Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Racial disparities in advance care planning and directives completion and end-of-life outcomes in the United States
(USC Thesis Other)
Racial disparities in advance care planning and directives completion and end-of-life outcomes in the United States
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Racial Disparities in Advance Care Planning and Directives Completion and End-of-Life
Outcomes in the United States
By
Yujun Zhu
A Dissertation Presented to the
FACULTY OF THE USC LEONARD DAVIS SCHOOL OF GERONTOLOGY
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(GERONTOLOGY)
August 2023
ii
Dedication
To my wife Yingying Wang, daughter Yizhen Joanna Zhu and son Youzhen Joshua Zhu for their
unconditional love and being with me up and down.
I would like to extend my heartfelt gratitude to my parents and in-laws for their understanding
and support in allowing me to pursue my doctoral studies on the other side of the Pacific. I am
profoundly grateful for their sacrifices that I may or may not know.
Lastly, to May Ng for her unselfish support throughout my years in the United States.
iii
Acknowledgements
I would like to thank Dr. Susan Enguídanos, my mentor and dissertation chair, for her
tremendous support and guidance throughout my doctoral journey. I cannot achieve this without
her encouragement and trust in me. I am grateful for her willingness and patience to guide me in
transferring research ideas into meaningful publications.
I want to thank my dissertation committee members, Dr. Mireille Jacobson and Dr. Kate Wilber,
for their valuable input throughout the entire research process, ranging from methodology
support to insightful discussions on policy implications of my studies.
I also thank my lab peers. We grew together and learned from each other over the past five
years.
iv
Table of Contents
Dedication ....................................................................................................................................... ii
Acknowledgements ........................................................................................................................ iii
List of Tables ................................................................................................................................. vi
List of Figures ............................................................................................................................... vii
Abstract ........................................................................................................................................ viii
Chapter 1: Introduction and Study Aims ........................................................................................ 1
Background ................................................................................................................................. 1
Literature Review........................................................................................................................ 2
Research Aims and Hypotheses .................................................................................................. 7
Theoretical Framework ............................................................................................................... 9
Data ........................................................................................................................................... 10
Chapter 2: Racial Differences in Trends in Advance Care Planning and Advance Directive
Completion among US Older Adults, 2000-2018......................................................................... 11
Introduction ............................................................................................................................... 11
Methods..................................................................................................................................... 12
Results ....................................................................................................................................... 14
Discussion ................................................................................................................................. 17
Chapter 3: The Association between Health Literacy and ACP/AD Completion ........................ 29
Introduction ............................................................................................................................... 29
Method ...................................................................................................................................... 30
Results ....................................................................................................................................... 33
Discussion ................................................................................................................................. 34
Chapter 4: Advance Directives Completion and Hospital Out-of-Pocket Expenditures .............. 41
Introduction ............................................................................................................................... 41
Methods..................................................................................................................................... 42
Results ....................................................................................................................................... 46
Discussion ................................................................................................................................. 49
Chapter 5: Discussion and Conclusion ......................................................................................... 57
v
Discussion ................................................................................................................................. 57
Conclusion ................................................................................................................................ 66
Bibliography ................................................................................................................................. 67
Appendices .................................................................................................................................... 79
Appendix A: Joinpoint Graphs ................................................................................................. 79
Appendix B: Propensity Score Model ...................................................................................... 81
vi
List of Tables
Table 2.1 Characteristics of the Study Population ........................................................................ 21
Table 2.2 Unadjusted and Adjusted Prevalence of ACP/AD Completion, 2000-2018 ................ 23
Table 2.3 Change in Trends of ACP/AD Completion by Race .................................................... 28
Table 3.1 Sample Sociodemographic Description Stratified by Health Literacy ......................... 38
Table 3.2 Logistic Regression Analyses of End-of-Life Planning ............................................... 39
Table 4.1 Sample descriptions ...................................................................................................... 53
Table 4.2 Predicted hospital out-of-pocket costs combining both parts of two-part model
(in 2014 US dollars) ...................................................................................................................... 55
Table 7.1 Appendix: Propensity score weighting and regression-adjusted estimate of AD
completion effects on hospital out-of-pocket costs (in 2014 US dollars) ..................................... 82
vii
List of Figures
Figure 1.1 Andersen’s Behavioral Model of Health Service Utilization with Selected
Variables ....................................................................................................................................... 10
Figure 2.1 Trends in Adjusted Prevalence of ACP/AD Completion ............................................ 26
Figure 3.1 Study Sample Selection Workflow ............................................................................. 30
Figure 4.1 Marginal effects of advance directive completion by age ........................................... 56
Figure 7.1Appendix A: Joinpoint Trends in Adjusted Prevalence of ACP/AD Completion ....... 79
Figure 7.2 Appendix B: Unweighted and IPW-weighted balance between AD completers
and non-completers ....................................................................................................................... 81
viii
1. Abstract
Advance care planning (ACP) is a process to plan for end-of-life care and may culminate
in the development of an advance directives (AD) that documents those plans. Studies have
documented positive quality of life outcomes including receiving care concordant with
preferences, fewer unwanted aggressive treatments, and dying in one’s preferred place of death.
Recently there has been some debate about the value of ACP/AD. The objective of this
dissertation was to better understand trends in engagement in ACP/AD among different racial
groups, factors associated with ACP/AD activities, and financial outcomes related to AD
completion. These aims were accomplished by analyzing data from the Health and Retirement
Study (HRS). The results showed stagnation in ACP/AD completion in recent years and different
trends across racial groups. I also found that health literacy was independently associated with
ACP/AD engagement. Finally, I found that completing an AD was related to lower hospital out-
of-pocket costs. Findings from this dissertation suggest current efforts at engaging Black
populations in ACP/AD may not be effective or culturally appropriate, but that efforts to engage
Hispanics have been successful with ACP/AD rates increasing across the years. Practice
implications include the need for healthcare providers to screen patients’ level of health literacy
and customize ACP/AD materials accordingly. Providers also should discuss costs related to
health interventions during ACP discussions.
1
Racial Disparities in Advance Care Planning and Directives Completion and End-of-Life
Outcomes in the United States
2. Chapter 1: Introduction and Study Aims
Background
There is growing debate regarding the value of continued research on advance care
planning (ACP) and advance directives (AD), with calls to instead focus on research that
advances palliative care practice.
1
Concern has focused on the fact that despite 30 years of effort
aimed at promoting ACP, some believe these efforts did not result in improved end-of-life (EOL)
care quality or goal concordant care.
2
However, there is broad recognition of the value of these
conversations for family members in determining what care patients desire at end of life. ACP
also is an important path to encourage people to think and talk about EOL healthcare options,
including palliative and hospice care, which have been shown to positively impact quality of
care.
3,4
Additionally, there are substantial racial disparities in ACP/AD completion rates as well
as in access to EOL care. Therefore, it is important to better understand the racial differences in
pathways that predict ACP/AD completion and healthcare preferences as well as quality
outcomes associated with ACP/AD.
Definitions
In 2017, a multidiscipline panel reached a consensus definition of advance care
planning, which is “a process that supports adults at any age or stage of health in understanding
and sharing their personal values, life goals, and preferences regarding future medical care. The
goal is to help ensure that people receive medical care that is consistent with their values, goals
and preferences during serious and chronic illness. For many people, this process may include
2
choosing and preparing another trusted person or persons to make medical decisions in the
event the person can no longer make this or her own decisions”.
5
Advance directives are “legal documents state a person’s wishes about receiving medical
care if that person is no longer able to make medical decisions because of a serious illness or
injury. An advance directive may also give a person (such as a spouse, relative, or friend) the
authority to make medical decisions for another person when that person can no longer make
decisions”.
6
An advance directive can be completed at any time and allows people to express
their values and desires in end-of-life situations. There are two main components to an advance
directive: a living will and a healthcare proxy. A living will contains written instructions that
help doctors know about treatment preferences should patients become unable to make decisions
for themselves. A durable power of attorney for healthcare is an assigned person that can make
medical decisions on behalf of the patient, also called healthcare proxy. In addition, medical
orders such as Do Not Resuscitate (DNR), Physician Orders for Life-Sustaining Treatment
(POLST) or Medical Orders for Life-Sustaining Treatment (MOLST) can be supplementary
forms of ADs to further fulfill patients’ wishes and act as medical orders that are recognized by
healthcare providers. However, in the Health and Retirement Study (HRS) dataset that was used
in this study, we are unable to distinguish the specific form participants completed.
Literature Review
History of ACP and AD
In 1967, the first advance directive was introduced by Luis Kutner who suggested an -
individual should indicate their healthcare preferences ahead of time. In 1976, California became
3
the first state to adopt this concept and put it into legislation. Although initial implementation of
advance directives was stalled due to ethical concerns and healthcare providers’ training that
focused on aggressive care to prevent death, by the end of 1986, forty-one states had adopted the
concept of living wills. To make up for the limitations of living wills, legislators came up with
another legal document, the durable power of attorney (also known as healthcare proxy). By the
end of 1997 every state had some version of a durable power of attorney. In 1990, Congress
passed the Patient Self-Determination Act (PSDA) that required healthcare providers to educate
and inform patients of their rights to make and document decisions for future healthcare
treatment in healthcare facilities such as hospitals, skilled nursing facilities, hospice
organizations and home health organizations. In 2016, the Centers for Medicare & Medicaid
Services provided new Current Procedural Terminology (CPT) codes to reimburse healthcare
providers for ACP discussions.
AD Completion Rates and Trends
AD completion rates are low but increasing, the 1995 GAO reported that studies
documented AD completion rate ranging from 5% to 29% in early 1990s.
7,8
In 2014, two
separate studies found the overall rate increased to 26% for adults 18 years above, and for those
aged 60 and older at death, 45% had a living will .
9,10
A recent systematic review found that
about one in three Americans had completed an AD of some type.
11
In addition, multiple studies
have found that Hispanics and non-Hispanic Blacks have significantly lower rates of AD
completion compared to non-Hispanic Whites.
12
Although studies have found an increase in AD
completion rates among Blacks,
13
these rates remain far lower than their White counterparts.
4
Although it is not the primary goal of the ACP conversation, the process of end-of-life
care discussions can lead to the development of a formal AD document. For example, Shepherd-
Banigan et al. found that patients with cognitive impairment who had had an ACP discussion
with their caregivers were more likely to have completed a formal AD.
14
However, the
prevalence of ACP discussion, especially with healthcare providers, was also low even among
patients 65 years old and over.
15,16
In addition, racial differences also exist in ACP. For example,
early studies found that Black and Hispanic patients who had metastatic lung cancer were less
likely than White patients to have discussions about hospice with their physicians.
17
More
recently, Palmer et al. found Black and Hispanic Medicare Beneficiaries were less likely to have
an ACP claim than Whites.
16
Given the growing concern around the effectiveness of ACPs and related research along
with sustained ethnic disparities in ACP/ADs, it is important to understand the trends of
ACP/AD completion among racial groups.
Health Literacy and Advance Directive Completion in Different Racial Groups
Previous studies have identified several factors such as age, education, and health
conditions that are associated with ACP/AD engagement.
18
Although race and ethnicity are well
studied and significantly associated with AD completion in early studies,
19
there are conflicting
findings about the role of racial disparities after controlling for covariates such as attitudes
toward AD and social support. Those differences may suggest potential unobserved factors that
have not been well captured in the dataset. In addition, some factors that have been associated
with ACP/AD engagement, such as education and health conditions are often difficult to change.
5
Efforts are needed to identified other important factors that may help inform effective strategies
to improve end-of-life planning.
One potential explanatory factor for lower ACP/AD completion is health literacy, a
relationship which has not been adequately explored. According to the Centers for Disease
Control and Prevention, health literacy is defined as “the degree to which individuals have the
ability to find, understand, and use information and services to inform health-related decisions
and actions for themselves and others”.
20
Studies have found an association between low health
literacy with lower understanding of ACP and with lower rates of ACP knowledge among non-
white participants.
21
These low levels of health literacy may contribute to low AD completion
among Hispanics and African Americans.
22
For example, one study that recruited patients from
local clinics and health centers found that race and low health literacy were significantly
associated with lower likelihood of completing an AD,
23
which may further affect end-of-life
healthcare access and treatment. Luo et al. found Medicare beneficiaries from high health
literacy areas were less likely to receive aggressive care.
24
However, to my best knowledge,
there is no study using nationally representative data that has explored the relationship between
health literacy and ACP/AD completion in the United States.
Hospital Out-of-Pocket Costs
Healthcare costs are high among older adults, largely due to increasing need for medical
services utilization with age. The average medical expenses among older adults between 70 and
90 years old have been found to be over two times the expenditures of those under 70.
25
A large
portion of healthcare costs are spent on end-of-life care and research also has documented that
minorities have higher end-of-life care costs than Whites, mostly due to opting for more
6
aggressive and life sustaining care.
26-29
In addition, studies have found that about one-fourth of
Medicare expenditures incurred were for care for patients in their last year of life.
29,30
The
average health care cost per capita in the last year of life was estimated at $80,000, and about
44% of these costs were related to hospitalizations.
31
Moreover, healthcare costs are generally accompanied by out-of-pocket costs insured
patients incur in the form of cost-sharing, such as deductibles, copayments, and coinsurance, and
expenses for services that are not covered by health insurance plans. For example, a study found
decedents with cancer experienced the highest level of out-of-pocket costs at the end of life,
mostly driven by inpatient hospitalizations.
32
Another study found Medicare beneficiaries aged
85 years and over had three times more out-of-pocket costs than those aged between 65 and 74.
33
Studies also have found that high out-of-pocket costs associated with hospitalizations can lead to
financial catastrophes.
34
In the United States, average out-of-pocket costs have been estimated at
$38,688 in the 5 years prior to death.
35
In the last year prior to death, total out-of-pocket costs for
health-related care were found to be $11,618,
36
and inpatient out-of-pocket costs were $2,176.
37
Further, research has found a strong association between aggressive care and end-of-life costs,
leading to increased risk of financial hardship.
38
Concerns about incurring high out-of-pocket copayments may influence patients’ health
care decision making.
39,40
In a study of cancer patients, researchers found that patients may be
less willing to pay high out-of-pocket charges when treatment results in modest clinical
benefits.
41
Other studies have found that those with lower socioeconomic status may forego
expensive treatment regardless of its effectiveness.
42-44
Moreover, terminally ill patients may
prefer to spend less on personal medical costs with limited benefits, and instead use the money
for other meaningful pursuits at end of life.
45
7
Advance directives (ADs) have been associated with improved care concordance with
care preferences, lower health care utilization, lower healthcare expenditures, and lower odds of
dying in a hospital.
27,46-51
In addition, researchers also found that overall, most patients
completing ADs elect limited or comfort care, but that individuals completing ADs in the last
three months of life had higher odds of preferring aggressive, life sustaining care.
52
However, absolute healthcare costs may not correlate with patient out-of-pocket costs, as
some insurances (e.g., Medicare or Medicaid) may bear the brunt of healthcare expenditures.
53
Although it is important to know the total costs of end-of-life health care, we believe hospital
costs are more directly related to costs that can be influenced by care choices made within an
AD. Thus, given this lack of correlation in expenditures, coupled with patient concern related to
incurring high out-of-pocket costs associated with hospital care at end of life, research is needed
to understand the relationship between AD completion and hospital out-of-pocket cost.
Research Aims and Hypotheses
This dissertation aims to 1) identify ACP/AD completion rate trends among different
racial groups between 2000 and 2018; 2) examine whether health literacy is associated with
ACP/AD completion; and 3) examine how AD completion status affects hospital out-of-pocket
expenditures.
Aim 1
Objective: To identify trends in prevalence of ACP/AD completion between 2000 and 2018
among decedents of different racial groups.
Hypotheses
8
a) Prevalence of ACP/AD completion rates will increase over time but change differently
across racial groups.
b) There will be a non-linear change in prevalence of ACP/AD between 2000 and 2018.
Aim 2
Objective: To explore the association between health literacy and end-of-life care planning.
Hypotheses
a) Non-Hispanic Black and Hispanic decedents will be more likely to have limited health
literacy compared to non-Hispanic White decedents.
b) Health literacy will be independently associated with ACP/AD completion after
controlling for other explanatory factors such as race, education, and religious
participation.
Aim 3
Objective: To understand the relationship between AD completion and hospital out-of-pocket
cost.
Hypotheses
a) Compared to decedents without ADs, those who completed ADs will have lower hospital
out-of-pocket costs in the end of life.
b) Completing an AD will be associated with higher hospital out-of-pocket reduction among
younger decedents.
c) Among those with ADs, timing of completion and documented care preferences will be
associated with hospital out-of-pocket costs.
