Describe, Locate, and Download Feasibility Files

Purpose

NHANES-CMS linked data feasibility files can be used to approximate maximum sample sizes. This module provides background on the structure, location, and use of the feasibility files and demonstrates how to download the full set of variables available in the files. It will also guide you through the process of downloading and saving the feasibility files and documentation. The variables can be used to identify whether a survey participant was eligible for the CMS linkage and, if eligible, whether the participant successfully matched to CMS administrative data. The files also include variables to indicate if a survey participant has a record on each of the CMS administrative data files described in Module 4, as well as the year of benefits coverage.

Modules 6–8 explain how to convert the files to SAS datasets and produce descriptive statistics using the data. You will use a subset of the variables in these files to assess demographic characteristics (e.g., race/ethnicity and gender) by linkage eligibility and match status. Not all programs and variables referenced in this module will be used in subsequent modules.

Task 1: Describe the Feasibility Files

NHANES-CMS linked data feasibility files can be used to approximate maximum sample sizes by demographic and health status variables. This task will describe the feasibility files and explain how they can be used to approximate a sample size prior to starting your research project.

Feasibility Files Overview

NCHS has created and released NHANES-CMS linked data feasibility files for public use. These files provide information about survey participants’ eligibility for CMS linkage, final match status, and the years of CMS linked data files available for each successfully linked survey participant. These files can be used to approximate the maximum available sample size to assess the feasibility of conducting analyses using the CMS Medicare and Medicaid linked data. These files make it possible to identify demographic and health status variables collected in the survey that may be of interest for subdividing the survey sample, estimating the number of people in categories of interest, and determining cell sizes by related categories. This process can help determine whether there will be sufficient numbers of individuals and a sufficient number of degrees of freedom (see the NHANES tutorial for key concepts about degrees of freedom) in the category of interest who have been successfully linked to CMS data for your NCHS Research Data Center (RDC) proposal.

Why the Feasibility Files are Necessary

Feasibility files allow analysts to determine whether sample sizes are large enough to answer their research questions. To reduce survey participant disclosure risk, the feasibility files contain a limited set of variables.

It is important to note that the variables in the feasibility file do not contain any information about Medicare or Medicaid benefits or payments. They simply provide linkage status variables which enable researchers to gauge whether their analytic samples will have sufficient numbers of sample persons with CMS linked data to warrant an application to the RDC.

Resources

Continuous NHANES Web Tutorial: Key concepts about degrees of freedom

An NHANES-CMS Medicare feasibility file has been created for the NHANES data in two year survey intervals, beginning with 1999-2000. The range of years of NHANES to which Medicare files were linked is described in Course 1, Module 4, Task 2 “Years of NHANES-CMS Linked Data.” Each file contains a linkage status variable, CMS_MEDICARE_MATCH, to indicate whether linkage of the NHANES participant to CMS Medicare records was possible based on linkage eligibility, and if so, whether the participant was successfully matched to at least one year of Medicare administrative data during the linkage interval.

CMS_MEDICARE_MATCH has four possible values:

Possible values for CMS_MEDICARE_MATCH
Values	CMS Linkage Status
1	Linkage-eligible and linked to at least one year of the CMS Medicare administrative data
2	Linkage-eligible but did not link to any year of the CMS Medicare administrative data
3	Linkage-eligible and linked to one or more years of the CMS Medicare administrative data. Child survey participant (less than 18 years of age at time of survey) and turned 18 during the administrative data coverage period. CMS Medicare administrative data available prior to the participant’s 18th birthday, but information indicating whether child linked after 18th birthday is not available.
9	Ineligible for CMS linkage

For the remainder of this tutorial, the term “eligible for linkage” will refer to those individuals represented by values “1,” “2,” and “3.” “Ineligible for linkage” will refer to individuals represented by value “9.”

Each file also includes variables to indicate if a survey participant has a linked data record for each of the seven CMS Medicare administrative record files (Master Beneficiary Summary File, the six fee-for-service claims files, and the Part D prescription drug event data) and for each year of Medicare benefit coverage.

ON_[FILE]_XXXX* has two possible values:

Possible values for linked data record indicator variable
Value	Description
0	Linkage-eligible, linked to one or more years of CMS Medicare administrative data, and does not have data on the file for year XXXX (OR) Linkage-eligible and not linked to any year of CMS Medicare administrative data (OR) Ineligible for CMS Linkage
1	Survey participant has data on the CMS Medicare administrative file for year XXXX

*The ON_[FILE]_XXXX variables will not be created by the sample SAS programs that you will download in Module 6. To create these variables, download, modify and run the programs referenced in Task 2, Step 4 of this module. Additional coding will be required to merge these variables with the NHANES demographic variables.

Information: The feasibility files do not contain any specific information about CMS Medicare benefits or reasons for medical care utilization, only whether at least one claim was found for that beneficiary in each claim type-year combination.

Structure

There is one feasibility file for each two-year NHANES data set. Each feasibility file has the same structure and number of variables. The link to the feasibility file codebook that lists all variables for the file is located in the Resources section.

