Back

Estimating Lower Bounds for COVID-19 Mortality from Northern Italian Towns

Coleman, T. S.

2020-06-12 epidemiology
10.1101/2020.06.10.20125005 medRxiv
Show abstract

For COVID-19 the Infection Fatality Rate or IFR - a crucial variable in epidemiological modeling - is difficult to estimate because many cases are asymptomatic and the overall infection rate is generally not known. Circumstances in the Italian provinces of Milano, Bergamo, Brescia, and Lodi allow estimation of lower bounds for age- and sex-specific all-cause excess mortality (a proxy for IFR) since anecdotal reports indicate some towns were close to fully infected. Using data from ISTAT on mortality from January 1 through April 15 for 2020 and the three preceding years, I estimate excess mortality by sex and age categories (0-14, 15-54, 55-64, 65-74, and 75+ years) while controlling for town-specific mortality that proxies for town-specific infection rate. The 99th percentile from the tail of the town distribution gives a lower-bound estimate for COVID-19 mortality. The overall population-weighted mortality at the 99th percentile is 1.09 percent (95% CI 1.06-1.14). The age- and sex-specific rates vary considerably: for men age 65-74 the estimate is 2.10 percent (95% CI 1.94-2.28) which is 3.5-times higher than men 55-64 and 2.7-times higher than women 65-74.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.