This section describes the context for county health — characteristics of a county’s population that have a potential effect on the amount and type of services used, health status, and resources available.

Data for Population Size, Poverty Level, Population by Age, and Population by Race/Ethnicity were obtained from the Current Population Survey (CPS) conducted by the U.S. Bureau of the Census. The CPS is an ongoing survey of states from which estimates for counties are derived. Due to the survey’s small sample size, the confidence intervals for these measures may be wide and should be consulted when making comparisons to Peer Counties. Information on survey methodology and confidence intervals are found at http://www.census.gov/cps/.

Population Size — This number is from “County Population Estimates by Sex, Race and Hispanic Origin by Five-Year Age Groups: July 1, 2005” obtained from the Population Estimates Program, Population Division, U.S. Census Bureau. These data are mid-year estimates of the resident population, 2005, and can be obtained at http://www.census.gov/popest/counties/asrh/CC-EST2005-alldata.html.

Population Density — this number is calculated by using the following formula: mid-year 2005 Population Estimate divided by 2000 Land Area (square miles). Land Area is from the Geographic Comparison Table GCT-PH1-R. Population, Housing Units, Area and Density: 2000, Census 2000 Summary File (SF1) 100-Percent Data, U.S. Census Bureau. “2000 Land Area by County” is a statistical abstract supplement published by the U.S. Bureau of the Census and obtained from the Area Resource File, Health Resources and Services Administration, 2005.

Poverty Level — the percentage of individuals living below the poverty level in 2003 is data obtained from the “Small Area Income Poverty Estimates (SAIPE)”, U.S. Bureau of the Census at http://www.census.gov/housing/saipe/estmod03/est03ALL.xls.

Population by Age — Age-specific population sizes are from “County Population Estimates by Sex, Race and Hispanic Origin by Five-Year Age Groups: July 1, 2005”. These data are estimates of the resident population for mid-year 2005, U.S. Bureau of the Census, at http://www.census.gov/popest/counties/asrh/CC-EST2005-alldata.html.

Population by Race/Ethnicity — Race- and ethnicity-specific population sizes are from “County Population Estimates by Sex, Race and Hispanic Origin by Five-Year Age Groups: July 1, 2005”. These data are mid-year estimates of the resident population, 2005, and reflect standard race and ethnicity categories in use by the U.S. Bureau of the Census, at http://www.census.gov/popest/counties/asrh/CC-EST2005-alldata.html . Note, percentages of white, black, Asian American/Pacific Islander, American Indian and percentages of Hispanic ethnicity do not total to 100% due to multiple race categories.

Peer County Range — the low and the high values found in each stratum of peers is provided. The low number is the tenth percentile of rates or percents in the stratum; the high number represents the ninetieth percentile. Eighty percent of all county values within the strata lie within this range.

FIPS (Federal Information Processing Standards) Code — these standard codes, which indicate State and county, are maintained by the National Bureau of Standards. The combined State and county codes create a unique county identifier. The State portion of the code is a two-digit number, while the county portion consists of a three-digit number. In general, numbering is sequential when States and counties are listed alphabetically. Some independent cities in Virginia have been combined into their original counties. However, in this edition of the CHSI reports, more local areas are represented in Alaska, Hawaii and Virginia. There are 3,141 counties for which data are presented in 2008.

Peer County Data Details

A distinctive aspect of this report is the ability to compare a county with its peers, those counties similar in population composition and selected demographics. Comparison of a county to its peers is thought to take into account some of the factors that make a difference in a community’s health.

Strata, or peer groups, were developed with input from an advisory committee composed of Federal, State, and local public health professionals and members of academia for CHSI 2000. The project goal was to develop strata of 20-50 counties each, providing several peers for each county. The relatively large number in each stratum allows counties to choose a few peers that they believe to be most like them. In CHSI 2008, stratum size averages 36 and ranges from 15 to 62 counties. There are a total of 88 strata. For the most part, counties are assigned to the same peer groups in CHSI 2008 as in CHSI 2000. New counties or newly reported entities (such as some Virginia independent cities, 27 Alaska counties) were (re)assigned to peer groups based on their frontier status, population size, poverty, age distribution, and/or population density in 2000.

To define the strata, the following five factors were used:

1. Frontier status - The National Committee on Rural Health recommended classifying areas as frontier if they had fewer than 7 persons per square mile. Source: Popper, F.J. (1986) The strange case of the contemporary American frontier. Yale Review: 76(1); 101-121;
2. Population size, using the National Association of County and City Health Officials’ population categories (less than 25,000; 25,000-49,999; 50,000-99,999; 100,000-249,999; 250,000-499,999; 500,000-999,999; 1,000,000 or more);
3. Poverty quartiles (less than or equal to 10.55%; 10.56-14.15%; 14.16-19.25%; 19.26% or more), based on the percentage of individuals in the county living below the poverty level (e.g., in 2000 and for a family of four, the poverty level is $17,603);
4. Median age categories, based on the percentage of children (percentage of persons age<18 less than 26.13%, or greater than or equal to 26.13%) and elderly (percentage of persons age 65+ less than or equal to 14.70%, or greater than 14.70%) in the county; and
5. Population density, as measured by half deciles (e.g., CHSI stratum 45 ranges between 44-187 persons per square mile).

Using an ordered, staged approach, counties were first grouped according to frontier status. Population size was used next. Then, as the number of counties in each category allowed, further groupings were made based on the remaining variables until the optimum stratum size was reached. Therefore, while all strata were classified according to the first two variables, only some were defined by factors of poverty, age, and population density. A schematic of the stratification process is in Appendix A. Each of the 88 strata is uniquely defined by two or more of the factors. It is possible that counties that are similar in several factors may not be in the same stratum due to category divisions.

The Strata Listing (see Appendix B) contains data on the number of counties and the ranges for population size, density, and poverty level by stratum. This table provides the demographic characteristics of all strata (based on data in CHSI 2008) and allows identification of other strata having similar characteristics. We encourage you to use this table to examine the strata generated using this method and to explore other counties’ reports. This is the project’s initial (2000) strategy to form peer groupings of counties and it was applied to the updated (2008) reports. Please feel free to comment at the CHSI Web site. (http://communityhealth.hhs.gov ). Future reports may use different methods and/or criteria for determining strata or provide an update to this process.

Finally, because the number of events for an indicator may be too small to report for a county for any single year, data were aggregated over several years (3, 5, or 10 years depending on county population) for several indicators to increase the number of counties with available data. Table 1 presents the number of years aggregated for measures of births and deaths as well as infectious diseases. Data from the Behavioral Risk Factor Surveillance System were also aggregated over time (see Appendix C for details), 2000-2006, for all counties.

Table 1: Years of data aggregated based on county population size applied to birth, death (not life expectancy), and infectious disease indicators:

County Population Size	# of Years Aggregated	Years	Peer County Stratum
<25,000	10	1994-2003	37-44, 54-76,78-88
25,000-99,999	5	1999-2003	8,17-36,45-53, 77
>100,000	3	2001-2003	1-7, 9-16