Information for this monitoring initiative is drawn from a wide range of available data sets. This section provides a brief overview of each of these data sets.
The Healthcare Cost and Utilization Project (HCUP) is a Federal-State-industry partnership to build a standardized, multi-State health data system, which is maintained by the Agency for Healthcare Research and Quality (AHRQ). The data on preventable hospitalizations and the concentration and distribution of uncompensated and Medicaid discharges used in this book come from the HCUP's 1999 State Inpatient Databases, which provided discharge abstract data for hospitalizations occurring in 28 participating States. Additional States are added each year.
The following HCUP partner organizations provided data that are used in this report:
In addition, hospital discharge data were obtained directly from the following States:
The Uniform Data System (UDS) is maintained by the Bureau of Primary Health Care (BPHC) at the Health Resources and Services Administration (HRSA). It provides data on health centers and their users, including health center and user characteristics, financing and revenues, encounters, and staffing. The health center data used in this book come from the 1999 UDS.
The National Health Interview Survey (NHIS) is a multipurpose health survey conducted by the National Center for Health Statistics (NCHS), Centers for Disease Control and Prevention (CDC). It is the principal source of information on the health of the civilian, noninstitutionalized household population of the United States. The NHIS has been conducted continuously since 1957, and data are released annually. The survey-based access to care data used in this book are a 2-year average of the 1999 and 2000 NHIS data.
Data on 1999 Medicaid expenditures per recipient were obtained from publicly available Form 2082 data provided by the Centers for Medicare & Medicaid Services (CMS). In addition, CMS staff supplied data on Medicare Disproportionate Share Hospital payments for 1999.
The Area Resource File (ARF) is a county-specific health resources information system designed to be used by planners, policymakers, researchers, and other professionals interested in the Nation's health care delivery system and the factors that may affect health status and health care in the United States. Assembled by HRSA, the database contains more than 7,000 variables for each of the Nation's counties, including information on health facilities, health professions, measures of resource scarcity, health status, economic activity, health training programs, and socioeconomic and environmental characteristics. All information in the file is derived from existing data sources. The health care delivery system data used in this book come from the 2001 ARF, which contains data for 1999.
The National Vital Statistics System (NVSS), maintained by the NCHS, CDC, is responsible for the Nation's official vital statistics. These vital statistics are provided through State-operated registration systems. While the registration of vital events-births, deaths, marriages, divorces, and fetal deaths-is a State function, standard forms for the collection of the data and model procedures for the uniform registration of the events are developed and recommended for State use through cooperative activities of the States and the NCHS. The birth outcome data in this book come from the 1999 Vital Statistics.
The HIV/AIDS Surveillance Report contains tabular and graphic information about U.S. AIDS and HIV case reports, including data by State, Metropolitan Statistical Area (MSA), mode of exposure to HIV, sex, race/ethnicity, age group, vital status, and case definition category. It is published semiannually by the Division of HIV/AIDS Prevention, National Center for HIV, STD, and TB Prevention, CDC. The HIV/AIDS data used in this book come from the 1999 HIV/AIDS Surveillance Report.
The Current Population Survey (CPS) is a monthly survey of approximately 50,000 households conducted by the Bureau of the Census for the Bureau of Labor Statistics. The CPS is the primary source of information on the labor force characteristics of the U.S. population. Estimates obtained from the CPS include employment, unemployment, earnings, hours of work, and other indicators. Supplemental questions to produce estimates on a variety of topics, including school enrollment, income, previous work experience, health, employee benefits, and work schedules, are often added to the regular CPS questionnaire. The health insurance data in this book are a 3-year average of data from 1999, 2000, and 2001.
The U.S. Census Bureau conducts a complete enumeration of the population of the United States every 10 years, as required by Article I of the Constitution for the purpose of reapportioning the U.S. House of Representatives. A limited number of questions are asked of the entire population, with a more detailed set of information gathered from approximately one in every six persons. The population and community context data used in this book come from the 2000 Census.
