NCHS Data Linkage

Frequently Asked Questions

1. What surveys have been linked to which administrative/follow-up data?

The following table presents the NCHS surveys and the administrative data source(s) each is linked to:

 

2. Where can I find a list of variables from the different restricted-use linked data files to construct my data dictionary for my RDC proposal?

 

3. Can files be linked across the different administrative data sources?

Yes, an analyst can request access to multiple linked data files across the different administrative data sources. For example, an analyst may be interested in Medicare claims data as well as mortality outcomes for NHANES III survey participants.  This analyst would request access to both the NHANES III linked Medicare files and NHANES III linked Mortality File in their RDC proposal.

 

4. I am interested in pooling multiple years of NHIS.  Are there any special issues I should be aware of?

Yes. There are people in the 1991 NHIS sample who were also included in the 1992 NHIS sample. If you are combining those two years of NHIS data, you should use the special 1992 NHIS file that excludes the participants who were also interviewed in 1991.  More information on how to select these individuals can be found on the NHIS website: ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Program_Code/NHIS/1992/readmenh.txtCdc-txt

Over time, NHIS has also undergone survey design changes that researchers may need to consider when pooling multiple years of NHIS data. Information on how to control for these survey design changes is available on the NHIS website: https://www.cdc.gov/nchs/nhis/index.htm

 

5. Can I use previous releases of the linked mortality files?

The updated Linked Mortality Files supersede any previous data releases.  Improvements to the matching algorithm make the latest mortality release more accurate than previous versions.  Therefore, we do not allow analysts access to previous releases of the restricted-use linked mortality files in the RDC.  Similarly, using outdated versions of the public-use files is not recommended.  Researchers accessing a prior release of the linked mortality data in the RDC will be allowed to finish work as necessary on a case by case basis.

 

6. What variables are used to merge the linked files with the individual surveys?

The variable to use to merge linked files with individual surveys varies by survey. The table below presents the merge keys.

NCHS Survey Survey Participant Identification Variable
National Health and Nutrition Examination Survey (NHANES) SEQN
Third National Health and Nutrition Examination Survey (NHANES III) SEQN
Second National Health and Nutrition Examination Survey (NHANES III) SEQN
NHANES Epidemiologic Follow-Up Study (NHEFS) SEQN
National Health Interview Survey (NHIS) PUBLICID*
Second Longitudinal Study on Aging (LSOA II) PUBLICID*
Supplement on Aging (SOA) PUBLICID*
National Nursing Home Survey (NNHS) RESNUM/FACID/CRID**

 

* PUBLICID must be created by the user for NHIS, SOA and LSOA II

** RESNUM is the merge key for the 2004 NNHS.  FACID is the merge key for the 1997 NNHS and CRID is the merge key for the 1995 NNHS.

7. How do I create PUBLICID?
PUBLICID is a function of several component variables and varies by year in NHIS. Please refer to the appropriate documentation (e.g. The Linkage of National Center for Health Statistics Survey Data to the National Death Index — 2015 Linked Mortality File (LMF): Methodology Overview and Analytic Considerations)Cdc-pdf for instructions for creating PUBLICID.

Page last reviewed: March 25, 2019, 09:35 AM