Study population and setting
The goal of this study was to estimate the weekly changes in deaths by any cause beyond what would have been expected if SARS-CoV-2 had never been introduced in the United States. Excess deaths were then compared with the reported number of deaths due to COVID-19. Data on deaths due to all causes (National Center for Health Statistics) and pneumonia, influenza and COVID-19, specifically (International Statistical Classification of Diseases Related Health Problems, Tenth Revision codes U07.1 or J09-J18) were all accessed from publicly available sources. Historical comparison data were also accessed from publicly available sources (CDC). Excess deaths, stratified by week and state, were determined by first estimating the baseline number of deaths in the absence of COVID-19 using historical data for the same time period from previous years, and then subtracting the expected number of deaths from the observed number of deaths between March 1 and May 30, 2020.
Summary of Main Findings
Between March 1 and May 30, 2020, there were a total of 781,000 total deaths in the United States. Compared with what would have been expected for this same period in the absence of COVID-19, there were an estimated 122,300 excess deaths (95% prediction interval: 116,800-127,00), or 18.5% increase in the number of deaths. Of these excess deaths, 95,325 were officially reported and attributed to COVID-19, indicating that the number of true excess deaths was 28% higher than what was reported. There was variability across states between the reported deaths attributed to COVID-19 and the estimated excess deaths, and in some states, these extra deaths were occurring prior to widespread availability of testing.
This is a relatively simple study design that utilizes publicly available data to better characterize mortality, particularly during the early phase of the COVID-19 outbreak in the United States when other data are limited. Additionally, the analyses presented here take into account delays in reporting.
Based on this analysis, there is not a clear delineation of the direct effects of SARS-CoV-2 the pathogen, and the indirect effects of the response to COVID-19. During this period, large changes in the form of lockdowns, physical distancing, and loss of employment also occurred. There is no discussion of what the breakdown of deaths from other causes captured here are.
This study provides a rigorous way to estimate excess deaths beyond what is expected. Estimates of excess deaths can aid in our understanding of the pandemic over time, in addition to viral testing.
This review was posted on: 28 July 2020