.‘J

   
  
 
         
   

Web“: ';'~4'.~ ' .

<. .4 fair in -'

rydanx; “a

5%"??? «114'

I: .3, '..'v

«m;

fix

‘1

i 21- .
1,1,;

3::

ficWL ‘m‘s

V ‘N
..n

 

max," '

. ~ u» A . -.
»'=?.2:'-2~::.~

w
a

. 5 VJ’"- 'h u

_:

4.
an:

 

 

:'
. E
. ‘
J ‘
-1
1
I 1
a V}
75
'~
i
.g,
-i‘

. rm
’2

.uw-u-J
mu

“.1, ;.u v 1r

«an

‘ , , '.
A .1
i ‘ ' ‘ ‘ I , ~
» ...
. . _ . 1. V 1:31;}
_ .. a. . ‘ .a':
i .

(a
41

 

 

J

: "é
‘ ._ c. :w
'. ~ Adam"

1'15:

 

 

TKBSIB

 

\3
(9LL/7
This is to certify that the
dissertation entitled
DECIDING TO ENGAGE IN ADVANCE CARE PLANNING:
A COMPARISON OF PARTICIPANTS’ EXPERIENCES

4% presented by

4—:
E m 1%
<1“. ..' 5
g; g) 2 KAREN JOY VANDER LAAN
.: "" C
_ l i, 3

2

has been accepted towards fulfillment

 

 

 

of the requirements for the

Ph.D. degree in Nursing

 

 

{law 7/ #4 fl"

Major Profesgzr’s Signature

22¢ '2‘ i": Z gag/é) Z
Date

MSU is an afiinnatn/e-action, equal-opportunity employer

 

—.-.--u--a---—n-

J-l-l-O-l-J—‘fiMQ-J-l-§-l-—<~d -.._A

.-_~—---_a-._l-r-D-I-l-C-'-0-'.'-.L-—‘

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DAIEDUE

DAIEDUE

DAIEDUE

 

MAR 0 7 2010

 

fialioa

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:ICIRC/DaIeDue.indd-p.1

 

DECIDING TO ENGAGE IN ADVANCE CAREPLANNING:
A COMPARISON OF PARTICIPANTS’ EXPERIENCES

By

Karen Joy Vander Laan

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
College of Nursing

2007

ABSTRACT

DECIDING TO ENGAGE IN ADVANCE CARE PLANNING:
A COMPARISON OF PARTICIPANTS’ EXPERIENCES

By
Karen Joy Vander Laan

Advance care planning (ACP) is a dynamic decision-making process that assists
people to construct and communicate their preferences for end of life care. Two decades
of research have shown that when preferences are not known, undesirable outcomes may
occur for individuals and other surrogate decision-makers. Barriers to advance planning
can be addressed through educational programs, especially through existing community-
based groups. The purpose of this study was to examine internal and external influences
within the decision-making context that may affect individuals’ decisions to engage and
re—engage in ACP. The Decision Process Model for Advance Care Planning provided
the conceptual framework for this secondary analysis of data from a prospective quasi-
experimental research project. Study participants were 147 adults from existing
community groups in the Midwestern United States who attended an ACP educational
program provided by certified ACP Facilitators. Participants completed pre- and post-
prograrn Participant Surveys, which included personal influences on decision
(information, individual characteristics, values, and prior experiences) and the personal
decision factor self-efficacy (ability and likelihood to engage in ACP conversations).
Results of this study describe the prevalence and associations of personal influence and
decision variables among participants. Statistically significant differences exist in most
variables when participants with and without previous ACP experience are compared.

After attending the ACP educational program, participants’ perceptions of their

knowledge, importance, ability, and likelihood to have ACP conversations were
significantly increased. When participants with and without previous ACP experience
were compared, the impact of the ACP program intervention remained significant in the
areas of knowledge, ability, and likelihood. Understanding the prevalence and
associations of internal personal variables and the impact of an external ACP intervention
on these internal variables can help health care professionals target community-based

educational interventions to promote ACP.

Cepyright by
KAREN JOY VANDER LAAN

2007

Wherever you go, may God go with you.

Whatever you need, may God provide.

Whenever you falter, may God forgive.

Whenever you stumble, may God save.

And when, at the end of your days, you lay yourself down
for the last time, may God raise you up for all time. Amen.

William B. Lawrence

 

 

 

for Li

I
3.0:;

.1
Cor...

 

 

 

ACKNOWLEDGMENTS

To my Committee, Gwen Wyatt, PhD, RN, Celia Wills, PhD, RN, Georgia
Padonu, PhD, RN, Patricia Mullan, PhD, and Clayton Thomason, MDiv, ID for the
opportunities and encouragement to learn and grow as a scholar;

To my Colleagues fiom the MMACPC Research Team and CAHA-ACPC for the
experiences of implementing and evaluating ACP research;

To my fellow ACP Facilitators for the passion to improve end-of-life decision-
making in our communities;

To my Colleagues and Students from the Calvin College Department of Nursing
for the encouragement to accept and complete the challenge of doctoral education;

To Doug, Aaron & Jonathon for countless sacrifices and unwavering support
along the way to this degree;

To my Family & Friends for their love, encouragement, and prayers; and

To my faithful Savior Jesus Christ, Who has done immeasurably more than all I
could ask or imagine by bringing this work to completion.

Funding sources for this project included the Blue Cross Blue Shield of Michigan
Foundation (Vander Laan); the Norman Kagan Endowed Scholarship in Graduate and
Professional Studies (V ander Laan, Michigan State University College of Human
Medicine); the Marion Peterson Nursing Research Endowment (V ander Laan, Calvin
College Department of Nursing); and Sigma Theta Tau, Kappa Epsilon Chapter-at-Large

(V ander Laan).

vi

LIST

CPL-I

PRO

cw
113*,

CO\

 

 

 

 

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. xi
LIST OF FIGURES ............................................................................... xii

CHAPTER 1

PROBLEM AND PURPOSE ...................................................................... .l

Barriers to Engaging in Advance Care Planning ........................................ 3

The Role of Advance Care Planning Research ........................................... 5

Influences on End of Life Planning .............................................. 6

Studies of End of Life Planning ................................................................... 7

Purpose of Study ...................................................................................................... 8

Research Questions .................................................................................................. 8

Overview of Dissertation Study ............................................................................... 9
CHAPTER 2

CONCEPTUAL FRAMEWORK AND DEFINITIONS ................................................... 10

Conceptual Models of ACP ................................................................................... 10

Models from Qualitative Studies ............................................................... 11

Models from Health Behavior Studies ....................................................... 11

Composite Model ....................................................................................... 12

Decision-Making Models ...................................................................................... 13

ACP Studies Based on Decision-Making Models ..................................... 14

Interactive Decision-Making Framework .................................................. 15

Decision Process Model ............................................................................. 16

The Decision Process Model for Advance Care Planning .................................... 17

The Decision to Engage in ACP ................................................................ 19

Behaviors of Engaging in ACP .................................................................. l9

Outcomes of Engaging in ACP .................................................................. 20

Decision-Making Context .......................................................................... 21

Dynamic Decision-Making ........................................................................ 22

Conceptual Definitions of Focused Model Concepts ............................................ 24

Decision to Engage in ACP ....................................................................... 24

Behaviors and Outcomes ........................................................................... 24

Decision-Making Context .......................................................................... 24

Personal Influences on Decision .................................................... 24

Personal Decision Factors .............................................................. 26

Community-Based ACP Program Intervention ............................. 27

Dynamic Decision-Making ........................................................................ 28

Chapter Summary ................................................................................................... 28

vii

 

 

CHA

LITE

 

CHI

ME:

 

CHAPTER 3

LITERATURE REVIEW & SIGNIFICANCE OF STUDY ............................................. 30
The Impetus for ACP ............................................................................................. 30
Key Features of ACP ............................................................................................. 33

A Deliberative Process ............................................................................... 33
A Communicative Process ......................................................................... 33
A Shared Decision-Making Process .......................................................... 34
A Dynamic Process .................................................................................... 36
Studies of Advance Care Planning ......................................................................... 36
The SUPPORT Study ................................................................................. 37
Studies related to the Patient Self-Determination Act (PSDA) ................. 37
Studies related to Special Populations ....................................................... 38
Studies of Interventions ............................................................................. 40
1990-1995: Randomized Controlled Trials Focused on
Physicians and Patients .................................................................. 40
1996-2001: Randomized Controlled Trials Focused on
Innovative Education Strategies ..................................................... 41
2001-2004: Randomized Controlled Trials Focused on
Resource Intense Strategies ........................................................... 42
Studies of Community-Based Approaches ................................................ 43
Results of ACP Studies relating to Research Questions ........................................ 44
Research Question 1 ................................................................................... 44
Personal Influences on Decision .................................................... 44
Personal Decision Factors .............................................................. 55
Research Question 2 ................................................................................... 57
Personal Influences on Decision .................................................... 58
Personal Decision Factors .............................................................. 60
Decision to Engage in ACP ........................................................... 60
Research Question 3 ................................................................................... 61
Personal Influences on Decision: Prior Experiences .................... 62
Research Question 4 ................................................................................... 63
Personal Influences on Decision .................................................... 63
Personal Decision Factors .............................................................. 64
Research Question 5 ................................................................................... 64
Significance of Study ............................................................................................. 65

CHAPTER 4

METHOD ........................................................................................................................... 67
Context of Study .................................................................................................... 67

Promotion of ACP ...................................................................................... 67
Development of Survey Instruments .......................................................... 68
Survey Data Collection Procedure ............................................................. 69
Data Management and Storage Procedures ............................................... 70
MMACPC Research Project IRB Approval .............................................. 70
Results Dissemination ................................................................................ 71
Approval of Secondary Data Analysis ....................................................... 71

viii

Study Methods ........................................................................................................ 71

Design ......................................................................................................... 71
Subjects ...................................................................................................... 72
Measure — Pre- and Post-Program Survey Instruments ............................. 72
Human Subject Considerations .................................................................. 72
Data Management ...................................................................................... 73
Data Analysis ............................................................................................. 74
Research Question 1 ....................................................................... 74
Research Question 2 ....................................................................... 75
Research Question 3 ....................................................................... 75
Research Question 4 ....................................................................... 76
Research Question 5 ....................................................................... 76
Chapter Summary .................................................................................................. 77
CHAPTER 5
RESULTS .......................................................................................................................... 78
Preliminary Analysis of Data ................................................................................ 78
Frequencies and missing data .................................................................... 78
Normality & Choice of Statistical Tests .................................................... 79
Recoding of Quality of Conversations Items ............................................ 80
Research Question 1 .............................................................................................. 81
Results ....................................................................................................... 81
Further analyses ......................................................................................... 86
Relationships between Having an AD, Having a HCA,
and Having Conversations ............................................................. 86
Participants’ Reported Conversation Partners ............................... 87
Perceptions of Quality of Conversations ....................................... 89
Summary ................................................................................................... 90
Research Question 2 .............................................................................................. 91
Correlations ............................................................................................... 91
Results ....................................................................................................... 93
Summary .................................................................................................... 96
Research Question 3 .............................................................................................. 97
Results ........................................................................................................ 97
Perceptions of Quality of Conversations ................................................. 104
Correlations .............................................................................................. 106
Summary .................................................................................................. 112
Research Question 4 ............................................................................................. 112
Results ...................................................................................................... l 14
Summary .................................................................................................. 116
Research Question 5 ............................................................................................. 116
Results ...................................................................................................... 1 17
Summary .................................................................................................. 126

ix

 

CHAP'
DISCL

APPET

 

CHAPTER 6

DISCUSSION .................................................................................................................. 127
Interpretation of Study Findings .......................................................................... 127
Research Question 1 ................................................................................. 127

Research Question 2 ................................................................................. 140

Research Question 3 ................................................................................. 147

Research Question 4 ................................................................................. 156

Research Question 5 ................................................................................. 157

Limitations of the Study ....................................................................................... 161
Broader Implications ............................................................................................ 162
Implications for Practice .......................................................................... 162

Implications for Education ....................................................................... 164

Implications for Research ........................................................................ 166

Implications for Future Nursing Research ........................................................... 167
Conclusion ........................................................................................................... 168
APPENDICES ................................................................................................................. 170
Appendix A: Participant Survey ......................................................................... 171
Appendix B: Information Letter / Consent Form ................................................ 176
Appendix C: Dissertation Study Items from the Participant Survey .................. 179
Appendix D: Resources for ACP Conversations ................................................. 181
Appendix E: Glossary .......................................................................................... 196
REFERENCES ................................................................................................................. 204

* 'lf 2.

Table 4.

Table 1.

IhAm.‘
TIA

Table I.

LIST OF TABLES

Table 1. Comparison ofValid Data Percentages..................................................79
Table 2. Frequencies of Pre-Program Survey Variables (Total) ................................. 82-83
Table 3. Spearman Rank Correlation Coefficients (Total)... ... ....................................... 92_
Table 4. Frequencies of Pre-Program Survey Variables (by ACP Experience) ........ 98—99
Table 5. Spearman Rank Correlation Coefficients (with ACP Experience). . . . 107
Table 6. Spearman Rank Correlation Coefficients (with No ACP Experience) ............. 108
Table 7. Frequencies of Pre- and Post-Program Survey Variables (Total)... ................1 13
Table 8. Frequencies of Pre- and Post-Program Survey Variables (by ACP Exp)..118-119
Table 9. Pairs of Influencing Variables in the Decision-Making Context. . . . . ........141-l42
Table 10. Composite Comparison of Selected Variables (by ACP Experience). . . 149

Table 11. Comparison of Pre- to Post-Program Ratings (by ACP Experience). . 159-160

Table Cl. Dissertation Study Items from Participant Survey 179-180

xi

LIST OF FIGURES

Figure 1. The Decision Process Model for Advance Care Planning. . ... ...................... 18
Figure 2. Focused model of the Decision to Engage in ACP ............................... 23
Figure 3. Percentage ofresearch study participants having an AD. . . . . .. . . . . . . . .49
Figure 4. Percentage of research study participants having a HCA. . . . . . . . . . . 52
Figure 5. Percentage of research study participants having conversations ............... 5 3
Figure 6. Percentage of participants who reported conversations with partners. . . . . 88
Figure 7. Percentage of participants reporting quality of conversations. . . . . . . . 89

Figure 8. Proportional percentage of participants reporting quality of conversations. 90
Figure 9. Percentages of participants from ACP Experience groups reporting

quality conversations .................................................................................. 105
Figure 10. Proportional percentages of participants from ACP Experience groups

reporting quality of conversations ................................................................ 106

xii

DA" fi’rgs _
J'LJL 1&U ‘.

r
W,‘ ""11 v1
rh’L‘k ~‘o‘

in‘fifi VH‘.‘)I
5

\‘1... u\;
.

-V1CQ

)u. .

.
Jr's.)
51 uni.

s.

 

limit.) ‘

Vt.

pleian
A“ \

 

CHAPTER 1
PROBLEM AND PURPOSE

Each year in the United States an average of 2.4 million people die (Centers for
Disease Control, 2006), most without their preferences1 for end of life care known or
honored (Kass-Bartelmes, Hughes, & Rutherford, 2003), When people’s preferences are
not known, undesirable outcomes may occur. For the individual, end of life care may be
incongruent with personal preferences (Gillick, 2001), impose burden and possible
suffering (Connors et al., 1995), or incur unnecessary health care costs (Field & Cassel,
1997). In addition, unawareness of preferences causes increased distress for others, such
as family members (Tilden, Tolle, Nelson, & Fields, 2001) and health care providers
(Daly, 2006; Hurley, Volicer, Rempusheski, & Fry, 1995). These people often become
surrogate decision makers when the individual is no longer able to participate in decision-
making.

Because end of life may take several trajectories, ranging from a sudden critical
injury or illness to a persistent decline from a recurring condition (Lunney, Lynn, Foley,
Lipson, & Guralnik, 2003), individuals may or may not have planned for end of life care.
When end of life care at last is needed, the capacity to make decisions may be severely
limited (Pierce & Hicks, 2001). Therefore, advance care planning (ACP) for end of life
preferences is highly desirable (Hammes & Briggs, 2002b).

ACP is defined, for the purposes of this proposal, as a dynamic decision-making
process that helps people construct and communicate their preferences for end of life
care. Because ACP is viewed as a decision-making process, further definitions of its

components are drawn fiom the decision-making literature. Decision-making is defined

 

' A glossary of key terms used throughout the proposal is included in Appendix E.

3.23.331 ct

'u-‘~ 33"
thA is”

e‘.-O‘\n

v
..--
iib \ ui‘L-IA-

season. i

in”; . -
LRQIUI‘AS

may be 3.?

SOCIO'Cul

Ot.‘ \H‘p‘

h"\,‘;1:~

“(a J
P1 5 ‘VITSL

 

 

as the evaluation of information about alternatives according to personal values (Mazur,
2003). The process of decision-making includes decisions, behaviors and outcomes
(Rothert et al., 1997; Wills & Holmes-Rovner, 2006). Decisions are defined as choices
between alternatives (Thompson & Dowding, 2002), that influence subsequent behaviors
and outcomes.

Decision-making occurs within a decision-making context, which includes the
decision, behaviors, outcomes and influences (internal and external) that shape the
decisions people make (Wills & Holmes-Rovner, 2006). ACP decisions, for example,
may be shaped by personal influences or decision factors (internal influences), broader
socio-cultural community decision factors or interventions (external influences), or
outcomes, such as initiated advance directives (AD) or congruence of decisions with
preferred care goals. The ACP decision-making process is dynamic because it may be
engaged in repeatedly, especially in response to changes within the decision-making
context. Individuals’ ACP preferences, however, may or may not change over time.

The goal of ACP is awareness of and respect for individuals’ preferences for their
future health care. The hope is that communication and consideration of preferences will
prevent avoidable distress and suffering for patients, families, and caregivers by honoring
individuals’ expressed wishes. This type of communication helps to foster the ofien
desired outcome of a “. . . good death——one that is free from avoidable distress and
suffering for patients, families, and caregivers; in general accord with patients’ and
families’ wishes; and reasonably consistent with clinical, cultural, and ethical standards”
(Field & Cassel, 1997). Based on these communication goals, ACP has been endorsed as

a domain of quality care at the end of life (Field & Cassel; Stewart, Teno, Patrick, &

docum:
plnnir
Lunnej
past If
l’t‘SSdrt
CXplar

fiduct

life pj
heak‘

(33’1ng

 

 

IS n._

 

 

Lynn, 1999) by public and private organizations such as the Michigan Commission on
End of Life Care (2001), the American Geriatrics Society (Lunney, Foley, Smith, &
Gelband, 2003), and Last Acts (a program of the Robert Wood Johnson Foundation)
(Kaplan & Peres, 2002).

Barriers to Engaging in Advance Care Planning

Although ACP has been encouraged and publicly supported, research summaries
document that at least 50% of the American population does not engage in advance
planning (Kaplan & Peres, 2002; Kass-Bartelmes et al., 2003; Lorenz et al., 2004;
Lunney, Foley et al., 2003). These results have persisted despite intense efforts over the
past 15 years to increase the prevalence of planning for end of life care. End of life
researchers and their colleagues who study shared decision-making offer a variety of
explanations, summarized by this investigator as four common barriers to ACP:
reluctance, relevance, resources, and restraints.

The first barrier is reluctance. People are often reluctant to consider their end of
life preferences. Weiner and Cole (2004) found this reluctance in both patients and
health care providers, related to personal discomfort with the topic, concern that it will
cause distress for others, or the perception that it will either take too much time or that it
is not needed at this time.

The second barrier is relevance, which may arise from a variety of sources. Some
people are not interested in participating in decision-making (Levinson, Kao, Kuby, &
Thisted, 2005; Say, Murtagh, & Thomson, 2006). Others perceive it as culturally
inappropriate (Makoul & Clayman, 2006). Still others may lack decisional capacity for

such planning to be relevant (Hammes & Briggs, 2002b; Pierce & Hicks, 2001).

harps
ACPis
ACPan
“hemp
depend
be reaii;
B a par
“ass:
SUPPOS

as 63581

OUICOI‘.‘

 

9031p}.
C0173}:-
“liter;
of the;
(Whit:
no 10:-

for“?
6“».

 

 

The third banier is availability of resources. Individually, some people may lack
the information, skills, or confidence to thoughtfully consider care options in light of
their personal values and communicate their decisions effectively. Within communities,
ACP is affected by the availability of ACP information and the perceived importance of
ACP among community members (Pearlrnan, Cole, Patrick, Starks, & Cain, 1995).
When people engage in ACP, their actual outcomes during end of life care often greatly
depend on the resources of their community health care system. Positive outcomes may
be realized in organizations that educate HCP’s to initiate and honor ACP conversations
as a part of routine adult health care (Ferrell et al., 2005; Robinson et al., 2004;
Weissman et al., 2005). An organizational culture of respect for patient preferences
supported by structures and processes to provide assistance with ACP has been reported
as essential (Hammes & Rooney, 1998). If these community resources are not available,
outcomes may be less optimal or nonexistent (Hammes & Briggs, 2002a).

The fourth barrier is perceived restraints—literally feeling held back by the
complexities of constructing and communicating preferences and keeping preferences
current. People who engage in ACP have much to consider: An array of choices,
uncertainties, and consequences which need to be understood and evaluated on the basis
of their preferences (which may be unclear) (Pierce & Hicks, 2001) so that fiJture care
(which may or may not be needed) is provided according to their wishes when they can
no longer make or communicate their decisions (Pearlrnan et al., 1995). Once
preferences are constructed, people then have to find a way to effectively communicate

them to others who may someday be involved in decision-making. This can be

.... ~

.,. .L

.
'- LA. §oh
..4. -

~

,_’,.‘; ‘f‘

i ' s .
9- ‘,.‘F-‘ I.
. —

’ ”-1: 15 f1
"3'”..‘x 5K u

1'..

'-...

.. ‘
b~\.

' 1
J';";‘ a F!
1 .h..& ‘.

I I

.9 I}; "" 1v- .

"‘-~~ “1111..
H

V F n
\‘ l 3‘
“‘41.“. ‘ .

I u
“’I- "in .
‘ l
isibhggh.“
\

f‘

‘A.

 

 

 

I‘C'S‘L‘ralr .

171n:~..

IL.”
Luld‘

 

 

challenging, particularly if others feel reluctant to discuss end of life issues or find them
irrelevant (Weiner & Cole, 2004).

Ideally, because ACP is a dynamic decision-making process—responsive to
changes in the decision-making context—people should regularly return to the decision
to engage in ACP. This provides the opportunity to validate or reconsider preferences in
response to the decision process itself or changes in individual or community influences.
However, re-engagement in ACP depends on peoples’ awareness of and willingness to
acknowledge or adapt to such changes (Grounds, 2004) and their abilities to still make
decisions, possibly under new constraints of physical or emotional stress or limited time
(Pierce & Hicks, 2001). If new preferences are constructed, effective communication of
those changes is also needed. In a review of the state of ACP research, Emanuel
summarizes these restraints as the “. .. great difficulty in knowing one’s own preferences,
articulating and recording them, and getting them understood and implemented...”
(2003)

The Role of Advance Care Planning Research

Through ACP research, changes within the decision-making context of ACP can
be examined and encouraged. For example, researchers may study how decisions lead to
outcomes or how changes in internal influences affect decisions or how an external
intervention affects peoples’ perceptions of reluctance, relevance, resources, and
restraints. Although in the beginning ACP research was focused on advance planning as

an intervention or an event, researchers are encouraged now to study ACP as a process

(Emanuel, 2003).

 

‘ I
. 0r- 3v
FITS: oIL.

COW?“

A tlhl

mere an
Advance
of influ-

interve:

 

IIIC Ifa‘:

RfSpc‘

 

”Emir:
SUPDQ.
851, L
“ her,
ETC-at;

COTE:-

 

Influences on End of Life Planning

Researchers have documented a variety of influences on end of life planning.
First there are internal individual influences—such as knowledge, demographics, values,
experiences, and self-efficacy (Cicirelli, 1997; Davison & Degner, 1998; Pearlman et al.,
1995; Pierce & Hicks, 2001). Secondly, there are external community influences, such as
access to ACP information and support of ACP processes within the health care
community (Byock, Norris, Curtis, & Patrick, 2001; Hammes & Rooney, 1998). Thirdly,
there are external interventions designed to influence and create a context in which
advance planning becomes more prevalent (Kass-Bartelmes et al., 2003). All three types
of influences have been studied by researchers using community-based educational
interventions that inform and encourage people to engage in ACP (Byock et al., 2001; P.
Clarke, Evans, Shook, & J chanson, 2005; Hamel, Guse, Hawranik, & Bond, 2002;
Hammes & Rooney, 1998; Waters, 2000).

Community-based approaches to ACP have been recommended as a way to
increase the number of people who engage in advance planning (Field & Cassel, 1997).
Respecting Choices®——a whole-community approach to ACP that includes professional
training, public education, and strategies for mobilizing organization and community
support—is an example of an intervention that increased advance plarming fi'om 15 to
85% by residents in the LaCrosse, Wisconsin community (Hammes & Rooney, 1998).
When people engage in ACP together, an environment of shared decision-making is
created (Kass-Bartelmes etal., 2003) that encourages continued shared consideration and

communication of end of life preferences.

"-339" ’3'...
th- ‘ "“‘

.2,
n:

:1.-
0
"r1

8?
(1 l
(J
F

as.-. an.
db\.3lki..

"“11 .‘
and. all:
\

.1 '
ma: pre

Instead
millage
Other .‘

OI d0;

 

rE-er, .

 

 

 

This dissertation examines the impact of a structured ACP educational program
intervention on selected personal influences on individuals’ decisions to engage in ACP.
The ACP program intervention, using the Respecting Choices® curriculum, was
designed to address common barriers to ACP and foster an environment of shared
decision-making to promote ongoing engagement in ACP. Programs were presented to
established groups of individuals in communities. During the programs, individuals were
encouraged to further engage in ACP; that is, to understand, reflect, identify and discuss
their preferences to help them formulate and make available plans for end of life care
(Hammes & Briggs, 2002b). Participants were given tools and strategies to continue
these conversations beyond the program with their significant others and health care
providers.

Studies of End of Life Planning

Previous ACP studies have not focused on ACP as a decision-making process.
Instead, researchers have focused almost exclusively on only one outcome of ACP: an
initiated advance directive (AD) that formally communicates an individual’s preferences.
Other features of the ACP decision-making process such as engaging in ACP behaviors
or deciding to engage in ACP are almost non-existent in the literature. The decision to
re-engage in ACP has also not been examined. Since the objectives of ACP are broader
than an AD (Kolarik, Arnold, Fischer, & Tulsky, 2002) and are best studied as a process
(Emanuel, 2003), this dissertation focuses on ACP as a dynamic decision-making process

that may be engaged in repeatedly.

:11

....

“cm-cc

b ' ...u‘-..
.

. .
7‘: .‘~fl\’1
an. all 1§1'

5" “.39 17‘
SLL-n-» ll.
- ..

Id
3
J‘

5,)
C7
0

 

 

 

Purpose of Study

The purpose of this dissertation is to add to the understanding of ACP as a
dynamic decision-making process by examining internal and external influences within
the decision-making context that may affect individuals’ decisions to engage and re-
engage in ACP.

Research Questions

This pre-post survey study of ACP is designed to answer the following questions,
as depicted in the Focused Model of the Decision to Engage in ACP in Chapter 2 (see

Figure 2, p. 23).

Before participation in a community-based ACP program:

1. What are the self-reported Personal Influences on Decision and Personal Decision
Factors of adults who have decided to engage in ACP?

2. Which variables from among the Personal Influences on Decision and Personal
Decision Factors are associated with the Decision to Engage in ACP?

3. Do the Personal Influences on Decision and Personal Decision Factor variables that
are associated with the Decision to Engage in ACP differ between participants with
previous ACP experience compared to participants without previous ACP
experience?

After participation in a community-based ACP program:

4. Are there changes from pre- to post-ACP program intervention in self-reported
ratings of enough knowledge, importance, ability, and likelihood to have an ACP

conversation?

'-
p.

extem;
decisic
Adi am
COHCCp
impetu
revieu

Undersz

“I” deg

 

5. Do the changes between participants’ self-reported pre- and post-ACP program
intervention ratings of enough knowledge, importance, ability, and likelihood differ
between participants with previous ACP experience compared to participants without
previous ACP experience?

Overview of Dissertation Study
Chapter 1 has presented the problems of approaching end of life without
preferences being known, common barriers to ACP, and the lack of studies that examine

ACP as a decision-making process. The purpose of this study is to add to the

understanding of ACP as a dynamic decision-making process by examining internal and

external influences on individuals’ decisions to engage and re-engage in ACP within the
decision-making context. Chapter 2 will introduce the Decision Process Model for

Advance Care Planning as the conceptual fi'amework for the study and define study

concepts based on this model. Chapter 3 will review the ACP literature to describe the

impetus for and key concepts of ACP as a dynamic decision-making process. Through a

review of descriptive and intervention ACP studies, this chapter will describe current

understanding of ACP, gaps in knowledge, and the significance of this study. Chapter 4

will describe the methods for the study: its context of the MMACPC Research Project,

its design, subjects, instruments, procedures, protection for human subjects, and plan for

data management and analysis. Chapter 5 will present the results of the study. Chapter 6

will provide a discussion of the results, including limitations and implications of the

study for nursing practice, education, and research.

I
model of
decision-
been 5m

ACP I35:

research
ladVance
concepu
Allied H
searched
Itffms Int
COnccprL

Planniné

dem0n5

 

impm‘t
Hamm.
dellber

model,

CHAPTER 2
CONCEPTUAL FRAMEWORK AND DEFINITIONS

The purpose of this chapter is to introduce a dynamic decision-making process
model of Advance Care Planning (ACP). Through analysis of existing ACP and
decisionsmaking models, the Decision Process Model for Advance Care Planning has
been synthesized by this investigator to provide a conceptual framework for ongoing
ACP research and clinical practice.

A literature search was performed to identify published journal articles reporting
research or a review of literature having a primary focus on ACP or end of life planning
(advance directives) in an adult population which included a depicted or described
conceptual model of ACP or decision-making. The Cumulative Index to Nursing and
Allied Health Literature (CINAHL), MEDLINE, and PsychINFO databases were
searched for peer-reviewed journal articles published from 1985-2006. Productive search

9, 6‘

terms included “advance directive + conceptual framewor , advance care planning +

9, 6‘

conceptual framework,” “advance care planning + theoretical models, advance care
planning + models,” and “advance care planning + decision-making + models.”
Conceptual Models of ACP
Interest in ACP gained momentum in the early 1990’s as research began to
demonstrate that a narrow focus on written advance directives was not effectively
improving communication or decision-making at end of life (Field & Cassel, 1997;
Hammes & Briggs, 2002b). In contrast, ACP emphasized the ongoing process of

deliberation and communication of end of life preferences (Emanuel, 2003). Conceptual

models of ACP began to appear in the literature around 1995.

10

Linus: V

Hz’van'
f. Bs.b¢.~

1‘10: IO 1
ICIL’IHC

the han

prefers

1999)

Models from Qualitative Studies

The earliest work used grounded theory and ethnographic methodologies to
describe the processes being used to make advance directive decisions. Hurley, Volicer,
Rempusheski, and Fry (1995) described the process of achieving consensus between
family members and health care providers related to treatment decisions for Alzheimer’s
patients. Rein et al. (1996) described the process used by inpatients to decide whether or
not to initiate an advance directive. Shidler (1998) described a system-focused,
interaction fi'amework to honor individuals’ treatment wishes that included all persons in
the long-term care resident’s treatment setting.

These qualitative studies were usefirl beginnings to describe components of the
process and structure of ACP decision-making. They also introduced the shared
decision-making context of ACP, in which individuals and others (such as invited family
members or health care providers) exchange information about medical conditions and
preferences to reach treatment decisions that are mutually agreeable (Frosch & Kaplan,
1999)

Models from Health Behavior Studies

The Health Belief Model (HBM), an example of a health behavior model, was
used as the conceptual framework for two ACP studies. The HBM focuses on
individuals’ motivations and beliefs about a threat to their health that influences their
likelihood of taking action to protect themselves. The hypothesis of the HBM is that
health-related action is dependent on three simultaneous personal factors: 1) having
sufficient motivation (or concern) so that the health issue becomes relevant; 2) believing

that there is a threat of a serious health problem; and 3) believing that following a health

11

reeommen-t
modiffing
only aliect
action (Petr
['51:
adults' pert
contrary to
an advance
banners an
medical co
influence 0
With findir
indicated I
dimension
AC
indll'ldual
HOWEVGI‘,
3Cli0n\\,‘
innUEnce-

 

combing:

recommendation will be beneficial (Rosenstock, Strecher, & Becker, 1988). In the HBM,
modifying factors such as cues to action or demographic, societal, or structural influences
only affect the perceived threat and do not have a direct effect on the likelihood of taking
action (Pender, 1996).

Using the HBM, Hamel, Guse, Hawranik, and Bond (2002) investigated older
adults’ perceptions of the likelihood and consequences of dying in a way that was
contrary to their wishes contrasted with the perceived benefits and barriers to completing
an advance directive. In an earlier study, VandeCreek and Frankowski (1996) examined
barriers and benefits to initiating a living will in a population of outpatients with ongoing
medical conditions. These studies found that perceived barriers were a significant
influence on people’s plans to complete an advance directive or living will. This concurs
with findings from a systematic review of a decade of studies based on the HBM, that
indicated that perceived barriers have been one of the HBM’s most significant
dimensions for prediction of health behaviors (J anz & Becker, 1984).

ACP studies using the HBM contribute to understanding the connection between
individuals’ perceptions—particularly of barriers—and their likelihood to take action.
However, influencing factors such as personal, societal, or structural influences or cues to
action—which are precursors to the concept of decision-making context—do not directly
influence the likelihood of action. Nor is there a link between actions (behaviors) and
outcomes.

Composite Model
Pearlman et al. (1995) proposed a composite conceptual model of ACP that

combined the concepts of the HBM with the concepts of stages of change (Prochaska &

12

J
1.; .
‘—’

4)

 

 

DiClemente, 1983), self-efficacy (Bandura, 1977), and human information processing
(Newell & Simon, 1972). In this model, ACP is viewed as having both deliberative and
communicative processes that lead to immediate and future outcomes. The environment
also plays a role, influencing an individual’s beliefs, information processing, behaviors,
and outcomes.

Like the HBM, this model made the link between personal influences and

subsequent behaviors. But it did much more:

1) It depicted ACP as a deliberative and communicative process, similar to
the models from the qualitative studies.

2) It identified aspects of the decision-making context and the influence of
stages of change on the process of ACP.

3) It introduced the role of beliefs about ACP, perceived self efficacy,
deciding to do ACP, and processing of information as additional personal
factors influencing the process.

4) It identified specific ACP behaviors of shared communication and
documenting preferences.

5) It linked influences, behaviors, and outcomes.

Although this model integrates many essential concepts of ACP, it does not fully
emphasize the decision-making focus of ACP.
Decision-Making Models
Decision-making assumptions and models were described in six ACP papers.
Two additional models of generic decision-making contribute to further understanding of

ACP as a decision-making process.

13

ACP Studies Based on Decision-Making Models

Hurley et al. (1995) examined the roles of health care professionals in the process
of surrogates’ decision-making. They described elements of the community context
(such as the patient, the surrogate, the health care provider, and health care organization),
preparation and catalysts for the interaction, and the outcomes and consequences of the
resulting decisions. Rein et a1. (1996) described an individual’s process of evaluating an
illness, establishing priorities, considering the implications of advance directives (ADS),
and finally coming to a decision. The decision could be initiating an AD, choosing not to
initiate an AD, or postponing making a decision about an AD.

In his paper describing individuals who may contemplate assisted suicide or
active euthanasia, Cicirelli (1997) posited the assumption that the decision made would
be related to the individual’s psychosocial characteristics and deliberation about options
and consequences. McAdam, Stotts, Padilla, and Puntillo (2005) modeled perceptual and
ethnic factors that affect Filipino individuals’ decisions to select or reject an AD. Decker
and Reed (2005) explored the effects of developmental stage and context on older adults’
plans for end-of-life treatment decisions.

These papers clearly acknowledge ACP as a decision-making process but also
demonstrate features seen in the composite model. Similarities between these papers and
the composite model include: 1) the role of the decision-making context; 2) the effect of
influences on the decision, interaction, or behavior; and 3) outcomes of decision-making.
These papers also add details about the steps of deliberation in ACP decision-making.
However, they do not capture the concepts that ACP decision-making involves both

construction and communication of preferences and that it is a dynamic process.

14

features <
hating- 31
Contents
provider
patient IS
Styles: e1
Double-i
dlnantic
factor of

Success (3

Pl 1’0! stud
deClSIOn-
increasin

pal'llcip“.

 

Oldecig
Slilll of

i”lilac: V

Interactive Decision-Making Framework

In 2001, Pierce and Hicks proposed an interactive decision-making framework,
based on expected utility and information processing theories. They described decision-
making as an interaction of patient and contextual factors that are catalyzed by the unique
features of the decision problem. In this model, the decision problem is described as
having alternatives (choices), complexity, probability, and outcomes or consequences.
Contextual factors include actual risk, urgency, and time frames as well as patient-
provider interaction, enviromnental stressors, cognitive demands and information. The
patient is described as having values (utilities); preferences for participation and decision
styles; expectations and perceptions of risk; and a physical and psychological state.
Double-headed arrows linking the decision problem, context, and patient demonstrate the
dynamic nature of decision-making. An assumption of the model is that the contextual
factor of the interaction between patient and health care professional is crucial to the
success of this shared decision-making process.

Briggs et a1. (2004) used this interactive decision-making model to frame their
pilot study of a patient-centered ACP intervention. Pairs of patients and their surrogate
decision-makers were interviewed by health care professionals with the intent of
increasing their knowledge of the decision problem, assessing their preferences for
participation and decision style, and evaluating the congruence of preferences and levels
of decisional conflict between patient and surrogate. This study also confirmed that the
skill of the health care professional interviewer was a critical link to promoting effective

interaction and successful shared decision-making.

15

...9 1r

‘5.& ..-
.o- --_~.;f1'“'<:
)__.L.<._ .‘a 55'

o. g
4 - '1.
273.373 “....

.
’ufi— 3-.
' perk.
1

(I:

OI".
ddlSLJE-lliir

9‘ ’. ‘
5.1.5. b1 B"=

an.
5

; D. '.
.om L115 11121

$0“ the rel:

A d5;
Bmmwkk
this? Ilium
infomazion
bfhftl‘jors all

'1 ” v .