9
Theoretical Framework
This dissertation will be guided by Andersen’s Behavioral Model of Health Service
Utilization, which is frequently used to guide healthcare service utilization studies due to its
simplicity and flexibility in selecting potential factors by providing theory based and straight
forward options. This conceptual framework also has been used to explore racial disparities in
healthcare services.
54,55
The main components of this model are 1) predisposing factors are fixed
individual characteristics such as biological gender and race; 2) enabling factors are conditions
such as income and insurance coverage that allow individuals to get access to healthcare
resources; 3) need factors are driven by perceived healthcare need usually indicated by self-rated
health status, disease diagnoses, symptoms and mobilities. These three factors may directly or
indirectly influence their 4) health behaviors such as personal health practices and use of
healthcare services; and ultimately affect 5) outcomes such as perceived health status and
healthcare satisfaction. This conceptual model has been applied in studies exploring factors
related to ACP/AD completion. Therefore, this theoretical framework will be guiding variable
selection to investigate the association between predisposing, enabling, need factors and
ACP/AD completion (health behavior) as well as racial disparities in healthcare outcomes.
(Figure 1.1)
10
Figure 2.1 Andersen’s Behavioral Model of Health Service Utilization with Selected Variables
Data
This study exclusively used a secondary dataset, the Health and Retirement Study (HRS),
a nationally representative longitudinal survey for American older adults. The HRS started in
1992 and replenished its sample in 1998, 2004, 2010, and 2016. The cumulative sample size
reached more than 42,000 participants. There are two main biennial data streams: 1) the HRS
Core datasets that collected information from living participants; and 2) the HRS Exit datasets
that focused on deceased participants. The Core interviews capture various population
characteristics, and the Exit interviews include rich information about decedents end-of-life
conditions from interviews conducted with their survey proxies. In addition, there are off-year
studies on specific topics that are not covered by the two main interviews such as an Internet
survey, a health care study, and a nutrition study. A subsample of HRS respondents were
randomly selected to participate in those studies.
11
3. Chapter 2: Racial Differences in Trends in Advance Care
Planning and Advance Directive Completion among US Older
Adults, 2000-2018
Introduction
Numerous studies have shown that engagement in advance care planning (ACP) and
advance directive (AD) completion is associated with more concordant care, less aggressive
treatment, and reduced family members’ stress when they need to make healthcare
decisions.
4,56,57
However, the rates of ACP/AD completion have been generally low, with only
one-third of US population having any type of ACP/AD completed.
11
It also is well-known that
Blacks and Hispanics are less likely to have completed an ACP/AD.
23,58,59
Although prior studies have investigated the prevalence of ACP/AD completion among a
variety of populations, only a handful of studies have focused on examination of the completion
trends over time, with mixing findings. For example, Narang and colleagues
60
in a study of
cancer patients, found that only healthcare proxy (HCP) assignment increased significantly from
2000 to 2012, while living will (LW) and ACP discussion did not increase. Another study
focused on the role of socioeconomic status found all ACP/AD components increased between
2002 and 2010.
61
Similarly, Block and colleagues also found gains in ACP/AD completion
among older adults who received intensive care at end of life between 2000 and 2015.
62
In a
recently published study, researchers found that none of the ACP/AD components increased
among dementia patients between 2000 and 2014.
63
In addition, only one study focused on
ACP/AD trends among racial groups, and they found Whites and Blacks, but not Hispanics, had
increasing use of living wills between 1998 and 2012.
64
The mixing findings may be partly due
to different study time frame, study populations, and various specific sub-populations examined.
12
However, it is also possible that the change in ACP/AD prevalence is non-linear, with variation
in trends across the years. These differences may have been overlooked in the analytic methods
used to examine the ACP/AD trends. Although traditional modeling such as linear and logistic
regressions are appropriate in many situations, this analytic approach may not reflect the
complex trends and may fail to capture changes in magnitude and direction in trends, particularly
when the data collected includes multiple years, allowing for significant longitudinal trend
analysis.
65,66
Therefore, this study was conducted to identify trends in prevalence of ACP/AD
completion among different racial groups. I sought to answer the following research questions:
(1) What are the differences in trends in ACP/AD completion across racial groups?
(2) What are the patterns of change in ACP/AD completion trends between 2000 and
2018?
Methods
Study Population
In this trend analysis, I used data from HRS Exit interviews, which collected deceased
participants’ information about their last two years of life. I limited the sample to those who died
between 2000 and 2018 (N = 12,675) as HRS began asking ACP/AD related questions in 2000
Exit interview and 2018 is the most current wave of data available for this dataset. Those who
had missingness on an ACP/AD measure (n = 1,320) and those identifying as a race/ethnicity
other than Non-Hispanic White, Non-Hispanic Black or Hispanic (n = 228) were excluded. The
final sample was 11,127.
13
ACP was defined as having had a discussion about end-of-life care preference. Survey
proxies of participants were asked whether the deceased ever discussed with anyone else the
treatment or care they wanted to receive in the final days of life. AD was defined as either having
a living will or healthcare proxy. Proxies were asked whether the deceased participants provided
written instructions or made arrangements for specific persons to make decisions about end-of-
life medical treatment.
Analysis
Bivariable logistic regression analyses were conducted to examine whether and how
prevalence varied by decedents’ end-of-life characteristics. Both raw numbers and weighted
prevalence were presented. Next, I conducted multivariable logistic regression models to
estimate the association between year of death and ACP/AD completion rates, controlling for
decedent characteristics such as age at death, gender, race/ethnicity, psychosocial variables
including education, insurance, and health conditions including cancer, heart conditions, lung
disease, memory problems, and stroke. To get adjusted prevalence, I performed estimated
probabilities of each end-of-life planning completion in each year and across each racial group.
Interaction effects between race and year of death were included to detect different ACP/AD
trends among different racial groups. Based on the adjusted prevalence results for each year of
decedents, Joinpoint Regression Software version 5.0 (National Cancer Institute) was used to
identify potential nonlinear trends by using Bayesian Information Criterion. This method
provides valid average percentage change (APC) in rates between trend-change points and
average annual percentage change (AAPC) over the entire study period from 2000 to 2018. APC
and AAPC are equal if a turning point is not detected.
67
14
In all analyses, respondent-level survey weights were taken into account for the complex
sampling design and p<0.05 (two-sided) was considered statistically significant.
Results
Table 2.1 shows the raw number and weighted percentages of decedents. Among the
sampled 11,127 deceased participants, 53.9%were female, 51.6% were aged 65 through 84 years
old, 38.2% were 85 years and over. In terms of race/ethnicity, 75.8% were non-Hispanic White,
16.6% were non-Hispanic Black and 7.6% Hispanics. Between 2000 and 2018, about 58.4% of
the deceased population had ACP discussion (95% CI, 56.7% to 60.0%), 47.2% completed a
living will (95% CI, 45.6% to 48.8%), 61.3% assigned a healthcare proxy (95% CI, 59.4% to
63.1%), 78.9% had at least one of the three ACP/AD components (95% CI, 77.7% to 80.0%) and
29.9% had all three ACP/AD components (95% CI, 28.6% to 31.3%).
Factors Associated with Having at Least One ACP/AD Component
All model variables were significantly associated with having at least one of the
ACP/LW/HCP components (Table 2.2). In details, decedents who were women, in older age
groups, non-Hispanic Whites, higher education, with chronic conditions, covered by Medicare,
and those did not have Medicaid were all significantly associated with higher probability of
having at least one of advance care planning component.
Prevalence and trends of Having at Least One ACP/AD Component
15
Figure 2.1 and Table 2.3 show trends in ACP/AD completion overall and by each racial
subgroup. Patterns of ACP/AD completion were different among each subgroup. Joinpoint
graphs are reported in Appendix A.
The adjusted prevalence of at least one advance care planning component increased from
68.6% to 82.6% at an AAPC of 1% (95% CI, 0.6% to 1.4%) from 2000 to 2018. The increase
was higher from 2000 to 2005, with an APC of 2.6% (95% CI, 1.3% to 6.4%), and then slowed
to an APC of 0.4% (95% CI, -0.5% to 0.7%) from 2005 to 2018.
Among non-Hispanic White decedents, the prevalence of having at least one component
of ACP/AD completion increased from 2000 to 2005 at an APC of 2.3% (95% CI, 0.7% to 8.6%)
but then increased at a much lower APC rate of 0.2% (95% CI, -5.0% to 0.6%) from 2005 to
2018 such that the AAPC increase from 2000 to 2018 was not statistically different from zero.
The prevalence of at least one ACP component increased consistently among both non-
Hispanic Black and Hispanic decedents at an AAPC of 1.9% (0.6% to 3.3%) and 2.4% (95% CI,
1.3% to 4.2%) respectively between 2000 and 2018.
Prevalence and trends of ACP discussion
From 2000 to 2018, adjusted prevalence of ACP discussion increased from 55.1% to
66.8%, an average annual percentage change (AAPC) of 1.5% (95% CI, 0.9% to 2.2%). In
addition, from 2000 to 2006 the increase was higher, an average percentage change (APC) of
3.7% (95% CI, 2.1% to 9.2%), and then increased at a much lower APC rate of 0.4% (95% CI, -
0.9% to 0.9%) from 2006 to 2018.
Among non-Hispanic White decedents, adjusted prevalence of ACP discussion increased
from 2000 to 2006, an APC of 3.1% (95% CI, 1.4% to 12.2%), and dropped to an APC of 0.3%
16
(-4.6% to 0.9%) between 2006 and 2018. The AAPC increase from 2000 to 2018 was 1.2% (95%
CI, 0.1% to 2.2%).
Among non-Hispanic Black decedents, the prevalence increased from 2000 to 2010 at an
APC of 4.5% (95% CI, 0.1% to 70%), then decreased at an APC of 2.5% (95% CI, -49% to
3.1%) between 2010 and 2018.
Hispanic decedents had a consistent increase in prevalence of ACP discussion, an APC of
3.0% (95% CI, 1.3% to 5.7%) across the period.
Prevalence and trends of LW Completion
From 2000 to 2018, adjusted prevalence of LW completion increased from 36% to
52.6%, an AAPC of 1.5% (95% CI, 0.7% to 2.2%). The increase was much higher between 2000
and 2004, an APC of 6.1% (95% CI, 2.1% to 14.7%) and then plateaued to an APC of 0.2%
(95% CI, -1.3% to 0.8%) from 2004 to 2018.
Among non-Hispanic White decedents, adjusted prevalence of LW completion increased
from 2000 to 2005, an ACP of 4.3% (95% CI, 0.8% to 24.9%), and stagnated at an APC of 0
(95% CI, -16.4% to 1%) between 2005 and 2018.
Among non-Hispanic Black decedents, the prevalence increased from 2000 to 2009, an
APC of 9.7% (95% CI, 3.9% to 97.1%), and decreased to an APC of 4% (95% CI, -32.8% to
0.9%) between 2009 and 2018.
The adjusted prevalence of LW completion among Hispanic decedents increased from
2000 to 2013 at an APC of 8.5% (95% CI, 4.5% to 25.8%) and then declined significantly at an
APC of -16.2% (95% CI, -49.9% to -2.3%) between 2013 and 2018. The change across the entire
study period (2000 to 2018) was not significant.
17
Prevalence and trends of HCP Assignment
From 2000 to 2018, adjusted prevalence of having an HCP assignment increased from
47.4% to 66.3% at an AAPC of 1.9% (95% CI, 1.2% to 2.7%). The rate increased by 5.4%
AAPC (95% CI, 2.6% to 13.3%) from 2000 to 2005, and then experienced a lower rate of
increase at an APC of 0.5% (95% CI, -0.8% to 1.1%) between 2005 and 2018.
Among non-Hispanic White decedents, adjusted prevalence of HCP assignment had
higher increase (5.1%, 95% CI, 1.7% to 17.2%) between 2000 and 2005, and plateaued (0.3%,
95% CI, -3.9% to 1%) thereafter. There was no significant AAPC increase from 2000 to 2018.
Among non-Hispanic Black decedents, the prevalence increased from 2000 to 2008 at an
APC of 6.9% (95% CI, 2.7% to 57.5%) and stagnated between 2008 and 2018.
The adjusted prevalence of HCP assignment among Hispanic decedents consistently
increased from 2000 to 2018 at an AAPC of 4.1% (95% CI, 2.1% to 7.7%).
Discussion
This work demonstrated that racial minority groups had not only lower rates of ACP/AD
completion, but also different rates of change over time. Based on the nationally representative
HRS data, the trend analysis found that the overall adjusted prevalence of ACP/LW/HCP
completion increased slowly from 2000 to 2018 after controlling for sociodemographic status
and health conditions. These results are somewhat consistent with a previous study that found
that gains in ACP among older adults’ population who received intensive care in end of life.
62
However, our results indicate the overall rates of each ACP/AD component had a higher rate of
increase from 2000 to about 2004-2006 but dropped off thereafter.
18
Trends differed by racial group. Non-Hispanic White decedents had far higher increases
in ACP/AD completion from 2000 to approximately 2005-2006, while non-Hispanic Black
decedents had a higher increase in later years, around 2008-2010. Non-Hispanic Black decedents
also experienced a significant decline in each component of ACP/AD completion thereafter,
which led to an insignificant change in prevalence of each ACP/AD component across the entire
study period of 2000 to 2018. Hispanic decedents experienced a steady increase in prevalence of
ACP discussion and HCP assignment from 2000 to 2018. Although they had a fast increase in
LW completion from 2000 to 2013, the prevalence decreased dramatically from 2013 to 2018.
Our study shed light on the varying trends in ACP-related behaviors across years and
demonstrates that the trends in end-of-life care planning are non-linear and differ across racial
groups, which may help to explain the mixing findings from previous studies about the growth in
ACP/AD completion over time. Although the overall trends of ACP/AD completion seemed to
increase across the past 19 years, the increases were largely experienced in the first 5 to 10 years
and have stagnated for many years since. These results in part support the recent argument that
we did not get enough benefit after decades of efforts.
1,68
The stagnation of ACP/AD completion
over the past decade may suggest a potential ceiling effect has been reached for non-Hispanic
White. Individuals from those groups who have not completed ACP/AD may be relatively
unresponsive to clinical and research efforts to increase end-of-life planning. As Dr. Periyakoil
used earthquake preparation as analogy, some Californians prefer not doing anything and accept
the consequence even though a variety of tool kits are available.
68
Similarly, non-Hispanic Black
decedents’ saw stagnation in HCP assignment and even experienced a dramatic decline in ACP
discussion and LW completion, suggesting it could be challenging to push the prevalence higher.
Potential barriers for this group include disparities in healthcare access and their distrust of
19
healthcare system due to the history of unethical conduct in medical treatment and research such
as Tuskegee study.
69
Based on this analysis, it seems unlikely that the prevalence among non-
Hispanic Black decedents will increase meaningfully in the short term.
In contrast, trends in the uptake of ACP/AD among Hispanics increased faster than non-
Hispanic Whites and Blacks. Hispanic decedents’ steady increase in prevalence of ACP
discussion and HCP assignment but dramatical decline in LW completion may reflect a family
decision-making preference demonstrated in previous studies. For example, Kwak and Haley
found Hispanics are more likely to have family-centered end-of-life decision than other racial
groups.
69
A qualitative study also found most Hispanic participants preferred family-oriented and
limited autonomy for end-of-life decision making.
70
Therefore, with the efforts of education
targeting Hispanic population, their prevalence of ACP discussion and HCP assignment are
likely to keep increasing if any of patients’ trusted family members adopt the concept of end-of-
life planning.
Beginning in 2016, CMS implemented separate payments to healthcare practitioners for
providing voluntary advance care planning services. The goal is to get healthcare providers paid
appropriately for the ACP consultation to benefit patients’ health outcomes in a long term. A
recent study found that the number of Medicare beneficiaries’ uptake of ACP grew steadily, but
the prevalence remained low based on the Medicare outpatient claims data from 2016 to 2019.
16
Our study also did not show substantial increases in ACP/AD after 2016, possibly due to the
short period of time since the reimbursement took effect and slow adoption by healthcare
providers. In addition, our sample included younger decedents who were not covered by
Medicare. However, further research is needed to track trends in the uptake of end-of-life care
20
planning after the implementation of CPT codes for ACP conversation, especially among under-
represented minorities.
