Connecting Surveys to Administrative Records: Data Linkage Opens Doors to New Longitudinal Research
A Valuable Resource for Researchers
NCHS’s record linkage program is designed to maximize the scientific value of the Center's population-based surveys. Linked data files enable researchers to take the “long view” and examine the factors that influence disability, chronic disease, health care utilization, morbidity, and mortality.
Producing Results Through Collaboration
One of NCHS’s major linkage projects combines various NCHS surveys with enrollment and claims data from the Centers for Medicare & Medicaid Services (CMS), and Retirement, Survivor, and Disability Insurance and Supplemental Security Income benefit data from the Social Security Administration (SSA). In other words, for NCHS survey respondents enrolled in one or more of these programs, either at the time of the survey, and in some cases prior to or after the survey, their survey responses are linked to their information from these administrative records.
Making the right connections between the various data sets involves many players and many moving parts. To produce the linked data files, it takes the full cooperation of NCHS, the Office of the Assistant Secretary for Planning and Evaluation, SSA, and CMS.
Through these many collaborative efforts, a number of reports have been published using linked CMS and SSA files. A few recent examples give an idea of the broad scope of research being undertaken:
- “Health Service Use Among the Previously Uninsured: Is Subsidized Health Insurance Enough?” Sandra L. Decker, Jalpa A. Doshi, Amy E. Knaup, and Daniel Polsky. The study used Medicare claims data linked to two different surveys – the National Health Interview Survey and the Health and Retirement Study (Institute for Social Research, University of Michigan) to describe the relationship between insurance status before age 65 and the use of Medicare-covered services beginning at age 65. The authors found that persons who were uninsured before age 65 have 16% fewer visits to office-based physicians than those who were previously insured, but make 18% (previously uninsured) and 43% (previously insured) more visits to hospital emergency and outpatient departments. The authors noted that a key question for the future may be why the previously uninsured continue to use the health care system differently from the previously insured – especially important as health coverage expansions are implemented. (Published in Health Economics, 2011)
- “The Predicted Effects of Chronic Obesity in Middle Age on Medicare Costs and Mortality.” Liming Cai, James Lubitz, Katherine M. Flegal, and Elsie R. Pamuk. The authors found that obese 45-year-olds had a smaller chance of surviving to age 65 and, if they did, incurred significantly higher average lifetime Medicare costs than normal-weight 45-year-olds. They concluded that Medicare costs may rise substantially in the future to meet the health care needs of today’s obese middle-aged population, and that active engagement by both the private and public sectors to prevent and reduce obesity are critically needed. (Published in Medical Care, Vol. 48, No. 6, June 2010)
- National Health Statistics Report No. 53, “Health Characteristics of Medicare Traditional Fee-for-Service and Medicare Advantage Enrollees: 1999-2004 National Health and Nutrition Examination Survey Linked to 2007 Medicare Data.” Lisa B. Mirel, Division of Health and Nutrition Examination Surveys, and Gloria Wheatcroft, Jennifer D. Parker, and Diane M. Makuc, Office of Analysis and Epidemiology. This recent National Health Statistics Report examines measured and reported health characteristics of Medicare enrollees and is the first full NCHS report analyzing data from the 1999-2004 National Health and Nutrition Examination Survey linked to the 2007 Medicare enrollment data. (Published in May 2012 and available for download from the NCHS website)
With a regulatory and legislative mandate to protect the privacy of those who participate in our programs, NCHS takes extraordinary measures to ensure confidentiality and prevent the disclosure of personal information.
Confidentiality is also an integral part of the proposal process. All requests to conduct research using the data linkage files are carefully scrutinized for potential disclosure risks. Proposals are also reviewed for technical feasibility, appropriateness, availability of the data, and public health benefits. Researchers are granted access only to the restricted variables necessary to undertake their work through the Research Data Center.
Looking to the Future
Linked files shed valuable light to help us understand new correlations. Program participation and utilization data expand the picture beyond the self-reported information collected in our surveys, just as survey information–individual and family socioeconomic status measures, self-reported limitations and health conditions, employment status, and more–add needed context to records data from CMS and SSA.
However, it is the additional ability to delve into the longitudinal aspect–to explore both the prospective and retrospective views–that enables researchers to provide the necessary framework for directing health policy.
Special Projects Branch analysts Cordell Golden and Donna Miller contributed to this article. They also make significant contributions to the successful execution of the data linkage program.
For more information on data linkage, including how to submit research proposals, contact Dr. Jennifer Parker, Special Projects Branch Chief, at firstname.lastname@example.org.
A session entitled “Linked NCHS–CMS Data: Analytic Methods and Research Examples,” will take place at the National Conference on Health Statistics, Tuesday, August 7, 10:30 a.m.-12:00 p.m Visit the Conference page to register.