Managing the Risk of COVID-19 Using Model Based Predictors: The Case of North Dakota

Abstract

Abstract Objective: North Dakota (ND) had the highest coronavirus disease 2019 (COVID-19) case and mortality rate in the United States for nearly 2 mo. This study aims to compare 3 metrics ND used to guide public health action across its 53 counties. Methods: Daily COVID-19 case and death totals in North Dakota were evaluated using data from the COVID-tracker website provided by the North Department of Health (NDDoH). It was reported as: active cases per 10,000, tests administered per 10,000, and test positivity rate (the North Dakota health metric). The COVID-19 Response press conferences provided data for the Governor’s metric. The Harvard model used daily new cases per 100,000. A chi-squared test was used to compare differences in these 3 metrics on July 1, August 26, September 23, and November 13, 2020. Results: On July 1, no significant difference between the metrics was found. By September 23, Harvard’s health metric indicated critical risk while ND’s health metric was moderate risk, and the Governor’s metric was still low risk. Conclusions: ND’s and the Governor’s metric underrepresented the risk of the COVID-19 outbreak in North Dakota. The Harvard metric reflected North Dakota’s increasing risk; it should be considered as a national standard in future pandemics. Public Health Implications: Model-based predictors could guide policy-makers to effectively control spread of infectious disease; proactive models could reduce risk of disease as it progresses in vulnerable communities.