Conclusion
In summary, this chapter found non-linear trends of ACP/AD completion over time and
rates of change vary among different racial groups. It is important to continuously monitor the
trends to evaluate the potential gains in ACP activities versus the resources expended to increase
rates of ACP activities. This information can also inform future analysis of cost effectiveness in
the new ACP billing codes.
21
Table 3.1 Characteristics of the Study Population
Unweighted sample size (%)
(N=11,127)
Weighted % (95%CI)
(N=30,834,192)
Gender
Female 5,998 (53.9) 52.7 (51.6-53.9)
Male 5,129 (46.1) 47.3 (46.1-48.4)
Age
Less than 65 1,133 (10.2) 13.7 (12.4-15.2)
65-84 5,745 (51.6) 48.1 (46.5-49.7)
85 and over 4,249 (38.2) 38.2 (36.5-39.9)
Race and ethnicity
Non-Hispanic White 8,432 (75.8) 82.9 (80.6-85.0)
Non-Hispanic Black 1,848 (16.6) 11.8 (10.2-13.7)
Hispanic 847 (7.6) 5.3 (3.9-7.1)
Education
Lower than high school 3,865 (34.7) 31.1 (29.7-32.6)
GED 499 (4.5) 4.6 (4.1-5.3)
High school graduate 3,318 (29.8) 30.9 (29.8-32.0)
Some college 1,981 (17.8) 19.1 (18.0-20.1)
College and above 1,462 (13.1) 14.3 (13.2-15.5)
Medicare
Yes 9,828 (90.4) 87.8 (86.7-88.9)
No 1,048 (9.6) 12.2 (11.1-13.3)
Medicaid
Yes 2,894 (27.5) 26.3 (24.9-27.8)
No 7,624 (72.5) 73.7 (72.2-75.1)
Cancer
Yes 3,994 (36.4) 37.3 (36.0-38.7)
No 6,993 (63.7) 62.7 (61.3-64.0)
Heart disease
Yes 5,954 (54.3) 54.3 (53.4-55.3)
No 5020 (45.7) 45.7 (44.8-46.6)
Lung disease
Yes 2,807 (25.6) 26.2 (24.8-27.5)
No 8,148 (74.4) 73.9 (72.5-75.2)
Memory problem
Yes 5,496 (50.0) 49.8 (48.5-51.1)
No 5,502 (50.0) 50.2 (48.9-51.5)
Stroke
Yes 3,065 (27.9) 27.1 (25.9-28.3)
No 7,920 (72.1) 72.9 (71.7-74.1)
Year of death
2000 352 (3.2) 3.1 (2.7-3.6)
2001 677 (6.1) 5.8 (5.2-6.4)
2002 601 (5.4) 5.3 (4.8-5.9)
22
2003 644 (5.8) 5.6 (5.1-6.1)
2004 649 (5.8) 5.5 (5.0-6.1)
2005 642 (5.8) 5.7 (5.2-6.2)
2006 623 (5.6) 5.7 (5.2-6.2)
2007 621 (5.6) 5.7 (5.2-6.1)
2008 637 (5.7) 5.7 (5.1-6.4)
2009 643 (5.8) 5.8 (5.2-6.5)
2010 620 (5.6) 5.5 (5.1-6.0)
2011 622 (5.6) 5.7 (5.2-6.2)
2012 619 (5.6) 5.6 (5.0-6.2)
2013 668 (6.0) 6.1 (5.5-6.8)
2014 589 (5.3) 6.1 (5.5-6.8)
2015 544 (4.9) 4.9 (4.4-5.5)
2016 587 (5.3) 5.3 (4.9-5.9)
2017 454 (4.1) 4.2 (3.9-4.7)
2018 335 (3.0) 3.3 (2.9-3.9)
ACP 6,211 (56.4) 58.4 (56.7-60.0)
Living will 4,947 (45.2) 47.2 (45.6-48.8)
Proxy 6,400 (59.2) 61.3 (59.4-63.1)
At least one ACP/AD 8,581 (77.1) 78.9 (77.7-80.0)
*Variables with missing data: living will 192 (1.7%), healthcare proxy 324 (2.9%), ACP
discussion 121 (1.1%), at least one of ACP component 212 (1.9%), Medicare coverage 251
(2.3%), Medicaid coverage 609 (5.5%), cancer 140 (1.3%), heart condition 153 (1.4%), lung
disease 172 (1.6%), memory problem 129 (1.2%) and stroke 142 (1.3%).
23
Table 3.2 Unadjusted and Adjusted Prevalence of ACP/AD Completion, 2000-2018
Unadjusted Prevalence, % Adjusted Prevalence (95% CI), %
ACP Living Will Proxy At Least 1 ACP/AD ACP Living Will Proxy At Least 1 ACP/AD
Gender
Female 60.1 50.1 65.2 81.3 62.1*
(60.3-64.0)
51.1*
(49.4-52.8)
64.3*
(62.7-65.9)
82.1*
(80.8-83.3)
Male (reference) 56.4 44.0 56.9 76.2 57.4
(55.4-59.4)
45.7
(43.7-47.6)
59.9
(57.4-62.4)
78.8
(77.3-80.2)
Age
Less than 65 (reference) 53.5 26.8 41.4 65.7 62.2
(57.5-66.9)
33.6
(28.7-38.5)
47.7
(42.5-53.0)
74.3
(70.4-78.2)
65-84 58.8 45.2 56.3 77.2 59.9
(58.0-61.7)
46.5*
(44.8-48.2)
58.1*
(56.1-60.0)
78.9*
(77.7-80.1)
85 and over 59.5 57.1 74.5 85.7 59.1
(57.2-61.0)
55.7*
(54.0-57.4)
72.5*
(70.2-74.9)
85.1*
(83.6-86.5)
Race and ethnicity
Non-Hispanic White
(reference)
61.8 53.1 66.3 83.4 62.5
(60.7-64.3)
52.6
(51.1-54.1)
65.7
(63.9-67.5)
83.9
(82.8-85.0)
Non-Hispanic Black 40.1 16.6 37.0 56.5 45.0
(42.1-47.9)
23.3*
(20.2-26.4)
44.1
(40.8-47.4)
64.4*
(61.2-67.5)
Hispanic 45.7 22.3 36.9 59.1 50.6
(44.9-56.3)
29.1*
(24.4-33.9)
44.1
(37.0-51.2)
66.7*
(61.6-71.9)
Education
Lower than high school
(reference)
51.5 36.5 52.4 71.9 56.1
(53.3-58.9)
40.7
(38.1-43.3)
55.2
(52.7-57.7)
76.2
(74.3-78.1)
GED 60.1 44.4 55.1 74.6 62.3*
(56.8-67.9)
48.5*
(42.6-54.3)
58.5
(53.1-64.0)
78.8
(74.3-83.3)
High school graduate 60.0 48.0 62.5 80.5 60.1*
(57.8-62.4)
47.8*
(45.9-49.7)
62.3*
(60.4-64.3)
80.8*
(79.2-82.3)
Some college 62.9 52.4 67.0 83.5 62.7*
(60.0-65.3)
52.2*
(49.5-54.9)
66.6*
(63.8-69.4)
83.4*
(81.5-85.2)
24
College and
above
63.1 63.0 72.5 85.9 63.5*
(60.2-66.7)
61.7*
(58.7-64.7)
72.4*
(69.2-75.6)
86.9*
(84.9-89.0)
Medicare 59.7 49.9 63.6 80.9 60.6*
(58.9-62.4)
49.3*
(47.8-50.7)
62.6
(60.9-64.3)
81.0*
(79.8-82.2)
Medicaid 52.3 36.9 55.4 73.8 56.3*
(54.1-58.6)
43.7*
(40.9-46.5)
59.9*
(57.0-62.7)
78.9*
(77.1-80.7)
Cancer 64.1 51.4 63.7 83.4 64.7*
(62.7-66.8)
52.5*
(50.4-54.5)
65.0*
(63.0-67.0)
84.5*
(83.2-85.9)
Heart disease 61.8 50.2 63.5 81.5 62.9*
(51.2-64.7)
50.1*
(48.5-51.7)
63.2*
(61.4-65.0)
82.2*
(81.0-83.3)
Lung disease 62.3 49.2 60.0 80.9 62.4*
(59.7-65.1)
51.2*
(48.9-53.5)
62.3
(59.5-65.2)
82.2*
(80.4-84.0)
Memory problem 59.8 51.3 68.9 83.5 60.6
(58.6-62.6)
51.1*
(49.5-52.7)
67.7*
(65.7-69.7)
83.5*
(82.4-84.7)
Stroke 59.7 49.2 65.1 80.9 61.3
(59.1-63.6)
50.0
(47.5-52.5)
64.6*
(62.4-66.9)
81.8*
(80.4-83.3)
Year of death
2000 (reference) 49.7 33.2 43.6 64.4 51.1
(45.1-57.0)
36.0
(31.4-40.6)
47.4
(41.8-53.0)
68.6
(64.5-72.7)
2001 46.7 41.3 50.7 70.9 49.1
(47.8-53.4)
44.0*
(40.1-48.0)
53.9*
(48.6-59.2)
74.5*
(71.0-77.9)
2002 53.0 43.2 51.5 74.5 53.3
(49.2-57.4)
43.5*
(39.5-47.4)
52.3*
(48.7-56.0)
74.6*
(71.0-78.3)
2003 55.4 44.2 56.7 76.1 55.7
(50.4-61.1)
45.5*
(41.6-49.3)
57.5*
(51.8-63.2)
76.9*
(73.4-80.3)
2004 57.6 48.2 58.9 78.5 57.4*
(53.3-61.5)
48.7*
(43.7-53.8)
59.2*
(53.9-64.5)
78.7*
(75.3-82.1)
2005 56.8 48.5 61.8 77.7 58.3
(54.0-62.7)
49.4*
(45.1-53.7)
63.3*
(58.9-67.7)
80.0*
(76.9-83.0)
2006 60.2 50.1 60.3 78.9 61.1*
(57.2-65.1)
49.2*
(44.9-53.5)
58.9*
(53.6-64.3)
78.9*
(75.1-82.7)
2007 58.3 45.6 61.2 79.6 59.5
(54.7-64.3)
47.2*
(43.3-51.1)
61.9*
(57.4-66.5)
81.0*
(77.0-85.1)
25
2008 60.9 51.0 65.4 82.4 63.1*
(58.6-67.6)
51.4*
(47.3-55.5)
66.1*
(61.9-70.4)
84.4*
(80.9-88.0)
2009 56.5 47.1 66.0 78.2 60.8*
(55.4-66.1)
50.3*
(44.8-55.8)
68.6
(64.6-72.7)
81.8*
(78.3-85.3)
2010 64.8 55.2 68.0 83.6 65.9*
(61.2-70.6)
55.8*
(51.2-60.3)
66.9*
(61.4-72.4)
84.3*
(80.3-88.3)
2011 62.4 47.9 61.6 81.6 62.2*
(57.1-67.3)
49.1*
(44.8-53.5)
61.9*
(56.9-66.8)
81.2*
(77.3-85.2)
2012 57.9 43.6 59.5 80.6 61.2*
(56.4-65.9)
45.7*
(39.6-51.8)
60.6*
(55.6-65.5)
82.3*
(79.0-85.6)
2013 60.3 48.1 64.9 78.8 60.5*
(55.7-65.4)
49.6*
(44.9-54.3)
66.4*
(62.4-70.4)
79.9*
(76.2-83.5)
2014 59.8 49.4 64.5 81.8 62.5*
(57.4-67.6)
50.9*
(45.6-56.2)
64.6*
(59.7-69.5)
83.7*
(79.5-88.0)
2015 63.4 50.6 68.7 84.7 64.6*
(58.9-70.3)
51.6*
(47.0-56.2)
68.9*
(64.0-73.8)
86.6*
(82.6-90.7)
2016 60.6 44.9 64.0 82.0 64.5*
(57.3-71.6)
47.1*
(42.3-51.9)
65.0*
(60.3-69.6)
85.0*
(81.2-88.7)
2017 61.4 52.1 69.3 82.0 62.0*
(56.6-67.4)
51.5*
(44.8-58.2)
69.7*
(63.2-76.3)
82.4*
(77.3-87.5)
2018 62.4 48.6 63.0 77.3 66.8*
(58.0-75.6)
52.6*
(45.7-59.4)
66.3*
(58.8-73.9)
82.6*
(75.3-89.8)
* P<0.05
26
Figure 3.1 Trends in Adjusted Prevalence of ACP/AD Completion
A.
B.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
ACP Discussion
Non-Hispanic White Non-Hispanic Black Hispanic
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Living Will Completion
Non-Hispanic White Non-Hispanic Black Hispanic
27
C.
D.
*Moving average smoothing were applied to get clearer trends
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Healthcare Proxy Assignment
Non-Hispanic White Non-Hispanic Black Hispanic
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
At Lease One ACP/AD Component
Non-Hispanic White Non-Hispanic Black Hispanic
28
Table 3.3 Change in Trends of ACP/AD Completion by Race
Trend 1 Trend 2 AAPC, %
(95% CI)
(2000-2018)
APC, %
(95% CI)
Years APC, %
(95% CI)
Years
ACP
Overall 3.7* (2.1 to 9.2) 2000-2006 0.4 (-0.9 to 0.9) 2006-2018 1.5* (0.9 to 2.2)
Non-Hispanic White 3.1* (1.4 to 12.2) 2000-2006 0.3 (-4.6 to 0.9) 2006-2018 1.2* (0.1 to 2.2)
Non-Hispanic Black 4.5* (0.1 to 70) 2000-2010 -2.5 (-49 to 3.1) 2010-2018 1.3 (-5.3 to 6.8)
Hispanic 3.0* (1.3 to 5.7) 2000-2018 NA NA 3.0* (1.3 to 5.7)
Living will
Overall 6.1* (2.1 to 14.7) 2000-2004 0.2 (-1.3 to 0.8) 2004-2018 1.5* (0.7 to 2.2)
Non-Hispanic White 4.3* (0.8 to 24.9) 2000-2005 0 (-16.4 to 1) 2005-2018 1.2 (-1.1 to 2.8)
Non-Hispanic Black 9.7* (3.9 to 97.1) 2000-2009 -4 (-32.8 to 0.9) 2009-2018 2.6 (-2.9 to 11.3)
Hispanic 8.5* (4.5 to 25.8) 2000-2013 -16.2* (-49.9 to -2.3) 2013-2018 1 (-4.7 to 6.8)
Proxy
Overall 5.4* (2.6 to 13.3) 2000-2005 0.5 (-0.8 to 1.1) 2005-2018 1.9* (1.2 to 2.7)
Non-Hispanic White 5.1* (1.7 to 17.2) 2000-2005 0.3 (-3.9 to 1) 2005-2018 1.6* (0.4 to 3)
Non-Hispanic Black 6.9* (2.7 to 57.5) 2000-2008 0 (-21 to 2.6) 2008-2018 3 (-0.2 to 7.3)
Hispanic 4.1* (2.1 to 7.7) 2000-2018 NA NA 4.1* (2.1 to 7.7)
At least one
component
Overall 2.6* (1.3 to 6.4) 2000-2005 0.4 (-0.5 to 0.7) 2005-2018 1.0* (0.6 to 1.4)
Non-Hispanic White 2.3* (0.7 to 8.6) 2000-2005 0.2 (-5.0 to 0.6) 2005-2018 0.8 (-0.2 to 1.4)
Non-Hispanic Black 1.9* (0.6 to 3.3) 2000-2018 NA NA 1.9* (0.6 to 3.3)
Hispanic 2.4* (1.3 to 4.2) 2000-2018 NA NA 2.4* (1.3 to 4.2)
29
4. Chapter 3: The Association between Health Literacy and
ACP/AD Completion
Introduction
According to the Centers for Disease Control and Prevention, health literacy is defined as
“the degree to which individuals have the ability to find, understand, and use information and
services to inform health related decisions and actions for themselves and others”. The
relationship between health literacy and health-related outcomes has been well studied. For
example, studies have found limited health literacy is associated with more healthcare utilization
such as emergency room visits and hospitalization, and lower use of preventive healthcare
services.
71,72
Health literacy also has been found to be associated with poor access to regular
healthcare, such as primary care and medication, and higher rates of mortality among older
adults.
73,74
Researchers have warned that limited health literacy could further lead to a more
injustice healthcare system if the needs of patients with limited health literacy are not met.
75
Literature is scarce in the context of health literacy and end-of-life care planning.
Volandes and colleagues suggested higher health literacy, not race, was associated with higher
likelihood of choosing comfort care at end of life.
76
Other studies found limited health literacy
was a barrier in patient-physician communication and mediated the effect of race on end-of-life
care preferences.
76,77
In addition, Nouri and colleagues found that limited health literacy was
strongly associated with poor knowledge of advance care planning.
21
Although Waite colleagues
found health literacy was independently associated with advance directive completion while
controlling for race, their sample was drawn from local healthcare settings and the majority were
White and Black participants.
23
30
To determine if previous findings related to health literacy can be generalized to the U.S.
population, while taking potential confounders into account, the goal of this study was to explore
the association between health literacy and end-of-life care planning by using a nationally
representative sample of older Americans.
Method
Data
Between 1992 and 2020 about 43,558 participants were interviewed as part of the HRS.
However, the health literacy screening question (confidence in filling out medical forms) was
asked among just a subpopulation of study participants across specific years. To obtain a robust
sample, all HRS datasets with the health literacy screening question were included. These
datasets include: 2009 Internet Survey, 2010 HRS Core, 2011 Health Care Mail Survey, 2013
Internet Survey, 2013 Health Care and Nutrition Study, and 2018 HRS Core. Of the 43,558
participants, about 16,156 were screened for health literacy and were included in this study.
Although this sample was purposefully selected based on the availability of health literacy and
ACP/AD completion variables, potential selection bias excluding participants without health
literacy measurement should not be an issue because those subgroups of participants were
randomly selected. Then, participants (n=614) reporting their race/ethnicity as other than Non-
Hispanic White, Non-Hispanic Black or Hispanic, and those who did not have any end-of-life
planning records (n=5,457) also were excluded. About 35% of the total sample was missing
ACP/AD information and dropped from this study. Lastly, cases with missing covariates (2.4%)
also were removed from the analytic dataset, leaving a total sample of 9848 (Figure 3.1).
Figure 4.1 Study Sample Selection Workflow
31
Measures
The primary dependent variable for this study was ACP/AD completion status. This
variable was captured by participants reporting to have an ACP/AD in any survey wave from
Core interviews (2012-2018) or Exit interviews (2002-2018). Specifically, proxies were asked
32
whether the deceased participants “ever discussed medical care with anyone” (ACP), if they had
“assign a durable power of attorney” (healthcare proxy) and/or had a “written living will for
healthcare” (living will).
The primary independent variable of interest was health literacy level. Participants were
asked how confident they were in filling out medical forms by themselves. A Five-item response
set included “not at all confident”, “a little confident”, “somewhat confident”, “quite confident”,
“extremely confident”. Previous studies have validated this single-item measurement as a
reliable screening tool for health literacy,
78
with “somewhat confident” or less suggested as the
optimal cutoff to identify patients with limited health literacy.
78,79
This variable was
dichotomized into “1” indicating adequate health literacy versus “0” indicating limited health
literacy.
Selection of Covariates were guided by the Andersen’s Behavioral Model of Health
Service Utilization and factors identified in previous literature, socioeconomic characteristics
included year of birth, gender, race/ethnicity, language (English or Spanish), number of years of
education, religious preference and religious service attendance. In addition, poverty status,
Medicare coverage and health conditions such as cancer, heart diseases, lung diseases, stroke and
dementia were extracted from the participants’ most recent core interview.
Analysis
To compare characteristics between participants with adequate and limited health literacy
I conducted a two-sample t test for continuous variables and χ
2
test for categorical variables.
Multivariate logistic regression models were conducted to examine if health literacy was
associated with ACP/AD completion after controlling for the covariates. Potential moderation
33
effects between health literacy and race were tested by applying the interaction term but were not
included in the final models due to insignificant results. Model diagnosis was performed to avoid
potential multicollinearity among independent variables. Sample weights (taken from when
participants first entered the study) were applied to get unbiased estimates of population
parameters. All analyses were conducted using STATA 16 (College Station, TX: StataCorp
LLC).
Results
This sample included 9,848 participants who were born between 1911 and 1968. About
75% were non-Hispanic White, 15% were non-Hispanic Black and the remaining 10% were
Hispanic. There were more women (57%) than men (43%). The average education level was 13
± 3 years. The majority of participants were Protestant (63%) and Catholic (27%), with 2%
Jewish, 7% none or no preference and 1% of other religious beliefs. About 10% of the
participants were living in poverty, and most (93%) had Medicare coverage. About one-third
(35%) of participants had a heart condition and 24% reported having cancer. In addition, there
was a relatively small portion of participants who had lung disease (14%), stroke (11%), or
dementia (7%).
In this sample, about 23% (n=2,283) of participants had limited health literacy.
Participants with adequate health literacy were more likely to be younger. Female participants
had slightly higher rates of having adequate literacy as compared to males (81% vs 79%
respectively), although this difference was not statistically significant. In addition, fewer non-
Hispanic Black (67%) and Hispanic participants (64%) had adequate health literacy as compared
to non-Hispanic Whites (82%). Participants with adequate health literacy had about 2.5 more
34
years of education than those with limited health literacy. Jewish participants had the highest rate
(92%) of adequate health literacy as compared to other religious backgrounds (Table 3.1).
The unadjusted regression model showed that adequate health literacy was significantly
associated with higher likelihood of engaging in advance care planning, completing a living will,
or assigning a healthcare proxy (Table 3.2, models 1-3). After adding sociodemographic factors
in models 4-6, adequate health literacy remained significance in end-of-life planning outcomes.
In addition, birth year, race and education were also independently associated with all end-of-life
planning. (Table 3.2, model 4-6). In the fully adjusted models adding religion, poverty, Medicare
coverage and health conditions, participants with adequate health literacy were 1.4 times more
likely to engage in advance care planning, 1.5 times more likely to have a living will, and 1.3
times more likely to assign a healthcare proxy than those who had limited health literacy (Table
3.2, model 7-9). Multicollinearity among variables using variance inflation factor (VIF) was not
detected in the fully adjusted models (Mean VIF = 1.42).
Discussion
To my best knowledge, there have been no previous investigations using US nationally
representative data that explore the association between health literacy and end-of-life care
planning. The results in this study were consistent with previous studies that found that about one
quarter of the participants had limited health literacy and minority populations had much higher
rates of limited health literacy. Non-Hispanic Blacks and Hispanics also were significantly less
likely to have end-of-life care planning. In addition, there was a strong association between
health literacy and ACP/AD completion.
35
To disentangle the association from other potential explanatory factors, we controlled for
sociodemographic characters such as religious participation and health conditions. Health
literacy remained significantly association with end-of-life care planning in these fully adjusted
models. Thus, health literacy was independently related to end-of-life care planning activities.
Previous studies found that racial differences were no longer significant after controlling
for health literacy, suggesting health literacy mediated race in end-of-life care preferences.
76
Our
results did not support this finding in end-of-life care planning, as racial disparities remained
significant in the fully adjusted models. In fact, our study found that the effect of health literacy
on end-of-life care planning was the same across different racial groups in the fully adjusted
models.
In addition, the results showed health literacy had a larger magnitude of effect size than
education attainment, and religious factors were not significantly associated with any end-of-life
care planning in the fully adjusted models. These findings have important implications for both
policy and practice. Health literacy is relatively easy to detect through the use of this one-item
screening question, facilitating identification of individuals that can be targeted for additional
education efforts. Thus, healthcare organization and practitioners should consider applying a
health literacy screening question in clinical settings before initiating advance care planning
conversations to ensure their patients understand important concepts and processes of advance
care planning conversations and directives. For those with limited health literacy, low literacy-
tailored language materials or other educational tools such as videos may help improve the
uptake of end-of-life care planning activities. For example, Sudore and her colleagues
redesigned the AD form to be written at a 5
th
grade reading level to accommodate the needs of
patients with low health literacy.
80
They also established an easy-to-use online tool offering step-
36
by-step guidance and explanation of ACP as well as help in completing the AD form.
81
Our
findings underscore the need for such approaches along with screening to more effectively target
patients who have limited health literacy.
Lastly, racial disparities remained significant in ACP/AD completion after controlling for
religion and health literacy suggesting healthcare providers need to use culturally tailored
approaches to introduce ACP/AD. Although it is important to maintain awareness of racial
disparities, healthcare providers also should avoid stereotyping patients decision-making
differences based on their race/ethnicity or make assumptions about their level of interest in
ACP/AD.
82,83
In addition, it is also important to improve healthcare providers’ communication
skills and increase their cultural competence and sensitivity.
84-87
Limitation
Our instrument measuring health literacy consisted of a one-item question. Although
HRS asked more detailed and objective questions focused on health literacy using segments of
validated instruments such as the Rapid Estimate of Adult Literacy in Medicine (REALM) and
Test of Functional Health Literacy in Adults (TOFHLA), those measures only occurred in the
2009 Internet Survey and among a sub-group of 2010 Core participants. In addition, HRS used
only an unvalidated subset of REALM and TOFHLA questions instead of using complete
instruments. Moreover, as stated previously, the one-item health literacy question has been
validated as an effective tool.
Lastly, other potential unobserved factors such as subjective value of autonomy, family
relationships, and trust of the healthcare system may exist and confound the results.
58,88-90
However, HRS did not capture these variables.
37
Conclusion
In summary, health literacy is a significant factor associated with end-of-life planning. To
ensure equity and quality when promoting ACP/AD, it is important to identify those who have
limited health literacy and use linguistical and cultural tailored approaches to explain the purpose
and goals of ACP/AD.
38
Table 4.1 Sample Sociodemographic Description Stratified by Health Literacy
Sample
Distribution
N=9,848
Limited
Health Literacy
N=2,283
Adequate
Health Literacy
N=7,565
P-value
Birth Year (Mean) 1942 1940 1943 <0.001
Gender 0.051
Male 49% 21% 79%
Female 51% 19% 81%
Race <0.001
Non-Hispanic White 82% 17% 83%
Non-Hispanic Black 10% 32% 68%
Hispanic 8% 36% 64%
Language <0.001
English 96% 19% 81%
Spanish 4% 46% 54%
Education 13.3 11.3 13.8 <0.001
Religion 0.034
Protestant 60% 21% 79%
Catholic 27% 22% 78%
Jewish 2% 9% 91%
None/no preference 9% 19% 81%
Other 1% 21% 79%
39
Table 4.2 Logistic Regression Analyses of End-of-Life Planning
(1) (2) (3) (4) (5) (6) (7) (8) (9)
ACP Living Will Proxy ACP Living Will Proxy ACP Living Will Proxy
Adequate health literacy 1.723*** 1.627*** 1.304*** 1.354*** 1.449*** 1.159 1.39*** 1.519*** 1.251**
(.111) (.111) (.083) (.111) (.111) (.101) (.112) (.114) (.105)
Birth year .944*** .917*** .909*** .953*** .927*** .919***
(.004) (.003) (.004) (.004) (.004) (.004)
Female 1.578*** 1.204*** 1.111 1.634*** 1.235*** 1.124**
(.11) (.066) (.064) (.115) (.07) (.065)
Race (Non-Hispanic White as reference)
Non-Hispanic Black .468*** .423*** .567*** .477*** .417*** .555***
(.054) (.04) (.045) (.052) (.041) (.049)
Hispanic .644** .418*** .502*** .709 .441*** .52***
(.124) (.055) (.062) (.147) (.064) (.072)
Language (Spanish) .581*** .64** .902 .608*** .68** .929
(.094) (.117) (.122) (.104) (.125) (.132)
Education 1.111*** 1.133*** 1.137*** 1.112*** 1.135*** 1.141***
(.013) (.014) (.017) (.013) (.014) (.017)
Religion (Protestant as reference)
Catholic .951 .974 .999
(.098) (.087) (.102)
Jewish 1.212 1.149 1.189
(.326) (.361) (.339)
None/no preference .934 .924 .946
(.123) (.101) (.109)
Other .828 .99 .77
(.223) (.296) (.2)
Religious attendance (More than once a week as reference)
Once a week .924 1.001 .993
(.12) (.133) (.132)
Two or three times a month 1.084 1.124 1.03
(.199) (.195) (.174)
One or more times a year 1.003 .934 .926
(.158) (.133) (.118)
Not at all .919 .836 .844
(.124) (.118) (.101)
Poverty .813 .747*** .912
(.097) (.076) (.121)
Medicare 1.521*** 1.895*** 1.52***
40
Heart disease 1.354*** 1.295*** 1.214***
(.1) (.093) (.07)
Cancer 1.455*** 1.397*** 1.421***
(.126) (.113) (.108)
Lung disease 1.212** 1.233** 1.291**
(.113) (.109) (.147)
Stoke 1 1.317** 1.342**
(.114) (.154) (.159)
Dementia 1.509*** 1.603*** 2.246***
(.194) (.227) (.328)
Standard errors are in parentheses
*** p<.01, ** p<.05,
41
5. Chapter 4: Advance Directives Completion and Hospital Out-
of-Pocket Expenditures
Introduction
Health care costs remain high at the end of life. However, the high end-of-life healthcare
costs may not correlate with improved quality of life. Additionally, aggressive treatment aimed
at prolonging life is often not consistent with patients wishes.
91-93
Previous studies found that
end-of-life conversations between patients and healthcare providers or AD completion was
associated with lower healthcare costs and better end-of-life outcomes such as quality and
satisfaction.
50,51,94
Ideally, healthcare costs can be reduced in hospital settings following by the
improvement on healthcare concordance and reduced aggressive treatment or hospital stays that
related to ACP/AD completion. However, it is worth noting that the findings on economic effects
are mixed, with some studies finding little or no significant differences in healthcare costs
associated with ACP/AD completion.
26,95,96
Moreover, some believe examining costs of
healthcare at end of life is a controversial goal care for ACP/AD. Therefore, there is a potential
ethical conflict between ACP/AD and healthcare cost savings.
97
Some opponents fear the focus
on cost containment as a motivation for ACP/ADs can lead to undertreatment.
98
However, out-of-pocket costs are a concern for most patients, who need more cost-related
information to help them make informed healthcare decisions.
45
For older adults with traditional
Medicare coverage only, they carry a 20% copayment for healthcare utilization and medication,
which can be incrementally high. Therefore, it is important to include out-of-pocket costs along
with prognosis when discussing treatment options.
99
Currently it is not known if there is a relationship between advance directive completion
and hospital out-of-pocket costs. This analysis addresses this gap in the literature and
42
investigates whether advance directive completion is associated with lower hospital out-of-
pocket costs at end of life. We hypothesized that, 1) compared to decedents without ADs, those
who completed ADs will have lower hospital out-of-pocket costs in the end of life. 2)
Completing an AD will have more hospital out-of-pocket reduction among younger decedents.
3) Among those with ADs, timing of completion and documented care preferences will be
associated with hospital out-of-pocket costs.
Methods
Data and sample
I used the Health and Retirement Study (HRS), which is sponsored by the National
Institute on Aging and is conducted by the University of Michigan. HRS isa longitudinal panel
study of United States adults aged 50 years or older and their partners. Participants are
interviewed every two years following enrollment, and among those who die, HRS Exit
Interviews are conducted with a survey proxy approximately two years following the
participant’s death. For this study, I used Harmonized HRS End-of-Life (1992-2014) data for
12,952 decedents’ death related details, health conditions, health care utilization including
hospital out-of-pocket costs and AD completion status.
100
I only included participants from wave
6 through wave 12 Exit Interviews (N=9,228) as earlier waves reported combined hospital and
nursing home out-of-pocket costs, making it impossible to separate these costs. I then merged the
RAND HRS Longitudinal File (1992-2016) for participant sociodemographic characteristics and
health insurance coverage.
101
Hospital out-of-pocket costs
43
During HRS Exit Interviews, proxies were asked to report the amount of hospital out-of-
pocket costs incurred since the previous interview or in the last 2 years prior to death. A value
was assigned to zero if decedents did not use any hospital services. Depending on the year of
death, the cost of hospital out-of-pocket expenses was adjusted to 2014 dollars in the
Harmonized HRS End of Life dataset.
102
Advance directive (AD) completion
Proxies were asked if decedents had written instructions about their preferences for
medical treatment during their final days of life, also known as living wills. Participants
indicating decedents had documented their preferences were further asked the year and month of
AD completion. The time from AD completion to death was also provided in the dataset. In
addition, proxies reporting ADs were in place at time of death were asked whether ADs included
instructions to limit care in certain conditions, withhold treatment (not initiating an intervention),
be kept comfortable (pain free but to forgo extensive intervention to prolong life), or prolong life
(receive all care possible under any circumstances).
