, to examine variations in the populations. GraphPad Prism was used to plot the data; we graphed the linear line-of-best fit using the least-squares method, and computed the Pearson correlation coefficient (r) and p-value to determine the strength of the linear fit.
To analyse trends in COVID-19 outcomes between cohorts, we compiled data from the NYC Department of Health (DoH) on the daily number of new cases, daily deaths at any location following a positive COVID-19 laboratory test, and daily hospitalisations during March and May 2020 (New York City Department of Health and Dental Hygiene, 2020). The number of daily deaths was subtracted from the number of daily hospitalisations on the same date to estimate daily recoveries as data on recoveries from COVID-19 infection were not reported (Martin, 2020; White, 2020; Hall, 2020). This method relies on the assumption that those who are hospitalised but do not succumb to COVID-19 have recovered. COVID-19 data collected by the NYCDoH are noted to be robust and having a three-day lag time (New York City Department of Health and Dental Hygiene, 2020). We calculated the daily cases per 10,000 people, and daily deaths and daily recoveries per 100,000 people, to examine variations in the populations. GraphPad Prism was used to plot the data; we graphed the linear line-of-best fit using the least-squares method, and computed the Pearson correlation coefficient (r) and p-value to determine the strength of the linear fit.
2.3 SVI-Stratified SIRD Model: An Innovative Approach
To model the evolution of COVID-19 spread within the vulnerability cohorts, we applied the SIRD model. The SIRD model contains the compartments susceptible (S), infected (I), recovered (R), and deceased (D) (Figure 6). This specific model was chosen because it was appropriate for the type of data we collected from the NYC DoH on daily cases, deaths, and hospitalisations.
Figure 6. Graphical representation of the interaction between different stages of the SIRD model.
S(t), I(t), R(t), and D(t) represent the fraction of the total population within the compartment. The parameters denoted by Greek letters are as follows: β denotes the transmission rate for susceptible individuals; ξ denotes the recovery rate for identified individuals; and τ denotes the mortality rate for identified individuals.
2.4 SVI-Stratified SIRD Model: An Innovative Approach
The model parameters were determined using data from March and May 2020 from the NYC DoH and Rt.live, a COVID-19 transmission tracker that analyses data from the Covid Tracking Project (New York City Department of Health and Dental Hygiene, 2020; Systrom & Vladeck, 2020). We determined the population of each cohort with 2019 US Census data (United States Census Bureau, n.d.). The transmission coefficient (β) was calculated with a sub-sampling technique developed by Kirkeby et al. (2017). This method estimates the transmission coefficient with greater accuracy than the traditional Poisson distribution method and only requires sub-samples for linear epidemiological models (Kirkeby et al., 2017). The mortality coefficient (τ) and recovery coefficient (ξ) were estimated using a data-based method for the SIRD model derived by Anastassopoulou et al. (2020). The Fourth Order Runge-Kutta method with a step size of one day was used to solve the differential equations to determine the number of individuals within a specific compartment on a given day. The fractions of the population within each compartment are converted into actual population numbers to account for variation in cohort populations. The initial parameters calculated with the above methods are put into the model for each simulation (Table 1).
Table 1. Calculated SIRD model parameters and initial values for each SIRD model simulation.
3. Results
3.1 Reported COVID-19 Cases and Deaths
The data for the daily cases per 10,000 individuals, deaths per 100,000 individuals, and recoveries per 100,000 individuals related to COVID-19 were compiled for each vulnerability cohort and graphed (Figure 7).
Figure 7. COVID-19 cases and deaths in low, moderate, and high vulnerability groups:
A. Reported daily cases by vulnerability cohort March 2020
B. Reported daily cases by vulnerability cohort May 2020
C. Reported daily deaths by vulnerability cohort March 2020
D. Reported daily deaths by vulnerability cohort May 2020
E. Reported daily recoveries by vulnerability cohort March 2020
F. Reported daily recoveries by vulnerability cohort May 2020
In March 2020, the three vulnerability groups (LV, MV, HV) displayed an increasing trend in daily COVID-19 cases per 10,000 people, daily COVID-19 deaths per 100,000 people, and daily COVID-19 recoveries per 100,000 people (Figures 7A, 7C, and 7E). In May 2020, figures for cases and deaths decreased for all three vulnerability cohorts (Figures 7B and 7D). It is important to note that the lack of correlation between vulnerability level and daily recoveries may by impacted by the techniques utilised to estimate recovery rate as opposed to a true weakening in relationship. The average numbers of daily COVID-19 cases, deaths, and recoveries in the MV and HV groups were greater than those of the LV group, suggesting that more socially vulnerable communities were disproportionately affected by COVID-19.
3.2 Moderate Vulnerability Population
Applying the SIRD model, the MV cohort was the most affected by COVID-19 across the infected, recovered, and deceased compartments compared with the other LV or HV cohorts (Figure 8). Evidence of this can be seen in the changes in the fraction of the population within each compartment. In the simulation, the MV group had the greatest percentage of the susceptible population that would contract and perish from COVID-19 (Figures 8A, 8B, 8G, and 8H).
Figure 8. SIRD models:
A. Susceptible population evolution March 2020
B. Susceptible population evolution May 2020
C. Infected population evolution March 2020
D. Infected population evolution May 2020
E. Recovered population evolution March 2020
F. Recovered population evolution May 2020
G. Deceased population evolution March 2020
H. Deceased population evolution May 2020
A. Susceptible population evolution March 2020
B. Susceptible population evolution May 2020
C. Infected population evolution March 2020
D. Infected population evolution May 2020
E. Recovered population evolution March 2020
F. Recovered population evolution May 2020
G. Deceased population evolution March 2020
H. Deceased population evolution May 2020
In addition, the calculated modelling parameters provided valuable information on rates of infection, mortality, and recovery. In March 2020, the LV observed the smallest transmission, mortality, and recovery rates of the three vulnerability cohorts (Figure 9). All parameters decreased in May 2020 with MV and HV nearing LV parameters. Thus, this suggests that the disparities in COVID-19 transmission and mortality were greater in March 2020 than in May 2020. Examining outcomes for the two time periods, we note the success of treatment and public health measures in not only minimising the spread of COVID-19, but also in reducing discrepancies between the different cohorts.
Figure 9. Calculated transmission, mortality, and recovery rates for SIRD model runs for low, moderate, and high vulnerability cohorts:
A. Transmission coefficient
B. Mortality coefficient
C. Recovery coefficient
3.3 High Vulnerability Cohort
Across all simulated months for cases and deaths, the HV group had the greatest cumulative number of cases and deaths followed by MV and LV groups (Figures 10A and 10B). By 31 May 2020, the cumulative numbers of COVID-19 cases for each of the vulnerability groups, in descending vulnerability order, were 100,352 (HV), 61,185 (MV), and 38,346 (LV). The cumulative numbers of COVID-19 related deaths, in descending vulnerability order, were 14,754 (HV), 9,788 (MV), and 5,378 (LV) (Figure 9C). The more vulnerable groups are simulated to experience the greatest number of cases and deaths due to the respective populations of each vulnerability group.
