Data Quality
The County Health Rankings team draws upon the most reliable and valid measures available to compile the Rankings.
Data reliability
The reliability of the data used is one of the primary concerns when estimating values for relatively small areas like counties. The County Health Rankings makes every effort to provide the most reliable data available, but users should be aware that reliability can vary by place and by measure. An easy estimate of reliability is the error margin for a measure. Larger error margins suggest lower reliability.
Although the reliability of some Couth Health Rankings’ measures varies, when multiple measures are used to capture an underlying concept, reliability improves. For example, each individual measure comprising the Quality of life Health Outcome sub-area may suffer from some reliability weaknesses, but together, the Quality of life measures provide a reliable indication of morbidity.
Statistical uncertainty
Where possible, we provide the margins of error (95% confidence intervals) for our measure values. In many cases, the values of specific measures i are not statistically different between counties. Error margins represent a range within which the true population experience may differ from our measure. However, we are 95% certain that the true population experience falls within that range. When the error margin ranges for a given measure overlap between two places, we can be less confident that the true population experiences (in those places) are different from each other.
Treatment of missing values
In many ranked counties, some individual measures do not have a large enough sample size to report data for that measure. In these counties, the state average is assigned in place of any missing measure value to generate the data needed to calculate a county rank. On the state map, counties with a missing value for a given measure are identified with N/A (not available).
Age-adjustment of measures
Age-adjustment is a useful strategy to increase the comparability of measure values between counties having different age structures, or within county comparisons over time if the county age structure changes This is especially important for measures of factors related to age. We adjust county values for factors known to differ by age groups (or change with aging) so that all counties reflect a standard age distribution. For example, Adult smoking is related to aging into adulthood. A county with a comparatively older population would be more likely to have a higher percentage of adults reporting smoking behavior because more adults live in that county.
However, age-adjustment can mask the true burden of a health need in a county – especially counties with many older residents. Measure data tables are available to communicate the absolute number of events occurring for many measures where the county value has been age-adjusted. County Health Rankings follows best practice to determine which measures are age-adjusted.
The following County Health Rankings measures are age-adjusted:
Health outcome measures
- Premature death (YPLL)
- Poor or fair health
- Poor physical health days
- Poor mental health days
Health factor measures
- Adult smoking
- Excessive drinking
- Preventable hospital stays
- Flu vaccinations
- Adult obesity
- Physical inactivity
Additional measures
- Premature age-adjusted mortality
- Life expectancy
- COVID-19 age-adjusted mortality
- Diabetes prevalence
- Frequent physical distress
- Frequent mental distress
- Insufficient sleep
- Suicides