    
  

and Vaers ;

making Sic:

The interactive decision-making framework is similar to the ACP qualitative
studies in its recognition of the importance of shared decision-making. Like the
composite model, it highlights the importance of patient and contextual factors
throughout the decision—making process. It also clearly identifies the aspects of the
problem which need to be deliberated during decision-making. This model adds an
emphasis on interaction; that is, communicating preferences with all those involved in
decision-making. It also recognizes the dynamic nature of such interaction. The pilot
study by Briggs et al. (2004) demonstrates this model’s fit with ACP concepts described
from the literature and encountered in clinical practice. However, it does not clearly
show the relationship of decision, behaviors, and outcomes in decision-making.

Decision Process Model

A decision process model (Rothert et al., 1997) with roots in the Lens Model
(Brunswick, 1956), expected utility theory (Shoemaker, 1982), and social judgment
theory (Hammond, Stewart, Brehmer, & Steinmann, 2002), examines how a patient’s
information and values combine to result in a decision that influences subsequent
behaviors and outcomes. Wills and Holmes-Rovner (2006) offered a simplified model of
patient decision-making based on the decision process model that shows how information
and values interact to create preferences that dynamically inform the remaining decision-
making steps: the decision, behaviors, and outcomes. These steps occur within a
decision-making context that surrounds the entire decision-making process. This
decision-making model is useful for guiding a study of ACP because it 1) demonstrates
the relation of influences, decision, behaviors, and outcomes, 2) recognizes preferences

as a key element, 3) captures the notion that decision-making is a dynamic process, and

16

concept.
relevant
to coast:
in] an;

I ‘I 1‘.‘
111:1: .1 .

Emphas
Decisio

'C‘Cisioi

 

4) acknowledges the decision-making context. Because the Decision Process Model
provided the simplest decision-making framework, it was chosen as the basis of the
conceptual model for this study.

In summary, each of the conceptual models reviewed in this chapter have some
relevance for a comprehensive model of ACP as a dynamic decision-making process used
to construct and communicate preferences. The structure and process of ACP decision-
making was described in the qualitative studies; individuals’ perceptions of barriers and
their likelihood to take action in the health behavior model studies; and the connection
between personal influences, behaviors, and outcomes from the composite model. The
ACP studies using decision-making models added details about the steps of ACP
decision-making while the ACP study based on the interactive decision-making model
emphasized communication and the dynamic nature of such interactions. Finally, the
Decision Process Model provided a useful fiamework to capture all of the elements of the
decision-making process within the decision-making context.

The Decision Process Model for Advance Care Planning

The Decision Process Model for Advance Care Planning (see Figure 1) is an
evolving adaptation of the decision process model applied to ACP. This conceptual
model has been synthesized over the past three years from reviews of ACP conceptual
models, decision-making models, and collective experiences in professional practice with
patients, families, and health care providers encountering end of life decisions. Building
on key concepts described for decision-making models, it depicts the breadth of how

individuals decide to engage in ACP with resulting behaviors and outcomes.

17

 

 

DECISION
DECISION TO ENGAGE IN ACP :

 

BEHAVIORS
: BEHAVIORS OF ENGAGING IN ACP

 

 

 

 

AA

 

   

. Understanding
decisions to be made

 

 

Key: ACP 2 Advance Care Planning

PERSONAL /NFLUENCES ON DEC/S/ON
(Internal Influences) 11

Information
. Knowledge of ACP

I Categorical definitions

- Enough knowledge to

discuss my views
0 Knowledge of life-sustaining treatment
- Knowledge of comfort care
. Knowledge of current and future
health states

Individual Characteristics

. Education

. Marital status

0 Race/ethnicity

. Health status

- Religious/spiritual preference
0 Religious/spiritual importance

 

Values

0 Important to have a
discussion

. Basic life values

0 Quality of life values

Prior Experiences
0 Making decision for
someone else
. Having an AD
. Having a HCA
- Having conversations &
Quality of conversations
-HCA
ISignificant others
IHCP

AD = Advance Directive

. Reflecting on values
and goals
. Identifying
preferences for
I HCA
- Instructions
- Goals
. Discussing
preferences with

 
   
    
  

- Significant
others
- HCP
. Formulating a plan
. Making the plan
available

    
 

 

I Assessing Motivators or Barriers

 

 

 

 

 

AI

 

._____.___> DECISION FACTORS

Personal Factors (Internal Influences),
Self—efficacy '
° Able to start discussion with others
- Likely to have conversation with others
- Decision Self Efficacy
Decisional Conflict
Anticipatory Decisional Regret
Stage of Decision
(‘- " Factors (External Influences)
Community Influence
- Family values
- Family behaviors
- HCP values
- HCP behaviors
- Peer values
- Peer behaviors

- Residency

- Health care systems

HCA = Health Care Agent

Figure l. The Decision Process Modelfor Advance Care Planning

Community Environment

. Shared decision—making
- Community-based group

- Health services research

HCP = Health Care Provider

 

 

OUTCOMES
OUTCOMES OF ENGAGING IN ACP

Products
Preferences
. Initiated AD
- HCA (DPOA-HC)
- Instructions for care
Communicatjflqfirefefinees
. Conversations about preferences
I HC
. Significant Others
- HCP
Presentecdflan
. Written plan
. Verbal plan

 

Process
. Satisfaction
I Satisfaction with Decision
- Satisfaction with shared decision—making
. Confidence
I Confidence in HCA ability to make decisions
- HCA confidence in ability to make decisions
0 Decision regret

Promises
o Congruence of decisions with preferred
care goals

. Satisfaction
I Satisfaction with shared decision—making
during end of life care
- HCA satisfaction with shared decision—
making during end of life care

. Decision regret in bereavement

 

 

.__.f

I

 

 

INTERVENTIONS I

 

(External Influences) J

I

DECISION-MAKING CONTEXT

 

 

 

 

 

 

I.
7‘ IL; 83!
1.11» ‘-
..
'Au-V ’) -
DeSIDaLL

irflaences

F0:
choice p30
Care. This
context. It
DnnmnF

Imen'E‘ntio

 

 

The major components of the model include: 1) the Decision to Engage in ACP,
2) the Behaviors of Engaging in ACP, 3) the Outcomes of Engaging in ACP, and 4) the
Decision-Making Context. The decision-making context includes internal (e.g., personal
influences on decision, personal decision factors) and external (e. g., community decision
factors, interventions) influences on the Decision to Engage in ACP. The arrows linking
the decision-making context influences, the decision to engage in ACP, the behaviors, the
outcomes, and the decision-making context demonstrate that ACP is a dynamic decision-
making process.

The Decision to Engage in ACP

For the purposes of this study, the Decision to Engage in ACP is defined as the
choice people make to involve themselves in exploring their preferences for end of life
care. This decision is shaped by internal and external influences of the decision-making
context. Internal influences include Personal Influences on Decision and Personal
Decision Factors. External influences include Community Decision Factors and
Interventions.

Behaviors of Engaging in ACP

The behaviors of engaging in ACP are understanding (Briggs et al., 2004),
reflecting (Pearlman et al., 1995; Pearhnan, Starks, Cain, & Cole, 2005), identifying
(Emanuel, 2000; Hammes & Briggs, 2002b), and discussing preferences (Barnard, 2002;
Emanuel, 2000) to help in formulating (Briggs et al., 2004; Pearlman et al., 1995) and
making a plan available for decision-making (Grimaldo et al., 2001; Richter et al., 1995;
Schneiderman, Pearlman, Kaplan, Anderson, & Rosenberg, 1992; Sulrnasy, Song, Marx,

& Mitchell, 1996; J Teno et al., 1997).

19

menu
Edd

ACPe

8UlOn0
OUICOr
Nth
ICIlUCa
dunfi
PIOCcs.

reSpo:

 

In Figure 1, the behaviors of engaging in ACP are depicted inside a circle to
reflect two observations from clinical practice. First, one behavior often leads to the
next, but all seem to be necessary to fully engage in ACP. If any of the behaviors are
missing, the resulting outcomes may be less optimal (Ditto et al., 2001; Hammes &
Briggs, 2002b; Kolarik et al., 2002). Second, these behaviors do not necessarily happen
sequentially. Instead, people engaging in ACP behaviors ofien move back and forth
between the steps over time, refining their thoughts and plans as they gain more
understanding, hear others’ perceptions, and experience the outcomes of ACP. Assessing
the motivators or barriers which affect these behaviors (Hamel et al., 2002; Rein et al.,
1996; VandeCreek & Frankowski, 1996) can provide useful direction for the design of
ACP educational program interventions which teach participants how to engage in ACP.

Outcomes of Engaging in ACP

In the composite model, Pearlman et al. (1995) described outcomes related to
ACP in two timefrarnes: immediate and at the time of mental incapacity. Immediate
outcomes included a coherent mental model, shared understanding, documents, sense of
autonomy and well-being, and decreased burden on proxy (decision-maker) and family.
Outcomes at the time of mental incapacity included a decreased burden on proxy, family,
and health care providers; having future medical care matching patient wishes; and
reduced health care expenditures. In Figure l, the outcomes of engaging in ACP are
classified by type rather than by timefrarne. Because ACP is a dynamic decision-making
process that may occur repeatedly over varied amounts of time, this model focuses on

responses that result from engaging in ACP behaviors.

20

the rne‘.L
1,"de i:
msnue
should I
note die

lfidh;

ofinvo'

IESpons

[he PIOI
abouth
PERU:

such as

 

 

The first response or outcome is Products: these are the actual preferences and
the method(s) used to communicate those preferences. Specific examples of such
products include an initiated AD, conversations about preferences, and a presented plan
in written or verbal form when health care is needed. In clinical practice, these products
should be discoverable during an assessment of ACP decision-making. It is important to
note that the initiated AD is not the only product of interest (Kolarik et al., 2002).
Ideally, all of these products are present following engagement in ACP.

The second outcome type is Process, which is defined as the affective responses
of involved individuals to the process of engaging in ACP decision-making. Such
responses may include satisfaction, confidence, or regret. ‘

The third outcome, Promises, refers both to effects of and affective responses to
the products and process of engaging in ACP. A desired effect of ACP is that decisions
about health care made after engagement in ACP are congruent with the goals of
preferred care (Briggs et al., 2004; Gillick, 2001). Affective responses to these decisions,
such as satisfaction or regret, provide an additional means of evaluation.

Decision-Making Context

The Decision-Making Context includes the decision, behaviors, outcomes and
influences (internal and external) that shape the decisions people make (Wills & Holmes-
Rovner, 2006). Personal Influences, internal to the individuals, include information
(Davison & Degner, 1998; Pearlman et al., 1995), individual characteristics (Cicirelli,
1997), values (Pearlman et al., 1995; Pierce & Hicks, 2001), and prior experiences (Rein
et al., 1996). Decision Factors, including internal personal and external community

factors (Pierce & Hicks, 2001), also influence the Decision to Engage in ACP.

21

kneri
decsz

PLJm

Inner

comn
techn
Chan
Succc

pnon

dnser

 

 

Interventions, which are external influences, may affect any of the other influences or the
decision, behaviors, and outcomes. The Decision Process Model for Advance Care
Planning includes the decision to engage in ACP, behaviors, outcomes, and influences
(internal and external), all within the decision-making context.
Dynamic Decision—Making
Optimal ACP is not a one-time-only process. It is ongoing deliberation and
communication that regularly reviews and possibly revises the plan in response to
technological, developmental or health status changes (Hammes & Briggs, 2002b).
Changes in the decision-making context influence the decision to engage again in ACP.
Successfirl ACP results in a plan that enables effective decision-making, even as
priorities or preferences change (Barnard, 2002; Emanuel, 2000; Pearlman et al., 1995).
The Decision Process Model for Advance Care Planning is ideal for guiding this
dissertation study for the following reasons:
1) It demonstrates the relation of the decision, behaviors, outcomes, and
decision-making context involved in ACP.
2) It identifies the modifying potential of personal influences and decision
factors on the Decision to Engage in ACP.
3) It delineates the internal and external influences included in the decision-
making context.
4) It demonstrates how ACP is an ongoing dynamic decision-making
process.
A focused model of the Decision to Engage in ACP was created to depict the variables

that are conceptually and operationally defined for this study (see Figure 2).

22

 

 

 

 

Noe. E owmmem 8 Samoan 06 no 338 consoom .N onE

San 5095.32. ecu 10.x. I c 323.5 0.30 5301 u no... E004. 200 5301 I (0.... 2502.0 cocu>u< I O< oEccuE 0.30 oocu>u< I n.0< “>0!

 

l ...xuhzoo oz.x<2-zo.2uuo _T
L

 

 

Ede—acct 10< game‘s—55500
Aoocoacc. .IEouxmv

ZOEZM> QWLZ‘

 

 

 

           

020 2.6050. con coioou Gains. .
cox.
8.50 Eco—.290.
(or.
ace—«Isaac ..o 3:030
d 0:0. .Eo>coo 05>. 0
(GI I .csiI
O< ca 9.3-...

     
 

    

    

 

4 w

c9050 5.3 coaogcoo go... 2 29.3 o
«9.050 5.3 :23:an :8. o. 0.2 .
g

 

 

 

 

5.0.33.8
a 0):: 2 «55.2.5. o
38:03:. .anE: a
89.05 29986 q<zommmm o buuoowu
«Egg

c250.)
>E canon—u on 023.305. zoaocw .
g

A V Sauces—E. .quEc

ZO‘QONQ ZO ”NUZMuncfi 4<20mmml

 

 

- - , no... 2. mo<ozm 9 206.03
02200.50 8.055%. 20.20!—

 

 

 

 

 

 

 

 

 

 

23

IHVC

part

this

pIO‘1

AC P

 

Conceptual Definitions of Focused Model Concepts

Decision to Engage in ACP

The Decision to Engage in ACP is defined as individuals’ willingness to
involve themselves in exploring their preferences for end of life care through
participation in a community-based, ACP education program. All participants in
this study made the decision to engage in ACP by virtue of attending the ACP

program intervention.

Behaviors and Outcomes
For the purposes of this study, neither the behaviors nor outcomes of engaging in
ACP will be examined.
Decision-Making Context
The Decision-Making Context is defined as the decision, behaviors, outcomes and
influences (internal and external) that shape the decisions people make (Wills & Holmes-
Rovner, 2006). The influences of interest in this study include internal influences of
Personal Influences on Decision and a Personal Decision Factor and the external
influence of a Community-Based ACP Program intervention.
Personal Influences on Decision
Personal Influences on Decision are defined as variables relating to the individual
that may act independently or interact with each other to modify the Decision to Engage
in ACP. These variables include Information, Individual Characteristics, Values, and
Prior Experiences.
Information. Information is defined as decision-relevant data inputs (Wills &

Holmes-Rovner, 2006). Participants’ information about ACP is measured as “Enough

24

knot

l
r

 

 

knowledge,” which is defined as the participants’ self-report of whether or not they have
sufficient information to discuss their preferences.

Individual characteristics. Individual characteristics of focus in this study include
“Age” in years and “Gender,” male or female.

Values. Values reflect the importance individuals place on information,
experience, and the context of decision-making (Wills & Holmes-Rovner, 2006). In this
study, participants’ value of ACP is measured as “Important to have discussion,” which is
defined as their self-report of their perception of the necessity of having conversations
about preferences for end-of-life care.

Prior experiences. Prior experiences are defined as events, circumstances, or
processes that have already occurred in individuals’ lifetimes. In ACP, relevant prior
experiences may include having an AD (Advance Directive), having a HCA (Health Care
Agent), participating in quality conversations about end-of-life care preferences, and
making a decision about life-sustaining treatment for another person. Participants in this
study will indicate the occurrences of these experiences prior to the time of their
participation in the ACP program.

Having an AD is defined as the participant’s self-report of executing a statutory

document for use at a future time when personal capacity to make informed health

care decisions is lost.

Having a HCA is defined as the participant’s self-report of choosing and legally

appointing a person to be the substitute health care decision-maker (Post,

Blustein, & Dubler, 1999).

25

Peri

mo

ally:

 

 

Having conversations is defined as the participant’s self-report of discussing
preferences for end-of-life care with others. The roles of these “others” may be a
health care agent (HCA), significant others, and a health care provider (HCP).
Quality of conversations is defined as participants’ self-reports of the
thoroughness oftheir discussions of preferences with others.
HCA. As noted previously, the HCA is a person who has been chosen and
legally appointed to be the substitute health care decision-maker (Post et
al., 1999).
Sigificant others are defrned as persons who play an important role in the
life of an individual (O'Toole, 1997).
A EC}: is defined as a licensed professional who provides health care to
individuals within a health care setting. ’ These licensed professionals may
include physicians, mid-level providers (e.g., Advance Practice Nurses,
Physician Assistants), registered nurses, dietitians, social workers,
therapists, and chaplains.
Making decision for someone else is defined as the participant’s self report of an
event or process of communicating a choice related to life-sustaining treatment on
behalf of another person who is incapable of making a choice.
Personal Decision Factors
Personal Decision Factors are defined as variables relating to the individual that
modify an individuals’ willingness or ability to make a choice between two or more

alternatives.

26

sum
Illi’OI'
Feldi

pros

thr
Depa
pEOp
cOmr
in th-‘

em?

 

 

Self-efi‘icacy. Self efficacy is defined as an individual’s perceived confidence to
perform certain behaviors (Bandura, 1986). In this study, participants’ self-efficacy will
be measured by their self-ratings of their perceived confidence in their “Ability to start an
ACP discussion” and the “Likelihood of having an ACP conversation.”
Community-Based ACP Program Intervention

The Community-Based ACP Program Intervention is an example of a decision
support intervention. Decision-support interventions are strategies used to promote
informed, values-based decision-making (A. Clarke, Jacobsen, O'Connor, Stilwell, &
Feldman—Stewart, 2005). This dissertation describes the impact of an ACP educational
program intervention that was presented to established community groups.

Community-based. A community-based intervention is directed toward
individuals and families within a community and designed to meet the needs of people
where they live, work, worship, go to school, and receive healthcare (Calvin College
Department of Nursing, 2003). The intervention used in this study was presented to
people in established community groups, which are defined as organizations located in a
community in which members naturally participate (P. Clarke et al., 2005). Participants
in this study were from established community groups (faith, health-related layperson, or
employee) in the Greater Lansing and Grand Rapids, Michigan communities.

ACP Program. Health education is defined as developing and providing
instruction and learning experiences to facilitate voluntary adaptation of behavior
conducive to health in individuals, families, groups, or communities (O'Toole, 1997).

For this study, the ACP program is defined as a health education information sharing

27

 

fill

session about ACP, presented by a certified Respecting Choices® ACP Facilitator to
voluntary participants who were members of an established community group.
Dynamic Decision-Making

Dynamic Decision-Making is defined as a decision-making process that may be
engaged in repeatedly in response to changes within the decision-making context. The
arrows linking the decision-making context influences, the decision to engage in ACP,
the behaviors, the outcomes, and the decision-making context depict ACP as a dynamic
decision-making process. However, participants’ reports of prior experiences with ACP
can also be used to demonstrate that pe0ple are re—engaging in ACP.

For this study, the dynamics of ACP decision-making will be represented as
participants’ ACP experience. As noted in Chapter 1, previous studies of ACP have
focused almost exclusively on initiation of an advancedirective or appointment of a
health care agent. Therefore, for this study ACP experience is defined as a participant’s
self-report of the prior experiences of having an Advance Directive (AD) or having a
Health Care Agent (HCA) or both.

Chapter Summary

The conceptual fiamework for this dissertation study was synthesized fiom
reviews of conceptual models of ACP, decision-making models, and professional practice
experiences with patients, family members, and health care providers. The Decision
Process Model for Advance Care Planning provides an ideal framework for this study
because it depicts ACP as an ongoing dynamic decision-making process involving a
decision, behaviors, outcomes, and internal and external influences within the decision-

making context. The focused model of the Decision to Engage in ACP identifies the

28

Desist
Decisr
nithir.

by AC

 

 

 

Decision to Engage in ACP; the internal Personal Influences on Decision and Personal
Decision Factor and external Community-Based ACP Program intervention influences
within the Decision-Making Context; and the dynamics of decision-making represented

by ACP Experience.

29

CHAPTER 3
LITERATURE REVIEW & SIGNIFICANCE OF STUDY

This chapter presents a brief overview of the literature related to the impetus for
ACP and key features of ACP, followed by a review of studies of ACP, characterized by
this investigator, that have laid the foundation for this dissertation. Results of ACP
studies, organized by this study’s research questions, will then be reviewed to
demonstrate current understanding of influences on the decision to engage in ACP. The
significance of this dissertation study to add to the knowledge base of ACP decision-
making will be discussed throughout the chapter.

Advance planning for end of life preferences has been studied intensively over the
past 15 years. ACP literature indexed in CINAHL and MEDLINE was reviewed to
identify peer-reviewed journal articles that reported systematic reviews and clinical trials
from 1985 to 2006. Govermnental agency websites (e.g., the Department of Health and
Human Services, the Center for Disease Control, the National Center for Health
Statistics) were also searched for national survey studies which included items about
advance directives (ADs) or ACP.

The Impetus for ACP

The impetus for ACP first came fi'om public dissatisfaction with the quality of
end of life care (Field & Cassel, 1997). Advances in technology and pharmacology have
influenced societal attitudes and beliefs about death, dying, and health care at the end of
life (Daly, 2006). While people hope for available treatments to cure their injuries or
illnesses, they often simultaneously fear the prospect of needless suffering at the end of

life, either from lack of care or fiom care that is ineffective or harmful (Field & Cassel,

30

1997). In America, with its predominant ethos of autonomy, these hopes and fears have
translated into a desire for rights of self-determination and control of treatment choices.

Control of treatment choices is also important to good stewardship of health care
resources. Americans fear the continuous increases in health care costs and the dire
predictions of economic and health care provider shortages as the population ages.
Because current treatments can prevent some immediate causes of death and effectively
treat slower causes of death, the aging baby boomer generation is predicted to live much
longer than previous generations (Field & Cassel, 1997). Costs related to end-of-life
care will not only be borne by health care organizations but also by the general public
through Medicare, Medicaid, the Veterans Administration, and other public programs
(Field & Cassel, 1997). Thus, it was hoped that documentation of preferences would lead
to reduced costs by preventing undesired treatments at the end of life.

In November 1991 the Patient Self Determination Act (PSDA) was implemented,
requiring all hospitals, nursing homes, and health care programs that receive federal I
fimding to ask patients about advance directives (ADS), make AD education available,
and incorporate AD information into medical records (Crane, Wittink, & Doukas, 2005).
ADS are statutory documents executed to communicate an individual’s health care
preferences for use at a future time when the individual’s capacity to make informed
health care decisions is lost (Post et al., 1999). Individuals, health care professionals,
and policy makers hoped that ADS would be an effective strategy to address concerns
about respect for individuals’ rights and dissatisfaction with patient and family treatment
at end of life (Galambos, 1998) as well as reduce resource use (Taylor, Heyland, &

Taylor, 1999).

31

However, summary reports from Last Acts (Kaplan & Peres, 2002), the Institute
of Medicine (Lunney, Foley et al., 2003), and the Agency for Healthcare Research and
Quality (Kass-Bartelmes et al., 2003; Lorenz et al., 2004) confirm that, on average, ADs
are unused by at least 50% of the population. Researchers have also provided evidence
that the single outcome of a completed AD has not been effective to identify and honor
most people’s end of life preferences (Ditto et al., 2001; Hofrnann et al., 1997 ; Kolarik et
al., 2002; J Teno et al., 1997).

Advance Directives have also not been successful as a financial mechanism to
reduce the costs of providing health care at the end of life. Although it was hoped that
advance directives would be a more ethical alternative for decreasing cost than rationing,
the actual relation of advance directives to cost savings has been mixed. In a systematic
review, Taylor, Heyland and Taylor (1999) reported that although three retrospective
studies Showed some cost savings, prospective studies—with highly controlled
experimental designs—demonstrated no evidence of cost savings attributable to advance
directives. They concluded that little evidence exists to support the hypothesis that

advance directives reduce resource use by hospitals.

These lackluster results provided an additional incentive to develop the concept of
ACP. In contrast to a focus on completion of a static legal document (AD), ACP is a
dynamic decision-making process with features of deliberation, communication, and
shared decision-making focused on achieving several outcomes (Field & Cassel, 1997).
In ACP, an AD can be a useful tool, but is only one of many means to achieve desired

goals (Field & Cassel, 1997; Kolarik et al., 2002).

32

Key Features of ACP

The key features of ACP are the following: 1) a deliberative process, 2) a

communicative process, 3) a shared decision-making process, and 4) a dynamic process.
A Deliberative Process

Engaging in ACP is a deliberative process. To understand decisions that may
need to be made, people require information about likely firture health conditions and the
benefits and burdens of various treatments for those conditions (Briggs et al., 2004).
Deliberation also includes personal reflection on values and goals to clarify which
benefits and burdens are acceptable (Pearlman et al., 1995; Pearlman et al., 2005).
Structured worksheets, such as living will or AD forms, can be helpful to identify
preferences (Emanuel, 2000), such as selecting a decision-maker or specifying
instructions (Hammes & Briggs, 2002b). The deliberative process, therefore,
encompasses the ACP behaviors of understanding, reflecting, and identifying
preferences.

A Communicative Process

Engaging in ACP is also a communicative process. People take the opportunity to
talk about end of life issues (Emanuel, 2000) and build up relationships with others in the
process (Barnard, 2002). In ACP, any form of communication (verbal, written, formal,
informal) is acceptable as a plan, as long as it clearly communicates the results of the
individual’s deliberative process (Pearlman et al., 1995). However, formulating legal
documents, such as an AD, does have a legitimate role (Emanuel, 2000). These
documents serve to “memorialize the conversation” about end of life preferences (C.

Thomason, personal communication, January 25, 2006). The goal of ACP

33

ICC :

Slut

Ifie

0p?

Opn

 

on

communication is shared understanding and a commitment to support the individual’s
preferences (Briggs et al., 2004). The communicative process of ACP includes the
behaviors of discussing preferences, formulating a plan, and making the plan available.

Many researchers overlook the element of making the plan available; however,
health care providers in professional practice readily identify this aspect of ACP
communication as important. McAdam et al. (2005) noted that people’s fears of not
receiving adequate care once an AD is presented are a deterrent to engaging in ACP. In
studies that do report a presented plan, the plan is usually located or documented in the
patient’s medical record or files (Grimaldo et al., 2001; Richter et al., 1995;
Schneiderman et al., 1992; Sulmasy et al., 1996; J T eno et al., 1997).

A Shared Decision-Making Process

Effective ACP is a cooperative effort between the individual and selected others
(Field & Cassel, 1997) to make decisions for which there are several available treatment
options, no obvious right or wrong choices, and often uncertainty about the impact of the
options on the well-being of those involved. Such decisions are defined by Frosch and
Kaplan (1999) as ideal shared decision-making opportunities.

Shared engagement in ACP ideally includes all four of the essential
characteristics of shared decision-making, as described by Charles, Gafni and Whelan
(1997). First, it involves at least two parties: the person and their invited others—often
family members and health care providers (Hurley et al., 1995). Secondly, the parties
participate in the process of decision-making through discussing the person’s preferences
(W einer & Cole, 2004) and formulating the plan. Thirdly, sharing of information must

occur; not only of medical information, but also of the person’s unique circumstances,

34

f0
,_
I:

C].
(V
.9.

IE?
or '
111.3

C02

-.1

P83

ch

 

 

values, and life priorities (Lee, Back, Block, & Stewart, 2002). Finally, both parties
agree to the decision by committing to support the plan, which is made available for
decision-making.

Conversations that occur during shared engagement in ACP are especially
important since advance planning decisions cannot be assumed to be clear, unambiguous,
or “ri t” (Barnard, 2002). Conversations may clarify perspectives, values and goals so
that decision makers have direction for situations not captured in a checklist or a
contrived scenario. Some people choose not to prepare written instructions or leave
instructions deliberately vague so that their appointed decision makers have permission to
respond to the situations at hand (Hawkins, Ditto, Danks, & Smucker, 2005). In some
cases ACP conversations reveal individuals who do not want to actively participate in
decision-making about their personal situations (Field & Cassel, 1997).

Active participation in decision-making is not always desired (Hawkins et al.,
2005). Lee et al. (2002) identify people for whom this may be true: 1) individuals who
prefer less control over decision-making, perhaps because of culture, ethnicity, or age;
2) patients who have had no experience with a condition or treatment; 3) acutely ill
patients without the physical or emotional resources to consider risks and benefits of
choices; and 4) patients with fears of regret. In addition, Tilden et al. (2001) found that
when decisions are burdensome, some people prefer to defer decision-making to others.

During shared engagement in ACP, the commitment is made to honor the best
available “portrait” of the person’s desires (Emanuel, 2000), recognizing that it may

evolve over the trajectory of the person’s illness, injury, or lifespan (Lunney, Foley et al.,

35

"i-i

r11
’3

2003; Weiner & Cole, 2004). This includes respecting people’s preferences to share and
sometimes defer decision-making responsibilities.
A Dynamic Process

Because people’s preferences and goals may change over time, ACP is a decision-
making process that is best engaged in repeatedly (Field & Cassel, 1997). Effective ACP
is thought to include ongoing deliberation and conversation that regularly reviews and
possibly revises the plan in response to technological, developmental or health status
changes. Successful ACP results in a plan that enables effective decision-making, even
as priorities or preferences change (Barnard, 2002; Emanuel, 2000; Pearlman et al.,
1995). However, studies of dynamic (repeated) ACP decision-making have not yet
appeared in the ACP research literature.

Studies of Advance Care Planning

For the past 20 years the shortcomings and the efforts of America’s health care
system to provide quality care at the end of life have been publicized (B. A. Brown,
2003; Field & Cassel, 1997; Kaplan & Peres, 2002; Kass-Bartelrnes et al., 2003; Lorenz
et al., 2004). This has led researchers, especially Since the 1991 federal Patient Self
Determination Act (PSDA) was enacted, to study and report the impact and outcomes of
efforts to promote ACP (Crane et al., 2005). The pre-PSDA five year Study to
Understand Prognoses and Preferences for Outcomes and Risks of Treatment
(SUPPORT) (Connors et al., 1995) is reviewed first because it represents a benchmark in

end of life decision-making research.

36

 

10

do

Ili

 

 

The SUPPORT Study

In 1989, data collection began for Phase I of the SUPPORT study, a two-year
prospective observational study to describe the process of decision-making and patient
outcomes for seriously ill, hospitalized patients (Connors et al., 1995). This study,
including 4301 patients from fiveteaching hospitals in the United States, documented
deficiencies in physician-patient-family communication and in honoring end of life care
preferences. Phase II, a two-year randomized controlled trial of an intensive intervention
to improve communication and decision-making, began in 1992 and included 4804
patients. The intervention included 1) sophisticated patient prognostic reports provided
to the patient’s physicians and 2) a skilled nurse, the SUPPORT nurse, to elicit and
document patient and family preferences for treatment. Surprisingly, although the Phase
II intervention resulted in anecdotal positive outcomes, it did not significantly improve
the primary outcomes related to physician-patient-family communication about end of
life preferences. The SUPPORT study validated public and professionals’ concerns that
patients’ preferences were not being communicated or honored and stimulated much
research activity to more effectively accomplish this goal.

Studies related to the Patient Self-Determination Act (PSDA)

In 1993, the Department of Health and Human Services conducted an
observational study of 72 randomly selected facilities in Six states to evaluate
implementation of the PSDA. It demonstrated that 21% of patients in hospitals, nursing
facilities, and home health agencies had ADS. Two thirds of patients interviewed had
some understanding of ADS and this understanding was influenced by receiving

information about ADS. Hopp (2000) reported data from the Asset and Health Dynamics

37

It

It;

 

 

of the Oldest Old (AHEAD) study, also conducted in 1993, which surveyed people born
in 1923 or earlier. Of the 520 participants, 95% stated they had someone they trusted to
make medical decisions but only 48.8% had talked with someone about their preferences
for care. Just over 37% reported having an AD (20% having both a Living Will (LW)
and Durable Power or Attorney for Health Care (DPOA-HC); 9% having a LW only; 8%
having a DPOA-HC only).

J. Teno et al. (1997) analyzed the SUPPORT study data from Phase I (pre-PSDA)
and Phase H (post-PSDA) and found no significant differences between patientS’
knowledge of a LW (62% pre; 67% post), knowledge of DPOA-HC (42% pre; 45% post),
or having an AD (20% pre; 24% post). Documentation of the AD in the medical record,
however, did increase significantly, particularly for Phase H patients who received the
SUPPORT intervention (6% pre; 35% post; 77% post + intervention). In a related study,
Bradley, Peiris, and Wetle (1998) examined the medical records of 600 randomly
selected nursing home residents admitted during 1990 and 1994, finding an average of
28.5% to have had a discussion about their preferences with a HCP (20% pre-PSDA;
36.7% post-PSDA). These studies related to the PSDA demonstrate that legislated
education cannot guarantee engagement in ACP.

Studies related to Special Populations

Next researchers targeted specific populations of patients who were hypothesized
to have a vested interest in ACP. In 2001, Wenger et al. reported data from a cross-
sectional survey of 2864 HIV patients in the United States (1996-1997) which showed

that 50% had discussed their end of life preferences with their HCP and 38% had

38

 

 

CC

IO

dt

dt

4)

in:

9.

 

completed an AD. This study found that patients were more likely to complete an AD
after conversation with their HCP.

The Assessing Care of Vulnerable Elders (ACOVE) project was a quality
assessment study conducted from 1998-1999 of 3207 randomly sampled community-
dwelling adults, ages 65 or older, who were enrolled in two managed care organizations
from 1998-1999. Five of the ACOVE End-of-Life Care quality indicators were specific
to ACP (Wenger & Rosenfeld, 2001) and reported by percentage of adherence: 1)
documentation of surrogate decision maker / AD in outpatient medical records (4%), 2)
documentation of surrogate decision maker / AD in inpatient medical records (25%), 3)
documentation of consideration of preferences for a patient with dementia (100%),

4) documentation of consideration of preferences for a patient admitted to ICU (17%),
and 5) documentation of a patient’s stated preferences or desire for an AD during the
interview (12%). The mean adherence to all 14 indicators for end of life care was only
9% (W enger et al., 2003).

In spite of Americans’ apparent interest in improving end of life decision-making
and health care providers’ increased understanding of ACP, these national studies
indicate that from 1989 — 1999 the knowledge, use, and outcomes of ACP were very
limited. In response, researchers tried a variety of interventions to increase the
prevalence of individuals’ ADS or ACP. Studies of ACP interventions can be divided

roughly into three time periods, each of which had studies using similar strategies.

39

 

t...
u

€11;

III:

the

DP

Sc}

inc

 

Studies of Interventions
1990-1995: Randomized Controlled Trials Focused on Physicians and Patients

In addition to Part H of the SUPPORT study, seven other randomized controlled
trials were conducted fi'om 1990-1995 to increase patients’ and health care providers’
engagement with ACP. A common denominator of these seven studies was that the
interventions were physician-driven. In the first three studies, physicians offered patients
the opportunity to discuss or initiate an AD. These resulted in increased consideration of
AD issues (Smucker et al., 1993), as well as 66% and 66.3% completed California
DPOA-HC forms respectively (Anderson, Kaplan, & Schneidermann, 1994;
Schneiderman et al., 1992).

In the next three studies, physicians initiated some sort of educational program to
increase the number of completed ADS. Rubin et al. (1994) provided an explanatory
pamphlet and DPOA—HC forms to 552 people with the result that 18.5% of the
intervention group initiated a DPOA-HC form vs. a 0.4% response by the control group.
However, Sachs, Stocking and Miles (1992) found that a 30-minute educational
intervention, including written and follow-up materials, with older patients at an
outpatient geriatric clinic did not result in a significant number of initiated ADS. Richter
et al. (1995) provided AD written materials (control) plus a structured discussion and
mailed follow-up reminder (intervention) to 176 patients with a result of 23%
(intervention) vs. 3% (control) ADS returned for the patients’ medical records within 6
months.

The final study in this group of seven was a comprehensive educational

intervention on health promotionwincluding counseling about ADS—provided to

40

 

a.
bi:

 

patients enrolled in a group health plan. Patrick et al. (1995) reported unexpected “excess
mortality” in the intervention group of 2558 older adults who had received the
intervention, which was thought to be related to the AD counseling.

In all of these studies, the physician-driven intervention produced an increase,
though not always significant, in completion of ADS. These studies demonstrate that
patients will respond to ACP interventions.

1 996-2001 .' Randomized Controlled Trials Focused on Innovative Education Strategies

From 1996 to 2001, researchers intervened with a variety of innovative education
strategies to try to increase patients’ engagement in ACP. Some researchers conducted
ACP education through counseling with hospital patient representatives (Meier et al.,
1996), nursing home admissions personnel (Molloy et al., 2000), or home health care
visits with patients (Patterson et al., 1997). Others used interactive seminars with patients
(Landry, Kroenke, Lucas, & Reeder, 1997); computer- generated reminders for health
care providers (Dexter et al., 1998; Tierney et al., 2001); and multimedia presentations
for patients (J. B. Brown, Beck, Boles, & Barrett, 1999; Siegert, Clipp, Mulhausen, &
Kochersberger, 1996; Yamada, Galecki, Goold, & Hogikyan, 1999). Still other
researchers targeted specific patient populations, e.g., hemodialysis (Singer et al., 1995),
pulmonary rehabilitation (Heffirer, Fahy, Hilling, & Barbieri, 1997), cardiac
rehabilitation (Heffiter & Barbieri, 2001), and preoperative patients (Grimaldo et al.,
2001). All of these interventions demonstrated at least modest effectiveness; however,
the prevalence of ACP or AD completion was no more than 15-40% in treatment or

control groups.

41

2001-2004: Randomized Controlled Trials Focused on Resource Intense Strategies

From 2001 to 2004, researchers intervened in more complex ways, e. g., on a
larger scale such as providing multimedia to 735 HMO members (Beck, Brown, Boles, &
Barrett, 2002) or by means of consultations with highly trained facilitators such as
Respecting Choices® nurses (Schwartz et al., 2002), home health social workers (Ratner,
Norlander, & McSteen, 2001), or a palliative care team (Rabow, Petersen, Schanche,
Dibble, & McPhee, 2003). These interventions demonstrated greater effects in the
intervention groups on knowledge of ACP (effect size = 22%, Schwartz et al., 2002) and
AD completion (46.2%, Beck et al., 2002; 83%, Ratner et al., 2001). However, such
interventions have been criticized, as was the SUPPORT intervention, as being too costly
for widespread use (Barnard, 2002).

In summary, ACP interventions—especially educational interventions—have
mostly resulted in some level of increased engagement in ACP, particularly in the form
of initiated ADS. A necessary prerequisite to engaging in ACP decision-making is
knowledge: not only of what is involved in end of life conditions or treatments and the
probabilities of different outcomes, but also self-awareness of preferences by the
individual and those with whom decision—making will be Shared (Briggs et al., 2004;
Gillick, 2001; Weiner & Cole, 2004). Through focus group research and public
discussions, Americans have reported that they value their rights of self-determination
and control of treatment choices, but need understandable information to guide decision-
making and planning (Lunney, Foley et al., 2003). Other studies have reported that, with
guidance, individuals are willing to formulate plans for end of life care (Kass-Bartelmes

etaL,2003)

42

Studies of Community-Based Approaches

In the past few years, some highly successful initiatives have focused on a “whole
community” approach to ACP. Community-based approaches to ACP have been
recommended by the Committee on Care at the End of Life (Field & Cassel, 1997) as a
way to increase the number of people who decide to engage in ACP. Both the LaCrosse
Advance Directive Study (LADS) (Hammes & Rooney, 1998) and the Missoula
Demonstration Project (Byock et al., 2001) documented the impressive effects of
extensive, whole-community ACP education. The LADS study documented a written
advance directive prevalence change from 15% to 85% in response to extensive
community advance directive education programs (now known as Respecting Choices®).
The Missoula Demonstration Project incorporated an interdisciplinary community-based
organization to both study and implement interventions to improve the quality of end of
life experiences for Missoula, Montana residents.

More recently, Clarke et al. (2005) reported the benefit of approaching people
through community-based organizations and groups in which they already participate.
Because these community groups were already valued by their members, people’s unease
toward engaging in ACP due to demographic, psychological, behavioral, or other barriers
was diminished. Capitalizing on this mode of entry, the researchers were able to provide
experts and resource materials to the community groups to encourage people to engage in
ACP. About one third of the participants requested the education materials and 30-50%
of each group’s participants subsequently completed an advance directive. In an
exploratory study of 27 Afiican Americans’ understanding and desire for a living will,

Waters (2000) reported that 82% of the participants stated they would consider writing a

43

living will after an educational group discussion compared to the baseline of 89% who

neither had nor wanted to consider a living will.

The ACP program intervention evaluated in this dissertation study was designed
to provide information and strategies to individuals so they could learn how to construct
and communicate their preferences for end of life care. This intervention incorporated
four of the community-based strategies found effective in other studies: 1) the
Respecting Choices® curriculum, 2) presentations offered to existing community groups,
3) certified Respecting Choices® ACP Facilitators (experts) conducting the ACP
programs, and 4) standardized content and resource materials to encourage participants to
continue their engagement in ACP.

Results of ACP Studies relating to Research Questions
Research Question I

Research Question 1: Before participation in a community-based ACP program,
what are the self-reported Personal Influences on Decision and Personal Decision Factors
of adults who have decided to engage in ACP? To answer this question, ACP studies
which reported baseline characteristics of their participants related to concepts in this
proposed study were reviewed.

Personal Influences on Decision

The focused model of the Decision to Engage in ACP (Figure 2, p. 23) shows four
Personal Influences on Decision: 1) Information, 2) Individual Characteristics, 3)
Values, and 4) Prior Experiences. At least one study in the reviewed ACP literature
reported baseline or other results for each variable included in the Personal Influences on

Decision section of the Focused Model of the Decision to Engage in ACP. However,

44

none of the studies examined the prevalence of all these variables simultaneously. This
dissertation study will contribute a more comprehensive picture of the variables included
in personal influences on the decision to engage in ACP.

Information. Ten studies, conducted from 1992-2004, describe people’s baseline
information about ACP. The percentage of people who “knew of” a living will ranged
from 25% (C. D. Moore & Sherman, 1999) to 62% (Singer et al., 1995; J Teno et al.,
1997) to 80% (Landry et al., 1997; Siegert et al., 1996) to 96% (Waters, 2000). In
Grimaldo et al.’s (2001) study, 80% stated they knew about DPOA—HC. Singer et al.
(1997) reported that 88.7% of AIDS patients had “heard of” an AD, but in Waters’
(2000) group of Afiican Americans, 78% had not. Monison et al. (1998) reported ethnic
differences in knowledge of ACP ranging from 57% (White) to 39% (Afiican American)
to 20% (Hispanic). In addition to these studies, other researchers have tried to determine
knowledge through true/false items (Ejaz, 2000; K. A. Moore et al., 1994) and asking for
a simple definition of “advance directive” (Office of the Inspector General, 1993).

Two studies focused specifically on participants’ knowledge. Silveira, DiPiero,
Gerrity, and F eudtner (2000) used a series of vignettes and questions about state statutes
to assess Oregon citizens’ knowledge of options for end of life care. Siegert et a1. (1996)
measured comprehension of ACP concepts (living will, cardiopulmonary resuscitation)
before and after an ACP intervention through open-ended and true/false interview
questions during a randomized cohort trial of elderly residents of an extended
care/rehabilitation center. Baseline comprehension of a “living will” was 75% in the
cohort that received an AD video and, surprisingly, 90% in the cohort that received a

general health-related video.

45

Baseline knowledge of ACP varied among participants in the ACP studies. None
of the ACP studies reporting knowledge of ACP included participants’ self-perception of
having “enough knowledge” to engage in ACP. In this dissertation study, participants
rated their perceptions of having “enough knowledge to discuss [their] views” using a
Single item on the Participant Survey.

Individual Characteristics. Individual characteristics considered in this pr0posed
study include age and gegd_e_r.

Age. In many studies, participants were of a specific age cohort (e.g., 65 or older)

so differences in age were not reported (Blackhall et al., 1999; Bradley et al.,

1998; Hamel et al., 2002; Hopp, 2000; Landry et al., 1997; Meier et al., 1996;

Sulmasy et al., 1996). Other studies included adult participants of varied ages (B.
A. Brown, 2003; Dexter et al., 1998; Hammes & Rooney, 1998; Morrison et al.,
1998; Patterson et al., 1997; Reilly et al., 1995; NS. Wenger et al., 2001).
Giggle; Participants’ gender was reported in most studies, with a mix of men and
women participants. None of the studies were exclusively of women or men, but
women were more prevalent in the majority of studies (Bradley et al., 2001; J. B.
Brown et al., 1999; Ditto et al., 2001; Doorenbos & Nies, 2003; Landry et al.,
1997; McKinley, Garrett, Evans, & Danis, 1996; Molloy et al., 2000; C. D. Moore
& Sherman, 1999; K. A. Moore et al., 1994; Morrison & Meier, 2004; Patrick et
al., 1995; Reilly et al., 1995; Sachs et al., 1992; Schneiderman et al., 1992;
Silveira et al., 2000; Smucker et al., 1993; Sulmasy et al., 1996; Teno et al., 2004;

Waters, 2000).

46

This dissertation study included men and women adult participants from young,
middle, and older age groups. Similar to other studies, women participants were more
prevalent.

Values. Only one study was found in which participants rated having plans for
firture care as important. Ditto et al. (2001), in a randomized controlled trial of 401
elderly (> 65 years old) outpatients and their designated surrogate decision-makers,
reported on a 1-5 point scale a mean of 4.49 (SD = 0.04) for the outpatients and a mean
of 4.55 (SD = 0.04) for the surrogates.

In this dissertation study, participants rated their perceptions of the perceived
importance of ACP discussions using a single item, “Important to have discussions,” on
the Participant Survey.

Prior Experiences. Prior experiences with ACP include having an AD, haying;
HCA, and having conversations about ACP (with HCA, significant others, and HCP), and
making a decision about life-sustaining treatment for someone else. The category of
having conversations includes both the indication that a conversation occurred as well as
the perceived quality (or thoroughness) of that conversation.

Having an AD. Baseline percentages of participants with advance directives

(ADS) were reported in 32 studies, ranging fi'om < 10% to > 70%. The relatively

large number of studies including these data reflects researchers’ predominant

emphasis on initiated ADS as an outcome of ACP interventions. Wide variations
in prevalence of ADS exist regardless of whether studies are organized by

numbers and types of participants or by date of publication. A depiction of the

47

percentage of research study participants having an AD organized by date of
study publication can be found in Figure 3.

In small exploratory studies (11 < 30 participants), Moore and Sherman
(1999) found that 6 (30%) of the minority and low-income seniors living in the
community they interviewed reported having an AD, while Waters (2000)
reported only 3 (11%) living wills (LW) in her interviews of community African
Americans of varied ages. Six larger studies (n = 30-100) reported percentages in
Specific patient populations. Curtis et al. (1999) found 65% of 57 AIDS patients
had LW while 60% had Durable Power of Attorney for Health Care (DPOA-HC).
Doorenbos and Nies (2003) reported a 9% prevalence of LW or DPOA-HC in a
survey of 45 Asian Indian Hindus. Heyland, Tranmer and Feldman-Stewart
(2000) found 24% written ADS among 37 hospitalized patients with serious
illnesses. Moss et al. (1996) reported that 76% of 50 patients with amyotrophic
lateral sclerosis on long-term ventilation had written ADS. In a survey of 36
rehabilitation/extended care patients, Siegert et al. (1996) reported that 8 (22%) of
them had an AD. Singer et al. (1995) surveyed 95 hemodialysis patients, finding
only 7% with a LW.

Thirteen studies, with numbers of participants ranging from 101 to 500,
documented percentages of ADS in various settings. Bradley et al. (2001) found
32.7% inpatients with cancer to have ADS while Schneiderman et al. (1992) found

only 4 (0.02%) of 337 patients with life-threatening illness (primarily oncology,

48

 

 

|\
:9, \\~\\‘S\\\\\\ \\ \\\\‘; see :Qo
o\\\\\s‘qé

75

e5 .\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

4

05 :\\\\\\\

 

 

 

 

:3 WW
3 ;\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
s es
5 es en. can. .\~:.~\.\\ .\ as
191 \S‘x“\\\\\\\:ess \\\e.

R] r; \\\:\.\\§ \ek ease

as \\‘:\\\\\\‘

 

8 seem \S\\\\\\\\<\\\.
8 \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ \\\\\\\\\\\\\\\\\\\\\\ \\\\\\\\\\\\\\\\\\\\
8 \‘s\\‘\\\‘ ‘ ‘ sea
3 “‘\\§\\\\\\\\\\\
3 senses“

.‘\‘\.\\\I\ \\\.\\‘\.‘\\\‘\.‘\‘ \‘

18.2

F ‘ ‘ \\\\\\
m \\\\.\ \\\\\‘. \\ 0‘\\\‘\\ \\\\\\\\\ \\\\\\\\\\\

:: se\\\\\

 

SI :.\.\.\\ was \‘h‘.\ \:

 

RW\W

:33

,5 it’s.

0%

S

:s,

4%

(a

s .‘Is

s ‘s

is

3' . . \\\§\\\\\\\\\\\‘.\\\\\\\\\\ Q” :X‘
%
is
.319»
fig,
‘11»
is
:2».
Qt»
is

.\\‘ \
* \\‘ \x“.

 

‘ \‘.\.‘~\‘ \‘\‘ \‘ \‘ \ \‘\ \'\‘\ ‘ \‘\ \‘ ‘\ \‘\
:‘\\\'\\$ \\‘ \\ \\\‘\‘\\\‘\ \\ \\\\\'\\ \\ \\\\\'\\§

35

.\\\

:‘s‘s \\“s\“§

12 12

\{\‘I\\\\\\\

a;
:22}
:b
“it"s,
:i:
:‘e
5,6
‘5,
2:,
he,
’“r.

N
o.
3 .
o

 

 

 

 

 

 

 

 

 

100

90
so
70
60
so
40
so
20
1o -

a.

49

First Author, Date

Figure 3. Percentage of research study participants having an AD, organized by date of study publication.

pulmonary, and autoimmune deficiency) to have ADS prior to their intervention.
Grimaldo et a1. (2001) found an average of 9% preoperative patients had an AD.

Reilly et al. (1995) surveyed recently discharged acute care internal
medicine patients and found 4% having ADS. Patients from outpatient clinics
were varied in their percentages of ADS; 1) urban family practice had 12% LW,
6% DPOA-HC, and 3% other AD (K. A. Moore et al., 1994; Smucker et al.,
1993); 2) ambulatory cancer had LWS in 3% of their black and 34% of their
White patients; 3) internal medicine had 31% (Landry et al., 1997), 2% (Richter et
al., 1995), 11% (Sulmasy et al., 1996); 4) geriatric had 1% (Sachs et al., 1992); 5)
HIV had an average of 18.6% (Singer et al., 1997); 6) cardiac rehabilitation had
an average of 43.5% LW and 30.5% DPOA-HC (Heffirer & Barbieri, 2001); and
7) mixed outpatient clinics had 46% AD or LW and 52% DPOA-HC (Ditto et al.,
2001)

Three studies surveyed more than 700 people: an HMO group had an
average of 20% ADS (J. B. Brown et al., 1999); senior citizens from New York
City senior centers had an overall prevalence of 35% ADS with non-significant
differences between White (41%), Afiican American (30%), and Hispanic (35%)
participants (Morrison & Meier, 2004); and Oregon adults at an internal medicine
/ family practice clinic reported 31% ADS (Silveira et al., 2000).

Seven research reports analyzed very large sets of data for AD prevalence.
From the SUPPORT study data, F reeboume, Lynn and Desbiens (2000) found
618 (20.2%) ADS from 3058 patients with 11.7% LW, 13.4% DPOA-HC, and

4.9% both LW and DPOA-HC and Teno, Lynn et al. (1997) found 14% of 4804

50

patients with serious illness had an AD during their hospital admission but only
4.7% had this prior to the hospital admission. In a study of SUPPORT data pre-
and post-Patient Self Determination Act (PSDA), Teno, Licks et al. (1997) found
that 20% patients had ADS pre-PSDA and an average of 22% patients had ADS
post-PSDA. In surveys of older adults, Molloy et al. (2000) found that < 25% of
nursing home residents had an AD and Patrick et al. (1995) found an average of
35% prevalence in HMO participants. Rubin et al. (1994) found an AD
prevalence of 5.8% in 1101 patients recently discharged fiom the hospital, with an
average of 4.4% having a DPOA-HC. Finally, Teno et al.’s (2004) survey of
relatives of deceased patients reported that 70.7% of the patients had a written AD
at their last place of care before their deaths.

Having an HCA. Baseline percentages of participants with designated HCAS
were reported in 7 studies, ranging from < 10% to > 40% (Figure 4). In
randomized controlled trials of hospitalized or recently discharged patients, Rubin
et al. (1994), Reilly et al. (1995), and Meier et al. (1996) documented HCA
prevalences as 42-47%, 12.3%, and 6—26%, respectively. In studies of
outpatients, HCAS were designated by 6% of elderly family practice patients
(Smucker et al., 1993); 45.6-46.3% of internal medicine patients (Landry et al.,
1997); 20% Hispanic, 31% African American, and 40% Non-Hispanic White of
elderly geriatric and internal medicine patients; and 23% of hemodialysis patients

(Hines et al., 1999).

51

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Smucker. 1993 Rubin. 1994 Reilly.1 1995 Meier. 1996 Landry. 1997 Morrison. 1998 Hines. 1999

Figure 4. Percentage of research study participants having a HCA, organized by
date of study publication.
Having Conversations. Baseline percentages of participants who have had
conversations about ACP were reported in 14 studies. Figure 5 displays the
percentages of conversations participants reported with their HCAS, significant
others, or health care providers.
Not specified. Landry et al. (1997) reported that 50% of internal medicine
outpatient participants in their study had previous discussions of end of

life issues, but do not specify with whom.

52

 

 

 

 

 

 

 

 

 

 

60

 

 

 

 

 

m \\ ‘\ \\ \ \\\ \\\\‘.\
E \\\\\\\\\\\\\\\\\\\\\\\\\\\\\

 

r.

 

 

 

 

 

IO

*~\\\i>x~: \ \\ \
(D \\\".~.\,\\\\\§\\ “i .\\\‘§~\\\\~;\ ‘

V

49"
f.

2

f I
N
.e

105

ations, organized by date of study publication.

 

 

 

fik

fi"

.\ ‘0 * \ ‘\\\~\V vs \\\\\ \\
‘ \ ,_ . \\ t -. . .‘\
:\ \' \\\\\ \‘.‘~:\\

ea..
«xx

.959

 

 

 

 

 

 

 

      

 

    

 

 

 

 

 

 

   

3
I...
.959

:55"

‘f‘x’f‘f

 

.3

 

 

:65”

   
    

.4?

..4?

 

.95"

 

100

D.
0
I ‘1
~ K

I g \\\\\\\\\\\\\\\\\\\\\\\\\\\\;\\\
g
5 s
o
a
5
5 o $§§§§i§$§§$§$§:1‘\\\\§:\:§§\\\~\§\\\\\§N\\
u:
<
O
I
D '0.

Q

N

U l ' v

0 O O O O o O O O o
O (D h (D a) V (‘0 N ‘—

da“ 9“" f a" f

53

Flat Author. Date

ts having convers

1c1pan

Figure 5. Percentage of research study part

HCA. Two studies reported the percentage of participants who had
conversations with their HCA, ranging from 50-60% (Hines et al., 1999;
Karel, Powell, & Cantor, 2004).

Significant Others. Five studies reported conversations with family
members, with percentages of participants that ranged from 30% to 70%
(Grimaldo et al., 2001; Heyland et al., 2000; Hines et al., 1999; Moss et
al., 1996; O'Brien et al., 1995). Only one study compared the percentage
of conversations had with an HCA with conversations had with family
members (Hines et al., 1999).

11932. The percentage of participants who had conversations with their
health care providers was reported in 13 studies, with a range of 2% to
40%. Bradley et al. (1998) found that the Patient Self Determination Act
was positively associated with the likelihood of having a discussion with a
HCP. Moore et al. (1994) found that the amount of time health care
providers spent talking with patients about AD-related issues was related
to overall comprehension of ADS. Morrison et al. (2004) noted that
patients who were only comfortable discussing ADs when the HCP
initiated the discussion were less likely to have a completed AD. Prior
conversation with a HCP was a predictor for completion of ABS in two
studies (Beck et al., 2002; NS. Wenger et al., 2001). Again, only the
Hines et al. study (1999) compared the percentage of conversations held
with a HCP to the percentages of conversations with a HCA or significant

others.

54

many conversations. None of the ACP studies reviewed reported participants’
perceptions of the quality of conversations about ACP.

Making Decisions. Three studies reported the percentage of participants who had

experience making decisions for someone else (Karel et al., 2004; Silveira et al.,

2000; Waters, 2000); one additional study reports the experiences of patients with

amyotrophic lateral sclerosis who made decisions for themselves (Moss et al.,

1996)

Although numerous studies have reported the percentages of participants who
have had prior experiences with some aspects of ACP—having an AD, having a HCA,
and having conversations, making decisions for someone else—none of the studies
examined all four aspects. This dissertation study examined all four aspects of the
participants’ prior experiences with ACP. This study also compared the prevalence of
conversations with HCA, significant others, and HCP, which was only reported in one
previous study (Hines et al., 1999). Finally, this study solicited participants’ perceptions
of the quality of their conversations, a finding not currently reported in the ACP
literature.

Personal Decision Factors

Referring again to the Focused Model of the Decision to Engage in ACP (see
Figure 2, p. 23), this dissertation study examined the Personal Decision Factor Self-
Efficacy through participants’ perceptions of their ibflijy and likelihood of having ACP
conversations with others. Only two studies examined self-efficacy or confidence related
to ACP. Three additional studies discussed changes in participants’ interest levels in

ADs, which may be a reasonable proxy for likelihood. The focus on ADS may reflect a

55

greater emphasis in earlier advance planning studies on people’s completion of AD forms
(Kolarik et al., 2002) than on engaging in “caring conversations concerning their
expectations and wishes for end of life care” (Michigan Commission on End of Life
Care, 2001). None of these five studies specifically explored participants’ perceptions of
their ability or likelihood of having conversations related to ACP. By examining the
variables of ability and likelihood, this dissertation will contribute to a beginning
understanding of how prevalent self efficacy is among participants in an ACP program
intervention.

Self Efi‘icacy. Two studies reported participants’ self-efficacy. Brown (2003)
noted that providing an older adult with printed education materials and with
personalized reinforcement of the content by a nurse may increase self-efficacy in
deciding preferences for end of life care. Silveira et al. (2000) reported their survey
participants’ confidence in their knowledge of legal choices available for end of life
decision-making in Oregon.

Able to start discussion with others. None of the studies reviewed for this

proposal addressed the perception of confidence in the ability to start an ACP

discussion with others. In this dissertation study, participants rated their
perceptions of their ability to start a discussion with others using a single item on
the Participant Survey.

lelv to have conversation with others. Three studies reported changes in

participants’ interest in ACP following interventions: an increase in thinking

about an AD from the immediate post-program survey to the follow-up survey

two weeks later (Smucker et al., 1993); a willingness to consider “getting an AD”

56

by 20.9% after hearing about them (Office of the Inspector Office of the Inspector

General, 1993); and an intent to complete a living will that varied by age

(McKinley et al., 1996). However, none of the studies reviewed for this study

addressed the perception of being likely to have an ACP conversation with others.

In this dissertation study, participants rated their perceptions of their likelihood to

have a conversation with others using a single item on the Participant Survey.

Research Question 2

Research Question 2: Which variables from among the Personal Influences on
Decision and Personal Decision Factors are associated with the Decision to Engage in
ACP? This question is difficult to explore in the literature because none of the studies
reviewed for this study specifically examined associations between internal influences
and the decision to engage in ACP. Therefore, ACP studies which examined the
associations of variables included in this dissertation study were reviewed.

In the Focused Model of the Decision to Engage in ACP (see Figure 2, p. 23), the
Decision to Engage in ACP is thought to be associated with variables from “Personal
Influences on Decision” (Information, Individual Characteristics, Values, and Prior
Experiences) and the “Personal Decision Factor” self-efficacy (Ability and Likelihood).
The Decision to Engage in ACP is thought to then influence the Behaviors of Engaging
in ACP, which subsequently influence the Outcomes of Preferences and Communication
of Preferences. Of the 21 studies examining associations between variables included in
the model of ACP as a decision-making process, only nine included findings related to
more than one of the personal influences on decision (information, individual

characteristics, values, prior experiences). This dissertation study examined associations

57

among all four personal influences on decision and the personal decision factor self-
efficacy with the decision to engage in ACP.
Personal Influences on Decision

Among variables in Personal Influences on Decision, associations were found
between information, individual characteristics of age and gender, and of prior
experiences of making decisions, having a HCA, and having discussions.

Information. Ejaz (2000) found that knowledge of an AD was associated with
having an AD and the likelihood of implementing an AD. Morrison et al. (1998) found
that knowledge gained through exposure to a fiiend of family member being on a
ventilator influenced completion of an AD. In the same study, knowledge of a HCA was
associated with designation of a HCA. Knowledge of advance directives was associated
with experience making a decision (Silveira et al., 2000). Comprehension of an AD was
increased when patients were younger and more educated but decreased if the HCP spent
less time talking about AD issues (K. A. Moore et al., 1994). No studies examined
“enough knowledge” as an associative variable.

Individual Characteristics. The associations of age and gender with other
variables and outcomes were not consistent: some studies found them to be influences
but others did not. When age and gender were influences, the findings were not
consistent between younger and older ages or men and women.

Age. In two studies, age was associated with having a discussion (Dexter et al.,

1998) and identifying preferences (B. A. Brown, 2003). In three studies, younger

age was associated with an intent to complete an AD (McKinley et al., 1996) or

completion of an AD (K. A. Moore et al., 1994; Patterson et al., 1997). However,

58

Hammes and Rooney (1998) found that younger age was associated with not
completing a written AD. In contrast, four studies found older age to be
associated with completing an AD (Dexter et al., 1998; Morrison et al., 1998;
NS. Wenger et al., 2001) or designating a HCA (Reilly et al., 1995). Age was
not found to be associated in six studies: withthe intent to complete an AD
(Sulmasy et al., 1996); with the prior experiences of having an AD (Hopp, 2000)
or having a discussion (Bradley et al., 1998; Hamel et al., 2002); with completing
an AD (Hamel et al., 2002; Landry et al., 1997); or designating a HCA (Hopp,
2000; Meier et al., 1996).
M. In two studies gender was associated with completion of an AD: women
in one (Doorenbos & Nies, 2003) and men in the other (N .S. Wenger et al., 2001).
However, seven studies found no association of gender: with the intent to
complete an AD (Sulmasy et al., 1996); the prior experience of having a
discussion (Bradley et al., 1998; Hamel et al., 2002); completing an AD (Hamel et
al., 2002; Landry et al., 1997; Molloy et al., 2000; Rubin et al., 1994); or
designating a HCA (Meier et al., 1996).
Values. No studies reported testing the association of the perceived importance
of ACP.
Prior Experiences. The prior experiences of having an AD, having a HCA,
having discussions, and making decisions were found to be associated variables.
Having an AD. Ejaz (2000) found that knowledge of AD was associated
with having an AD. Hopp (2000) found no relation between age and

having an AD.

59

Having a HCA. Morrison et al (1998) found an association between
having a HCA and completing an AD.

Having Discussions. Hamel et al. (2002) found an association between
having discussions in general and completion of an AD; Wenger et al.
(2001) found a similar association but more specifically with having
discussions with a HCP. Morrison et al. (2004) found that people who
were only comfortable having a HCP-initiated discussion were not likely
to complete an AD. Bradley et al. (1998) found that the Patient Self
Determination Act was associated with having a discussion about ADs but
age and gender were not. Hamel et a1. (2002) also found no association
between age and gender and having a discussion about ADs. Dexter et al.
(1998), however, did find that age was associated with having such a
discussion.

Making decisions. Silveira et al. (2000) found an association between
making a decision and knowledge of ADs, but no relation with completing

anAD.

Personal Decision Factors

None of the reviewed studies mention associations with the decision factor self-

efficacy or the perceptions of being able and likely to engage in ACP conversations.

Decision to Engage in ACP

No studies specifically examine associations with the Decision to Engage in ACP.

However, the intent to complete an advance directive may be a reasonable parallel. The

intent to complete an AD was not associated with age or gender in one study (Sulmasy et

60

al., 1996) but was associated with being younger, white, more educated, and having more
income in another study (McKinley et al., 1996). Having no intention to complete an AD
was associated with being black and having stronger religious beliefs (McKinley et al.).
Research Question 3

Research Question 3: Do the Personal Influences on Decision and Personal
Decision Factor variables that are associated with the Decision to Engage in ACP differ
between participants with previous ACP experience compared to participants without
previous ACP experience? None of the studies reviewed for this dissertation have
considered the role of the previous experiences of having an AD, having a HCA, or
having both an AD and a HCA as a variable which may have an impact on other personal
influences or decision factors associated with the Decision to Engage in ACP. However,
studies with baseline percentages of participants who report prior experiences of having
an AD or having a HCA are a reasonable substitute.

Referring once again to the Focused Model of the Decision to Engage in ACP
(see Figure 2, p. 23), participants’ prior experiences are considered to be Personal
Influences on Decision that affect the Decision to Engage in ACP. Within a group of
participants attending an ACP program, it is likely that some will have had prior
experience engaging in ACP, while others may have decided to engage in ACP for the
first time. The model depicts this dynamic as the variable “ACP experience.” For the
purposes of this dissertation study, ACP experience is defined as a participant’s self-
report of the prior experiences of having an AD, having a HCA, or having both an AD
and a HCA. Using this definition, participants in ACP programs can be separated into

those with and without prior ACP experience. Research is needed to describe the

61

potential influence of ACP Experience on participants’ decisions to engage (again) in
ACP. This dissertation study defines the variable ACP Experience and tests its
association with other personal influence and decision factor variables and the decision to
engage in ACP.

Personal Influences on Decision: Prior Experiences

Having an AD. As reviewed earlier in this proposal, researchers of 32 studies
published the percentages of their participants who had an AD at baseline (see Figure 3,
p. 48). Already having an AD was considered exclusion criteria for participants in four
studies (Hamel et al., 2002; Landry et al., 1997; Sachs et al., 1992; Schneiderman et al.,
1992). In a study of HIV positive patients, Singer et a1. (1997) documented that 19.0% of
the participants approached for the study already were satisfied with their previous ADS
and declined participation in the study. Only Grimaldo et al. (2001) analyzed differences
in the percentage of ADS found in the medical records of preoperative patients who
declined participation in the study versus patients who participated in the randomized
controlled trial. The refusal group had 8%, the control group 9%, and the intervention
group 11% at baseline; these differences were not significant.

Having a H CA. As reviewed earlier in this proposal, seven groups of researchers
published studies reporting baseline percentages of participants who had a HCA (see
Figure 4, p. 51). Having a HCA was not an exclusion criteria for any of these studies.

Having Both an AD and a H CA. None of the studies reviewed indicated

participants who had both an AD and a HCA.

62

Research Question 4

Research Question 4: After participation in a community-based ACP program,
are there changes from pre- to post-ACP program intervention in self-reported ratings of
enough knowledge, importance, ability and likelihood to have an ACP conversation? As
a beginning answer to this question, 23 studies of ACP educational interventions, which
included a pre- and post-intervention measure, were reviewed to determine which
outcomes were being measured and the impact of the intervention on those outcomes.

The Focused Model of the Decision to Engage in ACP (see Figure 2, p. 23)
depicts the Intervention as influencing the Personal Influences on Decision variables of
“Enough Knowledge” and “Importance” as well as the Personal Decision Factor self-
efficacy variables of “Ability” and “Likelihood.” These personal influence and decision
factor variables are thought to influence the Decision to Engage in ACP which influences
subsequent Behaviors and Outcomes of engaging in ACP. In more than 80% of the ACP
educational intervention studies reviewed, the measured outcomes were fiom the
Outcomes section of the model, which is not a focus of this dissertation study. Thus,
only studies relating to the personal influences on decision and personal decision factor
self-efficacy variables or the intervention will be discussed. This dissertation study
evaluates the impact of a community-based ACP educational program on participants’
ratings of their perceived information, values, and self—efficacy related to ACP
conversations.
Personal Influences on Decision

Only six studies reported an intervention’s impact on individuals’ information

(11 = 5) and value of ACP (n = 1).

63

Information. Five studies reported an impact on participants’ knowledge of ADS,
including Living Will (LW) and Durable Power of Attorney for Health Care (DPOA-HC)
documents. Grimaldo et al. (2001), Schwartz et al. (2002), and Waters (2000) reported
increases in participants’ knowledge of ACP following a HCP-initiated conversation
about ACP. Yamada et al. (1999) also noted increased knowledge of ACP after
providing handouts and a video to participants. Teno, Lynn et al. (1997) reported a non-
significant increase in knowledge of LW or DPOA-HC following implementation of the
Patient Self-Determination Act (PSDA).

Values. Only the study by Waters (2000) reported a dramatic Shift in perceived
importance of ACP from 89% of the 27 Afiican American participants having no LW and
not wanting to consider one at the beginning of the program to 85% of the participants
being willing to consider a LW at the end of the program. In addition, more than 85% of
the participants stated that a community-based ACP discussion was valuable.

Personal Decision Factors

No studies discussed the impact of an intervention on self-efficacy or the

variables of ability or likelihood to engage in ACP.
Research Question 5

Research Question 5: Do the changes between participants’ self-reported pre- and
post-ACP program intervention ratings of enough knowledge, importance, ability, and
likelihood differ between participants with previous ACP experience compared to
participants without previous ACP experience? Only one study could be found that
describes the impact of an ACP intervention on participants who already had an AD.

Brown et al. (1999) reported that a “substantial number” of subjects with previous ADS

64

added DPOA—HCS or made other changes to their ADS following ACP education. This is
an example of re-engagement in ACP.

This final research question, using elements of Research Questions 3 and 4,
evaluates the influence, if any, of the variable ACP Experience on the impact of the
community-based ACP educational intervention on personal influences of knowledge and
importance of ACP and the decision factor of being able and likely to engage in ACP.
Since there is no literature discussing the role of ACP Experience on the impact of an
ACP intervention, this dissertation study provides a comparison of the reports of
community-based ACP program participants with and without ACP experience.

Significance of Study

This dissertation study intends to make five contributions to the knowledge base
of ACP decision-making:

1. This study focuses on the Decision to Engage in ACP rather than
behaviors or outcomes of engaging in ACP. The emphasis on this
decision encourages investigation of ACP fi'om a decision-making
theoretical perspective.

2. This study examines the prevalence and association of an assortment of
personal influence and decision variables among participants who have
decided to engage in ACP. This contributes a more comprehensive
picture of the internal influences on individuals’ decisions to engage in
ACP than is currently found in the literature.

3. This study measures the impact of a community-based ACP educational

program intervention on selected personal influence and decision

65

variables related to having ACP conversations. The emphasis on
information, values, and self-efficacy underscores the key features of
ACP as a deliberative, communicative, shared decision-making, and
dynamic process.

. This study explores the association of an external intervention with
changes in internal influences on a decision. This association provides a
new target for the impact of an intervention that differs from the usual
product outcome (AD).

. This study provides a first look at ACP Experience as a variable that may
modify other personal influences and decision variables as well as the
impact of an ACP intervention. By examining the differences, if any,
between participants with and without ACP experience, this study
contributes to the understanding of individuals’ decisions to engage in
ACP for the first time or again. This promotes the understanding of ACP

as a dynamic decision-making process.

66

CHAPTER 4
METHOD

This chapter reviews the methodology of the dissertation study. The context of
the study—the Mid-Michigan Advance Care Planning Coalition (MMACPC) Research
Project—will be described first as an example of a collaborative effort to promote ACP in
communities. The method of this study, which was a secondary analysis of data collected
during the MMACPC Research Project, follows.

Context of Study

Promotion of A CP

In 2003, a university-affiliated outreach program, the Capital Area Health
Alliance - Advance Care Planning Coalition (CAHA-ACPC),2 formed to promote ACP
for Michigan residents. With funding fi'om the Robert Wood Johnson Foundation
Rallying Points Program (2002), CAHA-ACPC selected Respecting Choices®, a
nationally recognized, comprehensive, community-based ACP pro gram (Hammes &
Briggs, 2002a), to provide professional training in the skills of ACP facilitation and
community-based ACP strategic planning. Between March, 2004 and January, 2006, 52
ACP Facilitators were certified during CAHA-ACPC-Sponsored ACP Facilitator training
conferences. Each ACP Facilitator received Respecting Choices® training and resources
to promote ACP. In return, CAHA-ACPC requested that the ACP Facilitators provide at
least one ACP educational program for an established group of individuals in their

communities.

 

2 The Coalition was originally named the Mid-Michigan Advance Care Planning Coalition (MMACPC). It
is a collaborative interdisciplinary group of health care providers and researchers from two local health care
organizations (systems) and Michigan State University.

67

In April, 2004, CAHA-ACPC appointed a Research Subcommittee, the Mid-
Michigan Advance Care Planning Coalition (MMACPC) Research Team, to measure the
effectiveness of using ACP educational programs, conducted by ACP Facilitators in
established community groups, to promote engagement in ACP. This investigator, as
part of her doctoral studies, designed the MMACPC Research Project, acted as the
Research Coordinator, and led Research Team meetings from June, 2004 to June, 2006.
Funding for the MMACPC Research Project was awarded from the Blue Cross Blue
Shield Foundation of Michigan Student Award Program (2004-2005) and the Marion
Peterson Nursing Research Endowment (2005).

The MMACPC Research Project was a prospective descriptive quasi-
experimental study that included an ACP program intervention with pre- and post-
prograrn surveys. Certified ACP Facilitators planned ACP programs, using Respecting
Choices® resources, for established groups in their local communities. Approximately
400 people attended one of 25 ACP programs offered between July, 2004 and November,
2005, targeted to existing faith (n=16), health-related layperson (n=4), and employee
(n=5) community groups in Greater Lansing and Grand Rapids, Michigan. A
convenience sample of 347 adult men and women completed survey instruments (85 %
participation).

Development of Survey Instruments

The pre- and post-pro gram survey instruments were developed by this

investigator from Community Surveys provided in the Respecting Choices® Advance

Care Planning Program Quality Improvement Toolkit3 (Briggs, Havlik, & Hammes,

 

3 While Respecting Choices® is a trade-marked name, the Community Surveys were made available in the
public domain as tools to evaluate ACP programming.

68

2000; Hammes & Briggs, 2004). Items were selected and adapted based on a review of
ACP decision-making literature and experiences from professional practice. The
resulting Participant Survey included pre- and post-program surveys (see Appendix A).
Content validity of the Participant Survey was confirmed through independent reviews by
one of the Respecting Choices® founders (L. Briggs), CAHA-ACPC health care .
providers (L. Peterson, L. Harrison), and members of the MMACPC Research Team.
Neither the Participant Survey nor the earlier Respecting Choices® Community Surveys
have published psychometric reliability data. The Participant Survey was piloted in May,
2004 at an ACP Facilitator training conference and was deemed easy to comprehend and
complete.
Survey Data Collection Procedure

When an ACP program was scheduled, the ACP Facilitator received survey
materials for the expected number of participants from the Research Coordinator. At the
beginning of the ACP program, the ACP Facilitator distributed a survey packet to each
person in attendance and invited them to participate in the study. ACP Facilitators
assured attendees that they could attend the ACP program whether or not they took part
in the study. Participants completed a consent form (see Appendix B) and the pre-
program survey before the planned ACP program content was begun. At the conclusion
of the ACP program, participants completed a post-program survey. Following the ACP
program, the ACP Facilitator returned the completed consents, pre-prograrn surveys, and
post-program surveys, along with a Program Identification form, to the Research

Coordinator.

69

Data Management and Storage Procedures

To prepare for data entry, the Research Coordinator created 1) a tracking database
to log sending and receiving of data collection packets; 2) a password-protected database
for data entry; and 3) a set of data entry instructions and a codebook listing each variable
from the survey. The Research Coordinator hired a student research assistant to assist
with data entry and oriented her to procedures for data entry and how to safeguard the
data and the privacy of participants’ survey information. Participants’ privacy was
protected because the consent forms with identifying information were kept separate
fi'om the pre- and post-surveys. Only the pre- and post-survey data were exported into
SPSS for analysis; therefore, the files used for data analysis contained no personal
identifying information.

When a packet of research data arrived, the Research Coordinator logged the
contents of the consent, pre-program survey, and post-program survey envelopes into a
tracking database and assigned a program code (e. g., Faith 1, Employee 4). The collected
data were entered into the secure database by the Research Coordinator (n = 8 programs)
and the student research assistant (11 = 17 programs). Analysis of the data was canied out
by the Research Coordinator with direction from the Research Team.