Figure 10. SIRD model cumulative cases and deaths:
A. Cumulative COVID-19 cases and deaths per SIRD simulation March 2020
B. Cumulative COVID-19 cases and deaths of per SIRD simulation May 2020
C. Cumulative COVID-19 cases and deaths of per SIRD simulation March 1, 2020, to May 31, 2020.
A. Cumulative COVID-19 cases and deaths per SIRD simulation March 2020
B. Cumulative COVID-19 cases and deaths of per SIRD simulation May 2020
C. Cumulative COVID-19 cases and deaths of per SIRD simulation March 1, 2020, to May 31, 2020.
4. Discussion
The objective of this study was to utilise data on COVID-19 cases, deaths, and recoveries in NYC counties grouped by SVI to model trends in disparities of COVID-19 transmission and mortality rates. We found that socially vulnerable communities are more prone to COVID-19, as demonstrated by the higher rates of COVID-19 transmission and mortality. The MV appeared to exhibit the greatest percentage of individuals falling ill and perishing to COVID-19, and the HV observed the greatest cumulative COVID-19 death toll. Additionally, the differences in data for March 2020 and May 2020 suggest that government intervention on a local, state, and federal level helped to reduce inequality in COVID-19 outcomes between vulnerability cohorts. This data utilising a social vulnerability index, however, must be viewed with the understanding that environmental factors – such as air quality, availability of public green spaces, and other factors which are distinct from social factors – can potentially have an impact on COVID-19 transmission and mortality rates.
4.1 Social Vulnerability and Trends in COVID-19 Outcomes
Our study builds upon previous publications that have drawn connections between individual social vulnerability indicators and the efficacy of COVID-19 testing, treatment, and government assistance, by employing the CDC’s SVI to account for 15 social vulnerability factors (Flanagan et al., 2018; Chowkwanyun & Reed, 2020; Han et al., 2020). Significant r-values and p-values indicate that the most vulnerable communities experience the worst outcomes (Figure 7). As noted, more vulnerable communities lack access to preventative healthcare resources; individuals in these communities are therefore more susceptible to chronic disease, novel pathogens, and severe COVID-19 infection and comorbidities (Figure 11) (Oates et al., 2018; Newman et al., 2019; Abrams & Szefler, 2020).
Figure 11. Prevalence of adults with two or more chronic conditions in NYC from 2011-2016 (Newman et al., 2019).
Highly vulnerable populations are also more likely to receive poorer COVID-19 treatment; this was evident in the disparities in COVID-19 mortality within our simulations (Finch & Finch, 2020). As such, individuals in the most vulnerable communities are more likely to contract COVID-19 through community transmission, to develop a severe infection due to comorbidities, to receive poorer care at under-resourced local healthcare facilities, and, subsequently, to die of COVID-19.
4.2 Health Burden on MV Cohort
Our data suggest that in NYC, the MV group saw a greater percentage of its population affected by and perishing from COVID-19 (Figures 7 and 8). This may be due to the fact that Queens County, one of the MV group, has the greatest percentage of individuals in NYC with no health insurance, who are therefore more likely to suffer from poorer health (United States Census Bureau, n.d.). Health insurance coverage is not captured by the SVI. While the most vulnerable groups may receive health insurance coverage through government welfare programs such as Medicaid, recent trends suggest that the number of moderately vulnerable individuals who are uninsured is growing (Tanne, 2006; Komisar, 2013). The lack of healthcare coverage for moderately vulnerable communities means that they will struggle to access adequate healthcare, and this may drive the trend in high COVID-19 incidence and morbidity within these communities (Figures 7 and 8). The reasons for this trend are multifactorial and influenced by the unique circumstances and aspects of the microcosm, such as age structure, lifestyle, and attitude toward the pandemic. Additionally, this moderately vulnerable group may be composed of individuals just above the Medicaid qualification threshold, but less likely to have access to employer-based health insurance. This could present a barrier and/or deterrent to accessing care.
4.3 Health Surveillance Limitations
Data collected by the NYC DoH and SIRD simulations suggest that the impact of COVID-19 on the HV is less severe than what we had hypothesised (Figures 7 and 8). We offer one possible explanation for this: that the lack of health surveillance mechanisms made it difficult to collect data on COVID-19 cases and deaths in the most vulnerable cohort (Magnusson, 2017). In addition to this lack of infrastructure, shortages in COVID-19 testing equipment have disproportionately impacted under-resourced communities (Finch & Finch, 2020; Vesoulis, 2020). Testing costs, limited testing capabilities, slow result returns, inaccuracy of early tests, and the lack of health education may have contributed to cases and mortalities in the HV group going underreported (Lazar & Davenport 2018; Goldstein et al., 2020). Recent efforts to use antibody testing to retroactively determine the COVID-19 death toll have found that COVID-19 infections in New York may have been ten times higher than reported (Syal, 2020). Underreporting is no doubt a serious concern. If significant portions of the most vulnerable communities are unaccounted for, healthcare organisations, politicians, and the public will not be able to visualise the full extent of the inequality now pervading the healthcare system.
4.4 Public Health Intervention on Healthcare Inequality
Our investigation provides unique insights into the effects of healthcare interventions on COVID-19 transmission, mortality, and cumulative caseload in different populations in NYC; it examines figures before and after these interventions were implemented – figures in March 2020 and May 2020 respectively (Figure 2). Our analysis of NYC reported data and of SIRD model simulations revealed that the disparities in COVID-19 outcomes decreased from March 2020 to May 2020, particularly in mortality rates (Figures 7 and 8). At the early stages of the pandemic, socially vulnerable communities lacked crucial access to adequate COVID-19 testing and treatment facilities (United States Department of Health and Human Services, 2020). This is reflected in the changes in the SIRD model coefficients: the month of March 2020 saw significantly greater rates of COVID-19 transmission and mortality in the MV and HV cohorts when compared to the LV cohort (Figure 9). Several interventions in NYC sought to address this disparity. These included federal support for the overburdened healthcare system, and legislation that allocated greater resources to more vulnerable communities. Results from May 2020 indicate that individuals in the MV and HV were no more likely than LV to perish from COVID-19 – evidence that the public health interventions were successful in reducing caseload (Figures 9 and 2).
4.5 Limitations
Three limitations in this study should be noted:
- First, the discrepancies between the cumulative case and mortality numbers of the SIRD model simulations, and the actual data for March 2020, may be attributed to inadequate testing capabilities at the early stages of the outbreak; the U.S. government did not have the means to gather sufficient data on the spread of COVID-19. Testing shortages in NYC has led to inconsistent data on cases, deaths, and recoveries. Indeed, the government’s subsequent attempts to increase testing capabilities, coupled with a heightened awareness of COVID-19, has led to an increase in reported cases, which cannot be explained by one single transmission probability (Global Change Data Lab, 2020).
- The second limitation is that the SVI can only measure social vulnerability at the county-level; it is not able to measure differences between smaller geographical areas, such as neighbourhoods, even though heterogeneity does exist within. In addition, the variation in population size between the three cohorts may impact the given results. We note that ecological fallacy may be at play by which the larger the geographical region examined, the greater chance that the correlations noted may not necessarily result from the independent variable examined but perhaps other factors at play (Robinson, 1950; Selvin, 1958).
