Using excess deaths and testing statistics to determine COVID-19 mortalities

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Quantifying epidemic fatality is challenging due to varying mortality definitions, prevalence uncertainties, and sampling biases.
• Consistent estimation of infected populations and infection-related deaths is crucial for cross-regional mortality comparisons.

Purpose of the Study:

• To improve the estimation of epidemic mortality by integrating diverse data sources and modeling approaches.
• To provide a more accurate assessment of excess mortality compared to reported infectious disease deaths.

Main Methods:

• Combined historical and current mortality data.
• Employed a statistical testing model.
• Utilized an SIR (Susceptible-Infected-Recovered) epidemic model.

Main Results:

• The average excess death in the US (Jan 2020-Feb 2021) was 9% higher than reported COVID-19 deaths.
• Some regions, like New York City, saw weekly deaths increase up to eightfold.
• Peru, Ecuador, Mexico, and Spain showed significantly higher excess deaths than reported.
• Germany, Denmark, and Norway had insignificant or negative excess deaths for most of 2020.

Conclusions:

• The developed methodology offers a more robust approach to estimating epidemic-related mortality.
• Findings highlight significant undercounting of epidemic deaths in numerous global regions.
• Regional variations in excess mortality underscore the need for localized public health interventions and data collection.