RESEARCH PROJECTS view all projects →

A comprehensive analytic framework for COVID-19 mortality in long term care facilities and neighborhoods applicable to major metropolitan centers

Project statusOngoing
Contact Michael D Cailas
Institution web pagehttps://pubhealthgis.uic.edu/profiles/michael-cailas/
Host institution School of Public Health, University of Illinois at Chicago
Team members Greg Arling (Purdue University), Matthew Blaser, Michael D. Cailas, John R. Canar, Brian Cooper, Joel Flax-Hatch, Peter J. Geraci, Kristin M. Osiecki, Apostolis Sambanis
Funding information (if funded) This project has been funded by the Public Health Geographic Information Systems program at the Health Policy and Administration Division of the School of Public Health, University of Illinois at Chicago
Project Summary

The goal of our project is to develop and apply a comprehensive analytical framework for analyzing COVID-19 mortality locally within neighborhoods and long-term care facilities in a major metropolitan center. During the current public health crisis, information on COVID-19 mortality has typically been reported for the overall population, at single time points, and without regard to locations such as long-term care facilities or other congregate settings. In our study of COVID-19 mortality in Cook County IL (Chicago and suburbs) we examined patterns in COVID-19 mortality over time at the neighborhood level (postal Zip codes) in households and in nursing homes and other long-term care facilities (LTCFs). This framework provides:

  • reliable estimates of commonly quoted COVID-19 mortality indicators;
  • a better understanding of spatial and temporal distribution of COVID-19 deaths;
  • accurate depictions of the role of race, ethnicity, and socioeconomic status in COVID-19 mortality; and
  • population and organizational parameters that can inform strategies for public health interventions.

Cook County, the primary setting for this study is a large urban area, located in the US Midwest, with 5.15 million people and covering 1,635 square miles. It has rich racial and socio-economic diversity, and it has a multiplicity of neighborhoods and LTCFs spread across the county.

The major source of mortality data for this study is the Medical Examiner (ME) Case Archive of COVID-19-related Deaths. This archive is organized in a searchable online database format and contains residential address and other information about decedents (age, gender, race, and cause of death) for all deaths that occurred in Cook County from January 2020 to present.

Another source of data is the COVID-19 Nursing Home Data, maintained by the Centers for Medicare and Medicaid Services (CMS), which contains a count of COVID-19 cases and deaths, updated weekly, among residents and staff in each of the 15,000+ nursing facilities in the United States. Information about nursing facility characteristics comes from Medicare’s Nursing Home Compare and from Brown University’s LTC Focus data system. Data on overall mortality come from the Johns Hopkins Coronavirus Resource Center.

Data on population size, age, race/ethnicity, and socio-economic status are obtained from the US Census at the block group level, and then aggregated to the postal zip code level for much of the analysis.

Data preparation and statistical analysis are performed with the IBM® SPSS® Modeller 18.2.1. Geocoding, data projections, geospatial data integration, mapping, and initial spatial analysis are performed using ESRI’s ArcGIS Pro.

Outputs / Expected Outputs

In the first phase of our study, we examined the 1st wave of COVID-19 (Spring and early Summer) in order to identify temporal and spatial patterns of mortality over time in Cook County. Our framework provided a reliable estimation of high-risk LTCFs and neighborhoods, while avoiding the distortions caused by the commonly used overall population figures.

Chicago began experiencing a 2nd COVID-19 wave, along with the rest of the US in October. Our analysis of 2nd wave mortality shows a repeat of the patterns from the 1st wave. Mortality rates have been highest in many of the same high-risk neighborhoods containing disproportionately more people from racial/ethnic minorities, older people, and people with low SES. The surge in mortality in LTCFs tracked closely with the surge overall within the county. However, there was no significant association between neighborhoods experiencing high rates of mortality and neighborhoods containing LTCFs with the highest mortality rates. Wave 2 mortality in LTCFs has thus far been difficult to predict; neither 1st waive mortality rates nor facility characteristics were significantly associated with 2nd wave mortality in LTCFs.

Nonetheless, findings from this phase of our research point to the urgency of immediate action to prevent an acceleration of COVID-19 cases and consequent deaths both in LTCFs and high-risk neighborhoods, especially neighborhoods with concentrations of minority group residents. Moreover, distribution of vaccines should be prioritized not only to nursing home residents and staff, as currently planned, but also residents of high-risk neighborhoods.

Within this analytical framework, currently research is underway to examine the second wave within a large geographic region such at the North Central Region of the United States; commonly known as the Midwest; which is almost half of the surface area of the European Union countries.

In addition, as COVID-19 vaccines roll-out over the next several months, we will monitor their uptake and effectiveness by tracking mortality rates in neighborhoods and LTCFs in Cook County, and more generally across the Midwest.

Project website https://pubhealthgis.uic.edu/covid-19-dashboard-maps/
Supporting File 1 Analyzing_COVID-19_Methods_508-1.pdf
Supporting File 2 OJPHI_12.2020.pdf

KEYWORDS / CATEGORIES

Countries
Care setting
Funding type
Impact/outcomes
Intervention types
Methods
Groups/organisations

A comprehensive analytic framework for COVID-19 mortality in long term care facilities and neighborhoods applicable to major metropolitan centers