- The New York City Department of Health and Mental Hygiene did not report daily recoveries from COVID-19 infection. In addition, there are also many limitations to estimating this value such as asymptomatic infection and lack of access to health surveillance infrastructure. For this reason, our ability to estimate this value was hindered.
4.6 Future Work
We make four points on how research could be developed in this area:
- The approach we took of grouping communities by vulnerability and SIRD model compartments could be applied to future studies of COVID-19 – or of other diseases resurging in the US and beyond.
- Research could also be directed towards analysing the impact of COVID-19 on smaller geographical regions, such as sub-boroughs and zip codes, between which disparities in outcomes have been found.
- As more data on COVID-19 transmission, infection, mortality, and recovery become available, the accuracy of the SVI-stratified SIRD model and its parameters will improve.
- This study focused on social vulnerability rather broadly through SVI. Future studies could analyse social vulnerability at the granular level, by examining individual factors in the SVI.
5. Conclusion: an action plan for critical healthcare and equity
This study presents evidence that disparities in COVID-19 health outcomes are linked to social vulnerability and that public health interventions can successfully minimise these disparities. Given these findings regarding social determinants of health, it is important to note that adverse social determinants of health can be directly combatted with effective public health intervention. We call for more preventative mechanisms to combat the higher rates of disease transmission in vulnerable communities. Beyond basic healthcare access, it is necessary to consider such mechanisms can include disease surveillance, healthcare education, and community outreach; the provision of food, housing, and sanitary resources to vulnerable populations would also help to improve people’s health and material circumstances. In particular, a more robust health surveillance system would enable actors to respond to health emergencies in good time. This could be achieved by improving the infrastructure of critical care and increasing people’s access to this care. By increasing equity of health infrastructure and reallocating resources to resource-poor settings, we are able to reduce population vulnerability to disease outbreak like COVID-19. Below we suggest ways in which healthcare might be improved.
5.1 Critical Care Facilities
Constructing primary and critical healthcare infrastructures is necessary if we are to be prepared for public health emergencies. As this study has shown, the lack of critical care facilities and expertise in resource-poor settings correlates with greater COVID-19 morbidity. Building trust between healthcare professionals and underserved patients, basic healthcare centres and adequate intensive care unit (ICU) beds in moderate and highly vulnerable communities may help better prepare these communities for future disease outbreaks. This need for more critical care facilities extends beyond the United States; it is felt on a global scale where greater disparities exist. In countries and regions that lack public healthcare infrastructure, critical care is uncoordinated, poorly funded, and in many cases non-existent. For example, each of the Sub-Saharan African nations, besides South Africa, has just over 2,000 ICU beds (Ma & Vervoot, 2020). Critical care infrastructure could drastically reduce the numbers of deaths from critical illness.
5.2 Equitable Health Surveillance
Our study identified that weak health surveillance prevented the accurate tracking of disease spread and its impact on marginalised communities. The lack of rigorous health surveillance in the US (i.e. testing) at the early stages of the COVID-19 pandemic resulted in increased numbers of infections and deaths (Rosenthal, 2020). Using health surveillance infrastructure in moderate and highly vulnerable communities would enable policymakers to gather more reliable data. In turn, this data could help to inform public health interventions, model disease spread, and thereby minimise the impacts of health emergencies. The data might also reveal inequalities in healthcare as between vulnerable communities and less vulnerable communities.
5.3 Develop and Deliver Vaccines
The socially vulnerable groups identified in this study are often underrepresented in vaccine trials. This has already been noted in many COVID-19 clinical trials, including the Moderna vaccine, which saw a lack of minority group participants (Cerullo, 2020; Chastain et al., 2020). To engineer effective vaccines, vaccine trials must involve participants that are representative of the population, including those who come from historically marginalised communities (Clark et al., 2019). The distribution of vaccines must be equitable, too, if a government wishes to achieve herd immunity (Schmidt et al., 2021). Policymakers should therefore establish an infrastructure to ensure that the vaccine is made accessible to all citizens. Policymakers should also consider how they might finance and deliver these vaccines, identify high-risk groups, and establish a sustainable supply chain for an effective vaccination campaign (Weintraub et al., 2020).
5.4 Provide Community Outreach and Education
Another barrier to COVID-19 testing and treatment in socially vulnerable communities is the lack of health education: knowledge of what resources are available and how they can be utilised. The lack of education, and distrust towards public health entities, may negatively influence the effectiveness of interventions to quell disease spread (Corbie-Smith & Ford, 2006). Education is vital for ensuring that vulnerable groups have ready access to healthcare resources. Outreach programmes can help to engage these groups and build trust between them and public health organisations. In turn, this will lead to vulnerable communities being able to access health resources. Policymakers ought especially to prioritise working with local leaders and community organisations that can advise on and promote the implementation of COVID-19 campaigns in those vulnerable communities.
6. Acknowledgements
The authors of this work would like to acknowledge support from the United States Department of Defense, National Institutes of Health, National Science Foundation, World Federation of Intensive and Critical Care Medicine, University of Oxford Medical Sciences Division, and Center for Excellence in Education. Additionally, we thank Darrell Gaskin and Roland Thorpe of Johns Hopkins University and Alex Shalek of the Broad Institute and Ragon Institute of Massachusetts General Hospital, Massachusetts Institute of Technology, and Harvard University for guidance and contribution. The views presented are those of the authors and do not necessarily represent the views of the United States Government and Department of Defense or its components.
