Johns Hopkins Bloomberg School of Public Health – COVID-19 Excess Mortality Risk Index

Description:

Researchers at the Johns Hopkins Bloomberg School of Public Health and the University of Maryland created a statistical model of how likely an adult (age 18+) is to die of COVID-19 as compared to the national average, based on demographic information and underlying health conditions. The researchers developed the model based on two sources of information: (1) multivariate-adjusted estimate of risk associated with gender, social deprivation index and 12 pre-existing conditions from the recently published UK-based OpenSAFELY study, and (2) death rates associated with different age and racial/ethnic groups in the US published by the Centers for Disease Control and Prevention (CDC) after performing external covariate adjustment accounting for the correlation of these factors with other risk factors in the model.

The researchers applied their statistical model to publicly available demographic and health data to create (1) an Index of Excess Risk (IER) which describes how likely the average adult in a city (Census place) or state is to die from COVID-19 compared to the national average, (2) estimated proportion and size of the general adult population in a city or state that exceed different risk thresholds, and (3) estimated proportion of the projected deaths among general adult population in a city or state that exceed different risk thresholds. Validation analyses were conducted based on the projected risks and data on recent deaths in the US to show that the model is well calibrated for the US population. They also applied the model to CMS data for the Medicare population age 65 and older to arrive at county- and state-level estimates.

The researchers also published a calculator that assesses individual risk of death from COVID-19 here: http://covid19risktools.com/.