COVID-19 deaths: Which explanatory variables matter the most?

Area of Science:

• Epidemiology
• Public Health
• Data Science

Background:

• Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) has caused a global pandemic, with many transmission factors remaining unclear.
• Understanding drivers of COVID-19 transmission is crucial for managing current waves and future pandemics.

Purpose of the Study:

• To identify statistically significant and important variables driving COVID-19 deaths across US states.
• To analyze the impact of population density, mobility, and other factors on early COVID-19 mortality.

Main Methods:

• Compiled a database of over 28 potential explanatory variables for all 50 US states.
• Employed traditional statistical and modern machine learning techniques to analyze the data.

Main Results:

• Population-weighted population density (PWPD) and mobility metrics were the most significant factors influencing early COVID-19 deaths.
• Statistically significant variables included "stay at home" metrics, temperature, precipitation, race/ethnicity, and chronic low-respiratory death rate.
• PWPD and mobility were found to be the dominant drivers, suggesting location was more impactful than individual actions initially.

Conclusions:

• Early COVID-19 mortality was primarily linked to geographic factors like population density and mobility.
• Social distancing measures effectively reduce transmission by lowering the effective PWPD.
• Lifting lockdown restrictions should be locally tailored based on population-weighted population density.