The overall goal of the project is to expand the utility of National Center for Health Statistics (NCHS) linked data resources to support patient-centered outcomes research (PCOR) through the development of publicly available synthetic linked data files. To protect NCHS survey participants’ confidentiality, most of the linked datasets produced by the NCHS Data Linkage Program are only available as restricted-use files accessible through the NCHS or Federal Statistical Research Data Centers (RDC). Requirements to access restricted-use linked data through the RDCs have limited the utility of linked data.
To expand the utility of NCHS linked data resources in patient-centered outcomes research, NCHS has developed two synthetic linked data files; the first linking NCHS National Health Interview Survey (NHIS) data with Medicare enrollment and utilization data from the Centers for Medicare & Medicaid Services (CMS) and Department of Housing and Urban Development (HUD) federal housing program participation data; and the second linking patient encounter records from the National Hospital Care Survey (NHCS) with death certificate information from the National Death Index (NDI). The synthetic linked NCHS-HUD-CMS data file is available for public-use and the synthetic linked NHCS-NDI data file is expected to be released in 2026. The synthetic linked data were created using statistical techniques to maintain properties of the original restricted data while protecting survey participants' privacy.
To enhance the utility of the synthetic linked data, NCHS developed a verification process that allows data analysts to verify results from selected logistic regression models without having to access the original restricted-use linked data.