25
Among those with an AD, we dichotomized the time period between AD completion and
death to 3 months or less before death or more than 3 months before death. This allowed us to
explore the effect of differences in AD completion timing on hospital out-of-pocket costs. Three
months was selected as the cut-off period in line with previous studies that found different
patterns of healthcare use and care preferences among those completing ADs in the last three
months of life.
52,103
Covariates
44
Decedents’ sociodemographic status included gender and years of education.
Race/ethnicity was grouped as non-Hispanic White, non-Hispanic Black, and Hispanic.
Although the Harmonized End of Life dataset includes decedents’ total estate value, it may not
fully capture respondents’ entire financial situation, therefore, we chose to use household income
and total wealth from the last wave of the core interview prior to death.
Characteristics associated with end of life included age at death, whether death was
expected, the location of death (private home, hospital, nursing home, hospice, other), main
cause of death (cancer, cardiovascular diseases, other), duration of final illness (e.g. less than 1
year: over 1 month but less than 12 months), and number of days between the last core interview
and death.
Finally, we controlled for health care insurance coverage due to its potential impact on
out-of-pocket costs. In this analysis, whether or not the decedents had one of the main types of
insurance were included: Medicare fee-for-service, Medicare Advantage, Medicaid, veteran
benefits, and private and/or employer-based health insurance.
Analysis
To compare health insurance coverage, sociodemographic and death-related
characteristics between decedents with and without an AD, we conducted two-sample t test for
continuous variables, Wilcoxon rank-sum tests for the skewed distribution of out-of-pocket costs
data and χ
2
test for categorical variables.
We found 219 observations with missing data on AD completion as well as 2,359
missing information on covariates in the model. Thus, we conducted a multiple imputation by
45
chained equations approach to impute the missing data. Multiple imputation creates multiple
values using both categorical and count data during the imputation process.
104
Due to the skewedness and large quantity of zeros in our out-of-pocket costs data, we
elected to conduct a two-part model to analyze the data. In this model, hospital out-of-pocket
costs served as the dependent variable and AD completion as the main independent variable of
interest. All covariates including sociodemographic status, end-of-life characteristics, and
insurance coverage were included in the model. In the first stage of this analysis, we used logit
regression to predict the probability of whether the decedent had any hospital out-of-pocket
costs. In the second part of the model, we conducted generalized linear model (GLM) on those
who had at least some out-of-pocket costs (non-zero expenditures). Natural logarithm as link
function and Gamma type distribution were applied in the GLM analysis. Next, the post-
estimation marginal effects were predicted by using both parts of the model including those with
zero expenditures. This same two-part model approach was then applied to examine hospital out-
of-pocket costs by care preference among those who completed ADs. This two-part model has
been widely used in health economics, particularly for health care expenditure data.
105-111
Among
those with an AD, we investigated hospital out-of-pocket costs in relation to the timing of when
the ADs were completed. We used Wilcoxon rank-sum test to examine whether the hospital out-
of-pocket costs for decedents who developed ADs in the three months prior to death were
statistically different from those completing ADs earlier in the disease trajectory. Additionally,
we compared the cost results yielded from our model with 2014 (aligning with the adjusted 2014
out-of-pocket costs for our sample) federal poverty level to explore potential impact among low-
income populations.
46
Lastly, to improve the robustness of this study, we performed sensitivity analysis by
using propensity score weighting combined with regression analysis to address potential
confounders such as health insurance coverage. In this process, we included the same covariates
in the two-part model described previously to predict the propensity score of having completed
an AD. Propensity score weighting can improve comparability between control (without an AD)
and intervention (AD completed) groups and have been used to address potential selection bias
when estimating the average effect.
112
However, we ultimately chose to report main results
generated from the two-part model because non-linear health care expenditures can yield bias
and inefficiency under the propensity-based approach.
113
Additionally, results from the
propensity score weighting were consistent with those achieved via the two-part model.
All analysis was conducted in 2020 using STATA 14 (StataCorp, College Station, TX).
Results
Demographics
Sample characteristics before imputation are shown in Table 4.1. More than half the
decedents were female. Overall, 3,950 (44%) of decedents had completed ADs. Non-Hispanic
White decedents had higher AD completion rates (3,512 out of 6,710, 52%) compared to non-
Hispanic Blacks (259 out of 1,456, 18%) and Hispanics (124 out of 675, 18%). The average
number of days from decedents’ last interview to death was 420 days (range: 0-730, SD=230)
and there was no statistically significant difference in time to death between decedents who
completed ADs and those who did not complete ADs (P=0.54). The average hospital out-of-
pocket cost was $2,114 (SD=$20,507). On average, the top 1% of out-of-pocket costs among AD
completers was $67,093 (SD=$79,916), while the top 1% of out-of-pocket costs for those who
47
did not complete an AD was $169,636 (SD=$175,803). Bivariate analysis revealed that
decedents with completed ADs had significantly lower hospital out-of-pocket costs (mean
difference = -$1,137), lived longer (mean difference = 3.79 years), and had a higher education
level (mean difference = 1.55 years) than those without ADs. In addition, participants with ADs
were less likely to die in a hospital than those without ADs (33% [1,285 out of 3,950) vs 39%
[1,980 out of 5,059]) and more likely to receive hospice services (11% [423 out of 3,950] vs 7%
[360 out of 5,059]) than those without ADs. Differences in hospital out-of-pocket costs with and
without ADs varied by health condition; patients with ADs and diagnosed with cancer had the
greatest magnitude of out-of-pocket spending reduction, dropping from an average of $4,872
(SD=$36,041) without an AD to $1,835 (SD=$6,707) with a completed AD.
Analytic results
The logit model in the first part of the two-part model indicated that decedents with ADs
were more likely to have some hospital out-of-pocket spending compared with those without an
AD (OR=1.20, 95%CI, 1.08 to 1.34, P=0.001) after controlling for sociodemographic status,
health insurance coverage, and death-related characteristics. However, the second part of the
GLM model showed that for those who had at least some expenditures, participants with ADs
had significantly lower out of pocket costs than those without ADs (OR=0.62, 95%CI, 0.49 to
0.80, P<0.001).
The marginal effects predict the hospital out-of-pocket costs based on the combination of
both probability of spending (Logit model) and the amount of spending (GLM) from the two-part
model. Having an AD was associated with lower hospital out-of-pocket costs at $1,639 (95% CI,
$1,306 to $1,972) for those with ADs, compared with $2,312 (95% CI, $1,806 to $2,817) for
48
those that did not have ADs, a statistically significant difference of -$673 (95%CI, -$1,203 to -
$142, P = 0.01) after controlling for death related information, insurance coverage and other
sociodemographic characteristics (Table 4.2). Hospital out-of-pocket costs also declined with
older age at death (-$65, 95%CI, -$94 to -$36, P < 0.001). In addition, the average difference in
hospital out-of-pocket spending between patients with an AD and without an AD dropped from -
$1,646 (95% CI, -$3,028 to -$264, P=0.02) at age 50 to -$442 (95% CI, -$811 to -$73, P=0.02)
at age 90 (Figure 4.1).
Comparison among those completing an AD
Further investigation of hospital out-of-pocket costs by AD completion timing revealed
that decedents who had their AD documented within 3 months (N=473) experienced higher costs
than those who completed their AD more than 3 months prior to death (N=3,025). For
individuals who completed ADs more than 3 months prior to death, the mean hospital out-of-
pocket cost was $1,176 (SD=$5,437), while the top 5% of out-of-pocket expenditures for these
early AD completers was $10,116 (SD=$3,712). In comparison, those who completed ADs
within 3 months of death had an average of $1,854 (SD=$10,232) in out-of-pocket spending,
with the top 5% of this group spending $12,858 (SD=$3,794). The results from the Wilcoxon
rank-sum test revealed that hospital out-of-pocket spending was significantly different between
those who completed ADs closer to death and those that completed ADs earlier in the trajectory
(P=0.001).
In this sample, over 90% of decedents with an AD expressed a desire to limit care (3,506
out of 3,867) or to be kept comfortable (3,560 out of 3,859), 79% (3,000 out of 3,805) indicated
that they wanted to withhold treatment, and just 6% (217 out of 3,887) wanted to prolong life. In
49
addition, choosing to limit care was significantly associated with lower hospital out-of-pocket
costs (-$1,443; 95% CI, -$2,702 to -$185, P = 0.025).
Sensitivity analyses
We conducted sensitivity analyses to assess the analytic and modelling assumptions. We
used propensity score weighting with imputation among our sample to reach good balance
between decedents with and without ADs. Results of subsequent analysis of these propensity
weighted subsamples produced similar results as found in the two-part model (see Appendices)
in that having an AD was significantly associated with lower hospital out-of-pocket costs (-
$1,354; 95%CI, -$2,382 to -$326, P=0.010).
Discussion
We found AD completion was significantly associated with lower hospital out-of-pocket
spending after controlling for socioeconomic status, health conditions, and health care insurance.
One factor that most likely contributed to this finding was that patients with ADs were more
likely to elect to limit care and focus on comfort care, rather than elect aggressive, life
prolonging care that can involve expensive procedures and hospitalizations. Previous studies
have found that the majority of decedents with ADs receive the care they desired,
114
and our
finding suggested preferences for limiting care were associated with lower hospital out-of-pocket
spending. Thus, these lower out-of-pocket costs may be influenced, in part, by the desire for less
aggressive care expressed in ADs.
Previous studies have observed lower end-of-life health care costs among older age
groups.
115,116
For example, a Medicare expenditure study found that the youngest decedents (age
50
65-69) spent twice as much in the last year of life as those in the oldest strata (age 85 and
over).
115
In our study, we found similar trends in that older age was associated with lower
hospital out-of-pocket costs. In addition, compared with older decedents with ADs, younger
patients with ADs had a greater magnitude of out-of-pocket cost reduction (Figure 4.1). This
trend similarly aligned with earlier studies examining age-related differences in AD care
preferences. Hamel and colleagues found that older patients were less likely to desire and receive
aggressive care in the hospital.
117
Thus, ADs were associated with a greater reduction in hospital
out-of-pocket costs among younger patients who otherwise would most likely have received
costly life-prolonging care.
Among those who had ADs, early completers had lower hospital out-of-pocket costs than
those who completed an AD within 3 months prior to death. This finding also is consistent with a
study examining AD completion timing that found a higher prevalence of electing aggressive
care among decedents who completed ADs in the months prior to death.
52
Some have suggested
that late AD completion often involves decision-making “in the moment” of a health care crisis
and may lack thorough consideration, education, and discussion.
118-120
A qualitative study on
health care providers suggest ADs should be initially developed when individuals are healthy,
and then regularly modified as diseases develop and progress.
121
Lastly, our findings have policy implications for physician-patient communication about
costs of care. While some patients may not feel comfortable discussing out-of-pocket cost during
advance care planning conversations, a recent study found that the majority of patients did take
financial concerns into consideration in decision-making and want to have such conversations
with their health care providers to understand expected cost associated with treatment and care
decisions.
122
Thus, there is a growing call for transparency in treatment recommendations and in
51
requirements for out-of-pocket costs to be discussed as side effects to treatment
recommendations due to the negative impact financial burden has on patients and their family
members.
123
Although research reports that both patients and physicians think it is important to
discuss out-of-pocket costs, few physicians have the conversation with patients because of the
complexity of health care and cost prediction.
123-125
For patients with terminal illnesses, in
particular, several advocates suggest that patients may need to make trade-offs in less costly
treatment to avoid financial burden to their families.
45,123
Our findings may encourage such
conversation to help patients shape end-of-life decision making in light of their financial
circumstances. However, it is important for physicians to ask patients if healthcare out-of-pocket
costs would impact decision-making prior to initiating these discussions. Given that the federal
poverty level was about $973 per month for individuals in 2014, our predicted average $673
reduction on out-of-pocket costs may relieve substantial burden among financially disadvantaged
individuals.
This study has several potential limitations. First, we focused on hospital out-of-pocket
costs rather than all types of health care costs such as long-term-care and home-care costs, as
these costs may not be as amenable to change by having an AD due to their different goals of
care related to each care venue. Additionally, Exit Interviews are not weighted for national
representation as the HRS does not apply sample weight variables for decedents.
102
Depending
on the timing of death, participants may have had low out-of-pocket costs if they died close to
the last core interview, while others surviving the entire two years between interview windows
could have had relatively higher costs. However, we adjusted for this survival time in our
analytic models and found that AD completion status remained significantly associated with
lower hospital out-of-pocket costs. Finally, proxies may experience recall bias on AD-related
52
questions and misreport out-of-pocket spending during surveys.
36
However, French and
colleagues found the out-of-pocket expenditure reported in the HRS exit interviews were
relatively consistent with the data in the Medicare Current Beneficiary Survey (MCBS),
indicating reliable quality and validity of the HRS exit medical expenditure data.
126
To date, this is the first paper to examine the relationship between AD completion and
hospital out-of-pocket expenditures. Given that patients may consider out-of-pocket costs when
engaging in health care decision making, understanding the relationship between ADs and end-
of-life treatment choices may motivate some consumers to complete ADs. Additionally, early
(e.g., more than 3 months before death) AD completion resulted in even lower hospital out-of-
pocket costs, thereby potentially adding greater patient incentive to engage in early advance care
planning conversations and AD completion.
53
Table 5.1 Sample descriptions
Variables Total
*
(N=9,228)
AD Completed
(N=3,950)
No AD
(N=5,059)
Missing on AD
(N=219)
P value
‡
Hospital out-of-pocket cost,
mean (SD)
2,091.3
(18,667.1)
1,475.7
(10,915.4)
2,612.6
(25,599.7)
1,152.8
(78,540.3)
<0.001
Age at death, mean (SD) 79.7 (10.9) 81.8 (9.8) 78.0 (11.3) 80.6 (11.4) <0.001
Education in years
†
, mean (SD) 11.3 (3.5) 12.1 (3.0) 10.6 (3.8) 11.7 (3.2) <0.001
Total income
†
, mean (SD) 40,203.8
(90,334.3)
45,896.7
(116,142.9)
35,955.9
(82,634.1)
31,844.6
(34,149.6)
<0.001
Total wealth
†
, mean (SD) 34,0020.1
(1,102,521)
456,855.5
(1,496,045)
247,761.4
(902,525)
297,629.1
(552,839)
<0.001
Female, n (%) 4,929 (53.4) 2,210 (56.0) 2,592 (51.2) 127 (58.0) <0.001
Insurance coverage
†
, n (%)
Medicare fee-for-service 7,363 (81.8) 3,272 (84.0) 3,934 (79.8) 157 (90.8) <0.001
Medicare Advantage 810 (9.8) 375 (10.6) 432 (9.5) 3 (1.5) 0.093
Medicaid 2,399 (27.6) 794 (20.8) 1,561 (32.9) 44 (30.8) <0.001
Veteran benefits 543 (6.0) 258(6.6) 271(5.5) 14 (8.1) 0.026
Private or Employment 4,841 (52.5) 2,402 (60.8) 2,285 (45.2) 154 (70.3) <0.001
Race
†
, n (%) <0.001
Non-Hispanic White 6,876 (74.6) 3,512 (88.9) 3,198 (63.3) 166 (75.8)
Non-Hispanic Black 1,495 (16.2) 259 (6.6) 1,197 (23.7) 39 (17.8)
Hispanic 684 (7.4) 124 (3.1) 551 (10.9) 9 (4.1)
Other 166 (1.8) 54 (1.4) 107 (2.1) 5 (2.3)
Death expected
†
, n (%) 5,383 (58.7) 2,523 (64.2) 2,738 (54.5) 122 (57.3) <0.001
Death location
†
, n (%) <0.001
Private home 2,640 (28.7) 1,083 (27. 5) 1,503 (29.8) 54 (25.5)
Hospital 3,343 (36.3) 1,285 (32.6) 1,980 (39.2) 78 (36.8)
Nursing home 2,211 (24.0) 1,091 (27.7) 1,065 (21.1) 55 (25.9)
Hospice 802 (8.7) 423 (10.7) 360 (7.1) 19 (9.0)
Other 208 (2.3) 64 (1.6) 138 (2.7) 6 (2.8)
Cause of death, n (%) <0.001
Cancer 2,098 (23.7) 976 (25.3) 1,085 (22.6) 37 (18.9)
Cardiovascular diseases 3,080 (34.7) 1,245 (32.2) 1,753 (36.4) 82 (41.8)
Other 3,694 (41.6) 1,644 (42.5) 1,973 (41.0) 77 (39.3)
Duration of final illness, n (%) <0.001
No warnings 809 (9.0) 247 (6.4) 538 (10.9) 24 (11.9)
Less than 1 day 615 (6.8) 215 (5.5) 385 (7.8) 15 (7.4)
Less than 1 week 1,521 (16.9) 689 (17.8) 806 (16.4) 26 (12.9)
Less than 1 month 1,780 (19.8) 813 (21.0) 939 (19.1) 28 (13.9)
Less than 1 year 2,305 (25.6) 1,002 (25.8) 1,246 (25.3) 57 (28.2)
More than 1 year 1,971 (21.9) 913 (23.5) 1,006 (20.5) 52 (25.7)
Number of days from death to
last core interview, mean (SD)
419.9
(229.5)
420.8
(227.4)
417.8
(230.9)
456.1
(235.1)
0.537
54
*
Missing data on AD completion (N=219) were excluded in this table
†
Variables with additional missing data
‡
P value indicates significant difference between decedents who had AD completed versus no
AD at 0.05 level
AD: Advance directive
Insurance coverage is not mutually exclusive
55
Table 5.2 Predicted hospital out-of-pocket costs combining both parts of two-part model (in 2014
US dollars)
Predicted costs 95% confidence interval P value
*
AD completed -672.8 (-1,203.4, -142.2) 0.013
Age at death -64.6 (-93.7, -35.6) <0.001
Female 866.4 (304.7, 1,428.0) 0.003
Medicare Advantage -1,354.7 (-1,985.0, -724.4) <0.001
Medicaid -1,537.2 (-2,091.4, -983.0) <0.001
Veteran benefits -1,102.0 (-1,645.2, -558.9) <0.001
Private or Employer insurance -767.1 (-1,339.0, -195.2) 0.009
Death expected 1,019.6 (483.9, 1,555.3) <0.001
Death location (Private home as reference)
Hospital 1,569.6 (879.2, 2,260.0) <0.001
Nursing home 774.5 (158.2, 1,390.8) 0.014
Cause of death (Cancer as reference)
Other -1,148.3 (-2,016.4, -280.2) 0.010
*
Significant results were provided, other covariates were omitted in this table
AD: Advance directive
56
Figure 5.1 Marginal effects of advance directive completion by age
AD: Advance directive
57
6. Chapter 5: Discussion and Conclusion
Discussion
Contribution to the Field
The ongoing debate about the value of ACP supports the need for additional research to
better understand factors related to ACP/ADs. To provide more insight into engagement in
ACP/AD activities, this dissertation investigated three different aspects of ACP/AD completion
in the United States: (1) changes in trends of uptake across racial groups, (2) unique factors that
may hamper ACP/AD engagement, and (3) how end-of-life care planning activities affect
patients’ out-of-pocket costs related to hospital care. Chapter 2 examined the trends in
prevalence of each ACP/LW/HCP component and found significant changes in trends during the
study periods, with a rapid increase in the early 2000s followed by a stagnation in more recent
years. In addition, patterns of change varied among different racial groups. Chapter 3 reenforced
the independent role of health literacy in ACP/AD completion in the U.S. population by using
nationally representative data, controlling for a variety of covariates. Lastly, Chapter 4
demonstrated lower rates of hospital out-of-pocket costs related to AD completion.
