In the model
Use County Health Rankings’ model of health to explore the measures that influence how long and how well we live.
Number of infant deaths (within 1 year) per 1,000 live births. The 2022 County Health Rankings used data from 2014-2020 for this measure.
Infant Mortality represents the health of a vulnerable age group and can provide context to support interpretion of the years of potential life lost (YPLL) in a county. Infant mortality is also commonly used to examine global health differences, as well as to understand historic racial inequities in the United States.
Data and methods
National Center for Health Statistics - Mortality Files
Data on deaths and births were provided by NCHS and drawn from the National Vital Statistics System (NVSS). These data are submitted to the NVSS by the vital registration systems operated in the jurisdictions legally responsible for registering vital events (i.e., births, deaths, marriages, divorces, and fetal deaths). In prior years of the Rankings, Premature Death was calculated by the National Center for Health Statistics, but this year the Mortality-All County (micro-data) file was requested. This allowed us to calculate Premature Death and Life Expectancy ourselves. While most calculations of mortality rates can be downloaded from CDC WONDER, the calculation of Years of Potential Life Lost and Life Expectancy requires raw data files.
Key Measure Methods
Infant Mortality is a rate
Infant Mortality measures the number of deaths among children less than one year of age per 1,000 live births. Rates measure the number of events (e.g., deaths, births) in a given time period (generally one or more years) divided by the average number of people at risk during that period. Rates help us compare data across counties with different population sizes.
Infant Mortality is a rare event (statistically speaking)
Infant death is a relatively rare event in most counties. Counties with smaller populations can see a lot of change in their rates of infant death data from year to year. Such changes are usually due to normal variation and are not necessarily caused by any actual change in the underlying risk of infant death in the county. To help determine if the infant death change in a county is due to normal variation or real change, we recommend examining the provided error margins. Error margins are statistical tools that can aid interpretation of variation in measures. If the error margins overlap year to year, it’s less likely that the variation in Infant Mortality reflects real underlying changes.
What deaths count toward Infant Mortality?
Deaths are counted in the county of residence, regardless of where the death occurred.
Some data are suppressed
A missing value is reported for counties with fewer than 20 infant deaths in the time frame.
The numerator is the cumulative number of deaths occurring before one year of age.
The denominator is the total number of live births.
Can This Measure Be Used to Track Progress
This measure can be used to track progress with some caveats. Infant mortality is a long-term health outcome, such that effects of interventions might not be reflected in this measures for years or even decades after a program or policy is implemented. Infant death is also a relatively rare event, especially in small counties. Statistics depend on large numbers of events to detect small changes, meaning that small changes in small communities may be difficult to detect. Lastly, it is important to note that the estimate provided in the County Health Rankings is a 7-year average.
Finding More Data
Disaggregation means breaking data down into smaller, meaningful subgroups. Disaggregated data are often broken down by characteristics of people or where they live. Disaggregated data can reveal inequalities that are otherwise hidden. These data can be disaggregated by:
We recommend starting with the CDC Wonder database, which contains information on death rates by race, ethnicity, age, gender, geography, cause of death, and more. Rates can be exported as crude or age-adjusted. Small counties might need to combine multiple years of data to see rates, as CDC suppresses any rates when there are fewer than 10 deaths.