Paula Diehr, PhD; Donald L. Patrick, MSPH, PhD
Grant Support: By National Heart, Lung, and Blood Institute contracts N01-HC-85079 through N01-HC-85086, N01-HC-35129, and N01 HC-15103.
Potential Financial Conflicts of Interest:Honoraria: P. Diehr (National Institutes of Health, Veterans Administration); D.L. Patrick (National Center for Health Statistics, Centers for Disease Control and Prevention); Grants pending: P. Diehr (Centers for Disease Control and Prevention), D.L. Patrick (Centers for Disease Control and Prevention).
Requests for Single Reprints: Paula Diehr, PhD, Department of Biostatistics, University of Washington, Box 357232, Seattle, WA 98195; e-mail, email@example.com.
Current Author Addresses: Dr. Diehr: Department of Biostatistics, University of Washington, Box 357232, Seattle, WA 98195.
Dr. Patrick: University of Washington, Box 357660, Seattle, WA 98199-7660.
Diehr P., Patrick D.; Trajectories of Health for Older Adults over Time: Accounting Fully for Death. Ann Intern Med. 2003;139:416-420. doi: 10.7326/0003-4819-139-5_Part_2-200309021-00007
Download citation file:
Published: Ann Intern Med. 2003;139(5_Part_2):416-420.
The study of healthy aging is essentially the study of how health-related variables, such as cognition, change over time for persons with different characteristics. One of the best ways to demonstrate this change is a graph, such as a plot of average cognitive ability over time for men and women. Unfortunately, if some people die during the study, their longitudinal health variable (which we refer to generically as “health”) is “missing” after their death. The usual analytic approach is to perform a complete case analysis or an available data analysis, both of which restrict study to the healthiest subgroup of the study sample and may give an overly optimistic picture of the trajectory. Our goal is to develop a graph of health over time that accounts appropriately for death. In this paper, we demonstrate how omission of the deaths gives incorrect results and explain and illustrate two approaches for including deaths in longitudinal graphs.
to gain full access to the content and tools.
Learn more about subscription options.
Register Now for a free account.
Copyright © 2016 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use
This PDF is available to Subscribers Only