MMA CPC Research Project [RB Approval

The Michigan State University Committee on Research Involving Human
Subjects (UCRIHS) approved the MMACPC Research Project in June, 2004, With
renewals in June, 2005, May, 2006, and April, 2007. In addition, specific agency internal
review board approval (“Reliance agreements”) were obtained for Sparrow Health

System (December, 2004) and Ingham Regional Medical Center (March, 2005) to cover

70

employee group pro grams offered in those health care organizations. An IRB Revision
was approved in March, 2005 to add a personal identification number to the consent form
and survey instruments. This number, while still protecting the participants’ privacy,
allowed comparison of individual participants’ pre- and post-program ratings of
knowledge, importance, ability, and likelihood of engaging in ACP. Copies of the [RB
applications and approval letters are located in Appendix C.
Results Dissemination

Findings from the MMACPC Research Project have been presented regionally
five times (V ander Laan, 2005a, 2005b; K.J. Vander Laan, P.B. Mullan, & G.K. Wyatt,
2005; Vander Laan, Mullan, Wyatt, McPhail, & Thomason, 2007; Wyatt, McPhail,
Mullan, Thomason, & Vander Laan, 2005). National presentations have been made
three times (K.J. Vander Laan, P.B. Mullan, & G.K. Wyatt, 2005; Vander Laan, Mullan,
Wyatt, McPhail, & Thomason, 2006; Vander Laan & Wyatt, 2005).
Approval of Secondary Data Analysis

At its April, 2006 meeting, the MMACPC Research Team approved the use of the
project data for this dissertation study. The planned data analyses of the matched pre-
post survey data have not been previously done or reported.

Study Methods

Design

This dissertation study is a secondary analysis of a subset of data collected for the
MMACPC Research Project. The MMACPC Research Project included pre- and post-

program surveys of participants in an ACP program.

71

Subjects

The subjects for this study are a convenience sample of 147 men and women,
ages 18 and older, who participated in one of 11 ACP programs offered by an ACP
Facilitator to existing faith (11 = 6), health-related layperson (n = 1), and employee (n = 4)
community groups in Greater Lansing and Grand Rapids, Michigan.

The inclusion criteria for this study required participants who: 1) were 18 years
or older because a potential product outcome of the ACP program is a legal document;
i.e., an advance directive; 2) attended the entire ACP program (if only one session) or at
least the first and last session of the ACP program (if more than one session); 3) were
able to complete the surveys, which required the ability to read and write in English;

4) completed both the pre— and post-program surveys, and 5) created a personal
identification number for use on both pre- and post-surveys. Participants were excluded
from this study if they 1) did not complete both the pre- and post-program surveys or 2)
did not write their personal identification number on both pre- and post-program surveys.
Measure — Pre- and Post-Program Survey Instruments

The data used for this study are a subset of items from the Participant Survey.
Each item relates to a concept in the Focused Model of the Decision to Engage in ACP
(see Figure 2, p. 23). Table Cl lists the Participant Survey items, related conceptual
model variables, data collection timing, and item type (see Appendix C).

Human Subject Considerations

Since the MMACPC Research Project continues to have IRB approval, a revision

form was submitted to the Community Research Institutional Review Board (CRIRB) to

add personnel for data entry and analysis consultation and to outline data management

72

precautions for this study (see Appendix C). All personnel involved in conducting this

study have current IRB training certification.

Data Management

Preparation for Data Entry. This investigator prepared the data to be used in this
study to facilitate data entry and analysis. First, photocopies of the pre- and post-program
surveys were made and the items from the Participant Survey to be used in this project
were highlighted on the copies. The alphanumeric personal identification numbers were
matched for pre- and post-surveys and given a common 4-digit identification number
(PERSID). Then a codebook of survey questions, coding, and the corresponding SPSS
template fields was created. F inally, an SPSS template was constructed with variables
and the PERSID numbers for both pre- and post-surveys. The copies, codebook, and
SPSS template were brought to CSTAT for actual data entry and preliminary analysis by

student consultants.

Procedure for Verifying Data Integrit. The data used in this dissertation study
were entered into the SPSS database by two students not previously associated with the
MMACPC Research Project. This double data-entry procedure fully meets the best
practice standard for data management prescribed by the National Institutes of Health
(Stommel & Wills, 2004). Once both sets of data were entered into SPSS, the data were
checked for discrepancies, illegal, and missing values. Correlations between the two data
sets were checked for values of r < 1.0000; no errors were detected so no corrections
were required. Had discrepancies or illegal values been found, the original surveys
would have been checked to correct the data. Missing values were examined for patterns,

such as skipping patterns related to preceding questions on the survey. Once all the data

73

were verified, the pre- and post-survey data files were merged by means of the unique
personal ID, the number that was used to match each participant’s pre- and post-survey

data. These data comprise the master data set for this dissertation study.
Data Analysis

SPSS® version 12.0 statistical software was used for analysis of data in this
study. The master data file was used to generate frequency distributions for each
variable. Distributions were examined for symmetry and measures of central tendency
and variability. The assumptions of normality were checked to determine whether
parametric or non-parametric statistical tests would be used. The statistician provided
guidance related to appropriate data transformations, significance levels, and statistical

tests. Each research question was analyzed as discussed in the following sections.

Research Question 1. Prior to their participation in a community-based ACP
program, what are the self-reported Personal Influences on Decision and Personal
Decision Factors of adults who have decided to engage in ACP?

The frequency distributions of variables from the pre—pro gram survey were
reported for all of the participants (11 = 147). Assumptions of independent samples and
expected counts were checked and statistical significance adjustment, using Bonferroni’s
correction, was made when applicable. Then, using crosstabulation and chi-square
analysis, the frequencies of the prior experiences of having an AD and having a HCA
were compared with the frequencies of having conversations. Next, the reported
conversation frequencies were compared between the HCA, significant others, and HCP
conversation partners. Lastly, participants’ perceptions of the quality of conversations

were compared between the conversation partners (HCA, significant others, HCP).

74

Research Question 2. Which variables from among the Personal Influences on
Decision and Personal Decision Factors are associated with the Decision to Engage in
ACP?

The pre-program survey data was used for analysis. Correlations were calculated
between each variable from the “Personal Influences on Decision” and the “Personal
Decision Factor,” depicted in the focused model of the Decision to Engage in ACP (see
Chapter 2, p. 25). Each pair of variables was examined for correlation strength and
statistical significance. Since multiple comparisons were used for each variable,
statistical Significance was adjusted using Bonferroni’s correction.

Research Question 3. Do the Personal Influences on Decision and Personal
Decision Factor variables that are associated with the Decision to Engage in ACP differ
between participants with previous ACP experience compared to participants without
previous ACP experience?

The pre-program survey data was analyzed in subgroups on the basis of the ACP
Experience variable, which was derived from responses to the Having an AD and Having
a HCA survey items. Data from participants who did not answer both of these items was
excluded for this analysis. The frequency distributions of the interval level variables
(enough knowledge, importance, ability, and likelihood) were checked for normality and
homogeneity of variance and the non-parametric Mann-Whitney U-test was used to
compare the groups. Next, correlations of all the personal influence and decision factor
variables were calculated to determine correlation strength and statistical Significance of
the variable pairs. Statistical significance was set at p < 0.05, with Bonferroni’s

correction for multiple comparisons.

75

Research Question 4. After participation in a community-based ACP program,
are there changes fi'om pre— to post-ACP program intervention in self-reported ratings of
enough knowledge, importance, ability, and likelihood to have an ACP conversation?

A matched pre-program/post-program survey dataset was used for data analysis
and each pre- and post-program rating of knowledge, importance, ability, and likelihood.
After checking for assumptions of normality, each pre- and post-program variable was
compared using the non-parametric Wilcoxon Matched Pair Signed-Ranks test.

Research Question 5. Do the changes between participants’ self-reported pre- and
post-ACP program intervention ratings of enough knowledge, importance, ability, and
likelihood differ between participants with previous ACP experience compared to
participants without previous ACP experience?

The matched pre-program/post—program dataset was then analyzed by subgroups
on the basis of the ACP Participant Experience variable. The pre- and post-program
ratings of knowledge, importance, ability, and likelihood were compared for each ACP
Experience group using the Wilcoxon Matched Pairs Signed-Ranks test. Difference
scores were calculated for each group of participants to examine the impact of the ACP
education program intervention on each group’s participants’ perceptions of their
knowledge, importance, ability, and likelihood of having ACP conversations. Difference
score distributions were checked for normality and homogeneity of variance, then

compared using the non-parametric Mann Whitney U test.

76

Chapter Summary
The purpose of this dissertation study is to add to the understanding of internal
(e.g., self-efficacy) and external (e.g., ACP program) influences within the decision-
making context that may affect individuals’ decisions to engage and re-engage in ACP.
A secondary analysis of matched pre- and post-program survey data will contribute to an
understanding of associations between internal influences and the decision to engage in
ACP, the impact of an ACP program intervention on internal influences, and the potential

modifying effects of previous ACP experience.

77

Chapter 5
RESULTS

This secondary analysis of matched pre- and post-program survey data was
designed to describe the frequency and associations between influences within the
decision—making context that are present for participants who decided to engage or re-
engage in ACP by attending an ACP program. The participants in this study were 147
adult men and women who attended an ACP program offered by an ACP Facilitator to an
existing community group of which they were members. Following a brief overview of
preliminary data analysis, the results for each research question will be presented.

Preliminary Analysis of Data
Frequencies and missing data

The data were examined with univariate descriptive statistics to determine
frequencies of valid responses and patterns of missing data. For all but four variables,
less than three percent of the frequency counts in the pre-program survey data were
missing data. The valid responses for each variable were examined to identify
appropriate statistical tests for determining associations between variables in later
research questions.

Three patterns of missing data were found: 1) a survey item not answered which
could be validly scored from a response elsewhere on the survey, 2) survey items which
were answered with the possible response of “I’m not sure,” and 3) survey items which
were contingent upon a previous item’s “Yes” response. In the first case, one participant
who did not answer the gender survey item identified her gender in her personal

identification number. Her gender was included in the valid reported results.

78

For the second case, all of the Prior Experiences variables included the possible
response of “I’m not sure.” Except for basic reporting of frequencies, only the “Yes” and
“No” responses are treated as valid data and the “I’m not sure” responses are treated as
missing.

In the third case, four Prior Experience items—Having conversations with HCA
and the three Quality of Conversations items—were designed to be contingent upon a
“Yes” response to the items immediately preceding them, which explains the greater
frequency of missing data for these items. However, three of the follow-up questions
were answered by some participants even though they had answered “No” to the
preceding items (see Table 1). The decision was made to include these participants’
responses as valid data.

Table 1

Comparison of Valid Data Percentages with and without Contingency Items

 

 

 

 

Valid data Responses added Valid data
Focus Model Variable if responses contingent if not contingent if all responses
on previous item on previous item counted
% n %

Having Conversations with 61.9% 12 72.1%
HCA

Quality Conversations with 61.2% 2 62.6%

Significant Others

Quality Conversations with 15.0% 1 15.6%
HCP

 

 

 

 

 

 

Normality & Choice of Statistical Tests
The four ordinal/interval variables of Knowledge, Importance, Ability, and
Likelihood were examined for normality of distribution in both pre- and post-program

survey data. Because none of the variables meet the criteria for normality (based on

79

Shape, skewness, or kurtosis), central tendency will be reported as the median and
variance as the range. Bivariate analysis of variables in this study will be conducted
using non-parametric (or “distribution-free”) statistic tests for four reasons: 1) Non-
parametric tests do not depend on distributions meeting assumptions of normality.

2) Non-parametric tests can be used with small samples (although they tend to need a
larger sample to detect statistical Significance). 3) Non-parametric tests are able to used
with data that is distorted from a restricted range of scores. 4) Non-parametric tests are
able to test differences in central tendency without being affected by outliers

(Howell, 2002).

When more than one relationship will be calculated for a variable, Bonferroni’s
correction will be used to adjust the statistical significance to the number of comparisons
being made. This prevents making a Type 1 error—declaring statistical significance to
exist when it really does not—which may occur if multiple comparisons are being made,
each at an alpha level of 0.05. For example, in Research Question 2, when Spearman rho
correlations will be calculated for every pair of variables, statistical significance will be
set at p = 0.05 / number of comparisons for each variable. Since there are 12 variables
and each will be matched with all the others, the statistical significance will be set at
(0.05/11 = 0.004).

Recoding of Quality of Conversations Items

In the pre-pro gram Participant Survey, the Quality of Conversations items were
presented as S-point ordinal/interval scales with descriptors of 1 = few comments only; 3
= good conversation; and 5 = complete and thorough conversation. Because this scale

appeared to mix concepts of quantity (few comments only) with concepts of quality

80

(good, thorough), these items were recoded to two nominal/ordinal categories (1 and 2 =
few comments; 3, 4, and 5 = thorough conversation).
Research Question 1

Prior to their participation in a community-based ACP program, what are the self-
reported Personal Influences on Decision and Personal Decision Factors of adults who
have decided to engage in ACP?

The frequencies of all variables from the pre-pro gram survey data are presented in
Table 2, in the sequence of the Focused Model of the Decision to Engage in ACP (see
Figure 2, p. 25).
Results

Enough knowledge. The distribution of participants’ ratings of having enough
knowledge to discuss their views was examined using a histogram. The distribution was
bimodal with modes of 3 and 5 and a kurtosis coefficient = - 2.32, which is too flat to be
normal. The median score was 4.0 with a range of scores from 1 to 5. Thirty percent of
the participants reported very much knowledge (5) and 30% reported knowledge at the
midpoint (3). Less than 5% of the participants reported having no knowledge at all (1).

Age and gender. Crosstabulation was used to examine the age and gender of the
study participants. There were participants from every age range; however, those in the
18-39 years range comprised only 1.4% of the total. Participants who were in the 31-59
years range were the most frequent with 58.8% of the men and 78.4% of the women

participants included in this category. Participants who were > 75 years comprised 7.6%

81

Table 2

Frequencies of Personal Influence on Decision and Personal Decision Factor Variables

in the Pre-Program Survey Data for the Total Sample

 

 

 

 

 

 

 

 

Focus Model Variables Responses 11 %
Enough knowledge to discuss 1 = not at all 6 4.1
my views 2 22 15.0
3 44 29.9
4 30 20.4
5 = very much 44 29.9
Missing 1 0.7
Age 18-30 2 1.4
31-59 107 72.8
60-75 25 17.0
> 75 11 7.5
Missing 2 1.4
Gender Male 34 23. 1
Female 1 13 76.9
Important to have a discussion 1 = not at all 1 0.7
2 3 2.0
3 15 10.2
4 23 15.6
5 = very much 103 70.1
Missing 2 1.4
Having an AD No 81 55.1
(Advance Directive) Yes 55 37.4
I’m not sure 7 4.8
Missing 4 7.5
Having a HCA No 52 35.4
(Health Care Agent) Yes 91 61.9
I’m not sure 3 2.0
Missing 1 0.7
Having Conversations with No 20 13.6
HCA Yes 83 56.5
I’m not sure 3 2.0
Missing 41 27.9

 

82

Table 2 (continued)

 

 

 

 

 

 

 

 

 

Focus Model Variables Responses 11 %
Quality Conversations with Few comments 23 15.6
HCA Thorough 60 40.8
conversation
Missing 64 43.5
Having Conversations with No 51 34.7
Significant Others Yes 90 61.2
I’m not sure 4 2.7
Missing 2 1.4
Quality Conversations with Few comments 40 27.2
Significant Others Thorough 52 35.4
conversation
Missing 55 37.4
Having Conversations with No 124 84.4
HCP (Health Care Provider) Yes 22 15.0
I’m not sure 0 0
Missing . l 0.7
Quality Conversations with Few comments 4 2.7
HCP Thorough 19 12.9
conversation
Missing 124 84.4
Making decision for someone No 85 57.8
else Yes 59 40.1
I’m not sure 1 0.7
Missing 2 1.4
Able to start discussion with l = not at all 4 2.7
others 2 7 4.8
3 35 23.8
4 40 27.2
5 = very much 60 40.8
Missing 1 0.7
Likely to have conversation I = not at all 0 0
with others 2 8 5.4
3 29 19.7
4 30 20.4
5 = very much 78 53.1
Missing 2 1.4

 

83

of the total. Women comprised 76.6% of the participants and were the only gender
represented in the 18-39 years range. Participants from the > 75 years category were
almost equally divided between men and women.

Importance. The distribution of participants’ ratings of feeling it is important to
have a discussion was examined using a histogram. The median score for this
distribution was 5.0 with a range of 1 to 5; however < 1% of participants rated
importance as “not at all” (1) and only 2% of participants rated it as “2.” On the basis of
these four outlier scores, the distribution was skewed to the left (skewness coefficient = -
9.228); it was also too peaked to be normal (kurtosis coefficient = 7.65). Seventy percent
of participants reported feeling that having a discussion was very much important (5).
Responses to this variable were the most homogenous of all the distributions.

Prior Experiences. Participants’ reported prior experiences included having an
AD (37%); having a HCA (62%); having conversations with HCAS (56.5%), significant
others (61.2%), and HCPs (15%); and making a decision for someone else (40%).
Sixteen of the study participants (10.9%) reported they were “not sure” whether they had
had these prior experiences: Having an AD (11 = 7); Having a HCA (n = 3); Having
conversations With HCAS (n = 3), or significant others (n = 4); Making a decision for
someone else (11 = 1). One participant was “not sure” about three types of experiences,
two participants about two types, and the remaining 13 about one type. These data
indicate that some participants were unsure about these concepts prior to the ACP
education program intervention.

Of the 53 participants who reported making a decision for someone else, 96% of

the decisions were made for an adult, < 2% for a child, and < 2% for both an adult and

84

child. F ifiy-six participants reported their feelings of preparedness to make a decision:
73.2% felt prepared, 17.9% did not feel prepared, and 8.9% were not sure.

Crosstabulation was used to further describe participants’ prior experiences. Of
the 134 participants who answered the Having an AD and Having a HCA items, 38.8%
reported having both an AD and a HCA. Of the 100 participants who answered the
Having a HCA and Having conversations with their HCA items, 77.0% reported having
conversations with their HCA.

The quality of conversations with partners was reported by varying numbers of
participants. Quality of Conversations with Health Care Agent was reported by 83
participants with 72.2% rating the quality as “thorough.” Quality of Conversations with
Significant Others was reported by 92 participants, almost equally divided between
ratings of “few comments” (43.5%) and “thorough” (56.5%). Quality of Conversations
with Health Care Provider was reported by 23 participants with 82.6% rating the quality
as “thorough.”

Ability. The distribution of participants’ ratings of feeling able to start a
discussion with others was examined using a histogram. This distribution was skewed to
the left (skewness coefficient = -4.114), reflecting a few scores at the low end of the
rating scale. The median score was 4.0 with a range of scores from 1 to 5. Only 2.7% of
the participants felt not at all able (1), while 40.8% felt very much able (5) to start a
discussion with others. Ratings of 3 and above were reported by 91.8% of the
participants.

Likelihood. The distribution of participants’ ratings of their likelihood to have a

conversation with others was examined using a histogram. This distribution was skewed

85

to the left (skewness coefficient = - 4.274), with only 5.4% of the scores at the minimum
of the range from 2 to 5. This range is the most restricted of the distributions, with a
median score of 5.0. None of the participants rated themselves as “not at all” likely to
have a conversation (1); more than half (53. 1 %) rated themselves as “very much” likely
(5). Ratings of 3 and above were reported by 93.2% of the participants.

Further analyses

Relationships between Having an AD, Having a H CA, and Having Conversations.
Only two studies reviewed for this dissertation reported an association between the prior
experiences of having an AD or having a HCA with conversations. Therefore, the prior
experiences of Having an AD and Having a HCA were examined further individually for
relationships with the Having Conversations variables using the Chi-Square test. Since
three comparisons were being made for each variable, statistical Significance was
adjusted for each analysis using Bonferroni’s correction (0.05/3 = 0.017).

The data provide significant evidence (p < 0.017) to indicate there are
relationships between having an AD and having conversations, in order of strength, with
a HCPf(1,N= 135) = 23.54, p < 0.001; with significant others {(1, N= 133) = 12.94, p
< 0.001 ; and with a HCA f0, N = 94) = 7.71, p = 0.007. To further explore the degree
to which Having an AD influences the Having Conversations variables, odds ratios were
calculated for each relationship. For participants who reported having an AD, the odds of
having conversations with a HCP were 14.11 times greater (95% CI 3.919, 50.829); with
Significant others, 4.42 times greater (95% CI 1.908, 10.253); and with a HCA, 4.80
times greater (95% CI 1.417, 16.260) than for participants who did not report having an

AD.

86

Next the variable Having a HCA was examined for relationships with the Having
Conversations variables using the Chi-Square test. A comparison between participants
who reported Having a HCA and Having Conversations with HCA could not be done
because the assumption of independence was not met (77.0% of the participants who
answered “Yes” to having a HCA also reported having conversations with a HCA). The
comparison of Having a HCA and Having Conversations with Significant Others could
also be problematic Since it is possible that some participants’ health care agents might
also be their significant others. However, Since the survey item did not explore this
potential overlap, the comparison will be reported as if it is independent of the Having a
HCA variable. Since two Conversation variables are being compared with the Having a
HCA variable, Bonferroni’s correction will be used to adjust the value for statistical
Significance (0.05/2 = 0.025).

The data provided significant evidence (p < 0.025) to indicate there are
relationships between having a HCA and Having Conversations, in order of strength,
with Significant Others {(1, N: 138) = 23.092,p < 0.001 and with a HCP [(1, N= 142)
= 8.500, p = 0.004. Odds ratios were also calculated for each relationship. For
participants who reported having a HCA, the odds of having conversations with
Significant others were 6.03 times greater (95% C12.813, 12.917) and with a HCP, 7.14
times greater (95% CI 1.597, 31.953) than for participants who did not report having a
HCA.

Participants ’ Reported Conversation Partners. In the ACP literature reviewed
for this study, there were relatively few studies that reported the experience of having

conversations and only one that compared frequencies of conversations with three

87

different partners (HCA, significant others, and HCP). Thus, participants’ reported

conversations were examined by crosstabulation and are presented in Figure 3.

 

      
 

HCA+ $0 + HCP

W

17
SO+HCP only 0
HCA+HCPonIy .

H<=A+1~=°<m~\w

HCP only

0

 

l

SOonly W 7
“CW" \14

.... W"

O 10 20 30 40 50 60
Percentage of Participants

 

 

 

 

 

 

 

 

Figure 6. Percentage of participants who reported conversations with partners.

Note. HCA = Health Care Agent, SO = Significant Others, HCP = Health Care Provider

Ninety-eight participants answered the pre-program survey items about Having
Conversations. About 50% reported having a conversation with their HCA and
significant others. As discussed previously, there may be overlap between conversations
with the HCA and the Significant others. Another 17% reported having a conversation
with their HCA, significant others, and HCP. However, 12% reported having no

conversations at all; 14% reported conversations only with their HCAs; and 7% only with

88

their significant others. Except for the participants who reported having all three types of
conversations, participants’ conversations with Health care providers were almost non-
existent.

Perceptions of Quality of Conversations. Perceptions of the quality of
conversations have not been reported in the literature reviewed for this study. Figure 7

depicts participants’ reports of the quality of conversations had with HCAS, Significant

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

others, and HCPS.
45
D Few Comments
40
IThorough Conversation
35
so
5 25
t 20
15
10
5
o
HCP

 

Figure 7. Percentage of participants reporting quality of conversations with conversation
partners. Note. HCA = Health Care Agent, S0 = Significant Others, HCP = Health Care
Professional.

With all three conversation partners, participants reported the quality of their

conversations more often as “thorough” than “few comments.” However, because the

89

actual fi'equencies of the conversations varied significantly, the magnitude of the reported
quality for conversation types can be more accurately depicted using proportional
percentages. Figure 8 shows that the few participants who had conversations with a HCP

rated this conversation type to have the most thoroughness.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

90
0 Few Comments
80
Thorough Conversation
70
60
E w
40
30
20
1° \
o S

 

 

 

 

Conversation Partner

Figure 8. Proportional percentage of participants reporting quality of conversation with
conversation partners. Note. HCA = Health Care Agent, SO = Significant Others,
HCP = Health Care Professional.
Summary

All of the personal influences on decision and personal decision factor variables
were present for at least some of the participants who completed the pre-program
surveys. The distributions of Knowledge, Importance, Ability, and Likelihood were all
significantly different than a normal distribution. Importance and Likelihood also

demonstrated a restricted range of values. Statistically significant relationships were

90

found between the prior experiences of having an AD or having a HCA and having
conversations with all partners. Conversations with partners vary significantly in
fiequency, with Conversations with health care providers being very minimal. The
combination of Conversations with HCA and Significant Others is the most frequent but
may not be independent. Conversation quality was also explored, with participants rating

conversations with Health care providers to be proportionally the most thorough.

Research Question 2

Which variables from among the Personal Influences on Decision and Personal
Decision Factors are associated with the Decision to Engage in ACP?
Correlations

All participants in this study have, by definition, decided to engage in ACP by
attending the ACP program. Thus, the focus of this research question is on the
relationships between the variables which are thought to influence this decision. These
relationships were examined using the Spearman rank correlation coefficient (rs), a non-
parametric test for correlation (see Table 3). The Quality of Conversations variables
were omitted from this analysis because for each a significant percent of the data was
missing (with HCA = 43.5%; with Significant Others = 37.4%; with HCP = 84.4%).
Because multiple comparisons will be made between each variable and the remaining 11
variables, statistical Significance was adjusted using Bonferroni’s correction

(0.05/11 = 0.004).

91

333258 26: 3 33:4 u 8:4 838586 gm 3 012 u oS< mom? 05228
How :oEooQ wee—22 u can: 23380805 0.80 53on fit» ”838.8380 u @030 £050 385:me 55, 533.5300 u Om-O 3.0m
63, 5330280 u <UE-U 35309.5 n 9.3 ”c0850 u “6:00 momma—Bog nwsocm u Bog £335: gamma: n Go ”wood w a ... .682

 

 

83 .83 N83 .33 .53 .83 .83 «N3 .33 83 «3.8 .83 83
83 23 .83 .83 83 .83 .23 .83 888 2.8.8 .33 28..
83 .83 ~83 83 83 83 83 68.8 .83 83 83
83 83 S3 .33 .83 83 88.8 .83 .83 83-0
83 .23 .83 .23 .83 53 83 .33 8-0
83 .833 83 .83 83 83 33 5:6
83 .E3 .83 28.8 83 .83 SE
83 83 82.8 .83 .83 9..
83 83 62.8 .63 as
83 82.8 83 8.0
83 23 um...
83 .5qu
83 as... 882 mono 8-0 <83 <0: 9.. as 8.0 cm... 32.8

 

63.53% 388 xokwfinetek 3.83% noEomQ 33.26% ~88 eommmomQ :0 86:63.89 ~39.on I wuemwowwmob 26:33.80 «56% 23.53%.

m 2an

92

Results

There were 32 pairs of variables with statistically significant correlations (p _<_
0.004) and 34 pairs of variables which were not statistically significant. Descriptions of
correlation strength were drawn from Munro (2001). The results are discussed as
relationships between variables, in the order of greatest to least strength.

Two pairs of variables, with a significance of p < 0.001, were categorized as
“highly” correlated; that is, with a correlation value of 0.70 to 0.89. Both involve a
personal influence on decision variable and a personal decision factor variable:
Importance and Likelihood (rs = 0.744) and Knowledge and Ability (rS = 0.737). The
strength of the first correlation is certainly influenced by the restricted ranges of values
for both of the variables, Importance and Likelihood. However, it is still likely that there
is a strong relationship between this pair of variables based on concepts from decision-
making theory about values and intentions. The correlation suggests that participants’
intentions to have discussions are strongly related to the perceived value of having
discussions. The correlation between knowledge and ability suggests that participants’
who report having enough knowledge to have discussions also report feeling confident
that they will be able to start those conversations.

Three pairs of variables, all with a significance of p < 0.001, are categorized as
“moderately” correlated (rs = 0.50-0.69): Having an AD and Having a HCA (rs = 0.571);
Ability and Likelihood (rs = 0.529); and Ability and Importance (rs = 0.497). Since
initiating an AD often includes designating a HCA, it is not surprising that these variables

are associated. That the correlation is only “moderate” may be explained because the

93

perCentage of participants reporting an AD was only 37% compared to 62% who reported
having a HCA. The second pair—Ability and Likelihood—are the two variables in the
personal decision factor Self-Efficacy, so association would be expected. However, since
correlation is only moderate, they may each capture a distinct feature of self-efficacy.
Finally, Importance and Ability represent another pairing of variables from the personal
influences on decision and personal decision factors. The strength of a participant’s
value of having discussions may be tempered by the perception of one’s ability to start
these conversations.

Correlations are categorized as “low” if the correlation coefficient is between 0.26
and 0.49. Twenty-seven pairs of variables fall into this category. The four strongest
associations, at significance levels of p < 0.001, all contain 3 Having Conversations
variable: Having a HCA and Conversation with HCA (rs = 0.449); Having an AD and
Conversation with HCP (rs = 0.418); Conversation with HCA and Conversation with
Significant Others (rs = 0.412); and Having a HCA and Conversation with Significant
Others (rs = 0.409). The first, second, and fourth correlations are supported by the earlier
discussion of the odds ratio of having conversations when participants had either an AD
or a HCA. The third correlation is supported by the earlier discussion of participants’
reported conversations, where conversations with HCA and significant others accounted
for 50% of the conversations.

Four of the next six correlations in close range, all at significance levels of p <
0.001, contain the variable Knowledge: Knowledge and Likelihood (rs = 0.399),
Knowledge and Having a HCA (rs = 0.373), Knowledge and Conversation with HCP (rs =

0.363), and Knowledge and Importance (rs = 0.361). These may speak to the need for

94

knowledge to be able to construct and communicate preferences. Another correlation,
Having an AD and Age (rs = 0.362), demonstrates a directional association between
initiating an AD and older age. Finally, Conversation with HCP and Making a Decision
for someone else (rs = 0.358) demonstrate an association between communicating with
health care professionals for oneself or on behalf _of someone else.

Three of the next four correlations, all with significance levels of p < 0.001,
include the variable Ability: Conversations with HCP and Ability (rs = 0.340), Having an
AD and Ability (rs = 0.335), and Having a HCA and Ability (rs = 0.325). This may
indicate the importance of having confidence in one’s ability to be able to formally
construct and communicate preferences. The other correlation in this range is
Knowledge and Having an AD (rs = 0.336), again supporting the idea that knowledge
may be necessary to initiating a formal record of preferences.

The next four correlations, at significance levels of p S 0.001, include a
Conversation variable: Conversation with HCA and Likelihood (r8 = 0.326),
Conversation with Significant Other and Having an AD (rs = 0.313), Conversation with
HCP and Age (rs = 0.302), and Conversation with Significant Other and Ability (rs =
0.300). This may speak to the influence of feelings of self-efficacy, formal construction
of preferences, and older age on communicating preferences to others. A fifth correlation
in this range is Having a HCA and Likelihood (rs = 0.307), which may indicate the
influence that having formally designated someone to make decisions has on a person’s
intent to have discussions.

Of the last eight correlations, at significance levels of p 5 0.004, six include

conversation variables: Conversation with Significant Others (rs = 0.291) and

95

Conversation with HCP (rs = 0.291) are paired with Likelihood; Conversation with HCA
is paired with Importance (rs = 0.285); Conversation with HCA and Having an AD (rs =
0.276); Conversation with HCA and Knowledge (rs = 0.257); and Conversation with
Significant Others and Knowledge (rs = 0.285). These may speak to the influence of
feelings of self-efficacy, value, and knowledge on people’s willingness to communicate
preferences to others. The relationship of Age and Making a Decision for someone else
(rs = 0.286) likely represents the directional association of growing older with the
experience of needing to make decisions for others, such as one’s parents or
grandparents. Finally, there is a correlation between Importance and Having a HCA (rs =
0.263). This may speak to the need for persons to value having a discussion and actually
designating someone to speak for them.

The remaining 34 variable pairs fall into the “little, if any” category of correlation
strength (rs = 0.00 to 0.25), with none meeting statistical significance at the p 5 0.004
level. Fifty-three percent of these non-significant pairs included the individual
characteristic variables of gender (11 = 11) or age (n = 7). This may suggest that these
demographic variables are not likely to influence many of the other variables in the
model. Prior experience variables, Making a Decision for someone else (n = 8) and
Conversation variables (n = 8), accounted for another 47% of the non-significant pairs.
The two remaining pairs included the variables Importance or Likelihood.

Summary

Forty-seven percent of the variable pairs demonstrated statistically significant

correlations. The most highly correlated variables are participants’ perceptions of 1) the

importance and likelihood of having discussions and 2) having enough knowledge and

96

feeling able to start a discussion. Moderate correlations were found between the
perception of being able to start a discussion and the perceptions of the importance and
likelihood of the conversation. A moderate correlation was also found between the prior
experiences of having an advance directive and having a health care agent. Low but
significant correlations were found between 27 of the variables, with 63% of these
including a Conversation variable (11 = 17). More than 50% of the 34 non-significant
correlations included the demographic variables gender (n = 11) or age (11 = 7).
Research Question 3

Do the Personal Influences on Decision and Personal Decision Factor variables
that are associated with the Decision to Engage in ACP differ between participants with
previous ACP experience compared to participants without previous ACP experience?

. Participants’ answers to the Having an AD or Having a HCA survey items were
examined to determine who had previous experience in constructing or communicating
their preferences for end of life decision—making. Participants who reported having
neither were classified as having No ACP Experience while those having an AD, having
a HCA, or having both an AD and HCA were categorized as having ACP Experience.
Participants with missing data for either of the AD or HCA items (11 = 13) were excluded
from analysis for this research question. The frequencies of all variables are presented in
Table 4 for participants with (n = 85) or without (11 = 49) ACP Experience.

Results
Enough knowledge. The distributions of participants’ ratings of having enough

knowledge to discuss their views were examined using histograms. For the ACP

97

Table 4

Frequencies of Personal Influence on Decision and Personal Decision Factor Variables

in the Pre-Program Survey Data for Participants Classified by ACP Experience

 

 

 

 

 

 

 

 

 

ACP No ACP
Experience Experience

Focus Model Variables Responses (n = 85) (n = 49)
n % n %
Enough knowledge to discuss l = not at all 3 3.5 3 6.1
my views 2 10 11.8 12 24.5
3 17 20.0 19 38.8
4 21 24.7 8 16.3
5 = very much 33 38.8 7 14.3
Missing 1 1.2 0 0
Age 18-30 0 0 2 4.1
31-59 59 69.4 41 83.7
60-75 16 18.8 5 10.2
> 75 8 9.4 1 2.0
Missing 2 2.4 0 0
Gender Male ' 20 23.5 9 18.4
Female 65 76.5 40 81.6
Important to have a discussion 1 = not at all 1 1.2 0 0
2 3 3.5 0 0
3 3 3.5 11 22.4
4 9 10.6 9 18.4
5 = very much 67 78.8 29 59.2
Missing 2 2.4 0 0
Having an AD No 30 35.3 49 100
(Advance Directive) Yes 55 64.7 0 0
Having a HCA No 3 3.5 49 100
fiealth Care_Agent) Yes 82 96.5 0 0
Having Conversations with No 8 9.4 7 14.3
HCA Yes 74 87.1 4 8.2
I’m not sure 1 1.2 0 0
Missing 2 2.4 38 77.6
Quality Conversations with Few comments 18 21.2 2 4.1
HCA Thorough 56 65.9 2 4.1

conversation

Missing 11 12.9 4 8.2

 

98

Table 4 (continued)

 

 

 

 

 

 

 

 

ACP No ACP
Experience Experience
Focus Model Variables Responses (11 = 85) (n = 49)
n % n %
Having Conversations with No 17 20.0 30 61.2
Significant Others Yes 66 77.6 18 36.7
I’m not sure 1 1.2 1 2.0
Missing 1 1.2 0 0
Quality Conversations with Few comments 20 23.5 14 28.6
Significant Others Thorough 46 54.1 5 10.2
conversation
Missing 19 22.4 19 38.8
Having Conversations with No 64 75.3 47 95.9
HCP (Health Care Provider) Yes 20 23.5 2 4.1
I’m not sure 1 1.2 0 0
Missing 0 0 0 0
Quality Conversations with Few comments 4 4.7 0 0
HCP Thorough 1 7 20.0 2 4. l
conversation ,
Missing 64 75.3 47 95.9
Making decision for someone No 49 57.6 30 61.2
else Yes 34 40.0 19 38.8
I’m not sure 1 1.2 0 0
Missing 1 1.2 0 0
Able to start discussion with 1 = not at all 2 2.4 2 4.1
others 2 3 3.5 3 6.1
3 14 16.5 1 8 36.7
4 23 27.1 15 30.6
5 = very much 42 49.4 11 22.4
Missing 1 1.2 0 0
Likely to have conversation I = not at all 0 0 0 O
with others 2 5 5.9 2 4.1
3 9 10.6 16 32.7
4 12 14.1 14 28.6
5 = very much 57 67.1 17 34.7
Missing 2 2.4 0 0

 

99

Experience group, the median score was 4.0 with a range of scores from 1 to 5. The
distribution was skewed to the lefi (skewness coefficient = - 2.627), with only 15.3% of
the responses at a score of 1 or 2, but had a normal kurtosis. For the No ACP
Experience group, the distribution was normal (skewness coefficient = 0.615; kurtosis
coefficient = 0.095), with a median score of 3.00, and a range of scores from 1 to 5.
Because the ACP Experience distribution was skewed, the distributions were compared
between the groups using the Mann-Whitney test, which showed a significant difference
in perceptions of knowledge between the Experience groups, U(133) = 1305.5, p < 0.001.
Age and gender. Crosstabulation was used to examine the ages and genders of
the participants. There were men and women participants in both ACP Experience
groups with women outnumbering men (ACP Experience: 75.9% women to 24.1% men;
No ACP Experience: 81.6% women to 18.4% men). In the ACP Experience group, there
were no participants in the 18-30 age range but the No ACP Experience group had two
women participants (4.1%). In the > 75 age range, the ACP Experience group had four
men and four women (9.6%) but the No ACP Experience group had none. In both
groups, the 31-59 years range was the most frequent (ACP Experience = 71.1%; No ACP
Experience = 83.7%). Because some of the crosstab cells had no values, a Likelihood
Ratio was used to compare the groups in terms of age and there was a significant
difference, LR(3, n = 132) = 9.430, p = 0.041. Since both experience groups had men
and women participants, the Pearson Chi-Square test was used to compare the groups in
terms of gender and there was no significant difference, fa, N = 134) = 0.488, p =

0.485.

100

Importance. The distributions of participants’ ratings of feeling it is important to
have a discussion were examined using histograms. The median scores for both groups
were 5.0 but the range of scores was 1 to 5 for the ACP Experience group and 3 to 5 for
the No ACP Experience group. Although the range for the ACP Experience group
appears to be less restricted than the No ACP Experience group, this was not the case.
The ACP Experience group’s distribution was severely skewed to the left (skewness
coefficient = - 10.432), with only 8.2% of this group’s participants rating Importance at a
score 5 3. The ACP Experience group’s distribution was also very peaked (kurtosis
coefficient = 14.243). The No ACP Experience group’s distribution was also skewed to
the lefi (skewness coefficient = — 2.338) but not as severely (22.4% of participants rated
Importance as a 3) and kurtosis was normal. Ratings of “very much” (5) were given by
78.8% (ACP Experience) and 59.2% (No ACP Experience) of the participants. Because
both distributions were skewed, the distributions were compared using the Mann Whitney
test, which showed a significant difference in the perceptions of importance between the
ACP Experience groups, U(132) = 1599.5, p = 0.009.

Prior Experiences. The last category of personal influences on decision variables
includes the prior experiences of Having an AD, Having a HCA, Having conversations,
Having quality conversations, and Making a decision for someone else. The first two
variables, Having an AD and Having a HCA, were used to create the “ACP Experience”
variable. Therefore, only the participants in the ACP Experience group reported prior
experiences with having an AD (64.7%) and having a HCA (96.5%). Of the 85

participants who responded to these items, 61.2% reported having both. By definition,

101

none of the participants in the No ACP Experience group reported Having an AD or
Having a HCA.

According to the survey item instructions, only the ACP Experience group
participants should have indicated if they had conversations with their health care agents
and 87.1% reported they had. However, 22.5% of the No ACP Experience group
participants also answered this question with 8.2% reporting conversations with a health
care agent. None of the participants in either group reported being unsure about whether
they had had a conversation with a health care agent. As would be expected, a significant
difference was found between the EXperience groups related to the Conversation with
HCA variable using Fisher’s Exact test (p < 0.001).

In the ACP Experience group, 77.6% of the participants reported having a
conversation with significant others, compared to only 36.7% of the No ACP Experience
participants. One person from each Experience group reported they were not sure if they
had a conversation with their significant others. Using the Pearson Chi-Square test, a
significant difference was found between the Experience groups related to the
Conversation with Significant Others variable, {(1, N= 131) = 23.339, p < 0.001.

As seen in the results for Research Question 1, conversations with Health care
providers were very many fewer for both Experience groups: 23.5% for ACP Experience
and 4.1% for No ACP Experience. Only one person from the ACP Experience group was
unsure about this type of conversation. Using the Pearson Chi-Square test, a significant
difference was found between the Experience groups related to the Conversation with

HCP variable, {(1, All—- 133) = 8.725, p = 0.003.

102

Participants from both groups reported having the prior experience of Making a
Decision for someone else (40.0% ACP Experience; 38.8% No ACP Experience). In the
ACP Experience group, 32 participants reported a type of decision with 96.9% being for
an adult and 3.1% (n = 1) being for both an adult and a child. When these participants
were asked about feeling prepared to make such a decision, 87.5% indicated they were,
with one additional person (1.2%) who was unsure. In the No ACP Experience group, 18
participants reported making decisions, all for adults. Two thirds of these participants
(66.7%) felt prepared but three participants (6.1%) were unsure. Using the Pearson Chi-
Square test to compare the distributions, no significant difference was found between the
groups related to Making a Decision for someone else, 12(1, N = 132) = 0.061, p = 0.804.

Ability. The distributions of participants’ ratings of feeling able to start a
discussion with others were examined using histograms. For the ACP Experience group,
the median score was 4.5 with a range of scores from 1 to 5. The distribution was
skewed to the left (skewness coefficient = - 4.601), reflecting the 5.9% of participants
who had rated Ability at a l or 2, and too peaked (kurtosis coefficient = 2.058) to be
normal. For the No ACP Experience group, the median score was 4.0 with a range of
scores from 1 to 5 and a normal distribution (skewness coefficient = - 1.265; kurtosis
coefficient = 0.012). Because the ACP Experience distribution was skewed, the
distributions were compared using the Mann-Whitney test, which showed a significant
difference in perceptions of ability between the Experience groups, U(133) = 2604.0, p =
0.001.

Likelihood. The distributions of participants’ ratings of their likelihood to have

conversations with others were examined using histograms. For the ACP Experience

103

group, the median score was 5.0 with a range of scores from 2 to 5. The distribution was
skewed to the left (skewness coefficient = - 5.826), reflecting the 5.9% of participants
who had scored Likelihood at 2, and too peaked (kurtosis coefficient = 2.241) to be
normal. For the No ACP Experience group, the median score was 4.0, also with a range
of scores fi'om 2 to 5, but the distribution was normal (skewness coefficient = - 0.612;
kurtosis coefficient = - 1.744). Because the ACP Experience distribution was skewed,
the distributions were compared using the Mann-Whitney test, which showed a
significant difference in perceptions of likelihood to have conversations between the
Experience groups, U(132) = 1363.5, p < 0.001.

Perceptions of Quality of Conversations

Figure 9 depicts the two groups of participants’ reports of conversation quality
with conversation partners. In the ACP Experience group, participants reported the
quality of their conversations across all conversation partners more often as “thorough”
than “few comments.” However, in the No ACP Experience group, participants reported
an equal quality for Conversations with HCA and a lesser quality for Conversations with
Significant Others. For the two participants who reported a Conversation with HCP, both
reported it as a “thorough conversation.”

When the proportional percentages of conversations were graphed by
conversation types (see Figure 10), the conversations with Health care providers are rated
the most thorough regardless of Experience group. This is the same result as was seen in
the total sample in Research Question 1.

The distributions of conversation quality by conversation partner were compared

for the Experience groups. Since the No ACP Experience group was not expected to

104

report Conversation Quality with HCA’s, no comparison of groups was performed related
to this conversation partner. The comparison of distributions of Conversation Quality
with Significant Others using the Pearson Chi-Square test demonstrated a significant
difference between Experience groups, 12(1, N = 85) = 11.568, p = 0.001. The
Likelihood Ratio was used to compare the distributions of Conversation Quality with
Health care providers because some cells had an expected count < 1; no significant

difference was found between Experience groups, LR(l, N = 23) = 0.803, p = 0.370.

 

7O

 

DFow commonts- EXP

IFow comments- N0 EXP
IThorough conversation -EXP
GThorough conversation - N0 EXP

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10

 

 

l n

 

    

 

 

 

&

SO HCP
Conversation Partner

 

 

 

 

 

Figure 9. Percentages of participants from ACP Experience groups reporting quality of
conversations with conversation partners. Note. HCA = Health Care Agent, SO =

Significant Others, HCP = Health Care Professional.

105

 

120
UFew comments - EXP

IFew comments - NO EXP
100

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IThorouahconvematlon-EXP 7 w
UThoroughconversation-NOEXP
80
‘c"
é so ‘
O
a.
40.
20
0

 

 

 

 

 

 

 

 

 

 

 

 

 

HCA SO HCP
Conversation Partner

Figure 10. Proportional percentages of participants from ACP Experience groups
reporting quality of conversations with conversation partners. Note. HCA = Health Care

Agent, SO = Significant Others, HCP = Health Care Professional.

Correlations

Next the relationships between variables from the Personal Influences on
Decision (Information, Individual Characteristics, Values, and Prior Experiences) and
Personal Decision Factor Self-Efficacy were examined using the Spearman rank
correlation coefficient (rs) for each of the ACP Experience groups (see Tables 5 and 6).
The prior experience variables Having an AD and Having a HCA were omitted because
they are represented by the type of group. As in Research Question 2, the Quality of

Conversations variables were omitted from this analysis because of the significant

106

nonamuogoo 26: 3 33mg u 35 633586 gm 8 oE< u 03< 620 05088
com 5385 $532 .I. 892 “Enemmmomofl 0.80 530m £5 notafioéoo u mOmO £060 «:momemm 5m? noummugcoo u Om-0 ”<03
5:» macaw—3:00 0 «63-0 anathema: u 9.5 ““0650 n 280 632305— 5395 u 265m ”898:: gamma: H CO ”mood .V. a .. .802

 

 

o2: ammo 23 ...: 2 N2 .o , ..ng .83 $5.8 2 3 gas 33
83 :3 3:3 $3 Ed :32 9.8.8 :3 tag 2?.
83 1:2 . 83 23 Ed :3 ammo ...82 one:
83 23 £3 5 .o 68.8 .33 .32. mo: .. o
83 $2 23 Rod Sod . m: .o 8 t o
83 RS 42 .o 23 :2 <0: I o
83 28.8 290 $93 as
83 82.8 fled vane
83 Sod one.
83 32M
35 2i one: new I 0 cm I o «Gm t o as 200 ”we. 38M

 

«Eureka EDV ~33 flecksuxek
SKBESBS Shock zomumomQ 3:2.me nae nomflomQ :o “cutouts ~38er Laxanmmofimob 29:33.80 «Sax nuéemnw.

m 2an

107

3383268 023 3 38:4 0 8.3 328586 gm 8 oB< 0 05¢. 880 28088
.8.“ season maflmz n 89.4 ”8:380.“on 080 580m firs 5.839600 0 mUmO ”.850 ESMEME 55 8:88:00 0 Om-O 25$
~23 53880280 0 <Umto 350%: u 95 £850 u 250 632305. .3325 0 38M confine gunmen n Go Hood .v. a ... .802

 

 

83 8.3 8:8 «:8 Ed 888 82.8 .88 88.8 o: .o 83
83 3 :8 Sod m: .o 88.8 3.2 2.8.8 888 88.0 23.
83 858 ES 2 8.8 88 .8 38.8 82 88.8 882
83 v88 28 88.8 $8 .8 . 83 888 mom I o
83 83 8.0 :88 88 8.2 0m 1 o
83 83 888 42 .o 8.0 5m .. o
83 83 .888 488 95
83 88.8 Sod 88
83 2 _ 08 8<
83 388
83 use. . 882 mom t 0 8 t 0 <0: I o an: 85 one. 3qu

 

326.2»ch 85V 62 ~33 anemmomtem
SKQBESA Shock noEomQ 398.292 88 zofiomQ no 888352 322.82 Sexanomomfimob 2883200 «:32 refinemaw.

8289

108

percentages of missing data in both Experience groups. Statistical significance was
adjusted using Bonferroni’s correction to adjust for multiple comparison between each
variable and the remaining nine variables (0.05/9 = 0.005).

ACP Experience group. For participants in the ACP Experience group, there
were 14 statistically significant pairs of variables and 31 non-significant pairs. Just as in
Research Question 2, the variables of Knowledge and Able (rs = 0.783) and Importance
and Likely (rs = 0.750) were “highly” correlated with statistical significance of p < 0.001.
Four variables were “moderately” correlated, with statistical significance of p < 0.001:
Able and Likely (rs = 0.580), Important and Able (rs = 0.558), Conversation with HCP
and Making a Decision for someone else (rs = 0.518), and Knowledge and Likely (rS =
0.496). The first two pairs are similar to the moderately correlated pairs in the total
sample, although with more strength in the ACP Experience sample. Compared to the
total sample, in the ACP Experience sample the last two pairs increased in strength from
“low” to “moderate.”

Twelve pairs of variables were found in the “low” correlation category. The first
two include Knowledge, first with Importance (rs = 0.466) and then with Conversations
with HCP (rs = 0.424). The third pair is Conversations with HCP and Ability (rs =
0.419); all three of these pairs have a statistical significance of p < 0.001. The next four
pairs have a statistical significance of p 5 0.005. Age is correlated first with Ability (rs =
0.341) and then with Making a Decision for someone else (rs = 0.339). These directional
associations suggest that older participants feel more able to start discussions and have

had to make decisions for others. Likelihood is paired with Conversations with HCA (rs

109

= 0.332), then Conversations with HCP (r, = 0.314). These associations suggest that
participants who intend to have conversations do actually have conversations.

There are 31 pairs of non-significant variables (p > 0.005), but five of these pairs
have a significance very close to p = 0.005. The first four of these pairs include Making a
Decision for someone else with 1) Knowledge (rs = 0.309), 2) Importance (rs = 0.278),
3) Ability (rs = 0.274), and 4) Likelihood (rs = 0.255). This suggests that making a
decision for someone else may contribute to participants’ stronger feelings about their
information, values, and self-efficacy related to engaging in ACP. The final pair of
variables is Conversation with HCA and Conversation with Significant Others (rs =
0.267), which was the most frequently reported combination of conversations in the total
sample.

The following observations related to the 24 non-significant pairs of variables are
of interest: 1) Gender was not significantly correlated with any other variable. 2) Age
was not significantly correlated with any variables except Conversations with HCP and
Making a Decision for someone else. 3) Conversations with Significant Others were not
correlated with perceptions of Knowledge, Importance, or Likelihood or with
Conversations with HCP or Making a Decision for someone else. 4) Conversations with
HCA were not correlated with the prior experiences of Making a Decision or
Conversations with HCP or with Ability. 5) Conversation with HCP was not correlated
with Importance.

No ACP Experience group. For participants in the No ACP Experience group,
there were just four statistically significant (p 5 0.005) pairs of variables and 41 non-

significant pairs. The most highly correlated pair was Important and Likely (rs = 0.728),

110

which suggests that even participants without ACP Experience report a strong correlation
between their value of and intention to have discussions. Knowledge and Ability are
again correlated (rs = 0.645), but this time only “moderately.” Both pairs have a
statistical significance level of p < 0.001. The remaining two pairs are at a “low” level,
with p = 0.002 and p = 0.005 respectively. The most surprising is Gender and Likelihood
(rs = 0.430) since gender has not had significant associations in either the total sample or
the sample of experienced ACP participants. Age and Importance (rs = - 0.399) is the
second. These correlations suggest that there is a possibility that participants who have
No ACP Experience may be younger and female and may value discussions more than
their older counterparts.

Five variable pairs begin the non-significant correlations (p > 0.005). First is
Able and Likely (rs = 0.386), which has appeared in all three samples but is weaker in
this sample. Second is Age and Ability (rs = - 0. 376), which suggests that younger
people may have more confidence to engage in ACP. The last three variable pairs
include 1) Age and Making a Decision for someone else (rs = 0.355), 2) Importance and
Ability (rs = 0.344), and 3) Knowledge and Age (rs = 0.311). It is notable that age this
time appears in the direction of older age. Importance and Ability have appeared as
“moderately” correlated variables in previous samples, but in this group are only
correlated at a “low” level. The correlation category of “little if any" strength is not
represented in this sample.

The majority of variable pairs, then, are not significant in this sample. Some
observations that are similar to the ACP Experience group sample and total sample

include 1) Except for the pair with Likelihood, Gender is not significantly associated with

111

any other variable. 2) The remaining combinations of age and other variables are not
consistent in direction; older age is present with Likelihood but younger age is present
with all types of conversations. 3) None of the conversation type variables are correlated
with any other variable. 4) The remaining variable combinations with Making a Decision
for someone else are all negative in direction. 5) Knowledge is not associated with either
Importance or Likelihood in this participant group.
Summary

When participants with and without ACP Experience were compared, there were
significant differences in most of the personal influences on decision and personal
decision factor variables: Knowledge, Age, Importance, all types of Conversations,
Ability, Likelihood, and Conversation Quality. Participants with ACP Experience had
many more significant associations between variables than participants without ACP
Experience. The variables of Importance to have Discussions and Likelihood of having
Conversations are consistently highly correlated regardless of ACP Experience, but this

may be related to the restricted ranges of these distributions.

Research Question 4
After participation in a community-based ACP program, are there changes from
pre- to post-ACP program intervention in self-reported ratings of enough knowledge,
importance, ability, and likelihood to have an ACP conversation?
The frequencies of the four pairs of variables that participants rated in the pre- and

post-surveys are presented in Table 7.

112

Table 7

Frequencies and Measures of Centralizfy, Variance, and Normality of Variables Rated by

Participants in Pre- and Post-Program Surveys

 

 

 

 

 

 

Pre-Program Post-Program

Focus Model Variables Responses n % n %
Enough knowledge to 1 = not at all 6 4.1 0 0
discuss 2 22 15.0 4 2.7
my views 3 44 29.9 19 12.9
4 30 20.4 56 38.1
5 = very much 44 29.9 65 44.2
Missing 1 0.7 3 2.0
Median / Range 4.00/ 1 to 5 4.00 / 2 to 5
Normality Normal (- 1.591) Skewed (- 4.241)
Too flat (- 2.124) Normal (0. 257)
Important to have a 1 = not at all 1 0.7 0 0
discussion 2 3 2.0 0 0
3 15 10.2 8 5.4
4 23 15.6 28 19.0
5 = very much 103 * 70.1 110 74.8
Missing 2 1.4 l 0.7
Median / Range 5.00/ 1 to 5 5.00 / 3 to 5
Normality Skewed (- 9.315) Skewed (- 8.828)
Too peaked (8.213) Too peaked (5.646)
Able to start discussion 1 = not at all 4 2.7 0 0
with 2 7 4.8 4 2.7
others 3 35 23.8 11 7.5
4 40 27.2 45 30.6
5 = very much 60 40.8 85 57.8
Missing 1 0.7 2 1.4

Median / Range 4.00/ 1 to 5 5.00 / 2 to 5
Normality Skewed (- 4.128) Skewed (- 6.700)
Nonnam277) Too peaked (3.577)
Likely to have 1 = not at all 0 0 0 0
conversation 2 8 5.4 3 2.0
with others 3 29 19.7 9 6.1
4 30 20.4 44 20.9
5 = very much 78 53.1 89 60.5
Missing 2 1.4 2 1.4

Median / Range 5.00 / 2 to 5 5.00 / 2 to 5
Normality Skewed (- 4. 276) Skewed (- 7.581)

113

Normal (- 1.322)

Too peaked (5.564)

Results

Enough knowledge. The distributions of participants’ ratings from pre- to post-
program survey of having enough knowledge to discuss their views were examined using
histograms. In the pre-program survey data, as described in Research Question 1, the
distribution of enough knowledge was bimodal and too flat to be normal (kurtosis
coefficient = - 2.124), with a median score of 4.00 and a range of scores from 1 to 5. In
the post-program survey data, the median score was still 4.0 but the range was reduced to
2 to 5. This distribution was skewed to the left (skewness coefficient - 4.241), reflecting
post-program scores of 2 by only 2.7% of the participants. Because neither the pre- or
post-program distributions met the criteria for normality, the distributions were compared
using the Wilcoxon Signed Ranks test. The data showed significant evidence that
perceptions of having enough knowledge were greater in rank after the ACP program
intervention than before (2 = - 6.642, p < 0.001).

Important. The distributions of participants’ ratings from pre- to post-program
survey of feeling it is important to have a discussion were examined using histograms. In
both the pre- and post-program survey data, the distributions were skewed to the left
(skewness coefficient = - 9.315 [pre] and — 8.828 [post]) and were too peaked to be
normal (kurtosis coefficient = 8.213 [pre] and 5.646 [post]). These non-normal
distributions reflect the small number of participants who scored Importance at a level _<_
3 (12.9% pre-program; 5.4% post-program). The medians were the same for both
distributions (5.00) but the range decreased from 1 to 5 (pre) to 3 to 5 (post). Because
the pre- and post-pro gram distributions did not meet the criteria for normality, the

distributions were compared using the Wilcoxon Signed Ranks test. Although the data

114

showed significant evidence that perceptions of feeling it is important to have discussions
were not the same before and after the ACP program intervention (2 = - 3.068, p = 0.002),
the ranks of importance were tied from pre- to post-program.

Ability. The distributions of participants’ ratings from pre- to post-program
survey of feeling able to start a discussion with others were examined using histograms.
In both the pre- and post-program survey data, the distributions were skewed to the left
(skewness coefficient = - 4.128 [pre] and -— 6.700 [post]) and the post—program
distribution was also too peaked to be normal (kurtosis coefficient = 3.577). This reflects
the small number of participants who scored Ability at a level of S 2 (7.5% pre-; 2.7%
post-survey). The median scores increased from 4.0 to 5.0 from pre- to post-program
with a decrease in range fi'om 1-5 pre— to 2 to 5 post-program. Because the pre- and post-
program distributions did not meet the criteria for normality, the distributions were
compared using the Wilcoxon Signed Ranks test. The data showed significant evidence
that perceptions of feeling able to start a discussion with others were not the same before
and after the ACP program intervention (2 = - 5.612, p < 0.001); however, the ranks for
ability were tied from pre- to post-program.

Likelihood. Finally, the distributions of participants’ ratings of their likelihood to
have a conversation with others were examined using histograms. In both the pre- and
post-program survey data, the distributions were skewed to the left (skewness coefficient
= - 4.276 [pre] and — 7.581 [post]), reflecting the small number of participants who rated
Likelihood at a level 5 2 (5.4% pre-; 2.0% post-program). The post-program distribution
was also too peaked to be normal (kurtosis coefficient = 5.564). The medians of 5.0 and

the ranges of 2 to 5 were the same for both distributions. Because the pre- and post-

115

program distributions did not meet the criteria for normality, the distributions were
compared using the Wilcoxon Signed Ranks test. The data showed significant evidence
that perceptions of likelihood to have a conversation with others were not the same before
and afier the ACP program intervention (2 = - 4.235, p < 0.001). However, as with
Importance and Ability, the ranks were tied from pre- to post-program.
Summary

The ACP education program intervention appears to have had a significant effect
in changing participants’ perceptions in all four areas: feelings of having enough
knowledge, importance, ability, and likelihood to have ACP conversations. This suggests
that an ACP education program intervention may impact both participants’ personal
influences on decision, such as Information and Values, and their personal decision
factor, such as Self-Efficacy. However, only Knowledge actually increased in positive
ranks from pre- to post-program. This suggests that Importance, Ability, and Likelihood
may have had a ceiling effect, where pre-program scores are so high that it is hard to

detect any difference in scores at post—pro gram.

Research Question 5
Do the changes between participants’ self-reported pre- and post-ACP program
intervention ratings of enough knowledge, importance, ability, and likelihood differ
between participants with previous ACP experience compared to participants without

previous ACP experience?

116

The frequencies of the four pairs of variables that participants rated in the pre- and
post-surveys are presented in Table 8 for participants with (n = 85) or without (11 = 49)
ACP experience.

Results

Enough knowledge. The distributions of participants’ ratings from pre- to post-
prograrn surveys of having enough knowledge to discuss their views were examined
using histograms. For the ACP Experience group, both pre- and post-program
distributions were skewed to the left (skewness = - 2.627 [pre]; - 5.057 [post]) and the
post-program distribution was too peaked to be normal (kurtosis = 3.867). This reflects
that only 3.5% (pre) and 0% (post) of the participants rated their Knowledge at a level of
1 (not at all). The distributions for the No ACP Experience group were normal in shape
for both pre- and post-program ratings. Both pre- and post-program medians were one
point higher for the ACP Experience group (pre-EXP = 4.00; pre—NO EXP = 3.00; post-
EXP = 5.00; post-NO EXP = 4.00), but the ranges were the same for both pre-program (1
to 5) and both post-program distributions (2 to 5).

Because the pre- and post-program distributions did not jointly meet the criteria
for normality, the distributions were compared using the Wilcoxon Matched Pairs
Signed-Ranks test. In both ACP Experience groups, the data showed significant
evidence that perceptions of having enough knowledge were not the same before and
after the ACP program intervention (ACP Experience [2 = -4.901, p < 0.001]; No ACP

Experience [2 = -4.543, p < 0.001]).

117

 

 

 

3333 6N3 .3 3.3% 3.333

3338.30 Z 3338.52 03.383834 23.330qu

332 -3 38.3 -3 383.33 -3 3va3: -3 88333 \ 393m

3303 3 33% 83 3 883m 83 s 8308 3303 2 3.08% 3333982

m s N \ 8.3 m s m \ 8.m m 2 N 3 8.“ m 9 3 S3 8353 3 53832

3 3 o o o o 1N N 833332

....8 N3 N8 8N N33 8 33.8 S 3388 .8... u m

SN 33 3.333 a 3.: N3 333 .3 3.

3.8 m ...NN 33 2 3. on N N

o o o o o o 2 N N 8383833U
o c o o o 33 N3 3 33s a 38 n 3 a 25.3 9 3:88:33
383 -3 @833 -v 98.3 338.3 -3

35.52 33.502 Omaha—33 mmmotflv— 35.52.

:88 -3 938.8 €88 -V ANNoN -3 _ 28:33 \ 383m

38.52 38.52 #3 3 @0335 $3 3 @0335 5353.502 m
m 2 N \ 8.3. m 3 3 83 m 2 N \ o3 ... 2 3 33. 83am 3 533832

3 3 o 33 N3 3 N3 3 839.32

8.8 2 3.3 N 3.8 8 3.8 mm .385 be, u m

33.8 .33 $3 33 EN Nm 3N 3N 3,

EN N3 33m .33 3 3 0.8 N3 m 2.63 .38
3 3 SN N3 «N N 33 2 N 93023V
33 o 3.8 m o o 3 3 33a 333 38 n 3 3 83833083 885
o\o 3.3 .x. a c\o 33 .x. 33 8333.883” 83333.5» 303302 £38m
833335-383" Efiwofibfi 8835.88 :83meka

38 u 5 85896 339., oz

38 u 5 0088me n30<

oozotomam 03V 3 REES
253883863 33 3.383233% Eegmsmamom 388 6.3% 2.3 38333. 838.233 \o 3935.82 328 6626.38; 5.4332698 8.8.3.83 328 ammozozumgk

w 033$.

 

 

 

3NN33 @335 .3 33%.N3 N333N.N3

3338.30 Z 3.83.302 orb-283.5 ogoaoq

ASN.3 -3 3N3: -3 G33 -3 6N3: -3 33.33333 3 393m

3303 9 35% 3882 3303 9 332593.0- 3303 3 83% 33338333373

m 3 N 3 8.3. m 3 N 3 8.33 m 2 N 3 333 m 9 N 3 8.3 0333 3 533532

ON 3 33 o o o 3..N N N33333:

33. NN 3.33m N3 33.: 33 3.3 3 338333 .533 u 3

33.333. ON BWN 333 3..NN N3 333 N3 3.

N33 33 NNN 33 mm m 3.333 N 3 3338 33333
3.33 N 3.3. N o o 3 m N 83325238
o o 33 o o o 33 o 3333 3a 38 u 3 2633 3 33833-3
3333.33 @3333 923.333 3338.3 m
38.82 3338.30 Z 0.5383853 033-3833qu

382 -3 $3.3 -3 some -3 338.3. -3 3383333 3 333m

“mo— 9 @0395 35.302 “3 3 wogoxm £2 3 @0335 haw—«8.52

m 2 N 3 8.33 3.. 9 3 3 8.33 m 9 N 3 8.3 n 9 3 3 333- owfim 3 533302

oN 3 o 33 33 33 N3 3 33333332

33.333- ON 33.NN 33 N33 33 33.93 N33 332:33 be, u n

3N3 3N 3.8 m3 SN 3N EN MN 33

N333 m 53 333 mm m 33 3.3 N E3333
3.33 N 3.3 N 2 3 23 m N 33333
o o 3.33 N o o 3..N N 33.3 3a 323 u 3 333333333333333 Eu 3 0333.33.
o\o 33 3x3 33 3x. 33 3x. 33 320535 373.33, 303332 goo-.5
833335-385 Sacra-05 8305.385 Ean5-o5

83. u 333 8833233333 393 02

G33 u 333 3833233333 3393

3332333333383 33 0335

For the ACP Experience group, the pre- and post-program ranks were tied, while
for the No ACP Experience group, knowledge positively increased in ranks from pre- to
post-program ratings. By comparing difference scores (pre-program rating subtracted
from post-program rating) for the participants in each of the groups, the sources of the
tied or changed ranks can be determined. In the ACP Experience group, 49.4% of the
participants had no difference in their scores from pre- to post-program and another
12.0% had difference scores that cancelled each other (6.0% -1 and 6.0% +1). Thus,
61.4% of the ACP Experience group’s scores were considered tied, while only 38.5%
showed a positive difference. In the No ACP Experience group, 41.7% of the
participants had no difference in their scores from pre- to post-program and another 4.2%
had differences that cancelled each other (2.1% +1 and 2.1% -1). The No ACP
Experience group had 45.9% tied scores but 53.7% showing a positive difference.

The difference score distributions for both ACP Experience groups were then
examined to determine if there was a significant difference in the impact of the ACP
education program intervention on Knowledge. This investigator hypothesized that the
difference in pre- to post-program scores for all variables would be greater for
participants with No ACP Experience than for those with ACP Experience.

For the ACP Experience group, the median difference score from pre- to post-
survey rating of enough knowledge was 0.000 with a range of (— 1) to 4, compared to a
median difference score of 1.000 with a range of (-1) to 3 for the No ACP Experience
group. The No ACP Experience group’s distribution was normally distributed by the
ACP Experience group’s distribution was skewed to the right (skewness coefficient =

4.256), meaning more participants had a larger change in scores than would have been

120

expected, and too peaked to be normal (kurtosis coefficient = 3.357). Thus, the Mann
Whitney U test was used to compare the difference scores between the Experience
groups. There was no significant difference between the ACP Experience groups in their
difference scores between pre- and post-program ratings of having enough knowledge to
discuss their views, U(131) = 1655.5, p = 0.085.

Importance. The distributions of participants’ ratings from pre- to post-program
surveys of feeling it is important to have a discussion were examined using histograms.
In both Experience groups for both pre— and post-program survey data, the median score
was 5.0. For the ACP Experience group, the range of scores became smaller in the post-
prograrn ratings (Pre 1 to 5; Post 3 to 5), but for the No ACP Experience group, the range
of scores was 3 to 5 for both pre- and post-programs. For both groups in both pre- and
post-programs, the frequency distributions were skewed to the left, reflecting the very
few participants who rated Importance at a level of S 3 (ACP Exp pre = 8.2%; ACP Exp
post = 4.7%; No ACP Exp pre = 22.4%; No ACP Exp post = 6.1%). For the ACP
Experience group, both pre- and post-pro gram distributions were also too peaked to be
normal. Because the range of values in the distribution was significantly restricted and
the median was at the ceiling for both Experience groups, significant differences were not
expected in Importance from pre- to post-program.

Because none of the distributions for Importance met the criteria for normality,
the distributions were compared using the Wilcoxon Matched Pairs Signed-Ranks test.
In the ACP Experience group, the data did not provide significant evidence that the
ratings of feeling it is important to have a discussion were different before and after the

ACP education program (2 = -1.469, p = 0.142) and, in fact, the ranks were tied.

121

However, in the No ACP Experience group, there was significant evidence that the
ratings of importance had changed from pre- to post-program (z = - 2.676, p = 0.007),
even though these ranks were tied as well. Difference scores were then examined for
each experience group to determine the source of the tied ranks. For the ACP Experience
group, 79.5% of the participants had no change in their pre- to post-program score and
14.4% had scores that offset each other (7.2% -1; 7.2% +1) for a total of 93.9% ties. For
the No ACP Experience group, 70.8% of the participants had no change in their pre— to
post-program score and 8.4% had scores that offset each other (4.2% —1; 4.2% +1) for a
total of 79.2% ties.

The difference score distributions for both ACP Experience groups were then
examined to determine if there was a significant difference in the impact of the ACP
education program intervention on Importance. For the ACP Experience group, the
median difference score from pre- to post-survey rating of importance was 0.0 with a
range of (— 1) to 4, compared to a median difference score of 0.0 with a range of (-1) to 2
for the No ACP Experience group. Since both Experience groups’ distributions were
skewed to the right (skewness coefficients: ACP Experience = 11.628; No ACP
Experience = 3.125), meaning there were more positive differences in scores than
expected, and too peaked to be normal (kurtosis coefficients: ACP Experience = 33.006;
No ACP Experience = 2.635), the Mann Whitney U test was used to compare the
difference scores between the Experience groups. There was no significant difference in
the difference scores between pre- and post-pro gram ratings of the importance of having

discussions between the ACP experience groups, U(131) = 1724.0, p = 0.085.

122

Ability. The distributions of participants’ ratings from pre- to post—program
surveys of feeling able to start a discussion with others were examined using histograms.
The median score for the ACP Experience group was 4.5 with a range of 1 to 5 in the pre-
survey and 5.0 with a range of 2 to 5 in the post-survey. The median score for the No
ACP Experience group was 4.0 for both pre- and post-surveys but the range became
smaller, from 1 to 5 (pre) to 2 to 5 (post). For the ACP Experience group, both pre— and
post-program distributions were skewed to the left, reflecting the small number of
participants who rated Ability at S 2 (5.9% pre; 1.2% post) and too peaked to be normal.
The distributions for the No ACP Experience group were normal in shape for the pre-
program ratings and skewed to the left in the post-program ratings (only 4.1% of the
participants rated at S 2). Because the pre- and post-program distributions did not jointly
meet the criteria for normality, the distributions were compared using the Wilcoxon
Matched Pairs Signed-Ranks test. In both ACP Experience groups, the data showed
significant evidence that perceptions of feeling able to start a discussion with others were
not the same before and after the ACP program intervention (ACP Experience [2 = -
4.302, p < 0.001]; No ACP Experience [2 = - 3.688, p < 0001]).

For the ACP Experience group, the pre- and post-program ranks were tied, while
for the No ACP Experience group, Ability positively increased in ranks from pre- to post-
prograrn ratings. By comparing difference scores for the participants in each of the
groups, the sources of the tied or changed ranks were determined. In the ACP
Experience group, 60.7% of the participants had no difference in their scores from pre- to
post-program and another 9.4% had difference scores that cancelled each other (4.7% +1

and 4.7% -1). Thus, 70.2% of the ACP Experience group’s scores were considered tied,

123

while 29.8% showed a positive difference. In the No ACP Experience group, 39.6% of
the participants had no difference in their scores from pre- to post-program and another
20.8% had differences that cancelled each other (10.4% +1 and 10.4% -1). This brought
the No ACP Experience group to 60.4% tied scores and 39.6% showing a positive
difference, which seems to contradict the positive ranks result.

To determine whether or not there was a difference in impact of the ACP
education program intervention between the ACP Experience groups on perceptions of
Ability, pre- to post-program difference scores were calculated and examined. For the
ACP Experience group, the median difference score from pre- to post-survey rating of
Ability was 0.0 with a range of (— 1) to 4, compared to a median difference score of 0.5
with a range of (-1) to 4 for the No ACP Experience group. Since both Experience
groups’ distributions were skewed to the right (skewness coefficients: ACP Experience =
7.459; No ACP Experience = 2.399), meaning that positive differences were greater than
expected in a normal distribution, and the ACP Experience distribution was too peaked to
be normal (kurtosis coefficient = 11.426), the Mann Whitney U test was used to compare
the difference scores between the ACP Experience groups. There was no significant
difference in the difference scores between pre- and post-program ratings of feeling able
to start a discussion between the ACP experience groups, U(132) = 1745.0, p = 0.158.

Likelihood. The distributions of participants’ ratings from pre- to post-program
surveys of the likelihood of having conversations with others were examined using
histograms. The median score for the ACP Experience group for both pre- and post-
surveys was 5.0 with a range of 2 to 5 in the pre-survey and 3 to 5 in the post-survey.

The median score for the No ACP Experience group was 4.0 for both pre- and post-

124

survey ratings with a range of 2 to 5 at both times. In both experience groups, the range is
restricted but there is a difference in median values. For the ACP Experience group, both
pre- and post-program distributions were skewed to the left, reflecting a small number of
participants who rated Likelihood S 2 (5.9% pre; 0% post), and too peaked to be normal.
The distributions for the No ACP Experience group were normal in shape for the pre-
prograrn ratings and skewed to the left in the post-pro gram ratings (4.1% with score 5 2).

Because the pre- and post-program distributions did not jointly meet the criteria
for normality, the distributions were compared using the Wilcoxon Matched Pairs
Signed-Ranks test. In both ACP Experience groups, the data showed significant evidence
that perceptions of likelihood to have conversations were not the same before and after
the ACP program intervention (ACP Experience [2 = - 2.827, p = 0.005]; No ACP
Experience [2 = - 2.796, p = 0.005]). However, both groups demonstrated tics in the
ranks of likelihood from pre- to post-program.

Difference scores were then examined for each experience group to determine the
source of the tied ranks. For the ACP Experience group, 72.3% of the participants had no
change in their pre- to post-program score and 12.0% had scores that offset each other
(6.0% +1; 6.0% -1) for a total of 84.3% ties and 15.7% positive differences. For the No
ACP Experience group, 57.1% of the participants had no change in their pre- to post-
program score and 12.2% had scores that offset each other (2.1% - 2; 4.1% - 1; 4.1% + 1;
2.1% +2) for a total of 70.8% ties and 29.2% positive differences.

To determine whether or not there was a difference in impact of the ACP
education program intervention between the ACP Experience groups on perceptions of

Likelihood, pre- to post-program difference scores were examined. For the ACP

125

Experience group, the median difference score from pre— to post-survey rating of ability
was 0.0 with a range of (— 1) to 3, compared to a median difference score of 0.0 with a
range of (-2) to 2 for the No ACP Experience group. The No ACP Experience group
scores were normally distributed both pre- and post-program. Since the ACP Experience
group’s distribution was skewed to the right (skewness coefficient = 7.261), meaning
more positive differences than expected in a normal distribution, and too peaked to be
normal (kurtosis coefficient = 11.583), the Mann Whitney U test was used to compare the
difference scores between the Experience groups. There was no significant difference in
the difference scores between pre- and post-pro gram ratings of the likelihood to have
conversations with others between the ACP experience groups, U(131) = 1734.5, p =
0.138.
Summary

In summary, the data suggest that for participants with No ACP Experience, the
ACP education program intervention significantly impacts perceptions of Knowledge,
Importance, Ability and Likelihood, with detectable positive changes in rank in
Knowledge and Ability. For participants already having ACP Experience, the ACP
education program intervention significantly impacts perceptions of Knowledge, Ability,
and Likelihood, but without detectable changes in rank from pre- to post-program. This
lack of sensitivity to statistically significant changes in participants’ ratings is likely a

result of the very high rating medians and the restricted ranges of variables.

126

Chapter 6
DISCUSSION

The focus of this dissertation study has been an examination of factors fi'om the
decision-making context that influence people’s decisions to engage in ACP. In this
chapter, interpretations of the study findings will be discussed by research question, with
resulting implications for nursing practice, education, and research. Next the limitations
of the study will be discussed. Then broader implications for practice, education, and
research derived from this ACP study will be discussed, concluding with an overview of
ideas for firture research that build on this study.

Interpretation of Study Findings
Research Question 1

Prior to their participation in a community-based ACP program, what are the self-
reported Personal Influences on Decision and Personal Decision Factors of adults who
have decided to engage in ACP?

One of the intentions of this study was to contribute a more comprehensive
picture of the internal influences on individuals’ decisions to engage in ACP than is
currently found in the literature. Thus, this study examined the prevalence and frequency
of 15 variables included in personal influences on decision (Information (Knowledge),
Individual Characteristics (Age and Gender), Values (Importance), and Prior Experiences
(Having an AD, Having a HCA, Having Conversations [with HCA, Significant Others, or
HCP], Having Quality Conversations [with HCA, Significant Others, or HCP], Making a
Decision for someone else) and the personal decision factor Self-Efficacy (Ability and

Likelihood). Of the literature reviewed for this study, at least one study reported a

127

baseline or post-intervention result of each of the personal influence or decision
variables, but no study examined all of the variables simultaneously. All of the personal
influence and decision variables were present to some extent in this convenience sample
of adults who decided to engage in ACP through attendance at an ACP educational
program.