Recently, a few researchers have critiqued the focus of research on ACP, noting that we
have little data to support the benefits of ACP despite large investment into ACP research and
practice.
1
However, this position is controversial, and there have been mixed findings on the
quality of care and healthcare utilization outcomes related to ACP, which were in part due to
measurement challenges.
127
Considering the current low prevalence of ACP/AD engagement
among minority groups, as well as the disparate trends observed over the past 19 years, it would
be unjustifiable to discontinue investment in research and the exploration of potential benefits
58
associated with ACP/AD. Doing so could potentially perpetuate racial disparities in end-of-life
care. Finally, findings from this dissertation provide some support for patient-centered outcomes
arising from the development of ADs, namely in lowered out-of-pocket costs, an outcome
important to many.
99
Taken together, these studies provide new empirical evidence related to
ACP engagement.
Contribution to the Literature
This dissertation sheds light on new and valuable analytical approaches to track the
prevalence of ACP/AD completion over time, particularly with longitudinal data that may reflect
a dynamic pattern and shifting directions in trends, rather than assuming a simple linear path.
Using this analytic approach, we found that there was no substantial increase in trends in
ACP/AD activities in recent years. Although it is well known that the overall prevalence of
ACP/AD completion is low and racial minority groups have lower rates than their White
counterparts, only a handful of studies have focused on the trends and patterns in ACP
engagement and most have focused on targeted sub-population such as patients with cancer or
dementia or those who received intensive care.
60,62,63
In addition, these studies did not account
for sudden changes in trends over long study periods nor did they dive into differences in trends
across racial groups. The trend analysis in this dissertation revealed new trends that diverge from
previous research findings. Although the results cannot provide sufficient evidence to conclude
causal effect, the patterns in the overall trend in this study parallel historical events related to
end-of-life care planning. Specifically, the overall rapid ACP/AD uptake in the early 2000s may
have been influenced by the Terri Schiavo case that rose national debate and awareness of end-
of-life care planning. This case of a battle over the removal of a feeding tube of a young woman
59
in a vegetative state made national headlines as her husband and her parents battled in numerous
court proceedings between 1998 and 2005. This high-profile case may have stimulated a prompt
increase in AD completion. The following stagnation or decline among Blacks may partly reflect
an effect of the “death panel myth” that widely spread and undermined the goal of ACP/AD
conversations. The idea of “death panel” came out as a political term and falsely accused the
Affordable Care Act proposal, specifically the proposed voluntary ACP/AD counseling for
Medicare patients, to be serving as “death panels” where physicians would decide who got
treatment and who did not. The dissemination of this “death panel myth” may have further
exacerbated African Americans’ distrust of the healthcare system, and subsequently contributed
to the decline in their engagement in ACP/AD activities. It also seems likely that a ceiling effect
has been reached among the non-Hispanic White population, who currently have a rate of 87%
having at least one ACP/AD component. Efforts to target the remaining non-Hispanic White
population may have diminishing returns. One possible explanation is that patients who want
aggressive care to prolong life may not be motivated to develop an AD, as our standard default
care is an aggressive approach which aligns with their treatment preferences. Thus, they may not
see the need for indicating their healthcare wishes in an AD. The rate of engagement among non-
Hispanic Black participants (only 61% having at least one ACP/AD component) suggests that
underlying issues may have haltered the development in ACP/AD activities. Although we cannot
conclude what factors resulted in the decline from the trend analysis, it is important to identify
potential reasons that led to African Americans’ reduced interest in end-of-life planning in recent
years. A potential reason is that current ACP/AD approaches may not align with a combination
of Blacks’ cultural values and beliefs, including preferences for aggressive care, less comfort
discussing death, distrust of healthcare system, and spiritual beliefs.
128
60
Chapter 3 adds to the literature by using nationally representative survey data including
both non-Hispanic Black and Hispanic participants and found that health literacy was an
independent factor associated with each of the ACP/AD activities after controlling for potential
confounders such as race, education, and variety of health conditions. Previous studies have
found that limited health literacy was associated with poor health and healthcare access.
73,74,78
The disparity in healthcare access can lead to a lower likelihood of receiving information about
end-of-life planning. Moreover, should patients receive information about ACP/AD from
healthcare professionals, those with lower health literacy may continue to experience difficulties
in understanding end-of-life terminology that impact their decision making.
129,130
Therefore,
healthcare providers play an important role to identify those with limited health literacy in the
U.S. population who are less likely to have end-of-life planning and then use customized
information to help them understand the goal of ACP/AD and involve them in the process.
Finally, although some researchers have challenged the ethics of investigating cost-
related outcomes as part of goals of ACP/AD as stated in the previous chapter, cost-informed
goals of care are necessary to help inform patient decision-making. Such information can help in
providing needed information to consider treatment benefits and prognosis along with potential
improvements in quality of life, especially for patients with disadvantages in healthcare access
and financial resources.
99
In addition, a recent study of community-based ACP initiatives among
Black older adults found that adding financial planning in ACP conversations was critically
important.
131
Chapter 4 is the first study to investigate the relationship between ACP and out of
pocket spending. We found those who had ADs had significant lower hospital out-of-pocket
costs ($673) than those without an AD.
61
Policy Implications
This study also reiterates the need for policy makers to reconsider approaches to
developing ACP/AD among different racial groups. As experts agree that ACP is a complex
process for end-of-life care preparation rather than a list of check box forms,
68
our findings imply
current one-size-fits-all approaches may not be appropriate for all populations. Our findings also
support calls to re-evaluate resources spent on ACP/AD, particularly given the plateau of rates
among Whites and the decline among Blacks. However, these findings by no means suggest that
we should halt efforts focused on end-of-life care preparation that may perseverate racial
disparities in ACP/AD completion. Instead, the focus is to consider how to re-allocate the
resources and improve the effectiveness of ACP/AD uptake based on different characteristics. As
shown in Chapter 2, our findings among Hispanics of growth in prevalence of ACP discussions
and HCP assignment suggests that the efforts with this population are paying off. Although
Hispanics experienced decline in LW completion, this may be in part due to their cultural
preference for family engaged decision-making over self-autonomy, thus having an HCP may be
more important for Hispanics than completing a LW. In fact, one study that included Spanish-
Speaking older adults found the intervention group, who went through a patient-centered and
interactive online program, had higher rates of ACP documentation.
132
Additionally, racial minorities may suffer multifaceted and amplified disadvantage in
ACP/AD activities due to the intricate interplay between race and health literacy. A previous
study investigating racial differences in the effectiveness of interventions aimed at improving
AD completion found that Black participants with limited health literacy had significantly lower
rates of up taking an AD compared to their White counterparts, while the difference was not
observed among participants with adequate health literacy.
133
These findings suggest that racial
62
minorities may have greater improvement in ACP/AD engagement if access to healthcare and
understanding of the material are improved. For example, a recent study found Medicare
minority beneficiaries had equal or slightly higher rates of ACP claims overall, suggesting that
the new ACP billing codes may help to narrow the disparities in ACP discussion.
16
However,
minorities had lower rates of ACPs during an annual wellness visit, largely because they had
fewer annual wellness visits overall. Furthermore, although Medicare beneficiaries do not have a
co-payment if an ACP conversation occurs during Medicare’s Annual Wellness Visits, patients
will have Medicare Part B cost sharing if the ACP discussion occurs outside of Medical
Wellness Visits. This cost may create unequal financial burden for engaging in ACP because
minorities are less likely to have Annual Wellness Visits. Thus, policies aimed at increasing
access to annual wellness visits among minorities may indirectly improve rates of ACP for
minorities. Additionally, more efforts are needed to ensure equity of access to end-of-life care
information. Policy makers should consider expanding reimbursement for ACP conversations to
broader settings or targeting population such as low-income patients with Medicaid, or
reimbursing healthcare professionals for ACPs that take place outside of a clinical setting. For
example, studies found community-based approaches (e.g., ACP events initiated and facilitated
by local community members) have higher ACP engagement rates because provision of ACP in
a surrounding they are familiar with brings more trust and comfort when talking about such
topic.
131,134
It is worth noting that adding additional workload components to primary care practice,
such as health literacy screening and discussions related to costs of healthcare can dramatically
increase the workload and burden to an already overloaded primary care workforce. Increasing
the involvement of other support staff such as social workers may help to alleviate the burden on
63
healthcare providers. In fact, social workers widely agree that ACP is their professional duty and
responsibility, including initiating ACP conversations and documenting discussions or ADs.
135
Additionally, current CPT codes are based on time of ACP consultation; 99497 covers
the first 16 to 30 minutes and 99498 can be used for each additional 30 minutes. This
reimbursement for ACP discussion is relatively low, with $80 to $86 for paid for the first 30
minutes and $75 for each subsequent 30 minutes.
136
Since AD completion is associated with
lower end-of-life Medicare expenditures,
27
policy makers may consider increasing the rates of
reimbursement without adding more financial burden to patients to further stimulate the
conversation.
Clinical Recommendations
Findings from this dissertation support culturally and patient-centered approaches to
disseminate ACP/AD concepts. Therefore, clinical practitioners should apply diverse strategies
in their conversations with patients from varying cultural backgrounds. It is essential to
recognize that individual healthcare preferences and priorities can vary, and that engaging in
conversations about end-of-life care preferences can help ensure that patients’ wishes are known
and respected, regardless of their care preference. Healthcare providers also should be aware
that the documentation of aggressive care can substantially enhance the likelihood of receiving
the desired level of care, as found in previous study.
114
In addition, it may be feasible to reduce
racial disparities in ACP/AD engagement by widely adopting effective ACP interventions, such
as facilitated discussions (e.g. including lay health navigators, integrated ADs, question prompts
or videos) and enhance clinician training on communication and shared decision-making
64
guidance. A recent scoping review found these types of interventions had positive impacts on
communication and documentation in randomized controlled trials.
137
The findings in Chapter 3 identified another under-represented population that require
additional clinical attention: those with limited health literacy. Our findings also demonstrate that
understanding end-of-life care concepts and goals when talking to healthcare providers is
important in engaging in ACP/AD. This is consistent with a randomized controlled trial that
found that while their intervention did not increase overall AD completion rate, participants with
adequate health literacy were more likely to complete an AD.
133
One potential explanation for
this findings was that the standardized AD form (10
th
grade level) they applied may too difficult
for participants to understand.
133
These findings support our current study by demonstrating that
health literacy plays an important role in understanding and completing ADs across racial
groups. Healthcare providers should consider adopting health literacy screening questions in
their clinical practice and providing ACP information in a format tailored to the literacy level
and the culture of the patient. Studies in this dissertation support the need for patient-centered
approaches to ACP. Existing literature demonstrates that these types of interventions have been
found effective with patients with limited health literacy, and even for those with adequate health
literacy.
80,138,139
Lastly, our findings that AD completion was associated with lower patient out of pocket
spending related to hospital costs highlights empirical evidence on financial benefits of
completing an AD. Recent studies support the idea that patients’ financial cost is another aspect
we cannot avoid and should be included in ACP conversations, especially for minority groups
who are more likely to be financially disadvantaged.
131
However, Hanchate and colleagues found
that Black and Hispanic decedents had higher end-of-life Medicare expenditures primarily
65
contributed to opting for more intensive life-prolonging treatments.
28
This is consistent with
another study demonstrating that experiencing financial hardship was associated with increased
likelihood of receiving aggressive treatments.
140
Therefore, healthcare providers are encouraged
to discuss cost-related information that can benefit vulnerable population who potentially will
have higher healthcare expenditures, especially when curation of disease is not achievable and
alternative treatments are available to improve end-of-life quality. Although it is difficult to
accurately predict patients’ out-of-pocket costs due to complex treatment plans and healthcare
systems, healthcare providers can provide patients with accessible information about healthcare
charges to improve price transparency. For example, healthcare providers may seek other clinical
staff who can obtain cost estimated for medical procedures and services as well as patients’
healthcare insurance plans to evaluate potential out-of-pocket payments. They also can actively
screen patients’ financial challenges and minimize unnecessary procedures or use alternative
more affordable interventions when possible.
141
Future Research
Although it is too early to fully evaluate, the 2016 CMS reimbursement for ACP, it offers
an opportunity to potentially determine the ACP reimbursement effectiveness on outcomes such
as prevalence of ACP/AD uptake, end-of-life healthcare quality, and costs. Rigorous studies
including causal inference such as regression discontinuity design can be used to examine the
effect of this policy by using large datasets such as HRS and CMS claim data. It is also important
to keep tracking existing quality measures and cost effectiveness to assess outcomes of these
policies.
66
Conclusion
To avoid perpetuating racial disparities in ACP/AD activities, this dissertation suggests
that policy makers should continue to support ACP/AD related research and CMS should make
efforts to expand access to ACP consultation for vulnerable populations by reducing cost sharing
in ACP billings and expanding ability to bill for ACP conversation to additional settings and
healthcare providers without extra cost to patients. It also is important to ensure that patients
understand the context of ACP/AD conversations. This can be achieved through systematic
screening patients’ level of health literacy when they enter the healthcare system and provide
information consistent with their literacy level to expand engagement and quality of ACP
conversations. In addition, cost-informed goals-of-care also should be included in ACP
discussions. Our findings can be a potential information that healthcare providers can share and
discuss with patients. These proposed strategies suggest an increase in ACP investment rather
than discontinuing support for research and policy initiatives related to ACP/AD.
67
7. Bibliography
1. Sean Morrison R. Advance directives/care planning: clear, simple, and wrong. Journal of
Palliative Medicine. 2020;23(7):878-879.