Location

A link to the NHANES feasibility files can be found in the Resources section below. They are ASCII DAT files that can be freely downloaded.

Resources

An NHANES-CMS Medicaid feasibility file has been created for the NHANES data in two year survey intervals, beginning with 1999-2000. There is one feasibility file for each two-year NHANES data set. The range of years of NHANES to which Medicaid files were linked is described in Course 1, Module 4, Task 2 “Years of NHANES-CMS Linked Data.” Each feasibility file has the same structure and number of variables. The link to the feasibility file codebook that lists all variables for the file is located in the Resources section.

Each file contains a match status variable, CMS_MEDICAID_MATCH, to indicate whether linkage of the NHANES participant to CMS Medicaid administrative data was possible based on linkage eligibility, and if so, whether the participant was successfully matched to at least one year of Medicaid administrative data during the linkage interval.

CMS_MEDICAID_MATCH has four possible values:

Possible values for CMS_MEDICAID_MATCH
Values	CMS Linkage Status
1	Linkage-eligible and linked to at least one year of CMS Medicaid administrative data
2	Linkage-eligible but did not link to any year of CMS Medicaid administrative data
3	Linkage-eligible and linked to one or more years of the CMS Medicaid administrative data – Child survey participant (less than 18 years of age at time of survey) and turned 18 during the administrative data coverage period – CMS Medicaid administrative data available prior to the participant’s 18th birthday, but information indicating whether child linked after 18th birthday is not available
9	Ineligible for CMS linkage

Each file also includes variables to indicate if a survey participant has a linked data record for each of the five CMS Medicaid administrative record files (Person Summary, Inpatient Hospital, Long Term Care, Other Services, Prescription Drug) and for each year of Medicaid benefit coverage.

ON_[FILE]_XXXX* has two possible values:

Possible values for linked data record indicator variable
Value	Description
0	Linkage-eligible, but does not have data on the file for a specific year (YYYY) or ineligible for the NCHS-CMS Medicaid linkage
1	Linkage-eligible and has data on the file for a specific year (YYYY)

Information: The feasibility files do not contain any specific information about CMS Medicaid benefits or reasons for medical care utilization, only whether at least one claim was found for that beneficiary in each claim type-year combination.

Location

A link to the NHANES feasibility files can be found in the Resources section below. They are ASCII DAT files that can be freely downloaded.

Resources

Task 2: Locate and Download Feasibility Files and Documentation

You will need to locate and download the corresponding feasibility data files and documentation from the NCHS website for the NHANES cycles that are included in your research project.

There are three items you will need to download:

Documentation
Public-use ASCII data files
Sample SAS program*

*This program will not be used in this tutorial. Modules 6-8 of this course will use a subset of these variables to assess demographic characteristics (e.g., race/ethnicity and gender) by linkage eligibility and match status.

Step 1: Locate Files and Documentation

Go to the Data Linkage page on the NCHS website.
On the left navigation menu, expand CMS Medicare Data and click on Public-Use Feasibility Files.

Step 2: Download and Review Documentation

Open the data dictionary or save it to your computer.

Step 3: Download Public-use ASCII Data Files

Click the FTP site to NCHS-CMS Medicare Feasibility Files link and open it.
Right click on the feasibility file(s) needed for your analysis by the appropriate survey and year to save to your computer. Modules 6-8 will use MCARE_FEAS_NHANES_2011_2012.dat.

Step 4: Save Sample SAS Program**

Click the Sample SAS Program link and open it.
Copy the text into a SAS editor and save.

**Note that this program is provided for your reference, but a different program will be provided for the activities in Modules 6–8. Save each program using a unique name to avoid confusion.

Step 1: Locate Files and Documentation

Go to the Data Linkage page on the NCHS website.
On the left navigation menu, expand CMS Medicaid Data and click on Public-Use Feasibility Files.

Step 2: Download and Review Documentation

Open the introduction and data dictionary, or save them to your computer.

Step 3: Download Public-use ASCII Data Files

Click on the FTP NCHS-CMS Medicaid Feasibility Files link and open it.
Right click on the feasibility file(s) needed for your analysis by the appropriate survey and year to save to your computer. Modules 6-8 will use MEDICAID_FEASIBILITY_NHANES_2011_2012.dat.

Step 4: Save Sample SAS Program**

Click the Sample SAS Program link and open it.
Copy the text into a SAS editor and save.

**Note that this program is provided for your reference, but a different program will be provided for the activities in Modules 6–8. Save each program using a unique name to avoid confusion.

Task 3: Locate and Download NHANES Files and Documentation

The 2011-2012 NHANES files and their accompanying documentation are used for the tutorial going forward, but similar procedures apply for all cycles of the continuous NHANES.

Go to the National Health and Nutrition Examination Survey page on the NCHS website.
On the left navigation menu, expand Questionnaires, Datasets, and Related Documentation, then expand NHANES 2011-2012, and click on Demographics Data.
Click on DEMO G Data [XPT – 3.6 MB] and save to C:\NHANES or other designated folder.
Download file documentation.

Page last reviewed: March 3, 2021

Content source: