Robust Ratio Estimation with an Application to Covid-19 Data from Louisiana

The first case of Covid-19 was identified in Wuhan, China in December 2019. Since then, it has spread worldwide causing an ongoing pandemic. It has affected all nations profoundly, causing issues from crippling economies to mental health problems. Consequently, accurately estimating the Covid-19 infections and deaths have been crucial for epidemiologists, public health workers, and federal governments to be able to make public policies and combat the disease.

There are not many studies that use survey sampling procedures to estimate Covid-19 cases. Epidemiologists generally use infectious disease models to estimate cases and deaths. However, with novel infectious diseases, such as Covid19, the initial estimates regarding the cases and deaths can be erroneous. Besides, due to the evolving nature of Covid-19, it has been challenging to predict cases and deaths for different waves caused by different mutations. For example, while the Omicron variant was found to be more transmissible than the Delta variant, it was also observed to be less severe. Some of these infectious disease models, such as the agent-based model, have been criticized for their unrealistic assumptions. Therefore, health researchers might want to consider utilizing more simplistic approaches in estimating the cases or deaths, especially when there are too many unknowns at the beginning of pandemics.

Dr. Evrim Oral and her colleagues proposed a novel ratio estimator, demonstrated its efficiency and robustness, and used it to analyze the publicly available Covid-19 data reported by the New York Times.

Their newly published manuscript is available HERE