References
Abrams, E. M., & Szefler, S. J. (2020). COVID-19 and the impact of social determinants of health. The Lancet Respiratory Medicine, 8(7), 659–661. https://doi.org/10.1016/s2213-2600(20)30234-4
Anastassopoulou, C., Russo, L., Tsakris, A., & Siettos, C. (2020). Data-based analysis, modelling and forecasting of the COVID-19 outbreak. https://doi.org/10.1101/2020.02.11.20022186
Arino, J., & Portet, S. (2020). A simple model for COVID-19. Infect Dis Model, 5, 309–315. https://doi.org/10.1016/j.idm.2020.04.002
Baah, F. O., Teitelman, A. M., & Riegel, B. (2018). Marginalization: Conceptualizing patient vulnerabilities in the framework of social determinants of health-an integrative review. Nursing Inquiry, 26(1), e12268. https://doi.org/10.1111/nin.12268
Cerullo, M (2020, August 25). Moderna vaccine trial lacks Black, Latinx and Indigenous participants. CBS News. https://www.cbsnews.com/news/covid-vaccine-moderna-trial-lacking-black-latino-asian-native-volunteers/
Chastain, D. B., Osae, S. P., Henao-Martínez, A. F., Franco-Paredes, C., Chastain, J. S., Young, H. N. (2020). Racial Disproportionality in COVID Clinical Trials. New England Journal of Medicine, 383, p.e59. https://doi.org/10.1056/NEJMp2021971
Chowkwanyun, M., & Reed, A. L. (2020). Racial health disparities and COVID-19 — Caution and context. New England Journal of Medicine, 383(3), 201–203. https://doi.org/10.1056/nejmp2012910
Clark, L. T., Watkins, L., Piña, I. L., Elmer, M., Akinboboye, O., Gorham, M., Jamerson, B., McCullough, C., Pierre, C., Polis, A. B., Puckrein, G., & Regnante, J. M. (2020). Corrigendum to ‘Increasing diversity in clinical trials: Overcoming critical barriers’. Current Problems in Cardiology, 44(5), 148–172. https://doi.org/10.1016/j.cpcardiol.2020.100647
Compton, B., Barash, D. M., Farrington, J., Hall, C., Herzog, D., Meka, V., Rafferty, E., Taylor, K., & Varghese, A. (2018). Access to medical devices in low-income countries: Addressing sustainability challenges in medical device donations. NAM Perspectives, 8(7). https://doi.org/10.31478/201807a
Corbie-Smith, G., & Ford, C. L. (2006). Distrust and poor self-reported health. Canaries in the coal mine? Journal of General Internal Medicine, 21(4), 395–397. https://doi.org/10.1111/j.1525-1497.2006.00407.x
Economic Policy Institute. (2018). The new gilded age: Income inequality in the U.S. by state, metropolitan area, and County. https://www.epi.org/publication/the-new-gilded-age-income-inequality-in-the-u-s-by-state-metropolitan-area-and-county/
Finch, W. H., & Hernández Finch, M. E. (2020). Poverty and COVID-19: Rates of incidence and deaths in the United States during the first 10 weeks of the pandemic. Frontiers in Sociology, 5. https://doi.org/10.3389/fsoc.2020.00047
Flanagan, B.E., Hallsey, E.J., Adams, E., Lavery, A. (2018) Measuring community vulnerability to natural and anthropogenic hazards: The Centers for Disease Control and Prevention’s Social Vulnerability Index. J Environ Health. 80(10), 34–36. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7179070/
Furman Center for Real Estate and Urban Policy. (2011). The Changing State of the New York City’s Property Tax Racial and Ethnic Makeup of New York City Neighborhoods. https://furmancenter.org/files/sotc/The_Changing_Racial_and_Ethnic_Makeup_of_New_York_City_Neighborhoods_11.pdf
Giordano, G., Blanchini, F., Bruno, R., Colaneri, P., Di Filippo, A., Di Matteo, A., & Colaneri, M. (2020). Modelling the COVID-19 epidemic and implementation of population-wide interventions in Italy. Nat Med, 26, 855–860. https://doi.org/10.1038/s41591-020-0883-7
Global Change Data Lab. (2020). Total COVID-19 tests conducted vs. confirmed cases. Our World in Data. https://ourworldindata.org/grapher/covid-19-total-confirmed-cases-vs-total-tests-conducted
Goldstein, J., Ferré-Sadurní, K., & Randel, A. (2020, March 11). Coronavirus in N.Y.: Desperate for a test, they Couldn’t get one. The New York Times. https://www.nytimes.com/2020/03/11/nyregion/coronavirus-testing-newyork.html
Hall, T. (2020, March 27). Why aren't COVID-19 recoveries being reported? NBC15. https://www.nbc15.com/content/news/Why-arent-COVID-19-recoveries-being-reported-569177691.html
Han, J., Meyer, B. D., & Sullivan, J. X. (2020). Income and poverty in the COVID-19 pandemic. The Brookings Institution, 1–49. https://www.brookings.edu/wp-content/uploads/2020/06/Han-et-al-conference-draft.pdf
Hendryx, M., & Luo, J. (2020). COVID-19 prevalence and mortality rates in association with Black race and segregation in the United States April 1 to April 15, 2020. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3582857
Intarasuwan, K., Vazquez, J., Shea, T., Rajamani, M., & Price, B. (2020, March 5). Timeline: Tracking the spread of COVID-19 in tri-state. NBC New York. https://www.nbcnewyork.com/news/local/timeline-tracking-the-spread-of-covid-19-in-tri-state/2313123/
Ivorra, B., Ferrandez, M. R., Vela-Perez, M., & Ramos A. M. (2020). Mathematical modeling of the spread of the coronavirus disease 2019 (COVID-19) taking into account the undetected infections. The case of China. Commun Nonlinear Sci Numer Simul. 88, 105303. https://doi.org/10.1016/j.cnsns.2020.105303
Kirkeby, C., Halasa, T., Gussmann, M., Toft, N., & Græsbøll, K. (2017). Methods for estimating disease transmission rates: Evaluating the precision of Poisson regression and two novel methods. Scientific Reports, 7(1). https://doi.org/10.1038/s41598-017-09209-x
Komisar, H. (2013). The Effects of Rising Health Care Costs on Middle-Class Economic Security. AARP®. https://www.aarp.org/content/dam/aarp/research/public_policy_institute/security/2013/impact-of-rising-healthcare-costs-AARP-ppi-sec.pdf
Lazar, M., & Davenport, L. (2018). Barriers to health care access for low income families: A review of literature. Journal of Community Health Nursing, 35(1), 28–37. https://doi.org/10.1080/07370016.2018.1404832
Lee, H., Kim, S., DeMarco, R., Aronowitz, T., Mtengezo, J., Kang, Y., Yang, Y., Touch, C., & Fitzpatrick, J. J. (2015). Recognizing global disparities in health and in health transitions in the 21st century: What can nurses do? Applied Nursing Research, 28(1), 60–65. https://doi.org/10.1016/j.apnr.2014.09.004
Ma, X., & Vervoot, D. (2020, May 4). COVID-19 and the critical shortage in critical care. Global Health NOW. https://www.globalhealthnow.org/2020-05/covid-19-and-critical-shortage-critical-care
Magnusson, R. (2017). Advancing the right to health: the vital role of law. World Health Organization. https://www.who.int/healthsystems/topics/health-law/health_law-report/en/
Martin, C. (2020, May 19). Why aren’t coronavirus recoveries always reported? Dallas News. https://www.dallasnews.com/news/public-health/2020/05/19/why-arent-coronavirus-recoveries-always-reported/