The Federal Bureau of Investigation maintains the Uniform Crime Reports, which provide tabulations of crime statistics for use in law enforcement administration, operation, and management. Data are available for States, MSAs, cities with more than 10,000 inhabitants, suburban and rural counties, and colleges and universities. The index crime rate data used in this book come from the 1999 Uniform Crime Reports.
These data are for 1997 and are from the Disparities in Health Insurance Supplement, published by the UCLA Center for Health Policy Research at the University of California, Los Angeles.
Each year, the American Hospital Association conducts a survey of hospitals in the United States and associated territories. The database includes data elements containing demographic, utilization, financial, and other hospital characteristics of interest to health care researchers.
Claritas provides a variety of data for ZIP Code areas, based in part on the 2000 census. The data included in this book come from the 2001 update and, for 1999, were interpolated using a straight-line method based on the 1990 census.
Data on HMO penetration and competitiveness come from Competitive Edge: Regional Market Analysis, published annually by InterStudy.
Table A-1 describes all the measures included in the two data books, their definitions, data sources, and relevant methodological notes.
Table A-2 lists the ambulatory care sensitive (ACS) conditions that are used for the preventable hospitalization rates included in this book. As documented in Chapters 3 through 7, a strong association exists between the outcome and performance measures (ACS hospitalizations and birth outcomes) in this book and population characteristics, especially with regard to area income and race/ethnicity. Accordingly, comparing simple rates can often be misleading because much of the difference in rates may relate to area differences in these factors rather than access problems or the performance of the safety net. This is especially true in comparing central city and suburban rates, where large concentrations of low-income and minority residents affect the underlying outcome rates of the area.
To better examine the impact of access and performance on rates, "observed/expected" ratios are also provided. These ratios are based on a multivariate regression with the basic outcome/performance measure (ACS rates, late/no prenatal care rates, and so on) as the dependent variable and area income and racial composition included in the equation as independent variables. The observed/expected ratio is the actual rate divided by the predicted value from the regression.
Because large and well-documented differences exist in physician practice style among regions of the country, interpreting the ACS rates presented here can be difficult. Many ACS conditions involve substantial physician discretion as to whether to admit a patient or manage him/her on an ambulatory basis. Accordingly, area differences in rates may also reflect differences in practice style rather than the impact of access barriers or safety net performance issues. To better examine these differences, "adjusted" ACS rates are also presented. These "adjusted" rates were developed based on multivariate regressions, where the dependent variable is the underlying ACS rate for an area and the independent variable is the area admission rate for a group of high-variation hospital admissions that are not sensitive to differences in income and race. This group of high-variation admissions was developed by examining ZIP Code-level admission rates from six geographically dispersed States. Diagnostic Related Groups (DRGs) were identified in which there was no more than a 25 percent difference in area rates between high- and low-income areas and between areas with large and small minority racial/ethnic populations. Separate analyses were conducted for children, young adults, and older adults. These "reference" conditions included diagnoses such as DRG 324 (urinary stones without complications), DRG 153 (minor bowel procedures), and DRG 181 (gastrointestinal obstruction). The "adjusted ratio to the mean" data presented in these books consist of this adjusted rate divided by the mean rate for all areas.
For MSAs including only one county (e.g., Madison, WI/Dane County), in the ACS and birth data tables, the rates are identical for the MSA and the county, but the two ratio columns are not. This difference reflects the fact that the regressions are based on different units of analysis (MSA versus county/city/residual), and many of the differences among areas that are captured at the county level are masked at the MSA level.
| Source | URL |
|---|---|
| ARF | http://www.bhpr.hrsa.gov/healthworkforce/data/arf.htm |
| CMS | http://www.cms.hhs.gov/home/rsds.asp |
| CPS | http://www.bls.census.gov/cps |
| HCUP | http://www.ahrq.gov/data/hcup |
| HIV/AIDS Surveillance Report | http://www.cdc.gov/hiv/stats/hasrlink.htm |
| NHIS | http://www.cdc.gov/nchs/nhis.htm |
| NVSS | http://www.cdc.gov/nchs/nvss.htm |
| U.S. Census Bureau | http://www.census.gov/main/www/cen2000.html |
| UDS | http://www.bphc.hrsa.gov/uds/default.htm |
| Uniform Crime Reports | http://www.fbi.gov/ucr/ucr.htm |
The data presented in these books come from a wide variety of sources, which are described above. Some of these data represent actual values, particularly data on population size, which are collected in the full Census. Most of the data, however, are estimates. For example, the data collected in the Current Population Survey and the National Health Interview Survey are estimates rather than actual values. While it is beyond the scope of this project to provide standard errors for all the estimates presented, it should be noted that this variance exists.
