County Health Rankings & Roadmaps

County Health Rankings & Roadmaps, a collaboration between the Robert Wood Johnson Foundation and the University of Wisconsin Population Health Institute, provides data, evidence, guidance, and examples that youth development professionals can use in their work to improve community health and well-being. The Rankings use a population health model to describe the multiple factors that contribute to health and equity. The information and resources available at the County Health Rankings & Roadmaps website is useful to youth development organization leaders, as it provides guidance for taking action to improve the well-being of youth and communities. Key resources available at www.countyealthrankings.org are discussed.


BACKGROUND & CONCEPTUAL GROUNDING About us: County Health Rankings & Roadmaps
County Health Rankings & Roadmaps (CHR&R) is a program of the University of Wisconsin Population Health Institute. CHR&R is known for effectively translating and communicating complex data and evidence-informed policy into accessible models, reports, and products that deepen the understanding of what makes communities healthy and inspires and supports improvement efforts. CHR&R's work is rooted in a sincere belief in health equity, the idea that everyone deserves a fair and just opportunity to be as healthy as possible.
CHR&R provides data, evidence, guidance, and examples to build awareness of the multiple factors that influence health and support leaders in growing community power to improve health equity. CHR&R produces the County Health Rankings every year. The Rankings are unique in their ability to measure the health of nearly every county in all 50 states, and are complemented by guidance, tools, and resources designed to accelerate community learning and action.

History
The University of Wisconsin Population Health Institute (UWPHI) has been supported by the Robert Wood Johnson Foundation (RWJF) since 2008 to develop what is now known as the County Health Rankings & Roadmaps (CHR&R) program. The first national County Health Rankings were released on February 17, 2010.

Goals
• Build awareness of the multiple factors that influence health.
• Provide a reliable, sustainable source of local data and evidence to communities to help them identify opportunities to improve their health. • Engage and activate local leaders from many sectors in creating sustainable community change.
• Connect community leaders and grow community power to improve health.

The County Health Rankings Dataset
The County Health Rankings include about 100 measures that help communities understand how healthy their residents are today (Health Outcomes) and what will impact their health in the future (Health Factors). Annually, CHR&R updates these measures using the most recently available data for nearly all U.S. counties. Not all measures are used to calculate a county's rank. The CHR&R dataset contains: • Ranked measures, which are used in the calculation of state-specific county ranks. Ranked measures are combined using weights according to the CHR&R model to create a Health Outcomes Rank and a Health Factors Rank for each county within each state. • Additional measures, which provide helpful community context but do not contribute to the Ranks. Additional measures may be Health Outcomes, Health Factors, or demographics. Demographic data are included for every county to provide context for the place and its data.

Conceptual Model: The County Health Rankings Model of Health
The Rankings are rooted in a conceptual model of the social determinants of health. The ranks incorporate 35 ranked measures that help communities understand how healthy their residents are today (Health Outcomes) and what will impact their health in the future (Health Factors).
The County Health Rankings Model emphasizes the many factors that influence how long and how well we live. The Model also portrays the methodology by which counties in each of the 50 states are ranked.

CHR&R Measures
The CHR&R team goes through a careful and deliberate process when selecting measures for inclusion. We consider measures that meet our program's goals, ensuring they reflect important aspects of population health that can be improved, and we consider measures that are innovative to meet the needs of communities. Our measures are also selected based on their technical and analytical feasibility, ensuring they are not limited by availability, cost, validity, coverage, or other concerns.
CHR&R measures describe community conditions that can be modified, called Health Factors. Health Factors influence Health Outcomes, or how long and how well people live within that community.

Considerations for County Health Rankings Measures
Measures of Health Factors and unranked measures are expanded, pared, or revised annually based on the considerations below. These considerations ensure that the CHR&R dataset remains consistent, salient, legitimate, credible, and grounded in equity. Measures may not meet all considerations due to geographic, data source, and time limitations. To operationalize these considerations, we regularly evaluate data sources and methods and seek expert input and review from scholars, practitioners, and external advisors.