Mills, A. (2011). Health systems in low- and middle-income countries. The Oxford Handbook of Health Economics, 29–57. https://doi.org/10.1093/oxfordhb/9780199238828.013.0003
New England Journal of Medicine Catalyst. (2017). Social determinants of health (SDOH). New England Journal of Medicine. https://catalyst.nejm.org/doi/full/10.1056/CAT.17.0312
Newman, D., Levine, E., Kishore, S. P. (2019). Prevalence of multiple chronic conditions in New York State, 2011–2016. PLOS ONE, 14(2), e0211965. https://doi.org/10.1371/journal.pone.0211965
New York City Department of Health and Mental Hygiene. (2020). COVID-19: Data. https://www1.nyc.gov/site/doh/covid/covid-19-data.page
Oates, G. R., Jackson, B. E., Partridge, E. E., Singh, K. P., Fouad, M. N., & Bae, S. (2017). Sociodemographic patterns of chronic disease: How the mid-south region compares to the rest of the country. American Journal of Preventive Medicine, 52(1), S31–S39. https://doi.org/10.1016/j.amepre.2016.09.004
Palmer, R. C., Ismond, D., Rodriquez, E. J., & Kaufman, J. S. (2019). Social determinants of health: Future directions for health disparities research. American Journal of Public Health, 109(S1), S70–S71. https://doi.org/10.2105/ajph.2019.304964
Pheage, Tefo. (2016, December). Dying from lack of medicines. United Nations Africa Renewal. https://www.un.org/africarenewal/magazine/december-2016-march-2017/dying-lack-medicines
Preda, A., & Voigt, K. (2015). The social determinants of health: Why should we care? The American Journal of Bioethics, 15(3), 25–36. https://doi.org/10.1080/15265161.2014.998374
Robinson, W.S. (1950). Ecological Correlations and the Behavior of Individuals. American Sociological Review, 1(3): 351–357. https://doi.org/10.2307/2087176
Rosenthal, B. M., Goldstein, J., Otterman, S., & Fink, S. (2020, July 1). Why surviving the virus might come down to which hospital admits you. The New York Times. https://www.nytimes.com/2020/07/01/nyregion/Coronavirus-hospitals.html
Scheffler, R. M., Campbell, J., Cometto, G., Maeda, A., Liu, J., Bruckner, T. A., Arnold, D. R., & Evans, T. (2018). Forecasting imbalances in the global health labor market and devising policy responses. Human Resources for Health, 16(1). https://doi.org/10.1186/s12960-017-0264-6
Schellekens, P., & Sourrouille, D. (2020). COVID-19 mortality in rich and poor countries: A tale of two pandemics? Policy Research Working Papers. https://doi.org/10.1596/1813-9450-9260
Schmidt, H., Weintraub, R., Williams, M.A., Miller, k., Buttenheim, A., Sadecki, E., We, H., Doiphode, A., Nagpal, N., Gostin, L.O., & Shen, A.A. (2021). Equitable allocation of COVID-19 vaccines in the United States. Nature Medicine. https://doi.org/10.1038/s41591-021-01379-6
Selvin, Hanan C. (1958). Durkheim's Suicide and Problems of Empirical Research. American Journal of Sociology, 63(6): 607–619. https://doi.org/10.1086/222356
Siedner, M. J., Harling, G., Reynolds, Z., Gilbert, R. F., Haneuse, S., Venkataramani, A., & Tsai, A. C. (2020). Social distancing to slow the U.S. COVID-19 epidemic: Longitudinal pretest-posttest comparison group study. https://doi.org/10.1101/2020.04.03.20052373
Syal, A. (2020). CDC: Antibody tests show virus rates 10 times higher than reported. NBC News. https://www.nbcnews.com/health/health-news/cdc-antibody-tests-show-virus-rates-10-times-higher-reported-n1234480
Systrom, K., & Vladeck, T. (2020). Rt COVID-19. Rt COVID-19. https://rt.live
Tanne, J. H. (2006). Number of uninsured middle class US citizens grows. BMJ, 332(7549), 1047.2. https://doi.org/10.1136/bmj.332.7549.1047-a
Trump, D. J. (2020, March 22). Memorandum on providing federal support for governors' use of the national guard to respond to COVID-19. The White House. https://www.whitehouse.gov/presidential-actions/memorandum-providing-federal-support-governors-use-national-guard-respond-covid-19
United States Department of the Treasury. (n.d.). The CARES Act works for all Americans. https://home.treasury.gov/policy-issues/cares
United States Census Bureau. (n.d.). U.S. Census Bureau QuickFacts: New York City, New York; Bronx County (Bronx Borough), New York; Kings County (Brooklyn Borough), New York; New York County (Manhattan Borough), New York; Queens County (Queens Borough), New York; Richmond County (Staten Island Borough), New York. Census Bureau QuickFacts, https://www.census.gov/quickfacts/fact/table/newyorkcitynewyork,bronxcountybronxboroughnewyork,kingscountybrooklynboroughnewyork,newyorkcountymanhattanboroughnewyork,queenscountyqueensboroughnewyork,richmondcountystatenislandboroughnewyork/PST045219
United States Department of Health and Human Services. (2020). HHS Initiatives to Address the Disparate Impact of COVID-19 on African Americans and Other Racial and Ethnic Minorities. https://www.hhs.gov/sites/default/files/hhs-fact-sheet-addressing-disparities-in-covid-19-impact-on-minorities.pdf
Vesoulis, Abby (2020, March 11). Coronavirus may disproportionately hurt the poor—And that's bad for everyone. Time. https://time.com/5800930/how-coronavirus-will-hurt-the-poor
Viner, R. M., Russell, S. J., Croker, H., Packer, J., Ward, J., Stansfield, C., Mytton, O., Bonell, C., & Booy, R. (2020). School closure and management practices during coronavirus outbreaks including COVID-19: A rapid systematic review. Lancet Child Adolesc Health, 4(5), 397–404. https://doi.org/10.1016/s2352-4642(20)30095-x
Wangping, J., Ke, H., Yang, S., Wenzhe, C., Shengshu, W., Shanshan, Y., Jianwei, W., Fuyin, K., Penggang, T., Jing, L., Miao, L., & Yao, H. (2020). Extended SIR Prediction of the Epidemics Trend of COVID-19 in Italy and Compared With Hunan, China. Front Med, 7, 169. https://doi.org/10.3389/fmed.2020.00169
Weible, C. M., Nohrstedt, D., Cairney, P., Carter, D. P., Crow, D. A., Durnová, A. P., Heikkila, T., Ingold, K., McConnell, A., & Stone, D. (2020). COVID-19 and the policy sciences: Initial reactions and perspectives. Policy Sciences, 53(2), 225–241. https://doi.org/10.1007/s11077-020-09381-4
Weintraub, R., Yadav, P., & Berkley, S. (2020, April 2). A COVID-19 vaccine will need equitable, global distribution. Harvard Business Review. https://hbr.org/2020/04/a-covid-19-vaccine-will-need-equitable-global-distribution
White, R. (2020, March 29). Why Michigan isn't reporting recovered coronavirus cases yet. WZZM13.com. https://www.wzzm13.com/article/news/health/coronavirus/why-michigan-isnt-reporting-recovered-coronavirus-cases-yet/69-19a58dfe-8790-466a-aa48-048f20cbb450
World Health Organization (2008). Closing the gap in a generation: health equity through action on the social determinants of health. Final Report of the Commission on Social Determinants of Health. Geneva: World Health Organization http://whqlibdoc.who.int/ publications/2008/9789241563703_eng.pdf?ua=1