Every effort has been made to ensure that the data used for these books have been processed accurately. However, errors may still exist, due to either an oversight in the data processing or problems with the original data.
The geographic areas included in this book are those for which data concerning the health care safety net, particularly hospital discharge data, were readily available. Data are presented for five types of geographic areas:
The general concept of a metropolitan area is that of a large population nucleus, together with adjacent communities having a high degree of social and economic integration with that core. Metropolitan areas1,2 comprise one or more entire counties, except in New England, where cities and towns are the basic geographic units, and are determined by the U. S. Office of Management and Budget (OMB). An MSA is a metropolitan area that is not closely associated with other metropolitan areas, and is typically surrounded by nonmetropolitan counties. A PMSA consists of a large urbanized county or cluster of counties that demonstrate not only very strong internal economic and social links, but also close ties to other portions of the larger area. PMSAs exist only in metropolitan areas with a population of 1 million or more. Where MSAs cross State boundaries, they are shown with the first State listed in the official MSA name. These names are designated by the U.S. OMB. For example, the Kansas City, MO-KS MSA is shown with the Missouri data.
1 U.S. Census Bureau. About Metropolitan Areas. Available at: http://www.census.gov/population/www/estimates/aboutmetro.html. Accessed April 24, 2003.
2 U.S. Census Bureau. Census 2000 Geographic Terms and Concepts. Available at: http://www.census.gov/geo/www/tiger/glossry2.pdf. Accessed April 24, 2003.
Counties and major cities are based on their political boundaries. Major cities are defined by the U.S. Census Bureau as "places." "County Residuals" show data for a given county from which the major city data have been removed. For example, the Tampa-St. Petersburg-Clearwater, FL MSA shows data for all of Hillsborough County (County), Tampa (major city/place), and for "Residual-Hillsborough County," which shows data for all of Hillsborough County outside of Tampa proper.
The U.S. Census Bureau has defined four regions of the country. The States included in this book belong to the following regions:
Northeast: Connecticut, Massachusetts, Maine, New Jersey, New York, Pennsylvania, Rhode Island South: Arkansas, District of Columbia, Florida, Georgia, Maryland, North Carolina, South Carolina, Tennessee, Virginia Midwest: Iowa, Illinois, Kansas, Michigan, Minnesota, Missouri, Wisconsin West: Arizona, California, Colorado, Hawaii, Nevada, Oregon, Utah, WashingtonIn this book, certain counties are excluded from MSAs of which they are a part. Table A-3 describes the reasons for these exclusions.
Columbia, SC, and Atlanta, GA are each part of multiple counties, although their data are shown in only one county. Data for Columbia are shown as part of Richland County; Columbia is also part of Lexington County. Data for Atlanta are shown as part of Fulton County; Atlanta is also part of DeKalb County.
In general, a city appears indented on the line following the county of which it is a part. The following are exceptions:
Counties in New England are different than those in other areas of the country. Typically, the functional governmental unit in New England is at the township or municipality level. As a result, New England counties may fall in more than one metropolitan area or may fall partly in both a metropolitan and a rural county. In addition to using MSA designations, New England is organized into New England County Metropolitan Areas (NECMAs).3 For consistency with data from the rest of the country, we use the standard MSA designations. Table A-4 shows counties in the major metropolitan areas of New England that are included in this book, indicating the areas in which they exist and where they are included in this data book.
3 U.S. Office of Management and Budget. Metropolitan Areas 1995: Lists I-IV. Available at: http://www.whitehouse.gov/omb/bulletins/95-04attachintro.html. Accessed April 24, 2003.