Strategic Considerations
Alignment with CHR&R goals • The measure speaks to a current or emerging population health issue and increases the value of CHR&R tools.
• The measure reflects aspects of population health that can be influenced through local, state, or national policies, practices, and systems change.
• The measure provides quantitative or qualitative information to explain concepts in the County Health Rankings Model.
• The measure supports data fluency and alignment in the field of data-to-action initiatives (e.g., America's Health Rankings, City Health Dashboard).
• The measure is of interest to community members, leaders, advocates, community health activists, equity champions, and field actors in public health and health care.

Theoretical Considerations
Connection of the measure to health and equity • The measure and its association with population health are scientifically supported through peer reviewed literature or expert opinion and a strong evidence base.
• CHR&R internal analyses (quantitative and qualitative) support the measure's connection to health.
• The measure clarifies the existence of health disparities and the potential for unfair, unjust differences.
• The measure centers learning from the wealth of knowledge, experiences, and priorities of a socially marginalized group.

Assessment of data sources and their methodology
• The measure draws from a data source that has transparent methodology and underlying assumptions.
• Source data are available for free or low cost.
• Source methods are valid. Data quality is maintained and updated regularly (within the past 3-5 years), where applicable.

Feasibility of quantitative and qualitative analysis for evaluation and production
• The measure draws from data that are available at, or can be aggregated to, the county level.
• Data can be disaggregated among population groups with an emphasis on groups that have historically or currently experience social disadvantage (e.g., race, ethnicity, gender, sex, education, disability status, family type, neighborhood, income, or wealth).
• The measure and its association with health and health disparities are validated internally and consistent with scholarly literature or expert evidence.
• The measure is numeric, ordinal, or binary to quantify differences that capture advantage or disadvantage between counties.
• The measure uses data that are available for most counties nationwide.
• The measure uses data that are representative locally and comparable across jurisdictions within a state.

Communication Considerations
Ability to meaningfully communicate and apply the measure to improve health and equity • The measure and its association with health and equity can effectively be communicated.
• The measure is recognized and documented by public health, healthcare, adjacent fields, or marginalized communities to have the ability to make change or have influence within systems of oppression.
• CHR&R can communicate limitations of the data and methods to audiences who want to interpret and apply the measure.
• The measure reflects a distinct concept and "call to action."

Rankings Methods
How do we calculate the Rankings? First, we receive, review, and prepare data from dozens of data sources across the U.S. Then, we follow these steps: 1. Assign weights: First, counties are ranked within each of the 50 U.S. states using measures that are assigned relative weights within the Health Outcomes and Health Factors Areas. 2. Calculate Z-scores: Next, measures are standardized within each state using a Z-score. 3. Create composite scores: The ranks are then calculated based on weighted sums of the standardized measures within each state-we call these weighted sums a composite score. The county with the lowest composite score (best health) gets a rank of #1 for that state and the county with the highest composite score (worst health) is assigned the lowest rank.

Assign weights
Each ranked measure contributes weight to the Health Outcome or Health Factor rank calculation relative to other ranked measures. Generally, ranked measures and corresponding weights are not changed year-to-year to retain consistency in methods. Standardizing each of these measures transforms them to the same metric -a mean (average) value of 0 (zero) and a standard deviation (measure of spread) of 1. We refer to these as Z-scores, where: Z = (County Value) -(Average of Counties in State)

(Standard Deviation of Counties in State)
Each Z-score is relative to the other counties in that state -not compared to an absolute standard -and shown in the metric of standard deviations. A positive Z-score indicates a value for that county higher than the average of counties in that state; a negative Z-score indicates a value for that county lower than the average of counties in that state. For example, if a county has a Z-score on a measure of 1.2 that means the county is 1.2 standard deviations above the state average of counties for that measure. For counties with a population of 20,000 or less, any Z-score less than -3.0 or greater than 3.0 is truncated to -3.0 or 3.0, respectively.