Anastassopoulou, C., Russo, L., Tsakris, A., & Siettos, C. (2020). Data-based analysis, modelling and forecasting of the COVID-19 outbreak. https://doi.org/10.1101/2020.02.11.20022186
Arino, J., & Portet, S. (2020). A simple model for COVID-19. Infect Dis Model, 5, 309–315. https://doi.org/10.1016/j.idm.2020.04.002
Baah, F. O., Teitelman, A. M., & Riegel, B. (2018). Marginalization: Conceptualizing patient vulnerabilities in the framework of social determinants of health-an integrative review. Nursing Inquiry, 26(1), e12268. https://doi.org/10.1111/nin.12268
Cerullo, M (2020, August 25). Moderna vaccine trial lacks Black, Latinx and Indigenous participants. CBS News. https://www.cbsnews.com/news/covid-vaccine-moderna-trial-lacking-black-latino-asian-native-volunteers/
Chastain, D. B., Osae, S. P., Henao-Martínez, A. F., Franco-Paredes, C., Chastain, J. S., Young, H. N. (2020). Racial Disproportionality in COVID Clinical Trials. New England Journal of Medicine, 383, p.e59. https://doi.org/10.1056/NEJMp2021971
Chowkwanyun, M., & Reed, A. L. (2020). Racial health disparities and COVID-19 — Caution and context. New England Journal of Medicine, 383(3), 201–203. https://doi.org/10.1056/nejmp2012910
Clark, L. T., Watkins, L., Piña, I. L., Elmer, M., Akinboboye, O., Gorham, M., Jamerson, B., McCullough, C., Pierre, C., Polis, A. B., Puckrein, G., & Regnante, J. M. (2020). Corrigendum to ‘Increasing diversity in clinical trials: Overcoming critical barriers’. Current Problems in Cardiology, 44(5), 148–172. https://doi.org/10.1016/j.cpcardiol.2020.100647
Compton, B., Barash, D. M., Farrington, J., Hall, C., Herzog, D., Meka, V., Rafferty, E., Taylor, K., & Varghese, A. (2018). Access to medical devices in low-income countries: Addressing sustainability challenges in medical device donations. NAM Perspectives, 8(7). https://doi.org/10.31478/201807a
Corbie-Smith, G., & Ford, C. L. (2006). Distrust and poor self-reported health. Canaries in the coal mine? Journal of General Internal Medicine, 21(4), 395–397. https://doi.org/10.1111/j.1525-1497.2006.00407.x
Economic Policy Institute. (2018). The new gilded age: Income inequality in the U.S. by state, metropolitan area, and County. https://www.epi.org/publication/the-new-gilded-age-income-inequality-in-the-u-s-by-state-metropolitan-area-and-county/
Finch, W. H., & Hernández Finch, M. E. (2020). Poverty and COVID-19: Rates of incidence and deaths in the United States during the first 10 weeks of the pandemic. Frontiers in Sociology, 5. https://doi.org/10.3389/fsoc.2020.00047
Flanagan, B.E., Hallsey, E.J., Adams, E., Lavery, A. (2018) Measuring community vulnerability to natural and anthropogenic hazards: The Centers for Disease Control and Prevention’s Social Vulnerability Index. J Environ Health. 80(10), 34–36. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7179070/
Furman Center for Real Estate and Urban Policy. (2011). The Changing State of the New York City’s Property Tax Racial and Ethnic Makeup of New York City Neighborhoods. https://furmancenter.org/files/sotc/The_Changing_Racial_and_Ethnic_Makeup_of_New_York_City_Neighborhoods_11.pdf
Giordano, G., Blanchini, F., Bruno, R., Colaneri, P., Di Filippo, A., Di Matteo, A., & Colaneri, M. (2020). Modelling the COVID-19 epidemic and implementation of population-wide interventions in Italy. Nat Med, 26, 855–860. https://doi.org/10.1038/s41591-020-0883-7
Global Change Data Lab. (2020). Total COVID-19 tests conducted vs. confirmed cases. Our World in Data. https://ourworldindata.org/grapher/covid-19-total-confirmed-cases-vs-total-tests-conducted
Goldstein, J., Ferré-Sadurní, K., & Randel, A. (2020, March 11). Coronavirus in N.Y.: Desperate for a test, they Couldn’t get one. The New York Times. https://www.nytimes.com/2020/03/11/nyregion/coronavirus-testing-newyork.html
Hall, T. (2020, March 27). Why aren't COVID-19 recoveries being reported? NBC15. https://www.nbc15.com/content/news/Why-arent-COVID-19-recoveries-being-reported-569177691.html
Han, J., Meyer, B. D., & Sullivan, J. X. (2020). Income and poverty in the COVID-19 pandemic. The Brookings Institution, 1–49. https://www.brookings.edu/wp-content/uploads/2020/06/Han-et-al-conference-draft.pdf
Hendryx, M., & Luo, J. (2020). COVID-19 prevalence and mortality rates in association with Black race and segregation in the United States April 1 to April 15, 2020. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3582857
Intarasuwan, K., Vazquez, J., Shea, T., Rajamani, M., & Price, B. (2020, March 5). Timeline: Tracking the spread of COVID-19 in tri-state. NBC New York. https://www.nbcnewyork.com/news/local/timeline-tracking-the-spread-of-covid-19-in-tri-state/2313123/
Ivorra, B., Ferrandez, M. R., Vela-Perez, M., & Ramos A. M. (2020). Mathematical modeling of the spread of the coronavirus disease 2019 (COVID-19) taking into account the undetected infections. The case of China. Commun Nonlinear Sci Numer Simul. 88, 105303. https://doi.org/10.1016/j.cnsns.2020.105303
Kirkeby, C., Halasa, T., Gussmann, M., Toft, N., & Græsbøll, K. (2017). Methods for estimating disease transmission rates: Evaluating the precision of Poisson regression and two novel methods. Scientific Reports, 7(1). https://doi.org/10.1038/s41598-017-09209-x
Komisar, H. (2013). The Effects of Rising Health Care Costs on Middle-Class Economic Security. AARP®. https://www.aarp.org/content/dam/aarp/research/public_policy_institute/security/2013/impact-of-rising-healthcare-costs-AARP-ppi-sec.pdf
Lazar, M., & Davenport, L. (2018). Barriers to health care access for low income families: A review of literature. Journal of Community Health Nursing, 35(1), 28–37. https://doi.org/10.1080/07370016.2018.1404832
Lee, H., Kim, S., DeMarco, R., Aronowitz, T., Mtengezo, J., Kang, Y., Yang, Y., Touch, C., & Fitzpatrick, J. J. (2015). Recognizing global disparities in health and in health transitions in the 21st century: What can nurses do? Applied Nursing Research, 28(1), 60–65. https://doi.org/10.1016/j.apnr.2014.09.004
Ma, X., & Vervoot, D. (2020, May 4). COVID-19 and the critical shortage in critical care. Global Health NOW. https://www.globalhealthnow.org/2020-05/covid-19-and-critical-shortage-critical-care
Magnusson, R. (2017). Advancing the right to health: the vital role of law. World Health Organization. https://www.who.int/healthsystems/topics/health-law/health_law-report/en/
Martin, C. (2020, May 19). Why aren’t coronavirus recoveries always reported? Dallas News. https://www.dallasnews.com/news/public-health/2020/05/19/why-arent-coronavirus-recoveries-always-reported/
Mills, A. (2011). Health systems in low- and middle-income countries. The Oxford Handbook of Health Economics, 29–57. https://doi.org/10.1093/oxfordhb/9780199238828.013.0003
