Rurality and COVID-19 in Tennessee: Identifying metrics to assess and communicate pandemic spread

Authors: Ingrid Luffman*, East Tennessee State University, Andrew Joyner, East Tennessee State University, William Tollefson, East Tennessee State University, Abbey Mann, East Tennessee State University, Megan Quinn, East Tennessee State University
Topics: Health and Medical, Rural Geography, Spatial Analysis & Modeling
Keywords: Covid-19, rural, Tennessee, statistics, metrics
Session Type: Virtual Paper
Presentation File: No File Uploaded


Tennessee has 70 of 95 counties defined as mostly (>50%) or completely rural (US Census). Initially, rural counties experienced low incidence rates with cases in state prisons marking the first large outbreaks in congregate settings. This study aimed to identify metrics to assess COVID-19 transmission related to rurality from initial cases through school re-openings. Metrics under consideration included: days to first case, hospitalization and fatality; number of days between state and county peak incidence rates; and number of days to reach an incidence rate of 10/100,000. We evaluated different ways to classify counties by rurality, using classification schemes based on the binary HRSA classification, RHP Rural classification (3 classes), the US Census percent rural classification (3 classes), and various five-class schemes using population density. Using ANOVA, we determined that days to first case, hospitalization, and fatality were dependent on rurality for all classification schemes, with urban counties experiencing earlier cases, hospitalizations, and deaths than rural counties. The timing of county peak incidence rate vis-à-vis state peak cases was unrelated to rurality, and exclusion of counties housing state prisons (outliers) did not change the result. Surprisingly, the number of days for a county to reach an incidence rate of 10/100,000 did not depend on rurality, suggesting that rurality was important in initial spread but less important as COVID-19 became more widespread. Rurality classification scheme had little impact on whether metrics were statistically different across rurality classes.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login