Chapter 5: A Simple Method for Estimating the Number of Unconfirmed COVID-19 Cases in a Local Area that Includes a Confidence Interval: A Case Study of Whatcom County, Washington¹

Background: Along with many other data problems affecting the unfolding of the COVID-19 pandemic in the United States, virtually nothing is known about the number of positive, unconfirmed cases, especially in local areas.

Methods: We show that it is possible to estimate the number of positive, unconfirmed COVID-19 cases using a simple, long-established method employed by demographers to estimate a population in the absence of a census count. We go on to show how a confidence interval can be constructed around an estimate of positive, unconfirmed COVID-19 cases constructed from this method, using Whatcom County, Washington as a case study.

Results: Because of the lag of approximately 3 weeks between infection and death, the method produces estimates that are lagged.

Conclusions: The method’s simplicity is a strength and without too much added complexity, we propose that the next step in researching this method is the development of a scale that would adjust the estimate, one that takes into account the stage of the pandemic and the time lag between infection and death.