Another intention of this study was to contribute to a better understanding of the
communication aspects of ACP, rather than focusing primarily on the formal construction
of preferences (such as through initiating an AD or designating a HCA). Thus, four of
the survey questions asked participants for their perceptions of personal information,
values, and self-efficacy related to having conversations with others; these aspects of
ACP communication have not been studied extensively in the ACP literature.

Enough knowledge. Before the ACP program intervention began, there was quite
a lot of variation in participants’ reports of having enough knowledge to discuss their
views. On the 1 to 5 scale, only 5% of the participants reported having “[none] at all”
(1), while 30% of the participants rated their knowledge as the midpoint (3) and 30% as
“very much” (5). In the literature reviewed for this study, participants were not asked to
rate their knowledge related to having conversations, so no comparisons can be made.
The variation in reported knowledge seems credible for a group of participants drawn
from the community at large.

The implication of this result for both nursing practice and education is that, in
general, most people have some information already about end of life decision-making.
It is important to assess what knowledge they have and be able to clear up

misunderstandings. At the same time, people may not have as much knowledge as they

128

perceive they need to be able to implement discussions. The implication for nursing
research is that an investigator must try to determine participants’ previous knowledge
levels in order to measure the effect of an intervention. Soliciting the participants’
appraisal of the level or usefulness of their knowledge (e. g., “having enough knowledge
to. . .”) has merit, particularly since a concept of interest in the Decision Process Model
for ACP is self-efficacy. However, since there may be discrepancies in understanding of
ACP terms (e.g., AD, HCA, living will, etc.), it also seems important to capture
participants’ more specific understanding of end of life decision-making elements. This
could be done by asking participants to define key terms, as the Office of the Inspector
General (1993) did, by having them answer true/false items, as Ejaz (2000) and Moore et
a1. (1994) did, or by analyzing the categorization of participants’ definitions of ACP as
the MMCAPC Research team did (Vander Laan, Mullan, Wyatt, McPhail, & Thomason,
2006). Measuring participants’ knowledge with a combination of items will likely
provide better comparative data.

Age and gender. In this study, adult participants from every age range attended
the ACP education programs, with participants in the 31-59 years range being most
frequent. This finding is different from most studies in the literature which either
targeted a specific age cohort or had varied but usually older age participants.
Participants in the 18-39 years range comprised < 2% of the total sample while
participants > 75 years comprised 7.6% of the total. Women comprised 76.6% of the
total sample of participants. Having a majority of women participants in this ACP study

is similar to 19 studies of ACP reviewed previously.

129

There are three implications for nursing practice, education, and research that can
be drawn from the age ranges and genders of this study’s participants. First, the strategy
of including existing community-based faith, health-related layperson, and employee
groups was beneficial to acquiring a sample of participants who were of varied ages and
gender. Secondly, very few participants were in the 18-39 years range. This is of
concern because the goals of ACP—awareness of and respect for preferences—are not
limited to any specific adult age group. The two participants who were in the 18-30 years
range participated in an employee group ACP program. Targeting ACP education and
research towards existing community-based groups of employees may be a strategy to
engage younger adults in ACP. However, it is likely that additional strategies may be
needed to address barriers of reluctance and relevance for younger adults in a community.
Thirdly, the ratio of men to women participants was approximately 1:3. This may
indicate the need to target existing community groups in which primarily men participate
(e.g., Service Clubs such as Rotary, Kiwanis, Lions, etc.) or to target community groups
where men and women participate equally. More research is needed to explore the
potentially unique needs of younger adults and men related to engagement in ACP. The
results of such research may more effectively target ACP interventions to promote whole
community engagement in ACP (Field & Cassel, 1997).

Importance. In this study, prior to the ACP education program intervention, 70%
of all participants reported feeling it is very important to have a discussion of firture,
potentially life-sustaining treatment with those close to them. This is similar to the
findings of Ditto et al. (2001), in a randomized controlled trial of elderly outpatients and

their surrogate decision-makers who rated having plans for future care as important. A

130

key feature of ACP is that it is a shared decision-making process, most often with
significant others and health care professionals. It is interesting to note that of the
approximately 100 participants who rated Importance as “very much,” 68.9% reported
having a conversation with significant others and only 19.4% reported having a
conversation with their health care provider.

The implications for nursing practice and education are two. First, people
generally report feeling these types of end of life discussions are important (Lunney et al.,
2003). This is a positive finding because when something is valued, people are likely to
be receptive to more information about it. The second implication is that feelings of
importance do not necessarily translate into discussions. This may be related to the
barriers of resources and restraints described in Chapter 1. People may need resources
such as decision-making supports or discussion guides to equip them for these
discussions. In addition, they may need support and invitations to have these discussions
on a regular basis. As health care professionals, nurses can be particularly helpful to
alleviate perceived restraints by assessing barriers and offering resources to promote
these shared conversations.

The implications for nursing research related to rating Importance involve the
problems of a ceiling effect and possible group bias. In this study, only one item was
used to detect participants’ value of ACP. This item included five responses, ranging
from 1 = not at all to 5 = very much. The median score of this item in the pre-program
survey was 5, indicating that at least 50% of the participants already valued Importance
to the greatest extent possible. This ceiling effect prevents the researcher from being able

to detect changes in participants’ perceptions of values in response to the ACP education

131

program intervention. From a statistical analysis perspective, this homogeneity of
responses also restricts the types of statistical tests which can be used and how significant
any findings really are. The second implication for nursing research relates to the
setting—existing community groups—in which the surveys were completed. A potential
problem of surveying a personal influence like values in such a group is the potential for
group bias; that is, the assumption that everyone else probably believes the topic is
important which then influences how the participant feels about the topic. Possible
remedies for the problem of detecting values include 1) using more than one item to
detect values, 2) increasing the number of points on the scale of value items, and 3)
including items that may differentiate participants’ personal values from the values they
perceive others in the group to have.
Prior Experiences. This study examined four aspects of prior experiences with
ACP simultaneously: Having an AD, Having a HCA, Having Conversations, and
Making a Decision for someone else.
Having an AD & Having a HCA. Only 37% of the participants in this study
reported having an AD; this finding is within the range of the baseline
percentages of < 10 to > 70% reported in the literature reviewed for this study, but
was lower than expected.4 However, 62% of the participants reported having a
HCA, which was significantly greater than the baseline percentages of < 10 to >
40% reported in the reviewed literature. This finding is encouraging in that the
majority of participants in this study recognized and acted on the importance of

having a designated decision-maker.

 

’ In a previous analysis of the entire pre-program survey data set (11 = 347), 42% of the participants had
reported having an AD (Vander Laan et al., 2006).

132

These findings suggest two implications for nursing practice and
education. First, approximately 40% of people who present themselves for health
care services may have an initiated AD or a designated HCA. These formal
products of engaging in ACP are important to communicate individuals’
preferences so must be made available to all the health care professionals
involved in the person’s care. Secondly, in this study only 38.8% of participants
reported having both an AD and a HCA. Although the designation of a HCA is
often found within the AD, the results indicate that the presence or absence of a
designated HCA cannot be assumed on the basis of having an AD. This implies
that people seeking health care services should be asked both questions to assure
that their preferences for decision-maker and any additional instructions are
known.

In the same way, these findings have implications for nursing research
related to ACP. First, researchers should anticipate that at least some participants
in ACP studies may have prior experiences with initiating an AD or designating a
HCA. Secondly, participants should be asked about both prior experiences since
they may not equate Having an AD with Having a HCA.

Having Conversations. This study explored participants’ prior experiences of
having conversations with three different partners: their HCA, significant others,
and HCP. Conversation prevalence was compared among the three partners and
participants were also asked to rate the quality of conversations with each.

fl. Of the 100 participants who answered the Having a HCA and

Having Conversation with HCA items, 77% reported having conversations

133

with their HCA. This is a greater percentage than that reported by Hines
et al. (1999) in a study of hemodialysis patients (50%) or Karel et al.
(2004) in a study of male outpatient veterans prior to a discussion guide
intervention (60%). However, this finding does raise an alert for nursing
practice and education: Just because a HCA has been designated, it
cannot be assumed that conversations have occurred between the
individual and their HCA.

Significant others. Conversations with significant others were reported by
61% of this study’s participantss, which is within the range of 30-70%
reported in the literature reviewed for this study. Study participants in the
comparative literature were outpatients who were preparing for surgery
(Grimaldo et al., 2001), seriously ill hospitalized patients (Heyland,
Tranmer, & Feldman-Stewart, 2000), hemodialysis patients (Hines etal.,
1999), patients with arnyelotrophic lateral sclerosis (Moss et al., 1996),
and nursing home residents (O'Brien et al., 1995).

In this dissertation study, 35% of the participants reported not
having conversations with their significant others. This finding may
reflect all four barriers to ACP: reluctance, relevance, resources, and
restraints. The implication for nursing practice and education is that
individuals’ preferences for end of life care may not be communicated,
even to the individuals’ significant others. As professional nurses, we

have the responsibility and often the opportunity to encourage pe0ple to

 

5 It must be noted that this study did not ask participants to specify if conversations with significant others
included the conversation with their HCA, so the frequencies of conversations with these two partners may
not be independent.

134

have these caring conversations (Field & Cassel, 1997) and to provide
resources to help these conversations occur.

HQ. Only 15% of the participants reported having conversations with
their health care provider. This finding, while within the range of 2 - 40%
reported in the 13 studies reviewed for this dissertation, should cause great
alarm, especially since health care providers may find themselves in a
surrogate decision-maker role when an individual is no longer able to
participate in decision-making at end of life (Daly, 2006; Hurley, Volicer,
Rempusheski, & Fry, 1995). An implication for nursing practice and
education is that it can almost be assumed that people have not
communicated their preferences to their health care provider. Again it is
likely that barriers of reluctance, relevance, resources, and restraints are
responsible for these results, and probably perceived by both individuals
and health care professionals. As professional nurses, we have a
responsibility to not only encourage but also facilitate conversations
between individuals and their health care providers (Field & Cassel,
1997). It may be most effective for the health care provider to initiate this
conversation, as discussed by Morrison and Meier (2004), in their study of
elderly residents of New York City.

Comparing Prevalences of Conversatifls. In the review of literature for
this study, only Hines et al. (1999) reported the percentages of
conversations with HCAS (50%), family members (46%), and HCPs (6%)

for their study participants who were receiving hemodialysis. In this

135

dissertation study, 98 participants reported their prior experiences with
Having Conversations. Only 12% reported having no conversations at all.
Those who had conversations with only one partner included 14% with
their HCA and 7% with their significant others. Of participants reporting
conversations with two partners, about 50% reported having a
conversation with their HCA and significant others6 but only 1% reported
a conversation with both an HCA and HCP. About 17% reported having a
conversation with all three partners (HCA, SO, and HCP). It is interesting
to note that all of the participants who reported conversations with all
three partners also reported Having an AD and Having a HCA. Since
some people have not had any conversations and most people have not had
conversations with all three partners, the implication for nursing practice
and education is that conversations about end of life preferences must be
promoted.

Quality of Conversam. None of the literature reviewed for this study
reported participants’ perceptions of the quality of their ACP
conversations, so the results of this study provide a first glimpse of these
variables. With all three conversation partners, participants reported the
quality of their conversations more often as “thorough” than “few
comments.” Of the 83 participants who reported Conversation Quality
with HCA, 72.3% rated these conversations as “thorough” compared to
27.7% who rated them as “few comments.” Of the 92 participants who

reported Conversation Quality with Significant Others, the ratings were

 

6 The variables Conversation with HCA and Conversation with SO may not be independent of each other.

136

56.5% “thorough” and 43.5% “few comments.” Of the 23 participants
who reported Conversation Quality with HCP, 82.6% rated them
“thorough” compared to 17.4% “few comments.”

There are three implications for nursing practice and education
from these results. First, it appears that health care providers who are
having conversations with individuals are perceived as doing it
thoroughly. This may possibly be credited to the many education efforts
focused on improving communication between health care providers and
individuals about end of life preferences (Ferrell et al., 2005; Robinson et
al., 2004; Weissman et al., 2005). Second, it appears that the majority of
conversations between individuals and their designated HCAs are also
perceived as thorough. Both of these results should be encouraging health
care providers who are involved in education efforts about ACP
communication. Third, the quality of conversations with significant others
was reported as more evenly distributed between “thorough” and “few
comments.” This may indicate that more or more effective resources are
needed to facilitate conversations between individuals and their significant
others.

The implication of these results for nursing research is that ACP

conversations with all partners need more study. Based on a preliminary glimpse

of participants’ perceptions of conversation quality, more study is also warranted

to determine what the essential elements are of a quality conversation and how

they might best be measured.

137

Making decisions. In this study, 40% of the participants reported making
a decision for someone else (almost all adults) about life-sustaining treatment. Of
those who reported their feelings of preparedness to make the decision, 73.2% felt
prepared but 26.8% either did not or were not sure. In the literature reviewed for
this dissertation, four studies reported the percentages of participants who had
experience making decisions for someone else or themselves. Participants in
these studies included male veterans (Karel et al., 2004), adult outpatients living
in Oregon (Silveira, DiPiero, Genity, & Feudtner, 2000), community-dwelling
Afiican Americans (Waters, 2000), and patients with amyotrophic lateral sclerosis
(Moss et al., 1996). An implication for nursing practice and education is that an
assessment should be made of pe0ple’s perceptions of preparedness when they are
acting as the surrogate decision-maker for someone else. Again, by encouraging
people to have conversations about end of life preferences, professional nurses
can contribute to people’s feelings of preparedness for this role. Proactive
conversations can hopefully reduce the incidence of distress that may occur when
people are unaware of others’ preferences (T ilden, Tolle, Nelson, & Fields, 2001).
Self-Eflicacy. This study asked participants to rate their perceived self-efficacy
related to having conversations about life-sustaining treatment. None of the literature
reviewed for this study included results for self-efficacy related to having conversations.
Ability. Before the ACP program intervention began, 40.8% of the participants
rated their ability to start a discussion with others as 5 (very much). The median
score for the sample was 4.0 and ratings of 3 or more were reported by 91.8% of

the participants.

138

Likelihood. In a similar way, 53.1% of participants rated their likelihood of
having a conversation with others as “very much” (5). The median score for the
sample was 5.0 and ratings of 3 or more were reported by 93.1% of the
participants.

The implication of these findings for nursing practice and education is that, even
before receiving education, people may have some confidence to start and intention to
have discussions about life-sustaining treatments with others. The key to promoting ACP
is moving the person confidently and intentionally forward into the behaviors of engaging
in ACP. Health care providers may be able to influence this through ACP interventions
that increase participants’ resources for engaging in ACP.

The implication of these self-efficacy findings for nursing research is similar to
the previous discussion of values. Both of the self-efficacy variables have high median
values (Ability 4.0; Likelihood 5.0); likelihood has a restricted range; and the
participants’ responses are very similar. Thus, it is likely that the limitations of ceiling
effect and group response bias also apply to the measurement of these personal decision
factors. In future studies, a more specific and sensitive instrument should be used to
measure the concept of self-efficacy as a personal decision factor.

In summary, all of the personal influences on decision and decision factor
variables specified in the Focused Model of the Decision to Engage in ACP were
reported by at least some of the participants before the ACP education program began.
Awareness of these internal influences from the decision-making context provides

direction for health care professionals as they 1) assess people’s engagement in ACP;

139

2) promote people’s engagement in ACP through supportive education resources; and 3)
use multiple measures to detect participants’ perceptions of their engagement in ACP.
Research Question 2

Which variables from among the Personal Influences on Decision and Personal
Decision Factors are associated with the Decision to Engage in ACP?

This dissertation study examined associations among all four personal influences
on decision (Information, Individual Characteristics, Values, and Prior Experiences) and
the personal decision factor Self-Efficacy (Ability and Likelihood). Thirty-two pairs of
variables were found to be correlated at a statistically significant level (p 5 0.004). Table
9 depicts these variable pairs in groupings by correlation strength and statistical
significance.

High correlation. For the two most highly correlated pairs—Irnportance and
Likelihood, Knowledge and Ability—there are no comparative reports in the literature
reviewed for this study. However, both make sense when considered from a decision
process model perspective. In the simplified decision process model (Wills & Holmes-
Rovner, 2006), information and values interact to influence the subsequent decision,
behaviors, and outcomes. The Focused Model of the Decision to Engage in ACP shows
the personal influences of information and values interacting with elements of the
personal decision factor self-efficacy to influence the decision. In these highly correlated
pairs, increased value is associated with an increased intention to act and increased
information with increased confidence to act.

Moderate correlations. In the moderate strength correlation category, the first

pair of variables, Having an AD and Having a HCA, was also reported by Morrison et al.

140

€050 agomEmmm u Om Raowmmomoi 080 5303 u mum Eom< 030 530m .I. <9; 250th oo§>e< u Q< 28M

Q< 5 mafiam um 033399
ban? a. <om a macaw
ban? a. 9. a mafia

33¢. a mum 5E causescoo

02m 8888 8m 5389 wcflaz a. mum 5.3» 20:86:00
owe. a. o4 5 mafia

8509.5 on 03036qu

mom fits cot—3:550 ow omen—Beam

<om a mess; a. umeoqum

82:33 a. umeofiofi

Om 5E aofimaéoo a 5m a was“:

0m 55 nouaeéoo a. <0: 55 eouameéoo
mom 5? couaeéoo a. n2 8 mean

Sn 5.; 8:35:00 a. an a mega

 

28.0 v e 8:20:00 33

3%:an a. rape.
Bogs: e 53¢
5m a mega a. n2 5 gram

 

:86 v a 83250 swag:

£22 a. 0262305“
coofioxfi do Bantams:

 

28d v e 532.50 swam

 

moreombafifi Bethesda BE Safety. zthmtoD 3 333630 “28:60 M:&§<.=o.§omQ 2: E $33.23; MfionmaNxérNo 2.8%

m 2.3

141

€050 885:me u Om _mcoflmmomofl 030 530mm u mom “sumac. v.30 fifiom u <03 9.5025 oo§>n< M Q $0M

$5 a 8.5% a. 8583

85 ocoofiom com commmoofl wing a. om<
omen—33M a. Om 53» 5:35:80
$3150ch do <03 5.5» couamb>noo
9. 5 83% a. 5m 5? Evasive
00525: a. Sum fit» nonmagaoo
83:35 a. mom 5.5» 8:85:50
uoofioxfi 9% Om 53» 5:85.500
82:85 a. <om a were:

b=B< a. Om fir,» coummbéoo

ow< aw AGE 59» 558.6»:00

Q am 9:32 a. Om fit» nouamb>aou
coonzuxfi Q mom 5.5» aouamugaou

 

€85 .v. e 83280 33

903983 a uses

142

(1 998) and is frequently noted in clinical practice. That the correlation is only moderate
may be explained because the percentage of participants reporting an AD was only 37%
compared to 62% who reported having a HCA. The second pair—Ability and
Likelihood—are the two variables in the personal decision factor Self-Efficacy. Since
the correlation is only moderate, they may each capture a distinct feature of self-efficacy.
Finally, Importance and Ability are another pairing of variables from the personal
influences on decision and personal decision factor. The strength of a person’s value of
having discussions may be tempered by the perception of one’s ability to start those
conversations. For the second and third pairs, there was no comparative literature found

for this study.

An implication for nursing practice, education, and research fiom these high and
moderate correlations is that a decision-making model, such as the Decision Process
Model, helps conceptually in understanding the associations of personal influences and

decision factors on the decision to engage in ACP.
Low correlations. In the low strength correlation category, there are 14 pairs with
a statistical significance of p < 0.001 and 13 pairs with a significance of p _<_ 0.004.
Low correlations with significance of p < 0. 001 . Each of the first four pairs
contains a Having Conversation variable; three of these are paired with either
Having an AD or Having a HCA. This is similar to the findings of Hamel et al.
(2002) in their study of community-dwelling elders for Having an AD and
Wenger et al. in their study of persons with HIV for Having a HCA. The next
four pairs include the variable Knowledge. The first pair—Knowledge and

Likelihood—are somewhat supported by Ejaz (2000) who found in their study of

143

institutionalized elderly that knowledge of an AD was associated not only with
having an AD but also the likelihood of implementing an AD. In this dissertation
study, the focus was on having knowledge and likelihood to have discussions
rather than concern for initiating or implementing an AD. The second pair—
Knowledge and Having a HCA—are somewhat supported by the Morrison et al.
(1998) study which found that, in an older adult outpatient sample, knowledge of
a HCA was associated with having a HCA. For the third and fourth pairs—
Knowledge and Conversation with HCP, Knowledge and Importance—there is no
comparative literature.

The next pair—Having an AD and Age—does have comparative
literature, but the results are mixed. In this dissertation study, Having an AD was
directionally associated with older age. This is similar to the results of Dexter et
al. (1998), Morrison et al. (1998), Wenger et al. (2001) but not with the results of
Moore et al. (1994) or Patterson et al. (1997) who found it associated with
younger ages or with Hopp (2000) who found no relation between age and having
an AD.

The next pair—Conversation with HCP and Making Decision for
Someone Else—are not reported in the literature but make sense from a clinical
perspective. Persons who have needed to communicate with health care providers
while making decisions for others are also likely to be able to communicate with
health care providers when making decisions for themselves.

The next three pairs all contain the variable Ability, which did not appear

in the comparative literature reviewed for this study. These pairs suggest that

144

having confidence in one’s ability to start discussions is related to formally
constructing and communicating preferences.

The last pair in this grouping is Knowledge and Having an AD, which is
somewhat supported by Ejaz (2000) who found that increased knowledge of an
AD was associated with having an AD. Again, the difference in this study is the
emphasis on knowledge for having discussions versus the outcome of an initiated
AD.

Two patterns in these 14 pairs of variables were noticed that have
implications for nursing practice and education. First, eight of the pairs included
either the variable Having an AD or Having a HCA, which will be discussed more
thoroughly later in this chapter. Three of these pairs included Conversations
variables, each representing a different conversation partner (HCA, HCP, SO).
Because these were only correlated at a low level, it is important to specifically
ask people who report having an AD or HCA if they have had conversations
about their preferences and identify with which partners.

The second pattern noted is that eight of the pairs include either the
variable Knowledge (5 pairs) or Ability (3 pairs). This finding suggests that if
people’s knowledge or confidence in having discussions can be increased, other
influences in the decision-making context, at least to a low level, may be affected
as well. Thus, an implication for nursing practice and education is that health care
providers who are planning interventions to promote ACP may want to include

both information and possibly opportunities to practice discussions (e.g., role-

playing). These teaching and learning strategies will provide people with

145

resources to guide their construction and communication of preferences (Lunney
et al., 2003).

Low correlations with p _<_ 0. 004. Of the remaining 13 statistically
significant variable pairs, 10 include a Conversation variable, with all three
conversation partners represented (HCA, SO, HCP). Two of these conversation
variables are associated with increased Knowledge, one with increased
Importance, one with increased Ability, and three with increased Likelihood.
This suggests that having conversations is related to people’s perceptions of
having enough information, value, confidence, and intention to actually engage in
such discussions. Two of the remaining conversation variables were associated
with having an AD, which was also reported by Hamel et a1. (2002). The
remaining conversation variable was associated with age, which was also reported
by Dexter et a1. (1998) but not by Bradley et al. (1998) or Hamel et al. (2002).

Two of the three remaining variable pairs include the variable Having a
HCA, associated first with increased likelihood and then with increased
importance. These may suggest that when people designate someone to make
decisions for them, they perceive more personal intention and value in having
conversations with others. The final variable pair is Age and Making a Decision
for Someone Else. This represents the association of growing older with the
experience of being asked to make decisions for others. In clinical practice, this
phenomenon is seen frequently, as adult children are asked to make decisions for

or with their parents or grandparents.

146

The implications for nursing practice and education from these 13 variable
pairs are very similar to those discussed in the previous section. Although only at

a low correlation level, the links between having conversations and perceiving

enough knowledge, importance, ability, and likelihood to have these

conversations were affirmed. This may lend support to assure that interventions
designed to promote ACP not only provide information and opportunities to gain
confidence, but also help participants clarify their values and encourage
translation of intentions into actions.

In summary, just less than half of the personal influences on decision and decision
factor variables specified in the Focused Model of the Decision to Engage in ACP were
significantly correlated with each other, at least at a low level. Almost all of the variable
pairs included 1) the self-reported ratings of enough knowledge, importance, ability, or
likelihood to have an ACP conversation, 2) having an AD or HCA, or 3) Having
conversations. These findings support the use of a decision-making theoretical
perspective, such as the Decision Process Model for ACP, to help interpret the
relationships between variables that may influence the Decision to Engage in ACP.
Research Question 3

Do the Personal Influences on Decision and Personal Decision Factor variables
that are associated with the Decision to Engage in ACP differ between participants with
previous ACP experience compared to participants without previous ACP experience?

Another objective of this dissertation study was to examine the dynamic nature of
ACP as a decision-making process by comparing the personal influences and decision

factor variables of participants who reported previous ACP experience with those of

147

participants who reported no previous ACP experience. There were no reports of this
sort of comparison in the literature reviewed for this study.

Composite comparison. Table 10 provides a composite view of the personal
influences on decision and personal decision factor variables for each group. As
described in previous chapters, participants’ reports of the variables Having an AD and
Having a HCA were used to create subgroups of participants for comparison, so are not
reported. Neither is the variable Conversation with HCA nor the three Quality
Conversation variables, because each contained a significant fiequency of missing data.
Of the nine variables compared, seven were found to be significantly different between
the groups.

The first implication for nursing practice and education is recognizing that in a
group of participants who have decided to engage in ACP by attending an ACP education
program intervention, there may be people who already have ACP Experience! This can
be a mixed blessing to the ACP Facilitator presenting the program. On the positive side,
participants who are willing to share their experiences often address how they personally
overcame barriers to engaging in ACP, such as reluctance, relevance, resources, or
restraints. Such “testimonials” can be empowering and encouraging to other participants.
On the negative side, participants with experience may have misconceptions about ACP
or may not have engaged in the entire decision-making process of ACP, so may be

sharing a limited view.

148

Table 10

Composite Comparison of Selected Personal Influence on Decision and Personal

Decision Factor Variables for Participants Classified by ACP Experience

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Significant ACP No ACP
Focus Model Variables difference Experience Experience
n = 35 n = 49

Enough knowledge to discuss Yes Median 4.0 Median 3.0
my views Range 1 — 5 Range 1 — 5

38.8% rated as 5 14.3% rated as 5
Age Yes Older Younger
Gender No 23 .5% Men 18.4% Men

76.5% Women 81 .6% Women
Important to have a discussion Yes Median 5.0 Median 5.0
Rangel—S Range3—5

78.8% rated as 5 59.2% rated as 5
Having Conversations with Yes 77.6% had 36.7% had
Significant Others
Having Conversations with Yes 23.5% had 4.1% had
HCP (Health Care Provider)
Making decision for someone No 40.0% had 38.8% had
else
Able to start discussion with Yes Median 4.5 Median 4.00
others Range 1 — 5 Range 1 — 5

49.4% rated as 5 22.4% rated as 5
Likely to have conversation Yes Median 5.0 Median 4.0
with others Range 2 -— 5 Range 2 -— 5

67.1% rated as 5 34.7% rated as 5

 

Enough knowledge. Both groups of participants left room for growth in their pre-

program ratings of having enough knowledge to discuss their views. Although

the ACP Experience group reported more knowledge than the No ACP

Experience group, only one third of the experienced participants felt they had

149

 

“very much” knowledge. The implication of this finding for nursing practice and
education is that there are still many Americans who have limited knowledge of
ACP as a dynamic decision-making process.

Age. The group of participants with ACP experience was older than the group
with No ACP Experience. As mentioned in the previous discussion of correlated .
variables, the literature reviewed for this study was inconsistent in its findings of
the relationship between age and having an AD or having a HCA. Because of the
inconsistency of results across studies, no implications will be discussed.

Gender. There were no significant differences in gender between the groups. As
with age, there are inconsistencies in findings of the relationship between gender
and having an AD or having a HCA, so no implications will be discussed.
Importance. Both groups of participants rated their value for discussions very
high in their pre-pro gram ratings. Although more people in the ACP Experience
group rated Importance very high (5), this finding indicates that people may value
discussions of end of life preferences regardless of previous ACP experience.
The crucial question for nursing practice and education is whether or not pe0ple
value discussing their personal preferences for end of life treatment or whether
valuing discussion is just a general sentiment for all people or a biased response
based on membership in a community group. An implication of this result may
be the need to provide resources to help people clarify their personal values.
Conversations with Significant Others. There is a very significant difference
between the groups in the reported frequency of conversations with significant

others. Although the result for the ACP Experience group in contrast to the No

150

ACP Experience group looks significantly better, it must be noted that just under
25% of the participants did not report conversations with their significant others.
This indicates that although preferences may have been formally constructed,
communication of those preferences to others cannot be assumed. It is also
significant that more than one third of the participants with No ACP Experience
reported having a conversation with their significant others. This indicates that
preferences may be communicated to others even if they may not have been
formally (legally) constructed.

When the Quality Conversation with Significant Others data is examined
between the groups, participants with ACP Experience report their conversations
as being significantly more “thorough” than “few comments.” The result for
participants with No ACP Experience is the opposite.

These results suggest two implications for nursing practice and education.
First, all participants in ACP education programs, regardless of previous ACP
experience, should be encouraged to engage in caring conversations with their
significant others about their preferences for end of life treatment. Second, ACP
education pro grams should provide resources to participants to help them have
more thorough conversations with their significant others. Such resources should
candidly address the barriers of reluctance, relevance, resources, and restraints
and provide participants with strategies to start and carry on these conversations.

An implication for nursing research from these results is that an
examination of ACP that looks only at the outcomes of having an AD or having a

HCA will provide an incomplete picture of people’s engagement in ACP. As

151

Kolarik et al. (2002) and Emanuel (2003) assert, the objectives of ACP are
broader than an AD and best studied as a process.

Conversations with Health Care Providers. The results of this study also
demonstrate a significant difference in the frequencies of conversations with
health care providers between ACP Experience. groups of participants.
Frequencies of conversations that are less than 25% (23.5% ACP Experience
group; 4.1% No ACP Experience group) should be of great concern to health care
providers, particularly from a group that has already formally constructed their
preferences for end of life treatment. Considering that barriers to ACP may be
applicable to both individuals and their health care providers, the implication for
nursing practice and education is that health care providers must be proactive in
initiating and continuing conversations about their people’s end of life
preferences.

When the Quality Conversation with HCP data is examined between the
groups, there is a potentially surprising finding: 100% of the participants with No
ACP Experience report that their conversations with health care providers are
“thorough” compared to 81% of the participants with ACP Experience. The
implication this finding suggests for nursing practice and education is that the few
people who have conversations with health care providers find these
conversations to be mostly thorough. A second implication is that a casual
conversation (“few comments”) is unlikely to occur.

An implication for nursing research is the challenge to examine the idea of

conversation quality and its effects on the behaviors and outcomes of engaging in

152

ACP. Presumably a “thorough” conversation is more effective than a “few
comments,” but by what degree? Shared decision-making principles come into
play here as well—are these conversations desired? What quality will be
acceptable to the patient and health care provider?

Making decision for someone else. There was no significant difference between
the ACP Experience groups in the frequency of making a decision for someone
else. The literature reviewed for this study has nothing further to add so
implications will not be discussed.

Able to start discussion. There was a significant difference in participants’ ratings
of their abilities to start discussions, with the ACP Experience group rating its
ability higher than the No ACP Experience group. The percentage of people
rating their ability as “very much” was just less than 50% for the ACP Experience
group and less than 25% for the No ACP Experience group. As in the knowledge
variable, both groups leave room to increase their confidence in these pre-
program survey ratings.

An implication of this finding for nursing practice and education is that
participants’ confidence in their ability to start conversations about end of life
preferences cannot be assumed, even if they report already having ACP
Experience. This finding again reflects the need for more comprehensive study of
ACP than simply formally constructing preferences.

Likelihood to have conversations. There was a significant difference between the
groups in participants’ ratings of their likelihood to have conversations. Although

both groups’ median rating was 5.0, the ACP Experience group had almost twice

153

as many people who rated their likelihood as “very much.” There were no
participants in either group who reported not intending to have conversations at
all. This result may imply that all participants intend to have ACP conversations
and thus all may need resources to accomplish this. A second implication is that
one third of the participants with No ACP Experience very much intend to have
conversations even though they have not (yet) formally constructed their
preferences.
Correlations. When correlations were calculated for variable pairs for each of the
ACP Experience groups, the number of statistically significant pairs was greatly reduced
in comparison to the total sample and significantly different between the groups. The
ACP Experience group had 14 pairs of variables while the No ACP Experience group had
only 4 significant pairs. It is interesting to note that the one highly correlated pair in both
ACP Experience groups was Importance and Likelihood. Knowledge and Ability had the
second strongest correlation—“high” in the ACP Experience group but only “moderate”
in the No ACP Experience group. The remaining significant variable pairs in the ACP
Experience group were similar to pairs found in the total sample, which is not surprising
since two thirds of the study participants are in the ACP Experience group. The two
remaining significant variable pairs in the No ACP Experience group are Gender and
Likelihood, Age and Importance. It is likely that these two pairs of variables demonstrate
significant relationships because the rating variables in each pair have a restricted range
and the majority of participants in the No ACP Experience group were women in the 31-

59 years range.

154

There are two implications of these findings for nursing practice and education.
First, regardless of ACP Experience group, participants’ value of discussions is highly
associated with their intention to have discussions. Secondly, regardless of ACP
Experience group, participants’ knowledge is related to their confidence to be able to start
such conversations. An implication for nursing research is to identify measures of
Knowledge, Importance, Ability, and Likelihood that have more specificity and
sensitivity. With such measures, the significance of the relationships between these
variables can be better characterized.

In summary, when participants with and without ACP Experience are compared,
there are significant differences in most of the personal influences on decision and
personal decision factor variables. This finding has implications for the design and
implementation of ACP education pro gram interventions, since it is likely that
participants at the same program may have different levels of ACP experience.
Participants who have ACP Experience report more prior experiences with ACP and
more positively perceive their levels of knowledge, importance, ability, and likelihood to
engage in ACP conversations. However, there are significant numbers of participants
with ACP Experience who have not communicated their preferences with significant
others or health care providers, despite having formally constructed their preferences in
AD or HCA documents. In the No ACP Experience group, there are fewer reported
experiences with ACP and a less positive perception of knowledge, importance, ability,
and likelihood to engage in ACP conversations. Yet, there are some participants who
have communicated their preferences with significant others or health care providers

even though they have not formally constructed their preferences. An implication for

155

nursing practice and education is that it can be expected that all participants in ACP
education program interventions, regardless of previous ACP Experience, may need to be
better informed and encouraged to have conversations with others about ACP.

Research Question 4

After participation in a community-based ACP program, are there changes from
pre- to post-ACP program intervention in self-reported ratings of enough knowledge,
importance, ability, and likelihood to have an ACP conversation?

Most studies of ACP interventions have examined outcomes—such as an initiated
AD—as a measure of the intervention’s impact. In this dissertation study, however, the
impact of the community-based ACP education program intervention is being measured
by changes in two personal influences on decision variables (Knowledge, Importance)
and two personal decision factor variables (Ability, Likelihood) that are thought to
influence the Decision to Engage in ACP. In addition, all of these variables relate to
conversations the participants may have about their preferences for end of life treatment.

Enough knowledge. Participants in the ACP education program intervention rated
their knowledge as significantly increased from pre- to post-program. This change
resulted in an increased number of positive ranks. In the literature reviewed for this
study, the study by Yamada et a1. (1999) contained the only report of an increase in
knowledge following their intervention of information handouts and an ACP video.

Importance, Ability, and Likelihood. The changes in results of participants’
ratings of Importance, Ability, and Likelihood all were statistically significant from pre-

to post-program surveys; however, for all the majority of the ranks were tied rather than

156

changing positively. There were no studies in the literature reviewed for this study that
examined the impact of an ACP intervention on these variables.

In summary, the results of this study demonstrate that the ratings of all four
variables changed from pre- to post-program surveys at a statistically significant level.
Knowledge demonstrated a detectable change in positive ranks from pre- to post-program
surveys, but the other variables only demonstrated ties in ranks. There are two
implications for nursing practice and education from these results. First, the ACP
education program intervention used in this study had a positive impact on participants’
perceived knowledge. This may suggest that the strategies employed in this ACP
program intervention (Respecting Choices® curriculum, presentations offered to existing
community groups, certified ACP Facilitators conducting the programs, and standardized
content and resource materials) are effective in promoting increased knowledge for
participants who decide to attend. Secondly, the ACP educational program intervention
did have an impact on each of the four variables. This suggests that an external ACP
intervention can be used to impact internal influences within the decision-making
context.

Detecting a change in participants’ perceptions was impeded by ceiling effects
and possibly group response biases. An implication for nursing research is the
importance of evaluating the concepts included in the Decision Process Model for ACP
with multiple measures that are able to detect change with specificity and sensitivity.
Research Question 5

Do the changes between participants’ self-reported pre- and post-ACP program

intervention ratings of enough knowledge, importance, ability, and likelihood differ

157

between participants with previous ACP experience compared to participants without
previous ACP experience?

The final objective of this study was to investigate the dynamic nature of ACP as
a decision-making process. The differences in ratings of Knowledge, Importance,
Ability, and Likelihood reported by participants after an ACP educational program
intervention, were examined after creating subgroups of the participants by their previous
ACP Experience. This provided an opportunity to compare people who had made the
decision to engage in ACP for the first time to people who were deciding to engage in
ACP again. Table 11 provides a comparison of the variable ratings in pre- and post-
program surveys by participants in ACP Experience group.

Comparisons. For participants with No ACP Experience, the ACP education
program intervention significantly impacted perceptions of Knowledge, Importance,
Ability, and Likelihood, with detectable positive changes in rank in Knowledge and
Ability. For participants with ACP Experience, the ACP education program intervention
significantly impacts perceptions of Knowledge, Ability, and Likelihood, but without
detectable changes in rank from pre- to post-program. When difference scores are
calculated for pre- to post-program ratings for each variable and compared between the
ACP Experience groups, there is no significant difference detected between the two
groups for any of the variables.