Reverse Coding of Some Measures
For most CHR&R measures, a higher Z-score score indicates poorer health (e.g., Children in Poverty). However, for some of our measures (e.g., High School Completion) a higher Z-score indicates better health. For these measures, we take this into account when computing composite scores by computing the Z-score as usual and then multiplying it by -1, so that a higher Z-score indicates poorer health for all measures. The following ranked measures are reverse coded in this manner:

Create Composite Scores
The scores computed for individual counties are weighted composites of the Z-scores where the weights represent relative importance towards total county health as determined by the CHR&R model (Table 1, Figure 1). A weighted composite is computed by multiplying each Z-score by its assigned weight and then summing all weighted Z-scores.
Below is the formula we use for our weighted composite scores:

Composite=∑wi Zi
In this formula, the Zi values are the Z-scores of the Ranked Measures. The wi values are the measure-specific weights. The ∑ sign indicates summation of the resultant values.
All CHR&R composite scores use the formula above, standardized Z-scores for each measure (reverse coded when necessary-see above), and the weights described in previous sections. Composite scores are computed separately by state.

Calculating Ranks & Quartiles
To generate Ranks, composite scores are sorted from lowest to highest within each state. The county with the lowest composite score (best health) is ranked #1 in that state and the county with the highest composite score (worst health) is assigned the lowest rank in that state.
CHR&R does not suggest that individual ranks represent statistically significant differences between counties. That is, the top ranked county in a state (#1) is not necessarily significantly healthier than the second ranked county (#2).
Beginning in 2021, we have displayed data in quartiles on the county snapshot to support comparison of a county to a similar grouping in each state. Health Outcome and Health Factor rankings are grouped into four equally sized groups (quartiles), ranging from the least healthy to healthiest counties (Lowest 0-25%, Lower 25-50%, Higher 50-75%, or Highest 75-100%) within each state. A county with a rank of #1 lies in the Healthiest (Highest 75-100%) quartile.

Create Tools Materials: Document Trends
Examining changes in Health Outcomes over time can provide an overall sense of community progress toward better health. Trends in Health Factors can inform specific health programs and may reflect the impact of local efforts.
We conduct linear regressions using all years of data shown in the trend graph to calculate whether a trend is decreasing, increasing, or stable. For each measure with trend data available, a detailed trend graph can be viewed by clicking on the graph icons in the county snapshot.
Each graph icon is color-coded to communicate the direction of the trend: • Red -The county value is trending worse for this measure • Yellow -The county value shows no significant trend • Green -The county value is trending better for this measure • Grey -Additional information is needed to interpret the trend for this measure • Black -Trend graph is available, but no interpretation has been provided Trend data are available for: •

Create Supplemental Tools: Areas to Explore and Areas of Strength
Areas of Strength and Areas to Explore are calculated for ranked Health Factors measures to highlight where a significant improvement or decline will lead to a similar change in overall Health Factor Rank. Areas of Strength and Areas to Explore are intended to serve as a starting point for identifying areas of strength or improvement in your county.
We obtain Areas to Explore and Areas of Strength by comparing your county value to your state value and the national value (median of counties) for each ranked health factor measure. For measures where your county is doing meaningfully better than the state and national values, these are highlighted as Areas of Strength and for measures where your county is doing meaningfully worse than the state and national values, these measures are highlighted as Areas to Explore.

Responsible Data Use
CHR&R tries to rank all counties or county equivalents that have a Federal Information Processing Standard (FIPS) code. Data limitations such as data missingness can lead to special considerations for ranking methods.

Limitations of Data Comparability Across States
CHR&R uses data from many sources, each with different methods for collection and processing of the data. For most of our measures, county data is comparable between counties within states and also comparable across state lines. For a few of our measures, caution must be exercised when making comparisons between counties in different states. See Appendix 3 for a list of measures which should be compared with caution across states.