2. Morrison RS, Meier DE, Arnold RM. What’s Wrong With Advance Care Planning?
JAMA. 2021;326(16):1575.
3. Kavalieratos D, Corbelli J, Zhang D, et al. Association Between Palliative Care and
Patient and Caregiver Outcomes. JAMA. 2016;316(20):2104.
4. Brinkman-Stoppelenburg A, Rietjens JA, Van Der Heide A. The effects of advance care
planning on end-of-life care: A systematic review. Palliative Medicine. 2014;28(8):1000-
1025.
5. Sudore RL, Lum HD, You JJ, et al. Defining Advance Care Planning for Adults: A
Consensus Definition From a Multidisciplinary Delphi Panel. Journal of Pain and
Symptom Management. 2017;53(5):821-832.e821.
6. Advance directive. National Cancer Institute.
https://www.cancer.gov/publications/dictionaries/cancer-terms/def/advance-directive.
Accessed2023.
7. Office UGA. Patient Self-Determination Act: Providers offer information on advance
directives but effectiveness uncertain. In: US General Accounting Office Washington,
DC; 1995.
8. Perkins HS. Controlling death: the false promise of advance directives. In. Vol 147:
American College of Physicians; 2007:51-57.
9. Silveira MJ, Wiitala W, Piette J. Advance Directive Completion by Elderly Americans: A
Decade of Change. Journal of the American Geriatrics Society. 2014;62(4):706-710.
10. Rao JK, Anderson LA, Lin F-C, Laux JP. Completion of Advance Directives Among
U.S. Consumers. American Journal of Preventive Medicine. 2014;46(1):65-70.
11. Yadav KN, Gabler NB, Cooney E, et al. Approximately one in three US adults completes
any type of advance directive for end-of-life care. Health Affairs. 2017;36(7):1244-1251.
12. Bazargan M, Cobb S, Assari S, Kibe LW. Awareness of Palliative Care, Hospice Care,
and Advance Directives in a Racially and Ethnically Diverse Sample of California
Adults. American Journal of Hospice and Palliative Medicine®. 2021;38(6):601-609.
68
13. Koss CS, Baker TA. Race Differences in Advance Directive Completion. Journal of
Aging and Health. 2017;29(2):324-342.
14. Shepherd-Banigan M, Ford CB, DePasquale N, et al. Making the informal formal:
discussing and completing advance care plans in care dyads with cognitive impairment.
Journal of palliative care. 2022;37(3):289-297.
15. Amber E B. Use of Advance Care Planning Billing Codes for �Hospitalized Older
Adults at High Risk of Dying: A National Observational Study. Journal of Hospital
Medicine. 2019;14(4):229.
16. Palmer MK, Jacobson M, Enguidanos S. Advance Care Planning For Medicare
Beneficiaries Increased Substantially, But Prevalence Remained Low: Study examines
Medicare outpatient advance care planning claims and prevalence. Health Affairs.
2021;40(4):613-621.
17. Huskamp HA, Keating NL, Malin JL, et al. Discussions With Physicians About Hospice
Among Patients With Metastatic Lung Cancer. Archives of Internal Medicine.
2009;169(10):954.
18. Dening KH, Jones L, Sampson EL. Advance care planning for people with dementia: a
review. International Psychogeriatrics. 2011;23(10):1535-1551.
19. Orlovic M, Smith K, Mossialos E. Racial and ethnic differences in end-of-life care in the
United States: Evidence from the Health and Retirement Study (HRS). SSM-population
health. 2019;7:100331.
20. Health Literacy. Centers for Disease Control and Prevention.
https://www.cdc.gov/healthliteracy/learn/index.html. Published 2023. Accessed2023.
21. Nouri SS, Barnes DE, Volow AM, et al. Health Literacy Matters More Than Experience
for Advance Care Planning Knowledge Among Older Adults. Journal of the American
Geriatrics Society. 2019;67(10):2151-2156.
22. Collins JW, Zoucha R, Lockhart JS, Mixer SJ. Cultural aspects of end-of-life care
planning for African Americans: an integrative review of literature. Journal of
Transcultural Nursing. 2018;29(6):578-590.
23. Waite KR, Federman AD, McCarthy DM, et al. Literacy and Race as Risk Factors for
Low Rates of Advance Directives in Older Adults. Journal of the American Geriatrics
Society. 2013;61(3):403-406.
24. Luo Q, Shi K, Hung P, Wang S-Y. Associations Between Health Literacy and End-of-
Life Care Intensity Among Medicare Beneficiaries. American Journal of Hospice and
Palliative Medicine®. 2021;38(6):626-633.
69
25. De Nardi M, French E, Jones JB, McCauley J. Medical spending of the US elderly.
Fiscal Studies. 2016;37(3-4):717-747.
26. Emanuel EJ. Cost savings at the end of life: what do the data show? Jama.
1996;275(24):1907-1914.
27. Nicholas LH, Langa KM, Iwashyna TJ, Weir DR. Regional Variation in the Association
Between Advance Directives and End-of-Life Medicare Expenditures. JAMA.
2011;306(13):1447.
28. Hanchate A, Kronman AC, Young-Xu Y, Ash AS, Emanuel E. Racial and Ethnic
Differences in End-of-Life Costs. Archives of Internal Medicine. 2009;169(5):493.
29. Riley GF, Lubitz JD. Long-Term Trends in Medicare Payments in the Last Year of Life.
Health Services Research. 2010;45(2):565-576.
30. Hoover DR, Crystal S, Kumar R, Sambamoorthi U, Cantor JC. Medical Expenditures
during the Last Year of Life: Findings from the 1992-1996 Medicare Current Beneficiary
Survey. Health Services Research. 2002;37(6):1625-1642.
31. French EB, McCauley J, Aragon M, et al. End-Of-Life Medical Spending In Last Twelve
Months Of Life Is Lower Than Previously Reported. Health Affairs. 2017;36(7):1211-
1217.
32. Narang AK, Nicholas LH. Out-of-Pocket Spending and Financial Burden Among
Medicare Beneficiaries With Cancer. JAMA Oncology. 2017;3(6):757.
33. Juliette Cubanski CS, Anthony Damico, Trica Neuman. How Much Is Enough? Out-of-
Pocket Spending Among Medicare Beneficiaries: A Chartbook. KFF.
https://www.kff.org/medicare/report/how-much-is-enough-out-of-pocket-spending-
among-medicare-beneficiaries-a-chartbook/. Published 2014. Accessed2023.
34. Reid E, Ghoshal A, Khalil A, et al. Out-of-pocket costs near end of life in low- and
middle-income countries: A systematic review. PLOS Global Public Health.
2022;2(1):e0000005.
35. Kelley AS, McGarry K, Fahle S, Marshall SM, Du Q, Skinner JS. Out-of-Pocket
Spending in the Last Five Years of Life. Journal of General Internal Medicine.
2013;28(2):304-309.
36. Marshall S, McGarry K, Skinner J. The Risk of Out-of-Pocket Health Care Expenditure
at End of Life. National Bureau of Economic Research;2010.
37. Fahle S, McGarry K, Skinner J. Out-of-Pocket Medical Expenditures in the United
States: Evidence from the Health and Retirement Study. 2016;37(3-4):785-819.
70
38. Lentz R, Benson AB, Kircher S. Financial toxicity in cancer care: Prevalence, causes,
consequences, and reduction strategies. Journal of Surgical Oncology. 2019;120(1):85-
92.
39. Hirth RA, Greer SL, Albert JM, Young EW, Piette JD. Out-of-pocket spending and
medication adherence among dialysis patients in twelve countries. Health affairs.
2008;27(1):89-102.
40. Bestvina CM, Zullig LL, Yousuf Zafar S. The implications of out-of-pocket cost of
cancer treatment in the USA: a critical appraisal of the literature. Future Oncology.
2014;10(14):2189-2199.
41. Wong YN, Hamilton O, Egleston B, Salador K, Murphy C, Meropol NJ. Understanding
How Out ‐of ‐Pocket Expenses, Treatment Value, and Patient Characteristics Influence
Treatment Choices. The Oncologist. 2010;15(6):566-576.
42. Wong Y-N, Egleston BL, Sachdeva K, et al. Cancer Patients’ Trade-offs Among
Efficacy, Toxicity, and Out-of-Pocket Cost in the Curative and Noncurative Setting.
Medical Care. 2013;51(9):838-845.
43. Zafar SY, Peppercorn JM, Schrag D, et al. The Financial Toxicity of Cancer Treatment:
A Pilot Study Assessing Out ‐of ‐Pocket Expenses and the Insured Cancer Patient's
Experience. The Oncologist. 2013;18(4):381-390.
44. Kullgren JT. Health Care Use and Decision Making Among Lower-Income Families in
High-Deductible Health Plans. Archives of Internal Medicine. 2010;170(21):1918.
45. Donley G, Danis M. Making the Case for Talking to Patients about the Costs of End-of-
Life Care. Journal of Law, Medicine & Ethics. 2011;39(2):183-193.
46. Weeks WB. Advance Directives and the Cost of Terminal Hospitalization.
1994;154(18):2077.
47. Teno JM, Gruneir A, Schwartz Z, Nanda A, Wetle T. Association Between Advance
Directives and Quality of End-of-Life Care: A National Study. Journal of the American
Geriatrics Society. 2007;55(2):189-194.
48. Degenholtz HB, Rhee Y, Arnold RM. Brief Communication: The Relationship between
Having a Living Will and Dying in Place. Annals of Internal Medicine. 2004;141(2):113-
117.
49. Heffner JE, Barbieri C, Fracica P, Brown LK. Communicating Do-Not-Resuscitate
Orders With a Computer-Based System. 1998;158(10):1090.
50. Chambers CV. Relationship of Advance Directives to Hospital Charges in a Medicare
Population. 1994;154(5):541.
71
51. Molloy DW, Guyatt GH, Russo R, et al. Systematic Implementation of an Advance
Directive Program in Nursing Homes. JAMA. 2000;283(11):1437.
52. Enguidanos S, Ailshire J. Timing of Advance Directive Completion and Relationship to
Care Preferences. Journal of Pain and Symptom Management. 2017;53(1):49-56.
53. Kelley AS, Morrison RS, Wenger NS, Ettner SL, Sarkisian CA. Determinants of
treatment intensity for patients with serious illness: a new conceptual framework. Journal
of palliative medicine. 2010;13(7):807-813.
54. Bradley EH, McGraw SA, Curry L, et al. Expanding the Andersen model: The role of
psychosocial factors in long ‐term care use. Health services research. 2002;37(5):1221-
1242.
55. Lehning AJ, Kim MH, Dunkle RE. Facilitators of home and community-based service
use by urban African American elders. Journal of Aging and Health. 2013;25(3):439-
458.
56. Morhaim DK, Pollack KM. End-of-Life Care Issues: A Personal, Economic, Public
Policy, and Public Health Crisis. American Journal of Public Health. 2013;103(6):e8-
e10.
57. McMahan RD, Knight SJ, Fried TR, Sudore RL. Advance Care Planning Beyond
Advance Directives: Perspectives From Patients and Surrogates. Journal of Pain and
Symptom Management. 2013;46(3):355-365.
58. Carr D. Racial Differences in End-Of-Life Planning: Why Don'T Blacks and Latinos
Prepare for the Inevitable? OMEGA - Journal of Death and Dying. 2011;63(1):1-20.
59. Gerst K, Burr JA. Planning for end-of-life care: Black-White differences in the
completion of advance directives. Research on Aging. 2008;30(4):428-449.
60. Narang AK, Wright AA, Nicholas LH. Trends in Advance Care Planning in Patients
With Cancer. JAMA Oncology. 2015;1(5):601.
61. Khosla N, Curl AL, Washington KT. Trends in engagement in advance care planning
behaviors and the role of socioeconomic status. American Journal of Hospice and
Palliative Medicine®. 2016;33(7):651-657.
62. Block BL, Jeon SY, Sudore RL, Matthay MA, Boscardin WJ, Smith AK. Patterns and
Trends in Advance Care Planning Among Older Adults Who Received Intensive Care at
the End of Life. JAMA Internal Medicine. 2020;180(5):786.
72
63. Gotanda H, Walling AM, Reuben DB, Lauzon M, Tsugawa Y. Trends in advance care
planning and end-of-life care among persons living with dementia requiring surrogate
decision-making. Journal of the American Geriatrics Society. 2022;70(5):1394-1404.
64. Portanova J, Ailshire J, Perez C, Rahman A, Enguidanos S. Ethnic Differences in
Advance Directive Completion and Care Preferences: What Has Changed in a Decade?
Journal of the American Geriatrics Society. 2017;65(6):1352-1357.
65. Xu J, Lin Y, Yang M, Zhang L. Statistics and pitfalls of trend analysis in cancer research:
a review focused on statistical packages. Journal of Cancer. 2020;11(10):2957-2961.
66. Kohler BA, Ward E, McCarthy BJ, et al. Annual report to the nation on the status of
cancer, 1975–2007, featuring tumors of the brain and other nervous system. Journal of
the national cancer institute. 2011;103(9):714-736.
67. Kim HJ, Fay MP, Feuer EJ, Midthune DN. Permutation tests for joinpoint regression
with applications to cancer rates. Statistics in medicine. 2000;19(3):335-351.
68. Periyakoil VS, Gunten CFv, Arnold R, Hickman S, Morrison S, Sudore R. Caught in a
loop with advance care planning and advance directives: how to move forward? Journal
of Palliative Medicine. 2022;25(3):355-360.
69. Kwak J, Haley WE. Current research findings on end-of-life decision making among
racially or ethnically diverse groups. The gerontologist. 2005;45(5):634-641.
70. Kelley AS, Wenger NS, Sarkisian CA. Opiniones: End-of-Life Care Preferences and
Planning of Older Latinos. Journal of the American Geriatrics Society. 2010;58(6):1109-
1116.
71. Cho YI, Lee S-YD, Arozullah AM, Crittenden KS. Effects of health literacy on health
status and health service utilization amongst the elderly. Social science & medicine.
2008;66(8):1809-1816.
72. White S, Chen J, Atchison R. Relationship of preventive health practices and health
literacy: a national study. American journal of health behavior. 2008;32(3):227-242.
73. Sudore RL, Mehta KM, Simonsick EM, et al. Limited Literacy in Older People and
Disparities in Health and Healthcare Access. Journal of the American Geriatrics Society.
2006;54(5):770-776.
74. Sudore RL, Yaffe K, Satterfield S, et al. Limited literacy and mortality in the elderly.
Journal of General Internal Medicine. 2006;21(8):806-812.
75. Volandes AE, Paasche-Orlow MK. Health Literacy, Health Inequality and a Just
Healthcare System. The American Journal of Bioethics. 2007;7(11):5-10.
73
76. Volandes AE, Paasche-Orlow M, Gillick MR, et al. Health literacy not race predicts end-
of-life care preferences. Journal of palliative medicine. 2008;11(5):754-762.
77. Sudore RL, Landefeld CS, Pérez-Stable EJ, Bibbins-Domingo K, Williams BA,
Schillinger D. Unraveling the relationship between literacy, language proficiency, and
patient–physician communication. Patient Education and Counseling. 2009;75(3):398-
402.
78. Levy H, Janke A. Health Literacy and Access to Care. Journal of Health Communication.
2016;21(sup1):43-50.
79. Wallace LS, Rogers ES, Roskos SE, Holiday DB, Weiss BD. Brief report: screening
items to identify patients with limited health literacy skills. Journal of general internal
medicine. 2006;21:874-877.
80. Sudore RL, Landefeld CS, Barnes DE, et al. An advance directive redesigned to meet the
literacy level of most adults: A randomized trial. Patient Education and Counseling.
2007;69(1-3):165-195.
81. Sudore RL, Boscardin J, Feuz MA, McMahan RD, Katen MT, Barnes DE. Effect of the
PREPARE Website vs an Easy-to-Read Advance Directive on Advance Care Planning
Documentation and Engagement Among Veterans. JAMA Internal Medicine.
2017;177(8):1102.
82. Frey R, Raphael D, Bellamy G, Gott M. Advance care planning for Māori, Pacific and
Asian people: the views of New Zealand healthcare professionals. Health & Social Care
in the Community. 2014;22(3):290-299.
83. Braun UK, Beyth RJ, Ford ME, Espadas D, McCullough LB. Decision-making styles of
seriously ill male Veterans for end-of-life care: Autonomists, Altruists, Authorizers,
Absolute Trusters, and Avoiders. Patient Education and Counseling. 2014;94(3):334-
341.
