Federal Data Users

Hello all,

Within the last year, CDC Wonder stopped providing age-adjusted death rates at the county level. This is what is posted as the reason on CDC Wonder:

• Standard age-adjusted rates (calculated with standard populations) are only available for Ten-Year Age Groups.
• Age-adjusted rates are not available at the county level for analysis of mortality by Single Race, including analysis of Urbanization categories for counties, because the populations are weighted to the 10-year age groups, and county-level population estimates are not available for the "<1 year" and "1-4 years" age categories.

However, some states, particularly those more densely populated, have published multi-year (ex: 2020-2023) age-adjusted death rates for the leading causes of death, including stratified rates by race when the population is large enough. My questions are.

1. Is it a resource/staffing issue that allows some states to make these calculations and not others?
2. Are there other factors (ex: population density, local demographic data collection, use of other age proxys, etc.) that allow this to be possible in some places and not others?
3. Does CDC Wonder's description of the Census changes imply that age-adjusting death rates will not be reliable below the state level analysis during this 10 year period?

Any and all insights are greatly appreciated.

The NCI SEER program has single age populations for counties https://seer.cancer.gov/popdata/download.html and there are death files:

The 2023 death data is here:

https://ftp.cdc.gov/pub/Health_Statistics/NCHS/Datasets/DVS/mortality/mort2023us.zip

You can use this data to get county level (age) adjusted death rates.  You can find the formulas in most introductory  epidemiology text books.

I haven't looked at the files recently but they probably also have sex and race

Best,

Dave Dorer

Dear Danielle,

I took a look and the NCHS mortality vital stats dataset that I referenced and it only goes to the state level.  The SEER database has county level all cause mortality data.  You need this to calculate relative survival. You need to sign a data use agreement to access the SEER data.  I don't think that you can access the data for commercial use without special permission.

Dave Dorer

Dear Danielle

Here are the Populations Estimates Program (PEP) US Census components of change (births, deaths, and migration) by county.

The csv file is currently blocked.  You can read the pdf data dictionary.

https://www2.census.gov/programs-surveys/popest/datasets/2020-2024/counties/totals/co-est2024-alldata.csv

https://www2.census.gov/programs-surveys/popest/technical-documentation/file-layouts/2020-2024/CO-EST2024-ALLDATA.pdf

Dave

SEER  has death data for County x Age x Sex x Race x Hispanic for 2019-2023.  You can't download all variables at the same time.  You can do State x Age x Sex x Race x Hisapanic and then adjust the state level table by the county marginals for example by County x Age x Sex

If you are using the data in a for-profit commercial context you need a commercial license.  Email seerstat@imsweb.com and ask about details.

Dave

Some larger states or states with better data infrastructure can carry out their own demographic programs. (I think LA County has its own demographic program, FWIW.)

CDC is being extremely conservative, and since they are not responsible for the projections or estimates that states may be producing, they cannot endorse them. They basically say, "We did not produce this data so we don't trust it." That's quite fair.

Reductions in force at CDC will negatively affect their ability to produce statistical products. We have seen that other agencies in similar situations had to drop some series, some surveys, and/or some degree of detail in producing their reports and estimates (https://www.amstat.org/the-nations-data-at-risk-year-two-ongoing-monitoring).

Dave provided everything else there is to say about the problem.