New England Journal of Medicine Catalyst. (2017). Social determinants of health (SDOH). New England Journal of Medicine. https://catalyst.nejm.org/doi/full/10.1056/CAT.17.0312
Newman, D., Levine, E., Kishore, S. P. (2019). Prevalence of multiple chronic conditions in New York State, 2011–2016. PLOS ONE, 14(2), e0211965. https://doi.org/10.1371/journal.pone.0211965
New York City Department of Health and Mental Hygiene. (2020). COVID-19: Data. https://www1.nyc.gov/site/doh/covid/covid-19-data.page
Oates, G. R., Jackson, B. E., Partridge, E. E., Singh, K. P., Fouad, M. N., & Bae, S. (2017). Sociodemographic patterns of chronic disease: How the mid-south region compares to the rest of the country. American Journal of Preventive Medicine, 52(1), S31–S39. https://doi.org/10.1016/j.amepre.2016.09.004
Palmer, R. C., Ismond, D., Rodriquez, E. J., & Kaufman, J. S. (2019). Social determinants of health: Future directions for health disparities research. American Journal of Public Health, 109(S1), S70–S71. https://doi.org/10.2105/ajph.2019.304964
Pheage, Tefo. (2016, December). Dying from lack of medicines. United Nations Africa Renewal. https://www.un.org/africarenewal/magazine/december-2016-march-2017/dying-lack-medicines
Preda, A., & Voigt, K. (2015). The social determinants of health: Why should we care? The American Journal of Bioethics, 15(3), 25–36. https://doi.org/10.1080/15265161.2014.998374
Robinson, W.S. (1950). Ecological Correlations and the Behavior of Individuals. American Sociological Review, 1(3): 351–357. https://doi.org/10.2307/2087176
Rosenthal, B. M., Goldstein, J., Otterman, S., & Fink, S. (2020, July 1). Why surviving the virus might come down to which hospital admits you. The New York Times. https://www.nytimes.com/2020/07/01/nyregion/Coronavirus-hospitals.html
Scheffler, R. M., Campbell, J., Cometto, G., Maeda, A., Liu, J., Bruckner, T. A., Arnold, D. R., & Evans, T. (2018). Forecasting imbalances in the global health labor market and devising policy responses. Human Resources for Health, 16(1). https://doi.org/10.1186/s12960-017-0264-6
Schellekens, P., & Sourrouille, D. (2020). COVID-19 mortality in rich and poor countries: A tale of two pandemics? Policy Research Working Papers. https://doi.org/10.1596/1813-9450-9260
Schmidt, H., Weintraub, R., Williams, M.A., Miller, k., Buttenheim, A., Sadecki, E., We, H., Doiphode, A., Nagpal, N., Gostin, L.O., & Shen, A.A. (2021). Equitable allocation of COVID-19 vaccines in the United States. Nature Medicine. https://doi.org/10.1038/s41591-021-01379-6
Selvin, Hanan C. (1958). Durkheim's Suicide and Problems of Empirical Research. American Journal of Sociology, 63(6): 607–619. https://doi.org/10.1086/222356
Siedner, M. J., Harling, G., Reynolds, Z., Gilbert, R. F., Haneuse, S., Venkataramani, A., & Tsai, A. C. (2020). Social distancing to slow the U.S. COVID-19 epidemic: Longitudinal pretest-posttest comparison group study. https://doi.org/10.1101/2020.04.03.20052373
Syal, A. (2020). CDC: Antibody tests show virus rates 10 times higher than reported. NBC News. https://www.nbcnews.com/health/health-news/cdc-antibody-tests-show-virus-rates-10-times-higher-reported-n1234480
Systrom, K., & Vladeck, T. (2020). Rt COVID-19. Rt COVID-19. https://rt.live
Tanne, J. H. (2006). Number of uninsured middle class US citizens grows. BMJ, 332(7549), 1047.2. https://doi.org/10.1136/bmj.332.7549.1047-a
Trump, D. J. (2020, March 22). Memorandum on providing federal support for governors' use of the national guard to respond to COVID-19. The White House. https://www.whitehouse.gov/presidential-actions/memorandum-providing-federal-support-governors-use-national-guard-respond-covid-19
United States Department of the Treasury. (n.d.). The CARES Act works for all Americans. https://home.treasury.gov/policy-issues/cares
United States Census Bureau. (n.d.). U.S. Census Bureau QuickFacts: New York City, New York; Bronx County (Bronx Borough), New York; Kings County (Brooklyn Borough), New York; New York County (Manhattan Borough), New York; Queens County (Queens Borough), New York; Richmond County (Staten Island Borough), New York. Census Bureau QuickFacts, https://www.census.gov/quickfacts/fact/table/newyorkcitynewyork,bronxcountybronxboroughnewyork,kingscountybrooklynboroughnewyork,newyorkcountymanhattanboroughnewyork,queenscountyqueensboroughnewyork,richmondcountystatenislandboroughnewyork/PST045219
United States Department of Health and Human Services. (2020). HHS Initiatives to Address the Disparate Impact of COVID-19 on African Americans and Other Racial and Ethnic Minorities. https://www.hhs.gov/sites/default/files/hhs-fact-sheet-addressing-disparities-in-covid-19-impact-on-minorities.pdf
Vesoulis, Abby (2020, March 11). Coronavirus may disproportionately hurt the poor—And that's bad for everyone. Time. https://time.com/5800930/how-coronavirus-will-hurt-the-poor
Viner, R. M., Russell, S. J., Croker, H., Packer, J., Ward, J., Stansfield, C., Mytton, O., Bonell, C., & Booy, R. (2020). School closure and management practices during coronavirus outbreaks including COVID-19: A rapid systematic review. Lancet Child Adolesc Health, 4(5), 397–404. https://doi.org/10.1016/s2352-4642(20)30095-x
Wangping, J., Ke, H., Yang, S., Wenzhe, C., Shengshu, W., Shanshan, Y., Jianwei, W., Fuyin, K., Penggang, T., Jing, L., Miao, L., & Yao, H. (2020). Extended SIR Prediction of the Epidemics Trend of COVID-19 in Italy and Compared With Hunan, China. Front Med, 7, 169. https://doi.org/10.3389/fmed.2020.00169
Weible, C. M., Nohrstedt, D., Cairney, P., Carter, D. P., Crow, D. A., Durnová, A. P., Heikkila, T., Ingold, K., McConnell, A., & Stone, D. (2020). COVID-19 and the policy sciences: Initial reactions and perspectives. Policy Sciences, 53(2), 225–241. https://doi.org/10.1007/s11077-020-09381-4
Weintraub, R., Yadav, P., & Berkley, S. (2020, April 2). A COVID-19 vaccine will need equitable, global distribution. Harvard Business Review. https://hbr.org/2020/04/a-covid-19-vaccine-will-need-equitable-global-distribution
White, R. (2020, March 29). Why Michigan isn't reporting recovered coronavirus cases yet. WZZM13.com. https://www.wzzm13.com/article/news/health/coronavirus/why-michigan-isnt-reporting-recovered-coronavirus-cases-yet/69-19a58dfe-8790-466a-aa48-048f20cbb450
World Health Organization (2008). Closing the gap in a generation: health equity through action on the social determinants of health. Final Report of the Commission on Social Determinants of Health. Geneva: World Health Organization http://whqlibdoc.who.int/ publications/2008/9789241563703_eng.pdf?ua=1
A. Graphs of Simulated Daily COVID-19 Cases and Deaths
Figure 12.