In summary, the implication of these findings for nursing practice and education
is that the ACP education program intervention used in this study did have a significant

impact on participants’ ratings of their Knowledge, Ability, and Likelihood of having

158

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

we: 83am $3.8m eon g 63.05
3% oz
385% “50%me oofiemansfiouuna
oz
m 3 e38 $03 n 3 e88 xmdm m me e38 .xaéw .m we we?" $w.wn
mimowfim mimowfim mimomamm ml—owaam
ed 5:632 o.m :3on ed 5:32 ed 5602 56386 a 98: 3 33.895
@8855 Eczema Edam “mombpm out madam $3.8m
oofieefiafiofiaa oofismasaocwna
, 02
m 3 e88 .xbdm m 8 e88 $3: m we won—8 :93 m we no?" $w.mm
mimomnnm mlfiowcmm mimomqam ml_owcam 9503?:
ed 5:52 o.m :2on ed €602 ed 5:32 353% 3 03232.2 Amsocm
Eahwoiamom Eauwocmbum ageing Eacwofiéum 9mm mU< oz
.9»
855an mU< oZ ooaotoexm mU< mum m0< moEmta> Hove: msoom
335th
iguana
mozomgmmxm

06V 3 nmmbmwfib fleckaowtem 3 hoofififiq use $an .moeetoQEN .thmNSonRo awnuux EEMoxmédn. S -25? coatemsob

S 03.3.

159

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

not 92am “moméi 3.5 g $8-8m
no? 8.».
358wa “536me 8:20an pfiomfimwm
02
n an F58 {3.3 m mm U88 85% n mm 393 Rafi. m 8 UBE .x;.$
n I N omaam m I N omnmm m I m 098% m I N owfim mbfio 5.5,
o.v 5.602 ed 5632 od 5:62 9m 5:82 cosamugcoo 26; 3 buxfi
c8855 Evie—mom mxaam $8-on we: send “8&me
3% mo?
Quechua ESE—ma oofismfi “gamma
02
n ma uBE o\ow.ov n mm 9:8 .x.v.- n ma uBfi gnaw n ma Bug fix?
m I N omcmm m I # omcé m I N 093% m I # omcam mbnuo
o.v £602 3‘ $232 o6 5602 0v :3qu at? 5&3me tau 3 25¢
EEwobEmom Sawmibum anokfmom Mahmoiéfi nxm mU< oz
.m>
coautomxm mode. 02 oofitonxm m0< mxm mU< 875.5, 6on Soon
838th
Eéfia

 

 

$02583 : 2‘3

160

conversations, regardless of previous ACP Experience. This lends support for offering
ACP education program interventions broadly to participants, including those with and
without previous ACP Experience, because all appear to receive some benefit. It is
hoped that individuals’ decisions to participate in an ACP education program intervention
may trigger subsequent behaviors and outcomes of engaging in ACP because of new or
increased perceptions of personal information, value, or self-efficacy.

The implication of these findings for nursing research in ACP is that the dynamic
nature of participants’ decisions to engage or re-engage in ACP can be studied and
compared through a variable like ACP Experience. However, the measurement of
changes in perceptions of personal influence on decision or personal decision factor
variables must be both sensitive and specific enough to capture individuals’ changes over
time or in response to an intervention.

Limitations of the Study

While this study adds to the understanding of ACP as a dynamic decision-making
process, the limitations of the study must be acknowledged. First, the participants in this
study were not a random sample of individuals. Instead, they were a convenience sample
of people who voluntarily attended an ACP education program, presented to an existing
community group of which they were a member. Because the participants were from a
convenience sample, the ability to generalize from the analysis of their responses is
limited.

Secondly, some items on the Participant Survey were not specific or sensitive
enough to detect changes in participants’ perceptions from pre- to post-program. When

the MMACPC Research Project was designed, the Research Team decided to use or

161

adapt survey items from existing measurement instruments offered as part of a Quality
Improvement toolkit by Respecting Choices®. This decision offered the possibility of
comparing the MMACPC Research Project data to data collected by other groups who
were promoting ACP in communities using the Respecting Choices® curriculum and
assessment tools. Unfortunately, the pre-post survey scaled items were prone to ceiling
effects, which limited their usefulness to detect significant changes in participants’
perceptions.

Thirdly, this study examined a small segment of time and activity related to
individuals’ ACP decision processes. Although this study’s focus on the Decision to
Engage in ACP rather than the behaviors or outcomes of engaging in ACP was
intentional, it also limits the interpretation of the results to the timefi'ame of the ACP
intervention.

Broader Implications
Implications for Practice

The results of this dissertation study demonstrated that 63% of the participants
had ACP Experience, defined as initiating an AD or designating a HCA. Less than 40%
had an AD or both an AD and HCA. These results are similar to the findings of studies
reviewed for this dissertation which had a range of < 10% to > 70% ADS and < 10% to >
40% HCAS. It appears that the barriers toward engaging in ACP are still formidable for a
large number of individuals. How can health care providers—not only nurses but also
physicians, social workers, chaplains, and other interdisciplinary team members—-

respond to try to prevent the distress of not having preferences known or honored?

162

First, as interdisciplinary health care providers, we can initiate ACP conversations
as a part of routine adult health care. Hammes and Briggs (2002) outline the advantages
of this approach as having more professionals who can 1) identify the needs of patients
(in any health care setting) for ACP conversations, 2) address the need for ACP
conversations from different perspectives, and 3) share the responsibility for
communicating, collaborating, coordinating, and connecting patients with resources, that
can set the ACP process in motion. ACP should be viewed as a health promotion activity
(Patrick et al., 1995) that provides resources to help people prevent undesired outcomes
at a later time. It is important to know the resources available in community health care
organizations to promote and support ACP; e. g., educational materials or ACP
Facilitators.

Secondly, if an individual we are caring for has an AD or HCA, we can be sure
this information is made available to all members of the health care team. An
individual’s preferences can only be honored if the health care team knows about them.
This is not only the nurse’s responsibility—it is a responsibility of all health care
professionals and the health care organizations for which they work. Organizations need
to adopt a culture of respecting individuals’ preferences, supported by structures and
processes that enable the health care team to know and honor those preferences.
Organizational resources such as indicators in the electronic or paper medical record,
automatic referrals from admission data profiles, continuing education on end of life
decision-making or communication, and ethics committees are all evidences that

advocacy for ACP is important to the organization.

163

Thirdly, if an individual has not had conversations with others—their designated
HCA, significant others, or health care professionals—we need to encourage these caring
conversations to occur. A variety of resources are available to help individuals start and
maintain these conversations (see Appendix D). As in many other health promotion
activities, the health care providers’ abilities to speak from their own experiences with
engagement in ACP can be both valuable and motivating to encourage other persons’
behaviors.

Finally, if individuals have had conversations with others but have not chosen to
initiate an advance directive, health care providers, can encourage them to formulate a
thoughtful plan (verbal or written, formal or informal) that communicates some values
and basic preferences (Hammes & Briggs, 2002). Formulating a plan moves patients and
their surrogate decision-makers toward being informed and willing to enter the decision-
making process and enables them to make decisions in a timely way when needed.
Implications for Education

The results of this dissertation study also concurred with previous research reports
that ACP interventions—especially educational interventions—have mostly resulted in
increased preparation for engagement in ACP decision-making. A significant finding of
this study is that individuals who participate in ACP interventions will likely have
different levels of ACP experience. This has implications for the types of education
programs and resource materials that may be needed. For example, it could be important
to have resources that introduce the concept and terminology of ACP as well as resources

that help individuals clarify their changing values and preferences or successfully

164

navigate the complexity of communicating their preferences to others who may be in
conflict.

Another implication of this study is that individuals still need more knowledge
about ACP as a decision-making process and about the types of decisions that may arise
in the fiiture. As health care providers, we have a professional responsibility to try to
clearly communicate what is involved in end of life conditions and treatments and the
probabilities of different outcomes (Davis et al., 2003; Moumjid, Bremond, & Carrere,
2003). This type of communication can support decision-making, particularly shared
decision-making. At the same time, however, health care providers must be sensitive to
those individuals who do not want to actively participate in decision-making about their
personal situations (Charles, Gafni, & Whelan, 1997, 1999; Hawkins, Ditto, Danks, &
Smucker, 2005; Lee, Back, Block, & Stewart, 2002). Knowledge of preferences—the
individual’s and others’—makes supporting and implementing an advance plan for end of
life care much more achievable.

A third implication is that providing ACP education to existing community-based
groups may promote access to a wider variety of participants at an earlier stage in the
need for end of life decision-making than waiting for a health care encounter. An
additional benefit is reduced barriers of reluctance and relevance and increased access to
resources such as expert ACP facilitators and educational materials (Clarke, Evans,
Shook, & Johanson, 2005).

Finally, this study’s report of the perceived quality of conversations between
individuals and their health care providers—though few in number—implies that

education efforts aimed at improving health care providers’ skills in initiating and

165

conducting ACP conversations are effective. End of life decision-making and
communication are important topics to be included in the curriculum of professional
health care education programs and in continuing education offerings (Ferrell et al., 2005;
Field & Cassel, 1997; Weissman et al., 2005). Such educational efforts support
development of competencies outlined by the Pew Health Professions Commission
(1998) in their recommendations for the 21St century health care provider.

Implications for Research

The implications of this dissertation study for research can be organized by the
key features of the ACP process. First, as a deliberative process, readiness to engage in
ACP or the impact of interventions depends on an accurate appraisal of participants’
knowledge of ACP. There must also be a way to differentiate individuals’ personal
values from their perceptions of others’ values of ACP. Lastly, the process individuals
use to identify their preferences—whether constructing, eliciting, or revealing them—
needs to be examined.

Secondly, as a communicative process, strategies for promoting conversations
with all decision-making partners need to be studied. The essential elements of a quality
conversation need to be defined and measurement criteria created to determine what level
of quality conversation is acceptable and effective for the individual and the decision
partners. Next, creative formulations of advance plans should be examined for
prevalence of use and effectiveness of guiding future end of life decision-making.
Finally, individual, organizational, and community strategies to make plans available

should be evaluated.

166

As a shared decision-making process, the method of presenting choices,
probabilities, and condition-related decisions needs to be examined to determine whether
it enhances the ability of individuals and their chosen others to share decision-making.
ACP research should study decisional conflict as well as design and test strategies to
navigate such conflict successfully. Further examination is also needed of the influences,
both internal and external, that affect decision-making. In addition to exploring personal
influences, studies should also assess influences from the community and the community
environment that may affect decision-making. Future research should also attempt to
identify unique ACP needs for groups of individuals; e. g., young adults, men, diverse
race/ethnic groups, diverse religious/spiritual groups, diverse economic status, etc.

Finally, as a dynamic decision-making process, ACP research needs to examine
the process of re-engaging in ACP. The results of this dissertation study indicate that a
variable like ACP Experience is usefirl to create groups for comparison. Areas of further
inquiry could include examining the impact of the frequency or regularity of re-
engagement with ACP on product, process, and promise outcomes.

Implications for Future ACP Nursing Research

This dissertation study has laid the groundwork for this investigator’s future ACP
research. First, this study introduces a conceptual fiamework for studying ACP fi'om a
decision-making perspective, the Decision Process Model for Advance Care Planning.
This model—with its attention to decision, behaviors, and outcomes occurring within a
dynamic decision-making context—provides a wealth of opportunities for future study.
It is hoped that in future studies, relationships between variables can be analyzed and

modeled. It will be interesting to test variables within the decision-making context to

167

determine if they act as moderators or mediators of the decision, behaviors, or outcomes
in the model.

Secondly, this study presents a more comprehensive picture of some internal
influences on individuals’ decisions to engage in ACP than is currently reported in the
literature. In future studies it will be interesting to explore more internal influences and
also external influences that may affect individuals’ decisions to engage in ACP.

Thirdly, this study measured the impact of a community-based ACP educational
program intervention on selected personal influence and decision variables related to
having ACP conversations. Much more study is warranted to examine how people
construct and especially communicate their preferences for end of life care. Given the
disparity in frequencies of conversations with HCAS, significant others, and HCPs, it
would be interesting to design and test interventions that target the prevalence and quality
of conversations with each partner.

Finally, this study provided a first look at using ACP Experience as a variable to
explore individuals’ ongoing engagement in ACP. Since ACP has been defined in this
study as a dynamic decision-making process, it will be interesting to design studies that
specifically explore individuals’ re—engagement in ACP.

Conclusion

This descriptive study has provided an in-depth examination of individuals who
decided to engage in ACP by attending an ACP education program provided in existing
community-based groups. For each participant, personal influences on decision and
personal decision factor variables were identified within the decision-making context.

The effectiveness of an external ACP educational program intervention was measured in

168

terms of its impact on selected internal personal and decision factors. ACP educational
programs, provided by certified ACP Facilitators to existing community-based groups,
were found to have statistically significant effects on the knowledge, importance, ability
and likelihood of having conversations with others about end of life preferences.
Participants who decided to engage in ACP were compared with each other on the basis
of previous ACP experience and significant differences were identified. All of this adds
to the understanding of ACP as a dynamic decision-making process that assists people to
construct and communicate their preferences for end of life care. Future research will
continue to examine the intemal and external influences within the decision-making
context that affect individuals’ decisions, behaviors and outcomes of engaging and re—

engaging in ACP.

169

Appendix A:
Appendix B:

Appendix C:

Appendix D:

Appendix E:

APPENDICES

Participant Survey
Information Letter / Consent Form

Dissertation Study Items from the Participant
Survey

Resources for ACP Conversations

Glossary

170

Appendix A

Participant Survey
Pre-Program Participant Survey — Page 1

Your personal identifying number 9

 

Birth month Birth day Gender Color
Mid-Michigan Advance Care Planning Coalition Research Project
Pre—Program Participant Survey

Please check the answer that best represents your opinion for each question. Please fill in the
blanks where indicated. Additional comments are welcome at the end of the survey.

Date Program Site

 

1. What do you think “Advance Care Planning ” means? (please check all that apply)

13 Living Will Cl Power of Attorney for El Power of attorney for
health care financial issues
Cl A Will El A Patient Advocate D A Legal Guardian
E] A Trust [:1 Funeral planning Cl A conversation about your
wishes
Cl Instructions for use at a El Something else (specify) El I’m not sure

time when you are unable
to make healthcare
decisions for yourself

 

 

 

2. Do you have an Advance Directive?
El Yes [:1 No D I’m not sure
-) 2a. If yes, what kind?

 

9 2b. If yes, where is your Advance Directive currently located?
(please check all that apply)

El At home El At my attomey’s office Cl In my medical record
where I receive health care
El With my health care El With other family members [:1 Safe deposit box
agent or significant others
Cl Somewhere else (specify): El I’m not sure

 

171

3. Have you ever made medical decisions regarding the use of life-sustaining treatments
for another person?
El Yes D No El I’m not sure

9 3a. If yes, was this person [:1 an adult E] a child
9 3b. If yes, did you feel prepared to make these medical decisions?
El Yes El No Cl I’m not sure

 

4. Have you chosen a person (Patient Care Advocate) to make decisions for you if you ever

 

 

 

 

become too ill to make decisions for yourself?

Cl Yes El No Cl I’m not sure
Pre-Program Participant Survey — Page 2
Birth month Birth day Gender Color

Your personal identifying number 9

 

9 4a. If yes, have you had at least one conversation about your end-of-Iife wishes with
that person (Patient Care Advocate)?

El Yes [:1 No D I’m not sure
9 4b. If yes, please estimate the quality of that conversation (circle the number):
1 2 3 4 5
Few comments Good conversation Complete
only and
thorough
conversation

5. Have you discussed your expectations for future medical care, including end-of-life
decisions or wishes, with your physician or other primary health care provider?

1:! Yes D No El I’m not sure
9 5a. If yes, please estimate the quality of that conversation (circle the number):
1. 2 3 4 5
Few comments Good conversation Complete
only and
thorough

conversation

6. Have you discussed your expectations for future medical care, including end-of-life
decisions or wishes, with other members of your family or significant others?
Cl Yes El No El I’m not sure

172

 

 

9 6a. If yes, please estimate the quality of that conversation (circle the number):

1 2 3 4 5
Few comments Good conversation Complete
only and
thorough
conversati
on

7. The following demographic information is optional but very helpful to us:

7a. Your gender El Male El Female

7b. Your age CI 18-30 D 31-59 CI 60-75 El over
75

7c. Your zip code

 

7d. Your religious or
spiritual preference

(if any)

 

7e. Your civic
affiliation (if any)

 

Pre-Program Participant Survey — Page 3

Your personal identifying number 9

 

Birth month Birth day Gender Color

8. Please circle the number that best reflects your opinion at this time:

8a. I feel that I am able to start a discussion about the use of future, potentially life-
sustaining medical treatment with those close to me.
1 2 3 4 5
not at all very much
8b. I feel I have enough knowledge to discuss my views of future, potentially life-sustaining
medical treatments with those close to me.
1 2 3 4 5
not at all very much
8c. I feel it is important to have a discussion of future, potentially life-sustaining treatment
with those close to me.
1 2 3 4 5
not at all very much
8d. I am likely to have a conversation with those close to me about future, potentially life-
sustaining medical treatment.
1 2 3 4 5
not at all very much

173

9. Are there additional comments you would like to make regarding this topic?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Please give this survey to the Facilitator. Thank you for participating in this project!

174

Post-Program Participant Survey - Page I

Mid-Michigan Advance Care Planning Coalition Research Project

Post Program Participant Survey

Please check the answer that best represents your opinion for each question. Please fill in the
blanks where indicated. Additional comments are welcome at the end of the survey.

Date Program Site

 

 

1. Please circle the number that best reflects your opinion at this time:

la. I feel that I am able to start a discussion about the use of future, potentially life-
sustaining medical treatment with those close to me.
1 2 3 4 5
not at all very much
1b. I feel I have enough knowledge to discuss my views of future, potentially life-sustaining
medical treatments with those close to me.
1 2 3 4 5
not at all very much
1c. I feel it is important to have a discussion of future, potentially life-sustaining treatment
with those close to me.
1 2 3 4 ~ 5
not at all very much
1d. I am likely to have a conversation with those close to me about future, potentially life-
sustaining medical treatment.
1 2 3 4 5
not at all very much

2. Are there additional comments you would like to make regarding this topic?

 

 

 

 

 

 

 

Please give this survey to the Facilitator. Thank you for participating in this project!

175

Appendix B
Information Letter / Consent Form

Information Letter — Page 1
Mid-Michigan Advance Care Planning Coalition Research Project
Participant Survey — Pro-Program & Post-Program

Welcome to this program on Advanced Care Planning! We are very glad that you have come and
we look forward to sharing information with you. While you are waiting for this program to
begin, would you consider completing this Participant Survey?

This survey is being conducted by the Mid-Michigan Advance Care Planning Coalition to help us
understand what advanced care planning you are already doing. This will help us plan firture
advance care planning educational programs for our community. The Pre-Program Participant
Survey has eight questions about advance care planning and demographic data. It should take
approximately five to ten minutes to complete. At the end of the program, the Facilitator will
distribute a one question Post-Program Survey. This should take approximately two to five
minutes to complete.

Your participation in this research project does not involve any cost. Your participation is
voluntary. You may choose not to participate at all or you may choose not to answer certain
questions. You may also discontinue your participation at any time without penalty. Ifyou
prefer not to complete the surveys, simply place the blank forms in the Participant Survey
envelopes provided by the Facilitator. There are minimal risks involved in completing the
survey; however you may experience strong feelings about the subject. If you do experience
distressing feelings related to this survey, please talk to the Facilitator who will provide you with
a community resource contact number.

Following this Advance Care Planning program, the Facilitator will send the surveys to our
research team. The survey data from all participants will be entered into a database and analyzed
as groups by the researchers. Your data will not be reported individually but only within a group.
These group results will be shared with the Mid-Michigan Advance Care Planning Coalition and
its collaborating organizations.

In the future, we may want to collect some follow-up data from participants in these programs.
Please indicate if you would be willing to be contacted one additional time. Your contact
information will be detached from your survey and filed in a separate place so that your name
will not appear anywhere on the survey form. Your privacy will be protected to the maximum
extent allowable by law.

176

 

 

If you have any questions or concerns
regarding your rights as a study participant,
or are dissatisfied at any time with any
aspect of this study, you may contact—
anonymously if you wish—

Peter Vasilenko PhD

Chairperson of the University

If you have any questions about this study,
please contact the investigator:

Gwen Wyatt RN, PhD
B422 West Fee Hall
Michigan State University
East Lansing, MI 48824

 

Phone: (517) 432-5511 Committee on Research Involving
Fax: (517) 353-8612 Human Subjects (UCRIHS)
Email: ngatt@msu.edu 202 Olds Hall

Michigan State University

East Lansing, MI 48824
Phone: (517) 355-2180
Fax: (517) 432-4503

Email: ucrihs@msu.edu

 

 

 

 

(please turn the page)
Consent Form — Page 2

If you are willing to participate, please do the following:
1. Sign this consent form.

You indicate your voluntary agreement to participate by completing and returning this consent
form and the surveys.

 

 

Signature Date

2. Create a personal identifying number by writing:

 

 

 

 

 

 

The two digits of The two digits of M for Male or The first letter of your
your birth month your birth day F for Female favorite color
(e.g. January = 01, (e.g. 06 or 13) (e.g. Green = G)
October = 10)

 

Now write all these numbers and letters in a row onto the top of both the pre-program and
post-program surveys. This will help us identify your surveys without having your name

appear on them anywhere. (e.g. O713FG for July 13 Female Green)

177

 

 

 

 

 

 

3. Are you willing to be contacted in the future for a follow-up to this program?
If yes, how may we contact you?
D YES
Name:
1] N 0 Address:
Phone: Email:

 

 

 

 

 

Please DO complete the pre-program survey now.

Please DO NOT complete the post-program survey yet!

Please put all of these papers in the appropriate envelopes the Facilitator will pass
around.

Thank you so much for participating in this survey!

178

 

Appendix C

Dissertation Study Items fiom the Participant Survey

Table C1

Dissertation Study Items from Participant Survey

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Data Collection
Participant Survey Items Conceptual Pre- Post- Item
Model Program Program type
Variables
8.b. I feel I have enough knowledge to discuss Enough X X Ordinal/
my views of future, potentially life-sustaining knowledge Interval
medical treatments with those close to me. to discuss
l 2 3 4 5 my views
1 = not at all 5 = very much
7.b. Your age Age X Nominal/
- 18-30 0 31-59 Ordinal
0 60-75 0 > 75
7.a. Your gender Gender X Nominal
0 Male 0 Female
Sc. I feel it is important to have a discussion Important to X X Ordinal/
of future, potentially life-sustaining treatment have a Interval
with those close to me. . discussion
1 2 3 4 5
1 = not at all 5 = very much
3. Have you ever made medical decisions Making X Nominal
regarding the use of life-sustaining treatments decision for
for another person? someone
0 Yes 0 No 0 I’m not sure else
3.a. If yes, was this person Making X Nominal
0 An adult 0 A child decision for
someone
else
3.b. If yes, did you feel prepared to make Making X Nominal
these medical decisions? decision for
a Yes 0 No 0 I’m not sure someone
else
2. Do you have an Advance Directive? Having an X Nominal
Advance
0 Yes 0 No 0 I’m not sure Directive
(AD)
4. Have you chosen a person (Patient Care Having a X Nominal
Advocate) to make decisions for you if you Health Care
ever become too ill to make decisions for Agent
yourself? (HCA)
- Yes 0 No 0 I’m not sure

 

 

 

 

 

 

179

 

Table C1 (continued)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Conceptual Data Collection
Participant Survey Items Model Pre- Post- Item
Variables Program Prgram type
4.a. If yes, have you had at least one Having X Nominal
conversation about your end-of-life wishes Conversatio
with that person (Patient Care Advocate)? ns with
a Yes I o No I 0 I’m not sure HCA
6. Have you discussed your expectations for Having X Nominal
future medical care, including end-of-life Conversatio
decisions or wishes, with other members of ns with
your family or significant others? Significant
0 Yes 0 No 0 I’m not sure Others
5. Have you discussed your expectations for Having X Nominal
future medical care, including end-of-life Conversatio
decisions or wishes, with your physician or ns with
other primary health care provider? HCP
0 Yes a No 0 I’m not sure
4.b. If yes, please estimate the quality of that Quality of X Norninal/
conversation [with HCA]: conversatio Ordinal
l ---- 2 ----- 3 ---- 4 --- 5 ns with
1 = few 3 = good 5 = complete HCA
comments conversation & thorough
only conversation ‘
6.a. If yes, please estimate the quality of that Quality of X Nominal/
conversation [with significant others]: Conversatio Ordinal
1 2 3 4 ---- 5 ns with
l = few 3 = good 5 = complete Significant
comments conversation & thorough Others
only conversation
5.a. If yes, please estimate the quality of that Quality of X Nominal/
conversation [with HCP]: Conversatio Ordinal
l 2 3 4 ---- 5 ns with
1 = few 3 = good 5 = complete HCP
comments conversation & thorough
only conversation
8.a. I feel that I am able to start a discussion Able to start X X Ordinal!
about the use of future, potentially life- discussion Interval
sustaining medical treatment with those close with others
to me.
1 2 3 4 --- 5
1 = not at all 5 = very much
8.d. I am likely to have a conversation with Likely to X X Ordinal/
those close to me about fixture, potentially have Interval
life-sustaining medical treatment. conversatio
1 2 3 4 --- 5 n with
1 = not at all 5 = very much others

 

 

 

 

 

180

 

 

 

Appendix D

Resources for ACP Conversations

8 Essential Talking Points for
Advance Care Planning

1. Advance Care Planning (ACP) has 4 parts:

1. Selecting a decision maker who can make decisions for you if you
are unable.

This person may be called the Health Care Agent (HCA), the Durable
Power of Attorney for Health Care (DPOA-HC), or, in Michigan, the
Patient Advocate.

2. Having conversations about your end of life care preferences.
You should talk with your Patient Advocate, family or fiiends, and
health care providers.

3. Completing an Advance Directive.

Although you can informally talk about your preferences, a written
Advance Directive is the most legally reliable document. It may
include designation of a Patient Advocate (DPOA-HC) and/or
documentation of your instructions/preferences.

4. Making your completed Advanced Directive readily available
Keep your original in a safe place BUT be sure that you give key
people copies that are readily available in a time of an unexpected
emergency, (see Point # 7).

2. ACP is appropriate for adults of all ages and health conditions.
At any time, you might not be able to participate in making medical
treatment decisions; for example, you may be unconscious, or unable to
communicate. Whether this is a temporary or permanent loss, you need
someone (your Patient Advocate) who can communicate your
instructions/preferences to the health care providers. It is appropriate to
involve your school age and adolescent children in your conversations
about your end of life preferences.

3. ACP is an opportunity to express your wishes about the kinds of end
of life care you would like to have and not like to have.
In your conversations and your Advance Directive, you should talk about
the circumstances of the care you want or don’t want. For example,

181

0 you may want the health care providers to try a treatment at first to see
if you can recover but discontinue it if it is not helping you improve
your condition.

0 you may always want comfort measures.

0 you may never want an experimental treatment.

4. ACP is an ongoing process.
Your wishes may change as you grow older or if your health status
changes. Perhaps there will be a treatment for a disease that is now
available. Perhaps you have a change of heart about a treatment. It is a
good idea to review your Advance Directive annually. It is also
important to talk with your Patient Advocate, your family or fiiends, and
your health care provider if your wishes have changed.

5. Your Patient Advocate only makes decisions for you if you are
unable to make them.

Usually doctors decide if you are unable to make decisions for yourself

based on your health condition. If you are unable, they will call in your

Patient Advocate to make decisions for you.

0 In Michigan, the Patient Advocate must sign the Advance Directive to
show that s/he is willing to be your decision-maker.

o It is a good idea to appoint a back-up Patient Advocate in case the first
person is not available when needed.

0 Your Patient Advocate can make decisions for you about consenting
to or refusing medical treatment. This includes decisions to withhold
or withdraw life-sustaining treatment; these decisions could allow you
to die of the illness or injury you have. It is very important that you
have talked about these decisions ahead of time with your Patient
Advocate and best if you leave instructions about them in your written
Advance Directive.

6. An attorney is not needed to complete an Advance Directive.
There are many free or low-cost booklets and forms that can help you fill
out your Advance Directive. However, some people find it convenient to
create an Advance Directive with their lawyer when they are making out
a will or doing estate planning.

182

7. Once you have created your Advance Directive, be sure everyone
who needs to know about it has a copy.

Keep the original copy with your personal papers.

Give a copy to your Patient Advocate(s).

Let your family know you have an Advance Directive and/or give
them copies.

Give a copy to your personal physician / health care provider.

Give a copy to your attorney (if you have one).

Always bring a copy with you to the hospital where you will receive
treatment.

Keep a copy in your car.

Fill out a wallet card that lets people know you have an Advance
Directive.

If desired, contact an organization that will keep a copy on record and
send it to wherever it is needed.

8. ACP is a gift to yourself, your family, and your Patient Advocate.
By talking about your wishes for end of life care, you can let your family
and Patient Advocate know what is most important to you. This can
make their decision-making less difficult.

10/2005 gw/kjv/jdm
Capital Area Health Alliance — Advance Care Planning Coalition

183

Take time to personally reflect on your values and wishes for end of life
care. You may wish to consider the following questions from “Caring
Conversations:”

0 What beliefs do you hold that influence your thoughts about life and your
thinking about dying?

 

 

 

- With whom do you want to have a caring conversation about your values
and wishes?

 

- What do you most want to say to them?

 

 

 

0 When and where will you have your caring conversation?

 

 

 

o What concerns do you have about your health or future healthcare?

 

 

 

o What are your fears regarding the end of your life?

 

 

 

184

What do you most value about your physical or mental well being?

(e. g. Do you love to be outdoors? Does being able to read or listen to music bring
you pleasure? How important is it to be aware of your surroundings and the people
with you? How important is seeing, tasting, touching?)

 

 

 

Are there circumstances under which you would refuse or discontinue
treatment that might prolong your life? If so, describe those
circumstances.

 

 

 

If you could plan it today, what would the last day of your life be like?
(e. g. Where would you be? What would you be doing? What would you eat? What
music would you listen to? What would be your final words and your last acts?)

 

 

 

How do you want to be remembered?

 

 

 

What gives your life its purpose and meaning?

 

 

 

What do you need for comfort and support as you journey near death?
(e. g. Do you want to pray with someone? Be read to from spiritual or religious texts?
Listen to poetry or tapes? etc.)

 

 

 

185

Would you want to have a hospice team or other palliative (i.e. comfort)
care available to you?

 

 

 

Would you want treatments that might prolong your life if you were
(circle)

0 No longer able to think for yourself? Yes No
0 Comatose and not likely to regain consciousness? Yes No
0 Terminally ill or near death? Yes No
0 Of very advanced age? Yes No

Who would you want to make healthcare decisions for you if you
couldn’t make them for yourself?

 

o What is his or her relationship to you?

 

0 Who would be your second choice?

 

0 Could they (or would they) be willing to carry out your wishes?

Where do you want to be and what things do you need to be comfortable

as you die? (e.g. Would you like to be in a hospital? A special place? At home?
Is it important to have sunlight or fresh breezes? To be free of uninvited guests? To
be held? To be alone? To review family traditions? To listen to music?)

 

 

 

Would you want to be sedated if it were necessary to control your pain?

 

 

 

If you could no longer swallow, would you want tube feedings?

 

 

186

What does “very advanced age” mean to you?

 

 

 

0 Do you wish to donate your organs and tissues?

 

 

 

Do you wish to donate your body to science for health education?

 

 

 

Would you agree to an autopsy? Do you prefer burial or cremation?

 

 

 

Compiled 2005 by John McPhail MA, CRC, LPC
Greater Lansing Health Ministries Consortium

187

What Should You Do With Your Advance Directive Now?

It is extremely important that this document, that specifies your wishes, is
available. Therefore, you should:

4 444-44

Aééé

Let your family know you have completed this.

Keep the original document with your personal papers.

Give a copy to your doctors.

Give a copy to your attorney.

Give a copy to the person(s) designated as your Patient Advocate or
Durable Power of Attorney for Health Care.

Give a copy to the hospital where you will seek medical care. (Bring to
the Medical Records Department.)

Always bring a copy with you to the hospital.

Keep a copy in your car. ’

Complete a wallet card.

Contact any organization that will keep a copy on record and send a copy
to wherever it is needed.

Other:

 

 

Compiled 2005 by John McPhail MA, CRC, LPC
Greater Lansing Health Ministries Consortium

188

Making Choices: Goals, Benefit & Burdens

How do you determine if a treatment is right for you? These questions may
help you clarify the value of a treatment to you.

PP’NT‘

What is the goal of the treatment?

What are the benefits of the treatment?

What are the burdens of the treatment?

Are the benefits and burdens compatible with your individual goals?

 

A treatment my be beneficial if it

A treatment may be burdensome
if it

 

 

. Is effective in prolonging life

0 Is effective in restoring /
maintaining function

. Is effective in relieving
suffering

. Promotes the goals / values of
the person

. Is consistent with the person’s
religious/cultural/spiritual
beliefs

 

 

Results in more or intolerable
pain

Is damaging to body image or
function

Is psychologically harmful
Is physically restrictive
Carries an unacceptable cost

for the person (financially,
emotionally, spiritually)

 

Some more things to consider:

- Your preferences and goals may change as your health status changes.
Some goals may become higher priorities or short-term goals may be

balanced against long-terrn goals.

. You may need the assistance of others to determine your goals. Those
who might provide assistance are your healthcare providers, pain-
management specialists, hospice team, spiritual counselors, social

workers, etc.

189

 

. In order for your goals to be honored, you must communicate them to
your designated decision-maker (patient advocate or durable power of
attorney for health care), family members and friends, and healthcare
team members.

. Think about your goals and have these conversations as early as possible
and in nonstressful situations with your family members and healthcare
professionals.

Making Choices: Decisions & Definitions

Life-sustaining treatment: Any treatment that is used to sustain biological
life

Choices about life-sustaining treatments:
0 When to start
When to forego or withhold
When to withdraw
How and when to maintain comfort
When to mix management: making choices that extend life as well as
choices that prepare for death and provide comfort

Cardiopulmonary Resuscitation (CPR)

. It is a standing order to start CPR in a hospital and most other healthcare
organizations and by paramedics if a patient suffers a cardiac or
respiratory arrest—unless there are clear orders not to start (e. g. a
physician’s order that CPR not be attempted).

. Basic CPR involves pressing on the chest and blowing air into the lungs
when breathing and heartbeat has stopped or has become ineffective.
CPR typically lasts for 15 to 30 minutes. It may require administration of
medications, the insertion of a tube to assist with breathing (intubation),
and electrical stimulation of the heart.

. People who are at clear risk for requiring CPR include those with heart

disease, those in long-term care facilities, or those with end-stage
diseases.

190

Benefits:

CPR is most beneficial for a healthy person whose heart has stopped
suddenly from an accident or heart attack or in a person whose
underlying condition can be effectively treated. CPR can prolong life
with good fimctioning especially in patients who are healthy and
younger, and if it is initiated within five nrinutes of the cardiac arrest.

Burdens:

Less than 15% of hospitalized patients survive CPR and return to
previous fimctional status. Only 1-4% of patients with multiple
underlying medical conditions survive to leave the hospital. Rarely does
a patient with an advanced cancer diagnosis leave the hospital after CPR.

Pressing on the chest can cause soreness, broken ribs or collapsed lungs.

Many people who survive CPR may require ventilator support for a
period of time.

Artificial Ventilation

Intubation involves the insertion of a tube through the mouth or nose into
the lungs. This tube can then be connected to a breathing machine, or
ventilator, to artificially support breathing.

People with advanced lung disease, pulmonary infections, and
neuromuscular disorders who could become ventilator-dependent should
consider the type of medical interventions they would want if they went
into respiratory failure.

Benefits:

Adequate breathing or respiration (which is the process of providing
oxygen to the body and removing carbon dioxide) can be interfered with
by a variety of temporary and reversible conditions. These conditions
include pneumonia and the need for breathing support after surgery or
when a lung has collapsed after an accident. The breathing tube and
artificial ventilator provide adequate respiration while the lung is healing
or the body is recovering from another illness.

For people with chronic pulmonary diseases, artificial ventilation can be

used on a trial basis to see if they can improve enough to adequately
breathe on their own. The benefits are to rest the lungs while other parts

191

of the person’s health condition are being managed.

In both situations, artificial ventilation may prolong life.

Burdens:

The breathing tube produces discomfort from throat irritation, coughing
and the need to suction secretions from the airway. The artificial
ventilator may require some getting used to while the person learns to let
the machine do part or all of the breathing. Psychologically, the person
may be afraid or suffer sleep disturbances.

To treat the discomfort, some people may require medications such as
morphine and sedatives, which may alter level of consciousness.

When using artificial ventilation for chronic pulmonary situations, it may
be difficult or impossible to remove the ventilator. In these cases,
persons are never able to resume breathing on their own and become
ventilator-dependent for the rest of their lives.

In addition, it may prolong dying.

Artificial Nutrition and Hydration

Artificial nutrition and hydration involves the short-term or long-term
administration of a balanced mix of nutrients and fluids via tubes
(nasogastric, gastrostomy, and jejunostomy, intravenous) placed directly
into the stomach, intestine, or vein.

Short-term administration is needed to temporarily support a person
while the cause of their inability to take nutrition is corrected (e.g.
recovery from surgery).

Long-term administration permanently sustains nutritional needs in
persons who will never recover the ability to take nutrition on their own
(e. g. persistent vegetative states, irreversible neurological disorders).

Benefits:
0 Life may be prolonged.

192

Person’s personal and/or religious preferences are honored.

Administration prevents weakness, dry mouth and thirst related to
dehydration.

Burdens:

Feeding tubes are often associated with aspiration of nutrients into the
lungs, causing pneumonia (30% of cases) or throat, esophagus and
stomach irritation and discomfort.

Intravenous fluids increase the volume of secretions in the lungs, making
breathing more labored and necessitating more frequent suctioning.
Intravenous fluids increase congestion in other parts of the body, such as
around tumors and organs, causing pain and discomfort as well as
increased urination requiring frequent elimination needs and linen
changes.

For some confused persons who are in danger of self-harm from pulling
at tubes, physical or chemical restraints may be required.

Antibiotics

Infections used to be the way many people died, both young and old.
Today we have sophisticated antibacterial agents that can prevent death
in many cases, even with serious infections.

Benefits:

Antibiotics eliminate the source of infection and, therefore, the
accompanying side effects of an infectious process such as fever, chills,
and discomfort.

They may prolong life.

Burdens:

Many antibiotics to treat infection need to be given intravenously, which
requires starting and maintaining an IV site. There may be potential
discomfort associated with starting the IV and keeping it patent.

Antibiotics may delay the dying process by temporarily reversing a fatal

event in an incurable illness (e. g. a person who has pneumonia due to
end-stage lung cancer).

