Data Tool Kit: Useful Databases and Websites
Criteria for Database Inclusion in this Toolkit
For the purposes of this Toolkit, databases were judged useful for inclusion if they met the following criteria:
- The database is national and contains data for all states that can be retrieved at the state level.
- The database is free and easily accessible to users.
- The database contains some geographic indicator of “rurality” for each entry such as MSA, census tract, zip code, or county.
- The database contains health systems delivery or health/human services data.
During our search for databases meeting the above criteria we tried to locate databases with multi-year data available for trending purposes, but did not make this an essential criterion for inclusion in this Directory. Our purpose in developing this Toolkit was to create a user-friendly, on-line device that carefully describes databases that meet the inclusion criteria above so that every SORH has a clear idea of:
What the selected databases contain both with regard to content and dates of data collection
- How to access the databases
- What data can be assembled from the databases that would be useful to SORHs
- How the data can be compiled and suggestions for presentation using GIS mapping
- What limitations exist with regard to availability of nationally comprehensive rural data and why
Limitations on National Comprehensive Rural Data Available
For those of us who work with and for rural populations to improve the availability of, and access to a wide spectrum of quality health care services, the need for good rural data is apparent. Unfortunately, for many researchers in the country, including those at the Federal government level, this need is not apparent; or if it is, there has obviously been insufficient effort spent on securing the resources and policy changes necessary to obtain such data. In order to understand the limitations on rural data available, it is useful to describe some examples of issues we encountered when searching for useful databases to include in this directory.
Example: The Medical Expenditure Panel Survey (MEPS)
The Medical Expenditure Panel Survey (MEPS) is a set of large-scale surveys of families and individuals, their medical providers, and employers across the United States. The MEPS is administered by the Agency for Healthcare Research and Quality, which states that the MEPS is “the most complete source of data on the cost and use of health care and health insurance coverage.”
Like many other national databases, however, the MEPS data can tell us almost nothing on a state level about the cost and use of health care and health insurance coverage with regard to the almost 60 million rural people (21%) living in the U.S., according to the 2000 U.S. Census.
|For 2000 census data on the rural population as a percent of state total by state and within the U.S. see http://www.nemw.org/poprural.htm.|
Although the codebook for the MEPS data leads the reader to believe that Metropolitan Statistical Area (MSA) data was a variable collected for all data of interest, upon closer analysis, it turns out that for the Household Component and the Medical Component, the only two geographic indicators collected were census region (West, South, Midwest, Northeast – see map below) and MSA (MSA, non-MSA). Zip code, county, and even state level data are not available in these databases.
For the Insurance Component of the MEPS, county level information by state is available but only regarding the availability of employee sponsored health coverage. Even though it is estimated and not actual, more complete and usable rural data regarding all health insurance coverage is available from the Current Population Survey’s “Small Area Health Insurance Estimates for Counties and States (SAHIE)” which is included in Part Two of this Toolkit.
National Versus State Databases
Please note that the databases included in this Toolkit are national only, according to criterion 1 above. While some state databases that meet criteria 2-4 above exist in some states, not all states are homogeneous in this regard. In addition, geographic indicators that allow analysis of state rural data only and comparison with state urban data are frequently lacking, just as they are for many national databases. This means that comparison of state rural and urban data with national rural and urban data is often impossible.
Rural Geographic Indicators Used
The best geographic indicators that could be used to determine rural residence are zip code or census tract. Unfortunately, these indicators are seldom included in the data collected. In fact, the only rural indicators that could be found in databases for inclusion in this Toolkit, were county designation. While county-level breakdown is helpful, it is still insufficient since many counties contain both large metropolitan areas and very rural areas, particularly in states with large counties, like many states in the southwest and west.
Example: The 2003 National Assessment of Adult Literacy
For the 2003 National Assessment of Adult Literacy, which contained some questions to determine health literacy status, the only data collected was residence in a Metropolitan Statistical Area – yes or no – as with some of the MEPS data described above. Those respondents not flagged as residing in an MSA could be determined to be rural, providing at least a rudimentary statewide comparison of health literacy levels of rural versus urban populations; but the survey data do not provide any breakdown even to the county level within states.
Other limitations with the NAAL database, as with many other national databases, are that the survey was only conducted in 39 states and the MSA/non-MSA data are not readily available. It took the Arizona Center for Rural Health about 100 hours of work to:
- Obtain answers to our questions about what the complete database contained
- Determine that we could request special access to the complete database
- Fill out all the necessary forms declaring our research intent
- Have the forms signed by numerous people at the University of Arizona and notarized
- Submit the forms only to have them lost after being received in Washington
- Fill them out and have them notarized again
- Finally receive the data.
And we are licensed to analyze the data and publish results only under the close scrutiny of the federal agency where the database resides. While the protection of subject privacy is essential, the database we received has no individual identifiers, and still the process for obtaining it was incredibly onerous to say the least.
Example: National Hospital Discharge Survey
Another example of problems with rural geographic indicators is the National Hospital Discharge Survey (NHDS), a national probability survey designed to meet the need for information on characteristics of inpatients discharged from non-federal, short-stay hospitals in the U.S. The survey samples approximately 270,000 inpatient records from a national sample of about 500 hospitals. Although the hospital zip codes are collected by the survey, they are not available in the public use database due to confidentiality concerns.
Databases Included in This Toolkit
While we hoped to find many more databases that are readily available, free, and have useful state level information for State Offices of Rural Health, we were able to include in this Toolkit the following databases that provide some data at the county level for most states.
- Behavioral Risk Factor Surveillance System
- Youth Behavior Risk Factor Surveillance System
- Current Population Survey
- CDC Wide-ranging OnLine Data for Epidemiologic Research (WONDER)
In addition, in Part Three of this Toolkit we make some recommendations for working on policy change and the development of resources to ensure that more national, health-related surveys include rural geographic indicators at no less than the county level, but preferably at the zip code or census tract level. We hope this Toolkit and the recommendations will be useful to individual State Offices of Rural Health, the National Organization of State Offices of Rural Health, and other organizations whose mission is improving health care for all rural Americans.