Blogs: A brief introduction to the Harmonized HRS End-of-Life Dataset

A brief introduction to the Harmonized HRS End-of-Life Dataset

by Hannah Bae
Oct 1, 2020

If you are interested in aging research, you may have heard about the Health and Retirement Study (HRS). As introduced in a previous blog post by Hunter Green, HRS is an ongoing longitudinal survey on a nationally-representative sample of elderly people in the U.S. initiated in 1992. The core HRS interview is conducted biennially on participants included in the panel sample. For deceased HRS participants, a post-mortem interview (referred to as an exit interview) is usually conducted within two years of their deaths. For detailed documentation of the HRS End-of-Life (EoL) data files, refer to the Harmonized HRS EoL codebook .

The USC Gateway to Global Aging team, funded by the NIA, created the Harmonized HRS End of Life (EoL), a user-friendly version of a subset of the HRS Exit and Post-Exit Interviews (interviews taking place after the Exit Interview to obtain further information). Both the data file structure and naming of variables mimic the RAND HRS Longitudinal File, created by the RAND Center for the Study of Aging, and other Gateway Harmonized datasets. The Harmonized HRS EoL dataset is downloadable from the HRS website along with the Harmonized HRS EoL creation code, which is provided for transparency.

The Harmonized HRS EoL data file presents multiple exploration advantages for researchers. This blog post addresses two of these merits. First, it promotes usability for researchers and individuals who aim to explore the exit interview. Since the Gateway to Global Aging Data team provides detailed codebooks and documentation (with close descriptions of variable creation using the original HRS exit interview data), users can easily familiarize themselves with the original HRS Exit Interview. For instance, Figure 1 presents the list of variables in the original HRS Exit interview used to construct variables for hospital stays in the Harmonized HRS EoL.

Fig 1. Sample codebook entry: HRS Variables Used

Once familiar with the Harmonized HRS EoL, the user becomes ready to conduct comparative analyses across the HRS Family of Surveys, which is the second merit. Consider a user who aims to conduct a comparative study on out-of-pocket expenditures by the elderly before death using the Survey of Health, Ageing, and Retirement in Europe (SHARE) and HRS. To enhance usability, the Gateway team adds an entry titled “Differences with the Harmonized HRS End of Life” for each variable included in the Harmonized SHARE EoL codebook. The section informs a user on whether text from the questionnaire and the construction of variables in the Harmonized SHARE EoL differ from those in the Harmonized HRS EoL. For instance, Figure 2 shows a section for out-of-pocket medical expenditure variables. As you will notice, the question text to measure the cost of in-home medical care in the SHARE end-of-life survey is not identical to that of the HRS exit interview.

Fig 2. Sample codebook entry: Differences with the Harmonized HRS End of Life

There are more merits to using the Harmonized HRS EoL, such as the advantages of easily linking files from the core and exit interviews to construct trajectories for comorbidities of the deceased respondents. For more information, please visit the Gateway website to check the list of studies with Harmonized datasets and explore how they could be used for your project.

about the author
Hannah Bae is a doctoral student in the Economics Department at the University of California, San Diego.