Skip to main content

Characteristics of Persons Who Died with COVID-19 – United States, February 12-May 18, 2020

Our take —

This CDC report provides a detailed description of confirmed 10,647 COVID-19 deaths, though the generalizability of results is questionable because most of these deaths occurred early in the epidemic in specific geographic areas. The median age of death of those with supplemental clinical data available was 75 years, but varied by race/ethnicity; Hispanic and nonwhite persons were more likely to die at younger ages. Underlying medical conditions, such as cardiovascular disease and diabetes, were common, and they were more prevalent among younger age groups. At least 8% of deaths among people younger than 65 occurred at home or in the emergency department, reflecting a lack of timely care.

Study design

Case Series

Study population and setting

This report includes case-based surveillance data on 52,166 deaths (median age 78 years, 44% female) among individuals with laboratory-confirmed COVID-19 from 47 US jurisdictions, reported using the standardized CDC case-report form between February 12 and May 18, 2020. Supplemental data were collected from medical charts and death certificates via 16 public health departments; these included more complete data on race/ethnicity, underlying medical conditions, clinical course of disease, and location of death. The supplemental surveillance data included 10,647 (median age=75, 39% female) COVID-19 deaths between February 12 and April 24, 2020; 9,997 (94%) of these deaths occurred in New York City, New Jersey, or the state of Washington.

Summary of Main Findings

Among the 52,166 deaths with case-based surveillance, 40% were white, 21% were Black, 14% were Hispanic/Latino, 4% were Asian, 3% were multiracial or of other race, and race was unknown for 18%. Among the 10,647 deaths with supplemental surveillance data, 35% were white, 25% were Black, 24% were Hispanic/Latino, 6% were Asian, 3% were multiracial or of other race, and race was unknown 6%. Among those with supplemental surveillance, age at death varied by race/ethnicity, with Hispanic decedents having the lowest median age at death (median: 71 years, IQR: 59-81), followed by nonwhite, non-Hispanic decedents (median: 72 years, IQR: 62-81). Age at death was highest among white decedents (median: 81 years, IQR: 71-88). 76% of individuals with supplemental surveillance had at least 1 underlying medical condition, and the prevalence was even higher among younger ages. The most common comorbidities were cardiovascular disease (61%) and diabetes (40%). Median illness duration prior to death was 10 days (IQR: 6-15) in a subset of patients with illness onset date (N=3,021, 28%). Overall, 62% of individuals died in the hospital, 5% in long-term care facilities, 5% in the emergency department, 1% at home or in hospice, and the location of death was unknown for 26% of individuals.

Study Strengths

The report provides detailed demographic and clinical data on over 10,000 COVID-19 deaths from 16 public health departments.


Due to missing data on clinical characteristics, comorbidities, and clinical course of disease in the case-based surveillance (N=52,166), it is unclear how representative the supplemental surveillance data (N=10,647) is of the larger population. Supplemental surveillance data was only available for 16/56 public health departments contacted by CDC to request the additional data, and the majority of these public health departments were located in areas with the largest COVID-19 outbreaks (New York City, New Jersey, and Washington state). Thus, the generalizability of these results to the broader US population is questionable. Reporting standards varied across locations, preventing the ability to examine specific comorbidities in the full data. Even in the supplemental surveillance, missing data was pervasive, especially for comorbidities, duration of illness, and clinical course, and location of death, and estimates of underlying condition prevalence and hospitalizations/mechanical ventilation are likely underestimated.

Value added

The is one of the broadest characterizations of COVID-19 deaths in the United States to date.

This review was posted on: 12 August 2020