A. Simulated daily COVID-19 cases by vulnerability cohort March 2020
B. Simulated daily COVID-19 cases by vulnerability cohort May 2020
C. Simulated daily COVID-19 deaths by vulnerability cohort March 2020
D. Simulated daily COVID-19 deaths by vulnerability cohort May 2020
A. Simulated daily COVID-19 cases by vulnerability cohort March 2020
B. Simulated daily COVID-19 cases by vulnerability cohort May 2020
C. Simulated daily COVID-19 deaths by vulnerability cohort March 2020
D. Simulated daily COVID-19 deaths by vulnerability cohort May 2020
B. Graphs of Simulated Daily COVID-19 Cases per 10,000 and Deaths per 100,000
Figure 13.
A. Simulated daily COVID-19 cases per 10,000 by vulnerability cohort March 2020
B. Simulated daily COVID-19 cases per 10,000 by vulnerability cohort May 2020
C. Simulated daily COVID-19 deaths per 100,000 by vulnerability cohort March 2020
D. Simulated daily COVID-19 deaths per 100,000 by vulnerability cohort May 2020
A. Simulated daily COVID-19 cases per 10,000 by vulnerability cohort March 2020
B. Simulated daily COVID-19 cases per 10,000 by vulnerability cohort May 2020
C. Simulated daily COVID-19 deaths per 100,000 by vulnerability cohort March 2020
D. Simulated daily COVID-19 deaths per 100,000 by vulnerability cohort May 2020
C. Simulation of May 2020 More Accurately Reflects Reported Case and Death Data than March.
To evaluate the efficacy of our model runs on estimating cumulative cases and deaths, the simulated data were plotted against reported data from the NYC DoH. Findings indicate that model runs of May 2020 more closely modelled the actual evolution of COVID-19 than runs in March 2020 (Figure 14).
Figure 14. Simulated vs. actual data.
A. Simulated vs. actual cumulative COVID-19 cases per SIRD simulation March 2020
B. Simulated vs. actual cumulative COVID-19 cases per SIRD simulation May 2020
C. Simulated vs. actual cumulative COVID-19 cases per SIRD simulation March 2020
D. Simulated vs. actual cumulative COVID-19 cases per SIRD simulation May 2020
A. Simulated vs. actual cumulative COVID-19 cases per SIRD simulation March 2020
B. Simulated vs. actual cumulative COVID-19 cases per SIRD simulation May 2020
C. Simulated vs. actual cumulative COVID-19 cases per SIRD simulation March 2020
D. Simulated vs. actual cumulative COVID-19 cases per SIRD simulation May 2020
D. Modeling COVID-19
To better understand the spread of COVID-19, epidemiologists have attempted to apply a variety of modelling paradigms to the pandemic. Utilising reported data, researchers have been able to decipher information that may guide the response to COVID-19. Models of the COVID-19 spread in various regions have primarily aimed to better understand the pathology of the disease, forecast short-term effects and advise public health interventions (Arino & Portet, 2020). In doing so, researchers are able to better understand the disease and how specific social practices or factors impact disease spread.
The most common general purpose models that have been applied to COVID-19 are the SIR model and SEIR model (Wangping et al., 2020). Both consider the susceptible (S), infected (I) and removed (R) populations, with the addition of the exposed (E) group in the SEIR model. Due to their straightforward nature and simplicity, they are easily applicable to all illnesses as the states are typically present in all outbreaks.
The θ-SEIHRD is a novel model developed by epidemiologists to take into account COVID-19 specific traits. The model contains the compartments susceptible (S), infected (I), infected undetected (Iu), hospitalised will recover (Hr), hospitalised will die (Hd), dead (D), recovered from detected infection (Rd) and recovered from undetected infection (Rr) (Ivorra et al., 2020). This particular model is specifically tailored to the COVID-19 pandemic as it accounts for the impact of undetected infectious individuals, the effects of hospital sanitary conditions, and the need for hospital beds which are all significant issues for COVID-19.
In response to the COVID-19 pandemic, a group of researchers developed the SIDARTHE model. The SIDARTHE model is a compartmental disease model that is made up of eight states: susceptible (S), infected (I), diagnosed (D), ailing (A), recognised (R), threatened (T), healed (H) and extinct (E) (Giordano et al., 2020). The addition of key stages allows for a more accurate representation of asymptomatic spread which has been a large concern for public health officials. In addition, the new model takes into account the different recovery rates based on the severity of the illness and the likelihood of an individual to progress onto more severe symptoms.
Mathematical models have been used to study the significance of public health intervention on the pandemic spread. However, there is a lack of understanding in regard to the usage of pandemic modelling to quantify the impact of SDOHs such as household income, race, or environment crime rate on health outcomes during the COVID-19 crisis. Evidence of inherent disparities can help bring attention to the necessity for interventions to bolster public health outreach in at-risk communities.
The most common general purpose models that have been applied to COVID-19 are the SIR model and SEIR model (Wangping et al., 2020). Both consider the susceptible (S), infected (I) and removed (R) populations, with the addition of the exposed (E) group in the SEIR model. Due to their straightforward nature and simplicity, they are easily applicable to all illnesses as the states are typically present in all outbreaks.
The θ-SEIHRD is a novel model developed by epidemiologists to take into account COVID-19 specific traits. The model contains the compartments susceptible (S), infected (I), infected undetected (Iu), hospitalised will recover (Hr), hospitalised will die (Hd), dead (D), recovered from detected infection (Rd) and recovered from undetected infection (Rr) (Ivorra et al., 2020). This particular model is specifically tailored to the COVID-19 pandemic as it accounts for the impact of undetected infectious individuals, the effects of hospital sanitary conditions, and the need for hospital beds which are all significant issues for COVID-19.
In response to the COVID-19 pandemic, a group of researchers developed the SIDARTHE model. The SIDARTHE model is a compartmental disease model that is made up of eight states: susceptible (S), infected (I), diagnosed (D), ailing (A), recognised (R), threatened (T), healed (H) and extinct (E) (Giordano et al., 2020). The addition of key stages allows for a more accurate representation of asymptomatic spread which has been a large concern for public health officials. In addition, the new model takes into account the different recovery rates based on the severity of the illness and the likelihood of an individual to progress onto more severe symptoms.
Mathematical models have been used to study the significance of public health intervention on the pandemic spread. However, there is a lack of understanding in regard to the usage of pandemic modelling to quantify the impact of SDOHs such as household income, race, or environment crime rate on health outcomes during the COVID-19 crisis. Evidence of inherent disparities can help bring attention to the necessity for interventions to bolster public health outreach in at-risk communities.