Sources of Pediatric Mortality Data
CDC COVID Data Tracker – Demographic Trends: This is based on preliminary surveillance data submitted by states, presumably of deaths among Covid cases. However, but it is a inaccurate. It significantly overcounts pediatric deaths and undercounts adult deaths. Updated daily. Read more about the accuracy issues with the Data Tracker.
NCHS (National Center for Health Statistics): This is based on death certificates submitted to the National Vital Statistics System (NVSS) by the states, so there is a lag of at least 1-2 weeks, but up to 4-6 weeks (sometimes more) for pediatric death data to be processed. There are may different visualizations and data downloads for this data (see table below). The NCHS data includes deaths where Covid is listed anywhere on the death certificate. For about 25% of pediatric deaths, Covid is on the death certificate, but not listed as the underlying cause of death. This data is updated weekly.
CDC Wonder Provisional Mortality Statistics: This is another CDC database of mortality data similar to NCHS, but with a lot more detail on causes of death and more options for searching, filtering, and breaking down the cause of death data. However, it is only updated monthly, so it has a greater lag than NCHS. Due to its complexity and level of detail, it’s also not as user-friendly as the NCHS data sources. It allows you to filter for pediatric deaths based on whether it’s the underlying cause of death vs another multiple cause of death (to filter out the 25% of pediatric Covid deaths that have a different underlying cause).
AAP (American Association of Pediatrics) State-Level Data Report: The AAP releases a weekly Children and COVID-19: State-Level Data Report, tallying cases, hospitalizations, and mortality based on data from state dashboards. For pediatric mortality, they aggregate deaths from 46 states, NYC, Puerto Rico, and Guam. While they are missing a few states, the majority of states in their report include deaths of people over 17 as pediatric deaths (1 state is 0-18, 28 states/territories are 0-19, and 2 are 0-20). Including these older age groups should more than account for the 5 missing states – Montana, NY (outside of NYC), Rhode Island, Utah, and West Virginia – and for Florida, who uses a 0-15 age bracket. The AAP numbers have historically run a little bit higher than NCHS pediatric death numbers, but less than the CDC Data Tracker demographics.
|Data Source||Age buckets for children||Download||Notes|
|NCHS Provisional Deaths by month or year, jurisdiction of occurrence, sex and age – RECOMMENDED||0-17 Years||Download||Table view, easy to filter|
|NCHS Provisional Death Counts for Influenza, Pneumonia, and COVID-19 by week||0-17 Years||Download||Weekly pediatric time series,|
but download only
|NCHS Provisional Deaths by week, sex, and age||<1|
|Download||Graph view over time, |
age groupings not ideal
|NCHS Provisional Deaths by sex, race/ethnicity, and age||0-5|
|Download||Download only; |
incl.18 year-olds (not pediatric)
|NCHS Provisional Deaths by age in years||0-5 months|
1-17 years (each individual year)
allows granular data by year
|CDC WONDER Provisional Mortality Statistics, 2018 through last month||various options |
incl. single years
|Not as user-friendly|
|CDC COVID Data Tracker – Demographic Trends – NOT RECOMMENDED||0-4 Years|
|Surveillance data; flawed (overcounts pediatric deaths)|