Non-Response and Weighting Issues with the NHANES-CMS Linked Data

In the case of NHANES-CMS linked files, non-response occurs in the process of linking NHANES to administrative data from CMS, due to linkage ineligibility. Linkage eligibility, the criteria by which survey participants can be potentially linked to CMS data, is defined in Course 1, Module 3.

Information: The term eligibility is being used to indicate linkage eligibility, not eligibility for Medicare or Medicaid benefits. Criteria for program eligibility can be found on the CMS website.

Resources

• Centers for Medicare and Medicaid Services (CMS) website

The properties of the NHANES interview and examination weights for linked data files with incomplete linkage due to ineligibility for linkage and non-matches are unknown. In addition, methods for using the survey weights for some longitudinal analyses require further research. Because this is an important and complex methodological topic, ongoing work is being done at NCHS and elsewhere to examine the use of survey weights for linked data in multiple ways.

Until specific recommendations are available, preliminary guidance is for you to create adjusted survey weights that account for linkage ineligibility. One approach for adjusting the NHANES sample weights follows:

1. Using the whole sample (linkage-eligible and not linkage-eligible), obtain estimated population counts by summing the NHANES sample weights within subgroups defined by combinations of gender, age group, and race/ethnicity.
2. For the survey participants who are linkage-eligible, adjust the NHANES sample weight so that the total sum of the adjusted weights in each combination of gender, age group, and race/ethnicity is the same as that for the whole sample. The adjusted weights for the survey participants who are not linkage eligible are 0. This approach uses a ratio adjustment of the weights for linkage eligibility (non-response) (Lohr, 1999). The criteria by which survey participants can be potentially linked to CMS data as defined in Course 1, Module 3, Task 1. Check your work by making sure that the sum of the weights of the total sample is the same as the sum of the weights for only the linkage-eligible sample when using the eligibility-adjusted weights.

Using methods like those described above, you can produce adjusted sample weights using these two steps by programming directly in standard statistical software, such as SAS and Stata. When deciding on the demographic combinations, particularly age categories, there should be enough survey participants in each combination (about 30 is often suggested, though there are no definitive guidelines). Age categories that are consistent with the NHANES sample design can be found in the NHANES analytic guidelines (see Resources for link). It is recommended that race/ethnicity groups be consistent with NHANES sampling (Curtin, 2012). Another option for adjusting the weights is to use the PROC WTADJUST procedure in SUDAAN, version 10. The WTADJUST procedure computes non-response weight adjustments using a model-based calibration approach like a logistic regression and offers more flexibility by permitting the use of continuous variables, more main effects, and lower order interactions in the model. The WTADJUST procedure can also post-stratify to population totals. More detailed information on adjusting sample weights for linkage eligibility using SUDAAN can be found in Appendix III of the series report, “Linkage of NCHS Population Health Surveys to Administrative Records From Social Security Administration and Centers for Medicare & Medicaid Services,” linked in the References section below.

Other approaches for handling linkage-eligibility for the linked CMS data, including alternative methods for adjusting sample weights, may be used. Researchers should seek assistance from a statistician for guidance on their project. Additional information will be posted on the NCHS Data Linkage webpage as it becomes available. Please contact the NCHS Data Linkage Team (datalinkage@cdc.gov) for more information.

Resources

• Lohr S. Sampling Design and Analysis. Albany, NY, Duxbury Press, 1999.
• NHANES Survey Methodology and Analytic Guidelines
• Curtin LR, Mohadjer L, Dohrmann S, et al. The National Health and Nutrition Examination Survey: Sample design, 1999–2006. National Center for Health Statistics. Vital Health Stat 2(155). 2012.
• Research Triangle Institute (2008). SUDAAN Language Manual, Release 10.0 Research Triangle Park, NC: Research Triangle Institute.
• Linkage of NCHS Population Health Surveys to Administrative Records From Social Security Administration and Centers for Medicare & Medicaid Services