Economic security and health disparity in COVID-19: A computational modeling approach.

点击翻译按钮获取中文摘要

Project summary Job insecurity and disease risk are inextricably linked, and the SARS-CoV-2 pandemic has highlighted the interdependence between these two critical outcomes. On the disease transmission side, current models for disease transmission rest on variants of mass-action Susceptible – (Exposed) – Infected – Removed (SIR, SEIR) frameworks or curve-fitting models tuned to SIR dynamics. The best of these models use compartments that are deemed biologically relevant, such as age, but they typically do not include social relevance, effectively ignoring well-known segregation and social stratification barriers to interaction that likely channel infection. We urgently need models that accurately account for core population differences in risk and burden of disease. Disease exposure is deeply structured by the racial and ethnic segregation of communities, differences in living arrangements, and ability to avoid close personal contact with others, which are compounded by well-known health disparities and lead to poorer COVID-19 outcomes. By assuming away such features, we miss how unevenly the burden of disease and disease avoidance activity is shared across vulnerable populations. On the economic burden side, it is well-known that job insecurity is patterned by race and socioeconomic status in the United States. African Americans and Latinos are considerably more likely than whites to work in hourly-wage, precarious jobs, and as a result, these populations are particularly vulnerable to job loss, reductions in income and benefits, and other job-related cutbacks during economic retrenchments. Similarly, there are marked gradients along the wealth distribution in economic vulnerability resulting from deficits in savings needed to cover basic living expenses during periods of income reduction or loss. Importantly, the very same populations who are economically vulnerable are also at higher risk of contracting diseases like COVID-19. African Americans, Latinos, and other low-SES populations are at particularly high risk of becoming ill, being hospitalized, and dying of complications resulting from COVID-19. Importantly, behaviors resulting from job insecurity are likely to exacerbate disease risk; and disease is likely to exacerbate job insecurity. Most attempts to model these processes do not take this essential interdependence into account. We propose to build and test a fully integrated Agent Based Model (ABM) of disease spread and socio-economic outcomes. In Aim 1, we will build the ABM based on real social network and activity data that reflect the mix of strong ties, weak ties, and incidental personal contacts. In Aim 2, we fit the ABMs to observed epidemic patterns to identify key disparity-driving features. In Aim 3, we propose policy alternatives that can help identify inherent tradeoffs between public safety and economic hardship and how such outcomes are unequally distributed across working people in the country.