At a glance
- There are certain guidelines for handling geographic variables in National Center for Health Statistics restricted-use data.
- Guidelines include permissible analysis levels, merging procedures, and rules for using urban-rural classifications in U.S. counties.
- Gain insight into the criteria for obtaining state and regional estimates, ensuring data use aligns with the survey's design and approval protocols.
Background
The National Center for Health Statistics (NCHS) provides geographic variables, also called geocodes, for some of its surveys.
Most restricted-use geographic NCHS variables can only be used to merge external data or to categorize geographic areas with similar characteristics. Researchers should not estimate regions smaller than the specific survey's sample design would support, even when smaller levels of geography are available.
The RDC rarely approves projects involving geographic estimates at levels lower than the national level.
Regional estimates
The NCHS Research Data Center Review Committee will consider regional estimates for all surveys.
State estimates
The committee will only consider projects to produce state estimates if they pool multiple years of data from one of these sources—
- National Health Interview Survey
- State and Local Area Integrated Telephone Survey
- Vital Statistics Birth Certificates
- Vital Statistics Death Certificates
Lower-level estimates
The committee never approves projects to produce county-level or lower-level estimates from NCHS survey data.
Surveys providing geographic variables
Urban-rural classification
NCHS developed a six-level urban-rural classification scheme for U.S. counties and county-equivalent entities. Researchers can apply this classification scheme to NCHS surveys without an urban-rural variable.
If you want to use the urban-rural measure, you must indicate this in your proposal. Include state and county in your list of restricted variables and the NCHS urban/rural continuum as an additional NCHS data source. The RDC will remove state and county data after merging the urban/rural codes.