Disparities Between Patients' End-of-Life Care Preferences and Actual End-of-Life Care Received

Background: Data suggest that patients’ end-of-life (EOL) care preferences are not consistently followed. Based on these data, in 2015, the IOM called for research to identify the prevalence of the problem of EOL care preferences not honored. The current study used an ethical framework to study this problem.

Objective: The purpose was to identify the prevalence of patient’s not having their EOL care preferences followed using data from an existing nationally representative study.

Methods: A secondary analysis of data from seven biennial waves (2002-2014) of the Health and Retirement Study (HRS) was performed. A cross-sectional analysis was accomplished using primarily the HRS exit interview. All respondents indicating that a written EOL preference had been documented were included. Deceased patients’ proxies provided the information about the patient’s preferred EOL care and the actual care received. Because proxies could select more than one EOL care option, both a sample and a subsample of proxies who only selected one option were evaluated. Associations between mismatched EOL preferred vs. EOL received care options and respondent characteristics were also evaluated.

Results: 3,754 respondents died during the study period, of who 3,660 met the inclusion criteria of having written EOL care instructions. The subsample included 212 respondents. Analysis of the sample found that the most preferred type of EOL care was comfort care, followed by limit care. The frequencies of both samples revealed that in the four EOL care types for which HRS collects data (all care possible, limit care, withhold treatment, comfort care), patient preferences were not followed 53-72% of the time.

A question about providers verbally offering EOL care treatments to the patient’s proxy were asked and nearly 17% of the time those instructions were not followed.

Discussion: Using data from a nationally representative study, this study confirms that regardless of the type of EOL care patients’ prefer, and the presence of written instructions, patients do not receive that care greater than 50% of the time, and a significant number are receiving unwanted EOL care. Further data are needed to ascertain why patients’ autonomous preferences are not followed at the EOL.