Addressing Missing Data in Rank Calculations
If a county has sufficient data to be assigned a rank, but is missing data for a given ranked measure, we assign the state mean for that measure value to calculate the county's rank.

Unranked Counties
Some counties in the nation are too small to have reliable measurements for Health Outcome measures. These counties are not ranked.
Counties are not ranked if any of the following is true: 1. County had a missing value for Premature Death (i.e., there are less than 20 deaths during the time period and data are suppressed for privacy reasons). CHR&R methods increase the number of ranked counties through: • Careful data selection: Ranked measures are based on data which are available for the greatest number of counties.
• Imputation: In some cases, data are combined over multiple years of data. For several measures, CHR&R averages multiple years of data, giving equal weight to each observation year. This approach increases the number of small, sparsely populated counties with reliable data estimates.
• Use of modeled data: Some measures, including Adult Smoking, Adult Obesity, and Children in Poverty, are based not only on survey response, but depend on statistical modeling techniques that improve the precision of the estimates.

DATA USE Downloading the Data Guide to Files
County Health Rankings data are downloadable in .csv and .sas format for analytic use. You can find the files in two places on our website: • National files are available from download on our Rankings Data & Documentation webpage.
• State-specific files are available for download from the respective state data page on the CHR&R website.

Data Sharing
CHR&R data sharing is dependent on the data use regulations of the source data. If you are interested in making a data request, please use the Contact Us form available on the website. Please include details of your request including any specifications. A member of our team will follow up and notify you if we are able to fulfill the data request and if so, establish a timeline. Institutional Review Board (IRB) approval may be requested if applicable. Your use of the data may be subject to Data Use Agreements. Cite CHR&R when you publish your work. CHR&R has provided a suggested citation on our FAQ page. For more information, review CHR&R's Terms of Use.

Missing Data
If a value is displayed as missing (.) or blank that means data is unavailable for that county or race/ethnicity group. This could mean data is unavailable, unreliable, or has been suppressed due to small numbers and resulting privacy concerns. Data suppression guidelines are generally established by the data sources.

Data Operations Age-adjustment of Measures
• Race is a form of identity constructed by our society to give meaning to different groupings of observable physical traits. An individual may identify with more than one race group. • Ethnicity is used to group individuals according to shared cultural elements.
Racial and ethnic categorizations relate to health because our society groups people based on their perceived identities. These categorizations have meaning because of social and political factors, including systems of power such as racism. Understanding the variation among racial and ethnic groupings in Health Factors and Health Outcomes is key to addressing historical and current context that underlie these differences.
Data sources differ in methods for defining and grouping race and ethnicity categories. To incorporate as much information as possible in our summaries, CHR&R race/ethnicity categories vary by data source.
With a few exceptions, CHR&R adheres to the following nomenclature originally defined by The Office of Management and Budget (OMB): • American Indian & Alaska Native (AIAN): includes people who identify as American Indian or Alaska Native and do not identify as Hispanic. • Asian: includes people who identify as Asian or Pacific Islander and do not identify as Hispanic.
• Black: includes people who identify as Black or African American and do not identify as Hispanic.
• Hispanic: includes people who identify as Mexican, Puerto Rican, Cuban, Central or South American, other Hispanic, or Hispanic of unknown origin. • White: includes people who identify as white and do not identify as Hispanic.
Racial and ethnic categorization masks variation within groups. Individuals may identify with multiple races, indicating that none of the offered categories reflect their identity; these individuals are not included in our summaries. OMB categories have limitations and have changed over time; these limitations and changes reflect the importance of attending to contemporary racialization as a principle informing approaches to measurement. For some data sources, race categories other than white also include people who identify as Hispanic.
The above definitions apply to all measures using data from the National Center for Health Statistics (see Appendix 5). For this data source, all race/ethnicity categories are exclusive so that each individual fits into only one category.
Other data sources offer slight nuances of the race/ethnicity categories listed above: • The American Community Survey (ACS) only provides an exclusive race and ethnicity category for people who identify as non-Hispanic White.