84. Nakazawa K, Kizawa Y, Maeno T, et al. Palliative care physicians’ practices and
attitudes regarding advance care planning in palliative care units in Japan: a nationwide
survey. American Journal of Hospice and Palliative Medicine®. 2014;31(7):699-709.
85. Braun UK, Ford ME, Beyth RJ, McCullough LB. The physician's professional role in
end-of-life decision-making: Voices of racially and ethnically diverse physicians. Patient
Education and Counseling. 2010;80(1):3-9.
86. Hopp FP, Duffy SA. Racial Variations in End-of-Life Care. Journal of the American
Geriatrics Society. 2000;48(6):658-663.
74
87. McDermott E, Selman LE. Cultural Factors Influencing Advance Care Planning in
Progressive, Incurable Disease: A Systematic Review With Narrative Synthesis. Journal
of Pain and Symptom Management. 2018;56(4):613-636.
88. Krakauer EL, Crenner C, Fox K. Barriers to Optimum End ‐of ‐life Care for Minority
Patients. Journal of the American Geriatrics Society. 2002;50(1):182-190.
89. Blackhall LJ, Murphy ST, Frank G, Michel V, Azen S. Ethnicity and attitudes toward
patient autonomy. Jama. 1995;274(10):820-825.
90. Morrison RS, Zayas LH, Mulvihill M, Baskin SA, Meier DE. Barriers to Completion of
Health Care Proxies. Archives of Internal Medicine. 1998;158(22):2493.
91. Skinner J, Chandra A, Goodman D, Fisher ES. The Elusive Connection Between Health
Care Spending And Quality. Health Affairs. 2008;27(Suppl1):w119-w123.
92. Yasaitis L, Fisher ES, Skinner JS, Chandra A. Hospital Quality And Intensity Of
Spending: Is There An Association? Hospitals' performance on quality of care is not
associated with the intensity of their spending. Health Affairs. 2009;28(Suppl1):w566-
w572.
93. Teno JM, Fisher ES, Hamel MB, Coppola K, Dawson NV. Medical Care Inconsistent
with Patients' Treatment Goals: Association with 1-Year Medicare Resource Use and
Survival. Journal of the American Geriatrics Society. 2002;50(3):496-500.
94. Zhang B, Wright AA, Huskamp HA, et al. Health care costs in the last week of life:
associations with end-of-life conversations. Archives of internal medicine.
2009;169(5):480-488.
95. Dixon J, Matosevic T, Knapp M. The economic evidence for advance care planning:
Systematic review of evidence. Palliative Medicine. 2015;29(10):869-884.
96. Emanuel EJ, Emanuel LL. The Economics of Dying -- The Illusion of Cost Savings at the
End of Life. New England Journal of Medicine. 1994;330(8):540-544.
97. Klingler C, In Der Schmitten J, Marckmann G. Does facilitated Advance Care Planning
reduce the costs of care near the end of life? Systematic review and ethical
considerations. Palliative Medicine. 2016;30(5):423-433.
98. Piemonte NM, Hermer L. Avoiding a “Death Panel” Redux. Hastings Center Report.
2013;43(4):20-28.
99. Li J, Braun RT, Kakarala S, Prigerson HG. How Should Cost-Informed Goals of Care
Decisions Be Facilitated at Life’s End? AMA journal of ethics. 2022;24(11):1040-1048.
75
100. Health and Retirement Study, Harmonized HRS End-of-Life (1992-2014) public use
dataset. Produced and distributed by the University of Michigan with funding from the
National Institute on Aging (grant number NIA U01AG009740). Ann Arbor, MI, (2019).
In.
101. Health and Retirement Study, RAND HRS Longitudinal File (1992-2016) public use
dataset. Produced and distributed by the University of Michigan with funding from the
National Institute on Aging (grant number NIA U01AG009740). Ann Arbor, MI, (2020).
In.
102. USC Program on Global Aging, Health, and Policy. Harmonized HRS End of Life
Documentation.
https://hrsdata.isr.umich.edu/sites/default/files/documentation/other/Harmonized_HRS_E
nd_of_Life_A_1992-2014.pdf. Published 2019. Accessed October 11, 2020.
103. Gozalo P, Teno JM, Mitchell SL, et al. End-of-Life Transitions among Nursing Home
Residents with Cognitive Issues. New England Journal of Medicine. 2011;365(13):1212-
1221.
104. Azur MJ, Stuart EA, Frangakis C, Leaf PJ. Multiple imputation by chained equations:
what is it and how does it work? International Journal of Methods in Psychiatric
Research. 2011;20(1):40-49.
105. Finkelstein EA, Trogdon JG, Cohen JW, Dietz W. Annual Medical Spending Attributable
To Obesity: Payer-And Service-Specific Estimates: Amid calls for health reform, real
cost savings are more likely to be achieved through reducing obesity and related risk
factors. Health affairs. 2009;28(Suppl1):w822-w831.
106. Duan N, Manning WG, Morris CN, Newhouse JP. Choosing between the sample-
selection model and the multi-part model. Journal of Business & Economic Statistics.
1984;2(3):283-289.
107. Lê Cook B, McGuire TG, Lock K, Zaslavsky AM. Comparing methods of racial and
ethnic disparities measurement across different settings of mental health care. Health
services research. 2010;45(3):825-847.
108. Cawley J, Meyerhoefer C. The medical care costs of obesity: an instrumental variables
approach. Journal of health economics. 2012;31(1):219-230.
109. Afifi AA, Kotlerman JB, Ettner SL, Cowan M. Methods for Improving Regression
Analysis for Skewed Continuous or Counted Responses. 2007;28(1):95-111.
110. Deb P, Norton EC. Modeling Health Care Expenditures and Use. Annual Review of
Public Health. 2018;39(1):489-505.
76
111. Mihaylova B, Briggs A, O'Hagan A, Thompson SG. Review of statistical methods for
analysing healthcare resources and costs. Health economics. 2011;20(8):897-916.
112. Rosenbaum PR, Rubin DB. The central role of the propensity score in observational
studies for causal effects. Biometrika. 1983;70(1):41-55.
113. Basu A, Polsky D, Manning WG. Estimating treatment effects on healthcare costs under
exogeneity: is there a ‘magic bullet’? Health Services and Outcomes Research
Methodology. 2011;11(1-2):1-26.
114. Silveira MJ, Kim SYH, Langa KM. Advance Directives and Outcomes of Surrogate
Decision Making before Death. New England Journal of Medicine. 2010;362(13):1211-
1218.
115. Bird CE, Shugarman LR, Lynn J. Age and gender differences in health care utilization
and spending for medicare beneficiaries in their last years of life. J Palliat Med.
2002;5(5):705-712.
116. De Kok IMCM, Polder JJ, Habbema JDF, et al. The impact of healthcare costs in the last
year of life and in all life years gained on the cost-effectiveness of cancer screening.
British Journal of Cancer. 2009;100(8):1240-1244.
117. Hamel MB, Lynn J, Teno JM, et al. Age-Related Differences in Care Preferences,
Treatment Decisions, and Clinical Outcomes of Seriously Ill Hospitalized Adults:
Lessons from SUPPORT. 2000;48(S1):S176-S182.
118. Deep KS, Griffith CH, Wilson JF. Communication and decision making about life-
sustaining treatment: examining the experiences of resident physicians and seriously-ill
hospitalized patients. Journal of general internal medicine. 2008;23(11):1877-1882.
119. Sudore RL, Fried TR. Redefining the “planning” in advance care planning: preparing for
end-of-life decision making. Annals of internal medicine. 2010;153(4):256-261.
120. Anderson WG, Chase R, Pantilat SZ, Tulsky JA, Auerbach AD. Code status discussions
between attending hospitalist physicians and medical patients at hospital admission.
Journal of general internal medicine. 2011;26(4):359-366.
121. Otte IC, Jung C, Elger BS, Bally K. Advance directives and the impact of timing. A
qualitative study with Swiss general practitioners. Swiss medical weekly. 2014;144.
122. Richards OK, Iott BE, Toscos TR, Pater JA, Wagner SR, Veinot TC. “It’s a mess
sometimes”: patient perspectives on provider responses to healthcare costs, and how
informatics interventions can help support cost-sensitive care decisions. Journal of the
American Medical Informatics Association. 2022.
77
123. Ubel PA, Abernethy AP, Zafar SY. Full disclosure--out-of-pocket costs as side effects.
The New England journal of medicine. 2013;369(16):1484.
124. Alexander GC. Patient-Physician Communication About Out-of-Pocket Costs. JAMA.
2003;290(7):953.
125. Pham HH. Physician Consideration of Patients' Out-of-Pocket Costs in Making Common
Clinical Decisions. Archives of Internal Medicine. 2007;167(7):663.
126. French E, Jones JB, McCauley J. The Accuracy of Economic Measurement in the Health
and Retirement Study. Forum for Health Economics and Policy. 2017;20(2).
127. Sudore RL, Hickman SE, Walling AM. Controversies About Advance Care Planning.
JAMA. 2022;327(7):685.
128. Johnson KS, Kuchibhatla M, Tulsky JA. What Explains Racial Differences in the Use of
Advance Directives and Attitudes Toward Hospice Care? Journal of the American
Geriatrics Society. 2008;56(10):1953-1958.
129. Ladin K, Buttafarro K, Hahn E, Koch-Weser S, Weiner DE. “End-of-life care? I’m not
going to worry about that yet.” Health literacy gaps and end-of-life planning among
elderly dialysis patients. The Gerontologist. 2018;58(2):290-299.
130. De Vries K, Banister E, Dening KH, Ochieng B. Advance care planning for older people:
The influence of ethnicity, religiosity, spirituality and health literacy. Nursing Ethics.
2019;26(7-8):1946-1954.
131. Nouri S, Quinn M, Doyle BN, et al. “We’ve Got to Bring Information to Where People
Are Comfortable”: Community-Based Advance Care Planning with the Black
Community. Journal of General Internal Medicine. 2023.
132. Sudore RL, Schillinger D, Katen MT, et al. Engaging Diverse English- and Spanish-
Speaking Older Adults in Advance Care Planning. JAMA Internal Medicine.
2018;178(12):1616.
133. Barker PC, Holland NP, Shore O, et al. The Effect of Health Literacy on a Brief
Intervention to Improve Advance Directive Completion: A Randomized Controlled
Study. Journal of Primary Care & Community Health. 2021;12:215013272110002.
134. Van Scoy LJ, Levi BH, Witt P, et al. Association of Participation in an End-of-Life
Conversation Game With Advance Care Planning Behavior and Perspectives Among
African American Individuals. JAMA Network Open. 2020;3(5):e204315.
135. Wang C-W, Chan CLW, Chow AYM. Social workers’ involvement in advance care
planning: a systematic narrative review. BMC Palliative Care. 2018;17(1).
78
136. Barwise AK, Wilson ME, Sharp RR, Demartino ES. Ethical Considerations About
Clinician Reimbursement for Advance Care Planning. Mayo Clinic Proceedings.
2020;95(4):653-657.
137. McMahan RD, Tellez I, Sudore RL. Deconstructing the Complexities of Advance Care
Planning Outcomes: What Do We Know and Where Do We Go? A Scoping Review.
Journal of the American Geriatrics Society. 2021;69(1):234-244.
138. Kripalani S, Robertson R, Love-Ghaffari MH, et al. Development of an illustrated
medication schedule as a low-literacy patient education tool. Patient education and
counseling. 2007;66(3):368-377.
139. Gerber BS, Brodsky IG, Lawless KA, et al. Implementation and evaluation of a low-
literacy diabetes education computer multimedia application. Diabetes care.
2005;28(7):1574-1580.
140. Tucker-Seeley RD, Abel GA, Uno H, Prigerson H. Financial hardship and the intensity of
medical care received near death. Psycho-Oncology. 2015;24(5):572-578.
141. Arora V, Moriates C, Shah N. The challenge of understanding health care costs and
charges. AMA journal of ethics. 2015;17(11):1046-1052.
79
8. Appendices
Appendix A: Joinpoint Graphs
At Least One ACP/AD Activity
ACP Discussion
Figure 7.1 Appendix A: Joinpoint Trends in Adjusted Prevalence of ACP/AD Completion
80
Living Will Completion
HCP Assignment
81
Appendix B: Propensity Score Model
Our propensity score model was based on a probit regression on advance directives
(ADs) completion status with the same covariates as in the two-part model. Then we applied
inverse probability weights (IPW) making weighted decedents with ADs (treated group) and
those without ADs (control group) have similar distribution of measured covariates. Thus, these
two groups are more comparable. In addition, we ran regression with those weights to adjust for
direct effects of AD completion on hospital out-of-pocket costs, so called “double-robust” to get
more reliable estimates.
Figure 8.2 Appendix B: Unweighted and IPW-weighted balance between AD completers
and non-completers
82
Table 7.1 Appendix: Propensity score weighting and regression-adjusted estimate of AD
completion effects on hospital out-of-pocket costs (in 2014 US dollars)
Coefficients 95% confidence interval P value
*
AD completed -1,334.0 (-2,339.0, -329.1) 0.009
Age at death -98.5 (-167.6, -29.4) 0.005
Female 1,634.7 (625.9, 2,643.4) 0.001
Medicaid -1,742.3 (-3,067.1, -417.4) 0.010
Death expected 1,422.5 (557.2, 2,287.8) 0.001
Death location (Private home as reference)
Hospital 2,365.7 (802.0, 3,929.5) 0.003
Nursing home 1,615.2 (146.5, 2,813.9) 0.008
Cause of death (Cancer as reference)
Other -2,313.4 (-4,623.0, -3.9) 0.050
*
Significant results were provided, other covariates were omitted in this table
AD: Advance directive
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Eliciting perspectives on palliative care: outpatient visits, advance care planning, and the impact of COVID-19
PDF
Exploring end-of-life care experiences and disparities among Hispanic populations in the United States
PDF
Racial/ethnic variation in care preferences and care outcomes among United States hospice enrollees
PDF
What is success? an exploration of person-centered intervention of complex elder mistreatment
PDF
New directions for family caregiver interventions
PDF
Nursing home transitions: a new framework for understanding preferences, barriers and outcomes
PDF
Palliative care: breaking the knowledge barrier among community-based older adults
PDF
Investigating the effectiveness of a social work intervention on reducing hospital readmissions among older adults
PDF
Lifting up ourselves by lifting up others: examining cognitive-affective states linking generativity to well-being
PDF
Residential care in Los Angeles: policy and planning for an aging population
PDF
Estimating survival in the face of pain: evidence from the health and retirement study
PDF
Life course implications of adverse childhood experiences: impacts on elder mistreatment, subjective cognitive decline, and caregivers' health
PDF
Family caregiver well-being: generational differences and coping strategies
PDF
Thinking generatively versus acting generatively: exploring the associations of generative self-concept and generative activity with cognitive function among older adults
PDF
Longitudinal change in active life expectancy: the longitudinal studies of aging 1984-2000
PDF
Developing a better understanding of elder abuse multidisciplinary teams: addressing gaps for research, policy, and practice
PDF
Getting to end-of-life discussions in advanced cancer care: barriers and attitudes that limit end-of-life communication for disadvantaged Latinos
PDF
Health differences between the elderly in Japan and the United States
PDF
The dollars and sense of caregiving: exploring the financial financial realities of providing informal care to an older adult
PDF
Crisis and values at Angel Crest Manor: a case study in long-term care management
Asset Metadata
Creator
Zhu, Yujun
(author)
Core Title
Racial disparities in advance care planning and directives completion and end-of-life outcomes in the United States
School
Leonard Davis School of Gerontology
Degree
Doctor of Philosophy
Degree Program
Gerontology
Degree Conferral Date
2023-08
Publication Date
07/26/2023
Defense Date
07/07/2023
Publisher
University of Southern California. Libraries
(digital)
Tag
advance care planning,advance directives,OAI-PMH Harvest
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Enguídanos, Susan (
committee chair
), Jacobson, Mireille (
committee member
), Wilber, Kathleen (
committee member
)
Creator Email
yujunzhu@usc.edu,zhuyujun0901@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-oUC113290647
Unique identifier
UC113290647
Identifier
etd-ZhuYujun-12158.pdf (filename)
Legacy Identifier
etd-ZhuYujun-12158
Document Type
Dissertation
Rights
Zhu, Yujun
Internet Media Type
application/pdf
Type
texts
Source
20230728-usctheses-batch-1075
(batch),
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Repository Email
cisadmin@lib.usc.edu
Tags
advance care planning
advance directives