Skip to main content

Revealing fine-scale spatiotemporal differences in SARS-CoV-2 introduction and spread

Our take —

This study uses 247 viral sequences from SARS-CoV-2 patients in two different counties (Dane and Milwaukee) in Wisconsin to understand epidemic dynamics on a sub-state level. They found evidence for distinct epidemics with limited mixing between the two counties, and showed that introductions were more common in Dane county, though sampling bias may have affected this conclusion. They also found that the “Safer at Home” statewide order enacted on March 25, 2020, likely reduced case counts in both counties in April 2020.

Study design


Study population and setting

This study includes data from 247 SARS-CoV-2 genomes from two counties in the US state of Wisconsin, collected between January 30 and April 26, 2020. The two counties—Dane and Milwaukee—are both in Southern Wisconsin (127 km apart) and are the most populous counties in the state, despite having notably different demographics. Dane county was of particular interest because it was home to the 12th reported COVID-19 case in the United States. Additionally, a statewide “Safer at Home” order was enacted on March 25, 2020, and lasted throughout the duration of the study period, providing an opportunity for the authors to examine the effects of these social distancing requirements on the number of reported cases.

Summary of Main Findings

The authors found that the viruses in Dane and Milwaukee counties were genetically distinct, and that there was evidence of multiple introductions of SARS-CoV-2 into each county. However, they also found evidence for different epidemic dynamics in the two counties, with more introductions and viral diversity in Dane county. There was limited evidence for viral spread between the two counties, suggesting largely separate epidemics within the same state. They also claim that there was no onward transmission of the first SARS-CoV-2 case in Dane county (the 12th reported case in the United States), citing successful control practices. Finally, the study presents evidence for reduced transmission (due to a lower estimated reproductive number, R0) after the “Safer at Home” order enacted on March 25, 2020, suggesting this order was at least partially responsible for the decline in cases in April compared to earlier months.

Study Strengths

Data from two densely populated counties from within the same state allowed the authors to explore epidemic dynamics at a sub-state level. Genomic data as well as demographic data from both regions allowed the authors to hypothesize explanations for observed differences in transmission patterns.


This study has two problematic limitations: first, the conclusion of no onward transmission from the first case in Dane county relies heavily on the finding of an in-frame deletion in the SARS-CoV-2 genome isolated from this sample that has not been observed in any other samples. However, this deletion occurs in a poly-A homopolymer, and the sequencing technology used has known issues in homopolymer/low-complexity regions. The authors did not provide any details (e.g., read depth, validation on another platform) in support of this finding, and in general, provide minimal details on laboratory controls and sequence validation. Second, the authors only minimally addressed that biased sampling could artificially reduce the viral variation observed in Milwaukee county, did not provide details of their sampling strategy (which cases were selected for genomic sequencing) and did not control for population size/density in any way.

Value added

This study demonstrates that geographically close counties in the United States could have separate epidemics with limited mixing, and that non-pharmaceutical interventions such as the Wisconsin “Safer at Home” order likely played a role in reducing cases in April 2020.

This review was posted on: 9 December 2020