Rapid detection of COVID-19 clusters in the United States using a prospective space-time scan statistic
Alexander Hohl, Eric Delmelle, Michaël Desjardins
Abstract
Novel coronavirus disease 2019 (COVID-19), caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is a pandemic with 1,420,299 confirmed cases and 85,992 total deaths within the United States as of May 15th, 2020. As the number of cases continues to climb, detecting clusters of COVID-19 is critical to alleviate the strain on our public health system through improved resource allocation and decision-making. Here, we report on an analysis of daily case data at the county level using the prospective spatial-temporal scan statistic. In previous work, we performed the analysis for March 27th 2020 [1], and here we report updated results as of April 27th 2020, producing a new set of "active" and emerging clusters present. Our analysis resulted in sixteen significant space-time clusters of COVID-19 at the county level in the U.S. during the time span of March 22nd -- April 27th. The spacetime pattern of significant clusters mirrors active and emerging disease hot-spots at the end of our study period. The statistic can be rerun to support timely surveillance of COVID-19, as demonstrated here.