193

Dialysis

. Dialysis is a treatment provided to a person whose kidneys have stopped
working. The kidneys take the waste out of the blood stream put there by
the body’s cells. If this waste is not removed, it will build up and

ultimately cause death.

Benefits:
. Dialysis removes toxic waste products, allowing a person’s other vital
organs to function more normally.

. It may prolong life.

Burdens:

. Dialysis inVolves the insertion of catheters into the bloodstream and up to
several hours of- removing and filtering the person’s blood several times a
week.

. Once kidney function is gone, people are dependent on dialysis for the
rest of their lives, or they may receive kidney transplants.

. Dialysis is expensive.

. It may prolong the dying process.

Making Choices:
Withholding & Withdrawing Life-Sustaining Treatment

Many people fear that once a treatment is started, it will never be withdrawn.
This may lead people to make statements such as, “I don’t want to be
hooked up to machines,” or “I don’t want any tubes.”

Unfortunately, some of these fears may be translated into decisions that
make it more difficult for health professionals and your decision-maker to
attempt short-term trials of interventions that could be beneficial and
consistent with your established goals.

If a trial of aggressive treatment is not serving the purpose for which it was

started, it can be compassionately withdrawn without violating the rights of
the person.

194

There is no moral or legal distinction between withholding treatment and
withdrawing treatment. Both actions result in the same outcome: allowing
the person to die of their underlying disease process.

The central moral question is whether or not there is justification to start or
continue a treatment for a person given the circumstances of the situation
(i.e., the person’s diagnosis, prognosis, benefits and burdens of treatment,
goals and values).

Making Choices: Comfort Care

What does “comfort care” mean to you? Choosing comfort care measures
does not mean giving up or not providing potentially therapeutic treatment.
Comfort care measures can and should be offered and provided at any stage
of a person’s health.

How do you feel about the control of symptoms during the dying process?
Would you prefer to be as free as possible of your symptoms? Many
interventions are available but some may have unintended effects (e. g. pain
medicine may also cause sedation).

What other preferences do you have for comfort? Examples: music,
selected people around you, spiritual counseling, massage, pets,
environment, etc.

What present or future experiences are most important for you to live well at
this time in your life?

In what way could you make this time especially meaningful for you and
your loved ones? What legacy do you wish to leave?

Compiled from Respecting Choices® Advance Care Planning Facilitator’s Manual (2002), Chapter 3
KVL 2005
Capital Area Health Alliance — Advance Care Planning Coalition

195

Appendix E
Glossary

Advance Care Planning (ACP)
A dynamic decision-making process (including decisions, behaviors, and outcomes) that
helps people construct and communicate their preferences for end of life care.
Advance Care Planning (ACP) Experience
A participant’s self-report of the prior experiences of having an Advance Directive (AD)
or having a Health Care Agent (HCA) or both.
Advance Care Planning (ACP) Program
A health education information sharing session about ACP, presented by a certified
Respecting Choices® ACP Facilitator to voluntary participants who were members of an
established community group.
Advance Directive (AD)
A statutory document executed to communicate an individual’s health care preferences
for use at a future time when the individual’s capacity to make informed health care
decisions is lost. An AD may include designation of a health care agent and/or
instructions for health care decisions.
Community
Both formal associations of people who are valued by one another and informal social
structures and processes (e. g., discussions, activities, affiliations, responsiveness to one
another) that reflect or advance some recognized commonality of history, culture, or

perspective that is of value to the participants (Byock et al., 2001).

196

Community-Based Intervention

An intervention that is directed toward individuals and families within a community and
designed to meet the needs of people where they live, work, worship, go to school, and
receive healthcare (Calvin College Department of Nursing, 2003).

Culture

A set of attributes that collectively characterize a community: commonly held values,
attitudes, assumptions, history, expectations, hopes, fears, and customary modes of
professional, social, and personal interactions (Byock et al., 2001).

Decision

Choices between alternatives (Thompson & Dowding, 2002).

Decision Factors

Variables that modify an individuals’ willingness or ability to make a decision.
Decision-Making

The evaluation of information about alternatives according to personal values (Mazur,
2003).

Decision-Making Context

The decision, behaviors, outcomes and influences (internal and external) that shape the
decisions people make (Wills & Holrnes-Rovner, 2006).

Decision Support Interventions

Strategies used by health care providers to promote informed, values-based decision-

making (A. Clarke et al., 2005).

197

Decision to Engage in ACP

Individuals’ willingness to involve themselves in exploring their preferences for end of
life care.

Durable Power of Attorney — Health Care (DPOA-HC)

A person who has been chosen and legally appointed by a patient to be the substitute
health care decision-maker. (Also known as Health Care Agent (HCA), Patient Advocate,
and Proxy) (Post et al., 1999).

Dynamic Decision-Making

A decision-making process that may be engaged in repeatedly in response to changes
within the decision-making context.

End of Life

The period during which an individual copes with declining health from an ultimately
terminal illness, fi‘om a serious though perhaps chronic illness, or from the frailties
associated with advanced age—even if death is not clearly imminent (Lunney, Foley et
al., 2003).

End of Life Care

Both professional and informal care to treat medical conditions or meet basic human
needs (such as bathing or eating) at the end of life (Byock et al., 2001).

Engaging in ACP

Understanding, reflecting, identifying, and discussing alternatives to help in formulating

and making available plans for end of life care (adapted from Hammes & Briggs, 2002b).

198

Enough Knowledge

The participant’s self-report of the perception whether or not they have sufficient
information to discuss their preferences.

Established community groups

Organizations located in a community in which members naturally participate (P. Clarke
et al., 2005).

Family

A set of relationships—established by blood or emotional bonds—in which individuals
find identify and fulfill mutual obligations of domestic affairs, such as maintaining a
household, raising a child, or providing care for each other (Byock et al., 2001).

Having an AD

The participant’s self-report of executing a statutory document for use at a future time
when personal capacity to make informed health care decisions is lost.

Having a HCA

The participant’s self-report of choosing and legally appointing a person to be the
substitute health care decision-maker (Post et al., 1999).

The participant’s self-report of appointing an individual to make surrogate decisions at a
future time when personal capacity to make informed health care decisions is lost.

Having conversations

The participant’s self-report of discussing preferences for end-of-life care with others.

199

Health Care Agent (HCA)

A person who has been chosen and legally appointed to be the substitute health care
decision-maker. (Also known as Durable Power of Attorney — Health Care, Patient
Advocate, and Proxy) (Post et al., 1999).

Health Care Provider (HCP)

A licensed professional who provides health care to individuals within a health care
setting. These licensed professionals may include physicians, rrrid-level providers (e. g.,
Advance Practice Nurses, Physician Assistants), registered nurses, dietitians, social
workers, therapists, and chaplains.

Health Education

Developing and providing instruction and learning experiences to facilitate voluntary
adaptation of behavior conducive to health in individuals, families, groups, or
communities (O'Toole, 1997).

Important to have discussion

The participant’s self-report of the perception of the necessity of having conversations
about preferences for end-of-life care.

Informal Surrogate

A person who is asked by the medical team to help make treatment decisions because no
one has been appointed by the patient or legally authorized (Post et al., 1999).

Information

Decision-relevant data inputs (Wills & Holmes—Rovner, 2006).

200

Initiated Advance Directive

Legal documents executed by persons while they are able to make personal decisions for
use at a firture time when they have lost the capacity to make informed decisions.
Making decision for someone else

The participant’s self report of an event or process of communicating a choice related to
life-sustaining treatment on behalf of another person who is incapable of making a
choice.

Patient Advocate

A person who has been chosen and legally appointed by a patient to be the substitute
health care decision-maker. (Also known as Durable Power of Attorney — Health Care,
Health Care Agent (HCA), and Proxy) (Post et al., 1999).

Patient-Centered Care

Care that is respectful of and responsive to individual patient preferences, needs, and
values and ensuring that patient values guide all clinical decisions (Committee on Quality
of Health Care in America, 2001).

Patient Self Determination Act (PSDA)

The Patient Self-Determination Act requires hospitals, nursing homes, and health care
programs to ask patients about advance directives and then to incorporate the information
into medical records (Crane et al., 2005).

Personal Decision Factors

Variables relating to the individual that modify an individuals’ willingness or ability to

make a choice between two or more alternatives.

201

Personal Influences on Decision

Variables relating to the individual that may act independently or interact with each other
to modify the Decision to Engage in ACP.

Preferences

A relatively greater liking of one alternative compared to another (Wills & Hohnes-
Rovner, 2006).

Prior experiences

Events, circumstances, or processes that have already occurred in an individual’s lifetime.
Proxy

A person who has been chosen and legally appointed by a patient to be the substitute
health care decision-maker. (Also known as Durable Power of Attorney — Health Care ,
Health Care Agent (HCA), and Patient Advocate) (Post et al., 1999).

Quality of Care

Care that meets recognized standards of professional service and that conforms to the
values and preferences of the individuals, families, and professionals within the
community (Byock et al., 2001).

Quality of Conversations

The participant’s self-report of the thoroughness of discussions of preferences with
others.

Self efficacy

An individual’s perceived confidence to perform certain behaviors (Bandura, 1986).

202

Sense of Control in the Dying Process

A patient’s internal feeling of being able to die in a way deemed acceptable and that any
medical outcomes or states deemed unacceptable will be avoided (Kolarik et al., 2002)
Shared decision-making environment

A context in which individuals and others (such as invited family members or health care
providers) exchange information about medical conditions and preferences to reach
treatment decisions that are mutually agreeable (F rosch & Kaplan, 1999).

Significant others

Persons who play an important role in the life of an individual (O'Toole, 1997).
Surrogate

A person whose authority to make health care decisions for someone else is based on
state statute or case law (Post et al., 1999).

Values

The importance individuals place on information, experience, and the context of

decision-making (Wills & Hohnes-Rovner, 2006).

203

REFERENCES

Anderson, J. P., Kaplan, R. M., & Schneidermann, L. J. (1994). Effects of offering
advance directives on quality adjusted life expectancy and psychological well-
being among ill adults. Journal of Clinical Epidemiology, 47(7), 761-772.

Bandura, A. (1977). Self-efficacy: toward a unifying theory of behavioral change.
Psychology Review, 84, 191-215.

Bandura, A. (1986). The explanatory and predictive scope of self-efficacy theory.
Journal of Social and Clinical Psychology, 4(3), 359-373.

Barnard, D. (2002). Advance Care Planning is not about "Getting It Right". Journal of
Palliative Medicine, 5(4), 475-481.

Beck, A., Brown, J., Boles, M., & Barrett, P. (2002). Completion of advance directives
by older health maintenance organization members: the role of attitudes and
beliefs regarding life-sustaining treatment. Journal of the American Geriatrics
Society, 50, 300-306.

Blackhall, L. J ., Frank, G., Murphy, S. T ., Michel, V., Palmer, J. M., & Azen, S. P.
(1999). Ethnicity and attitudes toward life-sustaining technology. Social Science
& Medicine, 48, 1779-1789.

Bradley, E. H., Hallemeier, A. G., Fried, T. R., Johnson-Hurzeler, R., Cherlin, E. J ., Kasl,
S. V., et al. (2001). Documentation of Discussions about Prognosis with
Terminally Ill Patients. American Journal of Medicine, I I I , 218-223.

Bradley, E. H., Peiris, V., & Wetle, T. (1998). Discussions About End-of-Life Care in
Nursing Homes. The Journal of the American Geriatrics Society, 46, 1235-1241.

Briggs, L. A., Havlik, P., & Hammes, B. (2000). Respecting Choices® Advance Care
Planning Program Quality Improvement Toolkit. LaCrosse, WI: Gundersen
Lutheran Medical Foundation.

Briggs, L. A., Kirchhoff, K. T., Hammes, B. J ., Song, M., & Colvin, E. R. (2004).
Patient-centered advance care planning in special patient populations: a pilot
study. Journal of Professional Nursing, 20(1), 47-5 8.

Brown, B. A. (2003). The History of Advance Directives. Journal of Gerontological
Nursing, 29(9), 4-14.

Brown, J. B., Beck, A., Boles, M., & Barrett, P. (1999). Practical methods to increase use
of advance medical directives. Journal of General Internal Medicine, 14, 21-26.

204

Brunswick, E. (1956). Perception and the representative design of psychological
experiments (2nd ed.). Berkeley, CA: University of California Press.

Byock, I., Norris, K., Curtis, J. R., 8c Patrick, D. L. (2001). Improving End-of-Life
Experience and Care in the Community: A Conceptual Framework. Journal of
Pain and Symptom Management, 22(3), 759-772.

Calvin College Department of Nursing. (2003). Conceptual Framework. Retrieved July
27, 2006, 2006, from
httpJ/wwwczdvin.edu/academic/nursinglconceptualframehtm

Charles, C., Gafiri, A., & Whelan, T. (1997). Shared decision-making in the medical
encounter: What does it mean? (or it takes at least two to tango). Social Science &
Medicine, 44(5), 681-692.

Cicirelli, V. G. (1997). Relationship of psychosocial and background variables to older
adults' end-of-life decisions. Psychology and Aging, 12(1), 72-83.

Clarke, A., J acobsen, M. J ., O'Connor, M. A., Stilwell, D., & Feldman-Stewart, D.
(2005). Section A: using a systematic development process. In A. O'Connor, H.
Llewellyn-Thomas & D. Stacey (Eds.), IPDAS Collaboration Background
Document (pp. 3-5): International Patient Decision Aid Standards (IPDAS)
Collaboration.

Clarke, R, Evans, S. H., Shook, D., & J ohanson, W. (2005). Information Seeking and
Compliance in Planning for Critical Care: Community-Based Health Outreach to
Seniors About Advance Directives. Health Communication, 18(1), 1-22.

Committee on Quality of Health Care in America. (2001). Crossing the quality chasm: a
new health system for the 21 st century. Washington, DC: The National
Academies Press.

Connors, A. F., Jr., Dawson, N. V., Desbiens, N. A., Fulkerson, W. J ., Jr., Goldman, L.,
Knaus, W. A., et al. (1995). A controlled trial to improve care for seriously ill
hospitalized patients: the study to understand prognoses and preferences for
outcomes and risks of treatments (SUPPORT). Journal of the American Medical
Association, 20, 1591-1598.

Crane, M. K., Wittink, M., & Doukas, D. J. (2005). Respecting end-of-life treatment
preferences. Am Fam Physician, 72(7), 1263-1268.

Curtis, J. R., Patrick, D. L., Caldwell, E., Greenlee, H., & Collier, A. C. (1999). The
Quality of Patient-Doctor Communication about End-of-Life Care: a study of
patients with advanced AIDS and their primary care clinicians. AIDS, 13, 1123-
1 13 l .

205

Daly, B. J. (2006). End-of-life decision making, organ donation, and critical care nurses.
Critical Care Nurse, 26(2), 78-86.

Davison, B. J ., & Degner, L. F. (1998). Promoting patient decision making in life-and-
death situations. Seminars in Oncology Nursing, 14(2), 129-136.

Decker, I. M., 8:. Reed, P. G. (2005). Developmental and contextual correlates of elders'
anticipated end-of-life treatment decisions. Death Studies, 29(9), 827-846.

Dexter, P. R., Wolinsky, F. D., Gramelspacher, G. P., Zhou, X.-H., Eckert, G. J .,
Waisburd, M., et al. (1998). Effectiveness of Computer-Generated Reminders for
Increasing Discussions about Advance Directives and Completion of Advance
Directive Forms: A Randomized, Controlled Trial. Ann Intern Med, 128(2), 102-
1 10.

Ditto, P. H., Danks, J. H., Smucker, W. D., Bookwala, J ., Coppola, K. M., Dresser, R., et
al. (2001). Advance directives as acts of communication: a randomized controlled
trial. Archives of Internal Medicine, 161(3), 421-430.

Doorenbos, A. Z., & Nies, M. A. (2003). The Use of Advance Directives in a Population
of Asian Indian Hindus. Journal of T ranscultural Nursing, 14(1), 17-24.

Ej az, F. K. (2000). Predictors of Advance Directives in Institutionalized Elderly. Journal
of Gerontological Social Work, 33(4), 67-89.

Emanuel, L. L. (2000). How Living Wills can help doctors and patients talk about dying.
British Medical Journal, 320, 1618-1619.

Emanuel, L. L. (2003). Research in advance care planning. In R. K. Portenoy & E.
Bruera (Eds), Issues in Palliative Care Research (pp. 261-273). New York:
Oxford University Press.

Ferrell, B. R., Virarri, R., Grant, M., Rhome, A., Malloy, P., & Bednash, G. (2005).
Evaluation of the end-of-life nursing education consortium undergraduate faculty
training program. Journal of Palliative Medicine, 8(1), 107-1 14.

Field, M. J ., & Cassel, C. K. (Eds). (1997). Approaching death: improving care at the
end-of-life. Washington, DC: The National Academies Press.

Freebome, N., Lynn, J ., & Desbiens, N. A. (2000). Insights about Dying from the
SUPPORT Project. Journal of the American Geriatrics Society, 48, $199-$205.

Frosch, D. L., & Kaplan, R. M. (1999). Shared decision-making in clinical medicine: past

research and future directions. American Journal of Preventive Medicine, I 7(4),
285-294.

206

Galambos, C. M. (1998). Preserving end of life autonomy: the PSDA and the UHCDA.
Health Social Work, 23(4), 275-281.

Gillick, M. R. (2001). Choosing appropriate medical care for the elderly. Journal of the
American Medical Directors Association, 2, 305-309.

Grimaldo, D. A., Wiener-Kronish, J. P., Jurson, T., Shaughnessy, T. E., Curtis, J. R., &
Liu, L. L. (2001). A randomized, controlled trial of advance care planning
discussions during preoperative evaluations. Anesthesiology, 95(1), 43-50.

Grounds, V. (2004). A personal perspective. In C. B. Mitchell, R. D. Orr & S. A.
Salladay (Eds.), Aging, Death, and the Quest for Immortality. Grand Rapids, MI:
William B. Eerdmans Publishing Company.

Hamel, C. F ., Guse, L. W., Hawranik, P. G., & Bond, Jr., J .B. (2002). Advance directives
and community-dwelling older adults. Western Journal of Nursing Research,
24(2), 143-158.

Hammes, B. J ., & Briggs, L. A. (2002a). Respecting Choices® Organization and
Community Advance Care Planning Course. LaCrosse, WI: Gundersen Lutheran
Medical Foundation.

Hammes, B. J ., & Briggs, L. A. (2002b). Respecting Choices® Advance Care Planning
Facilitator's Manual. LaCrosse, WI: Gundersen Lutheran Medical Foundation.

Hammes, B. J ., & Briggs, L. A. (2004). Respecting Choices® Advance Care Planning
Program Quality Improvement T oolla't. LaCrosse, WI: Gundersen Lutheran
Medical Foundation.

Hammes, B. J ., & Rooney, B. L. (1998). Death and End-of-Life Planning in One
Midwestern Commrurity. Archives of Internal Medicine, 158, 383-390.

Hammond, K. R., Stewart, T. R, Brehmer, B., & Steinmann, D. O. (2002). Social
judgment theory. In M. Kaplan & S. Schwartz (Eds.), Human judgment and
decision processes (pp. 271-312). New York: Academic Press.

Hawkins, N. A., Ditto, P. H., Danks, J. H., 8: Smucker, W. D. (2005). Micromanaging
death: Process preferences, values, and goals in end-of-life medical decision
making. . Gerontologist., 45(1), 107-117.

Heffner, J. E., & Barbieri, C. (2001). Effects of advance care education in cardiovascular

rehabilitation programs: a prospective randomized study. Journal of
Cardiopulmonary Rehabilitation, 21, 387-391.

207

Heffirer, J. E., Fahy, B., Hilling, L., & Barbieri, C. (1997). Outcomes of advance
directive education of pulmonary rehabilitation patients. American Journal of
Respiratory and Critical Care Medicine, 155(3), 1055-1059.

Heyland, D. K., Tranmer, J ., & F eldman-Stewart, D. (2000). End-of-Life Decision
making in the Seriously Ill Hospitalized Patient: an Organizing Framework and
Results of a Preliminary Study. Journal of Palliative Care, I 6(Supplement), S31-
S39.

Hines, S. C., Glover, J. J., Holley, J. L., Babrow, A. S., Badzek, L. A., & Moss, A. H.
(1999). Dialysis Patients' Preferences for Family-Based Advance Care Planning.
Annals of Internal Medicine, 130(10), 825-828.

Hofrnann, J. C., Wenger, N. S., Davis, R. B., Teno, J ., Connors, A. F ., Jr., Desbiens, N.,
et al. (1997). Patient preferences for communication with physicians about end-
of-life decisions. SUPPORT Investigators. Study to Understand Pro grroses and
Preference for Outcomes and Risks of Treatment. Ann Intern Med, 127(1), 1-12.

Hopp, F. P. (2000). Preferences for Surrogate Decision Makers, Informal
Communication, and Advance Directives Among Community-Dwelling Elders:
Results from a National Study. The Gerontologist, 40(4), 449-457.

Howell, D. C. (2002). Statistical Methods for Psychology (5th ed.). Pacific Grove, CA:
Wadsworth Group, Duxbury.

Hurley, A. C., Volicer, L., Rempusheski, V. F ., & Fry, S. T. (1995). Reaching consensus:
the process of recommending treatment decisions for Alzheimer's patients.
Advances in Nursing Science, 18(2), 33-43.

J anz, N. K., & Becker, M. H. (1984). The Health Belief Model: a decade later. Health
Educ Q, 11(1), 1-47.

Kaplan, K. 0., & Peres, J. R. f. L. A. (2002). Means to a better end: A report on dying in
America today: Last Acts.

Karel, M. J ., Powell, J ., & Cantor, M. D. (2004). Using a Values Discussion Guide to
Facilitate Communication in Advance Care Planning. Patient Education and
Counseling, 55, 22-31.

Kass-Bartelrnes, B. L., Hughes, R., & Rutherford, M. K. (2003). Advance care planning:
preferences for care at the end of life. (No. AHIRQ Pub No. 03-0018): Agency for
Healthcare Research and Quality.

Kolarik, R. C., Arnold, R. M., Fischer, G. S., & Tulsky, J. A. (2002). Objectives for
advance care planning. Journal of Palliative Medicine, 5(5), 697-704.

208

Landry, F. J ., Kroenke, K., Lucas, C., & Reeder, J. (1997). Increasing the use of advance
directives in medical outpatients. Journal of General Internal Medicine, 12, 412-
41 5.

Lee, S. J ., Back, A. L., Block, S. D., & Stewart, S. K. (2002). Enhancing physician-
patient communication. Journal of the American Society of Hematology, 464-483.

Levinson, W., Kao, A., Kuby, A., & Thisted, R. A. (2005). Not all patients want to
participate in decision-making: a national study of public preferences. Journal of
General Internal Medicine, 20, 531-535.

Lorenz, K., Lynn, J ., Morton, S. C., Dy, S., Mularski, R., & Shugarman, L. (2004). End-
of-life care and outcomes. (Sunrrnary, evidence report/Technology assessment
No. 110 No. AHRQ Publication No. 06-#004-1). Rockville, MD: Agency for
Healthcare Research and Quality

Lunney, J. R., Foley, K. M., Smith, T. J ., & Gelband, H. (Eds). (2003). Describing death
in America: what we need to know. Washington, D.C.: The National Academies
Press.

Lunney, J. R., Lynn, J ., Foley, D. J ., Lipson, S., & Guralnik, J. M. (2003). Patterns of
functional decline at the end of life. Journal of the American Medical Association,
289(18), 2387-2392.

Makoul, G., & Clayman, M. L. (2006). An integrative model of shared decision making
in medical encounters. Patient Education and Counseling, 60(3), 301-312.

Mazur, D. J. (2003). The New Medical Conversation: Media, Patients, Doctors and the
Ethics of Scientific Communication. Lanharn, MD: Rowman & Littlefield
Publishers, Inc.

McAdam, J. L., Stotts, N. A., Padilla, G., & Puntillo, K. (2005). Attitudes of critically ill
Filipino patients and their families toward advance directives. American Journal
of Critical Care, 14(1), 17-25.

McKinley, E. D., Garrett, J. M., Evans, A. T ., & Danis, M. (1996). Differences in End-of-
Life Decision Making Among Black and White Ambulatory Cancer Patients.
Journal of General Internal Medicine, 11, 651-656.

Meier, D. E., Fuss, B. R., O'Rourke, D., Baskin, S. A., Lewis, M., & Morrison, R. S.
(1996). Marked improvement in recognition and completion of health care
proxies: a randomized controlled trial of counseling by hospital patient
representatives. Archives of Internal Medicine, 1 5 6, 1227-1232.

Michigan Commission on End of Life Care. (2001). Report to the Governor. Lansing,
MI: Michigan Department of Community Health.

209

Molloy, D. W., Guyatt, G. H., Russo, R., Goeree, R., O'Brien, B. J ., Bedard, M., et al.
(2000). Systematic Implementation of an Advance Directive Program in Nursing
Homes: A Randomized Controlled Trial. Journal of the American Medical
Association, 283(11), 1437-1444.

Moore, C. D., & Sherman, S. R. (1999). Factors that Influence Elders' Decisions to
Formulate Advance Directives. Journal of Gerontological Social Work, 13(1/2),
21-39.

Moore, K. A., Danks, J. H., Ditto, P. H., Druley, J. A., Townsend, A., & Smucker, W. D.
(1994). Elderly outpatients' understanding of a physician-initiated advance
directive discussion. Archives of Family Medicine, 3(12), 1057-1063.

Morrison, R. S., & Meier, D. E. (2004). High rates of advance care planning in New York
City’s elderly population. Archives in Internal Medicine, 164, 2421-2426.

Morrison, R. S., Zayas, L. H., Mulvihill, M., Baskin, S. A., & Meier, D. E. (1998).
Barriers to Completion of Health Care Proxies: An Examination of Ethnic
Differences. Archives of Internal Medicine, 158, 2493-2497.

Moss, A. H., Oppenheimer, E. A., Casey, P., Cazzolli, P. A., Roos, R. P., Stocking, C. B.,
et al. (1996). Patients with Amyotrophic Lateral Sclerosis Receiving Long—Term
Mechanical Ventilation: Advance Care Planning and Outcomes. Chest, 110(1),
249-255.

Munro, B. H. (2001). Statistical methods for health care research. Philadelphia:
Lippincott.

Newell, A., & Simon, H. (1972). Human problem solving. Englewood Cliffs, NJ:
Prentice-Hall.

O'Brien, L. A., Grisso, J. A., Maislin, G., LaPann, K., Krotki, K. P., Greco, P. J., Siegert,
E. A., et al. (1995). Nursing Home Residents' Preferences for Life-Sustaining
Treatments. Journal of the American Medical Association, 274(22), 1775- 1 779.

O'Toole, M. T. (Ed.). (1997). Miller-Keane Encyclopedia and Dictionary of Medicine,
Nursing, and Allied Health (6th ed.). Philadelphia: W. B. Saunders Company.

Office of the Inspector General. (1993). Patient Advance Directives: Early
Implementation Experience. In H. a. H. Services (Ed).

Patrick, D. L., Beresford, S. A. A., Ehreth, J ., Diehr, P., Picciano, J ., Durham, M., et al.

(1995). Interpreting excess mortality in a prevention trial for older adults.
International Journal of Epidemiology, 24(3), S27-S33.

210

Patterson, C., Molloy, D. W., Guyatt, G. H., Bedard, M., North, J ., & Hassard, J. (1997).
Systematic implementation of an advance health care directive in the community.
Canadian Journal of Nursing Administration, 10(2), 96—108.

Pearlman, R. A., Cole, W. G., Patrick, D. L., Starks, H. E., & Cain, K. C. (1995).
Advance care planning: eliciting patient preferences for life-sustaining treatment.
Patient Education and Counseling, 26(1/3), 353-361.

Pearhnan, R. A., Starks, H., Cain, K. C., & Cole, W. G. (2005). Improvements in advance
care planning in the Veterans Affairs System: results of a multifaceted
intervention. Arch Intern Med, 165(6), 667-674.

Pender, N. J. (1996). Health promotion in nursing practice (3rd ed.). Stamford, CT:
Appleton & Lange.

Pierce, P. F., & Hicks, F. D. (2001). Patient decision-making behavior: an emerging
paradigm for nursing science. Nursing Research, 50(5), 267-274.

Post, L. F., Blustein, J ., & Dubler, N. N. (1999). The Doctor-Proxy Relationship: An
Untapped Resource. Journal of Law, Medicine & Ethics, 27(1), 5-12.

Prochaska, J. O., & DiClemente, C. C. (1983). Stages and processes of self-change of
smoking: toward an integrative model of change. Journal of Consulting Clinical
Psychology, 51(3), 390-395.

Rabow, M. W., Petersen, J ., Schanche, K., Dibble, S. L., & McPhee, S. J. (2003). The
Comprehensive Care Team: A Description of a Controlled Trial of Care at the
Beginning of the End of Life. Journal of Palliative Medicine, 6(3), 489-499.

Rallying Points. (2002). Rallying Points Coalition Assistance & Information. Retrieved
December 3, 2003, fi'om http://wwwrallyingpointsorg

Ratner, E., Norlander, L., & McSteen, K. (2001). Death at Home Following a Targeted
Advance-Care Planning Process at Home: The Kitchen Table Discussion. Journal
of the American Geriatrics Society, 49, 778-781.

Reilly, B. M., Wagner, M., Ross, J ., Magnussen, C. R., Papa, L., & Ash, J. (1995).
Promoting completion of health care proxies following hospitalization. A
randomized controlled trial in a community hospital. Archives of Internal
Medicine, 155(20), 2202-2206.

Rein, A. J ., Harshman, D. L., Frick, T., Phillips, J. M., Lewis, S., & Nolan, M. T. (1996).

Advance directive decision making among medical inpatients. Journal of
Professional Nursing, 12(1), 39-46.

211

Richter, K. P., Langel, S., Fawcett, S. B., Paine-Andrews, A., Beihler, L., & Manning, R.
(1995). Promoting the use of advance directives: an empirical study. Archives of
Family Medicine, 4, 609-615.

Robinson, K., Sutton, S., von Gunten, C. F., Fenis, F. D., Molodyko, N., & Martinez, J.
(2004). Assessment of the education for physicians on end-of-life care (EPEC)
project. Journal of Palliative Medicine, 7(5), 537-645.

Rosenstock, I., Strecher, V. J ., & Becker, M. H. (1988). Social Learning Theory and the
Health Belief Model. Health Education Quarterly, 15(2), 175-183.

Rothert, M. L., Holmes-Rovner, M., Rovner, D., Kroll, J ., Breer, L., Talarczyk, G., et al.
(1997). An educational intervention as decision support for menopausal women.
Research in Nursing and Health, 20, 377-387.

Rubin, S. M., Strull, W. M., Fialkow, M. F., Weiss, S. J ., & Lo, B. (1994). Increasing the
Completion of the Durable Power of Attorney for Health Care: A Randomized,
Controlled Trial. Journal of the American Medical Association, 271(3), 209-212.

Sachs, G. A., Stocking, C. B., & Miles, S. H. (1992). Empowerment of the older patient?
A randomized, controlled trial to increase discussion and use of advance
directives. Journal of the American Geriatrics Society, 40(3), 269-273.

Say, R., Murtagh, M., & Thomson, R. (2006). Patients' preference for involvement in
medical decision making: A narrative review. Patient Education and Counseling,
60(2), 102-114.

Schneiderman, L. J ., Pearlman, R. A., Kaplan, R. M., Anderson, J. P., & Rosenberg, E.
M. (1992). Relationship of general advance directive instructions to specific life-
sustairring treatment preferences in patients with serious illness. Archives of
Internal Medicine, 152(10), 2114-2122.

Schwartz, C. E., Wheeler, H. B., Hammes, B. J ., Basque, N., Edmunds, J ., & Reed, G.
(2002). Early intervention in planning end of life care with ambulatory geriatric
patients: result of a pilot trial. Archives of Internal Medicine, 1 62(14), 1611-1618.

Shidler, S. (1998). A systemic perspective of life-prolonging treatment decision-making.
Qualitative Health Research, 8(2), 254-269.

Shoemaker, P. J. (1982). The expected utility model: its variants, purposes, evidence and
limitations. Journal of Economic Literature, 20, 529-563.

Siegert, E. A., Clipp, E. C., Mulhausen, P., & Kochersberger, G. (1996). Impact of

advance directive videotape on patient comprehension and treatment preferences.
Archives of Family Medicine, 5, 207-212.

212

Silveira, M. J ., DiPiero, A., Gerrity, M. S., & Feudtner, C. (2000). Patients' Knowledge
of Options at the End of Life: Ignorance in the Face of Death. Journal of the
American Medical Association, 284(19), 2483-2488.

Singer, P. A., Thiel, E. C., Naylor, C. D., Richardson, R. M. A., Llewellyn-Thomas, H.,
& Goldstein, M., et al. (1995). Life-sustaining treatment preferences of
hemodialysis patients: implications for advance directives. Journal of the
American Society of Nephrology, 6(5), 1410-1417.

Singer, P. A., Thiel, E. C., Salit, I., Flanagan, W., & Naylor, C. D. (1997). The HIV-
Specific Advance Directive. Journal of General Internal Medicine, 12, 729-735.

Smucker, W. D., Ditto, P. H., Moore, K. A., Druley, J. A., Danks, J. H., & Townsend, A.
(1993). Elderly Outpatients Respond F avorably to a Physician-hritiated Advance
Directive Discussion. Journal of the American Board of Family Practice, 6, 473-
482.

Stewart, A. L., Teno, J ., Patrick, D. L., & Lynn, J. (1999). The concept of quality of life
of dying persons in the context of health care. Journal of Pain and Symptom
Management, 17(2), 93-108.

Stommel, M., & Wills, C. E. (2004). Clinical Research Concepts and Principles for
Advanced Practice Nurses. Philadelphia: Lippincott Williams & Wilkins.

Suhnasy, D. B, Song, K. Y., Marx, E. S., & Mitchell, J. M. (1996). Strategies to promote
the use of advance directives in a residency outpatient practice. Journal of
General Internal Medicine, 11, 657-663.

Taylor, J. S., Heyland, D. K., & Taylor, S. J. (1999). How advance directives affect
hospital resource use. Canadian Family Physicians, 45(2408-2413).

Teno, J ., Lynn, J ., Wenger, N., Phillips, R. S., Murphy, D. P., & Connors, A. F. (1997).
Advance directives for seriously ill hospitalized patients: effectiveness with the
Patient Self-Determination Act and the SUPPORT intervention. Journal of the
American Geriatrics Society, 45(4), 500-507.

Teno, J. M., Clarridge, B. R., Casey, V., Welch, L. C., Wetle, T., Shield, R., et al. (2004).
Family Perspectives on End-of-Life Care at the Last Place of Care. Journal of the
American Medical Association, 291(1), 88-93.

Teno, J. M., Licks, S., Lynn, J ., Wenger, N., Connors, A. F ., & Phillips, R. S. (1997). Do

advance directives provide instructions that direct care? Journal of the American
Geriatrics Society, 45, 508-512.

213

Thompson, C., & Dowding, D. (2002). Decision making and judgement in nursing -- an
introduction. In C. Thompson & D. Dowding (Eds.), Clinical Decision Making
and Judgement in Nursing (pp. 2-20). New York: Churchill Livingstone.

Tierney, W. M., Dexter, P. R., Gramelspacher, G. P., Perkins, A. J ., Zhou, X. H., &
Wolinsky, F. D. (2001). The Effect of Discussions About Advance Directives on
Patients' Satisfaction with Primary Care. Journal of General Internal Medicine,
16, 32-40.

Tilden, V. P., Tolle, S. W., Nelson, C. A., & Fields, J. (2001). Family decision making to
withdraw life-sustaining treatments fiom hospitalized patients. Nursing Research,
50(2), 105-115.

VandeCreek, L., & F rankowski, D. (1996). Barriers that predict resistance to completing
a living will. Death Studies, 20, 73-82.

Vander Laan, K. J. (2005a). Advance Care Planning: Partnering with Faith Communities
to Promote Advance Care Planning, Calvin College Christian Perspectives on
Science. Grand Rapids, MI.

Vander Laan, K J. (2005b). Approaches to Advance Care Planning, 2nd Annual
Psychiatry Update Ingham Regional Medical Society. East Lansing, MI.

Vander Laan, K. J ., Mullan, P. B., & Wyatt, G. K. (2005). Evaluation of Advance Care
Planning in Faith Based Communities. Paper presented at the American
Association of Cancer Educators, Cincinnatti, OH.

Vander Laan, K. J ., Mullan, P. B., & Wyatt, G. K. (2005). Partnering with Faith
Communities to Promote Advance Care Planning, Psych 992 Course Conference.
East Lansing, MI.

Vander Laan, K. J ., Mullan, P. B., Wyatt, G. K., McPhail, J. D., & Thomason, C. (2007).
Partnering with Communities to Promote Advance Care Planning, Community
Research Day. East Lansing, MI.

Vander Laan, K. J ., Mullan, P. B., Wyatt, G. K., McPhail, J. D., & Thomason, C. L.
(2006). Partnering with Communities to Promote Advance Care Planning. Paper
presented at the Annual Assembly of the American Association of Hospice &
Palliative Medicine and Hospice & Palliative Nurses Association, Nashville, TN.

Vander Laan, K. J ., & Wyatt, G. K. (2005). Partnering with Faith Communities to

Promote Advance Care Planning. Paper presented at the 29th Annual Research
Conference Midwest Nursing Research Society, Cincinnatti, OH.

214

Waters, C. M. (2000). End-of-Life Care Directives Among African Americans: Lessons
Leamed--A Need for Community-Centered Discussion and Education. Journal of
Community Health Nursing, 1 7(1), 25-37.

Weiner, J. S., & Cole, S. A. (2004). Three Principles to Improve Clinician
Communication for Advance Care Planning: Overcoming Emotional, Cognitive,
and Skill Barriers. Journal of Palliative Medicine, 7(6), 817-829.

Weissman, D. E., Mullan, P. B., Ambuel, B., von Gunten, C. R, Block, S. D., & Milch,
R. (2005). National residency end-of-life project. Journal of Palliative Medicine,
8(3), 646-664.

Wenger, N. S., Kanouse, D. E., Collins, R. L., Liu, H., Schuster, M. A., Gifford, A. L., et
al. (2001). End-of-Life Discussions and Preferences Among Persons with HIV.
Journal of the American Medical Association, 285(22), 2880-2887.

Wenger, N. S., & Rosenfeld, K. (2001). Quality Indicators for End-of-Life Care in
Vulnerable Elders. Ann Intern Med, 135(8_Part_2), 677-685.

Wenger, N. S., Shekelle, P. G., & the, A. I. (2001). Assessing Care of Vulnerable Elders:
ACOVE Project Overview. Ann Intern Med, 135(8_Part_2), 642-646.

Wenger, N. S., Solomon, D. H., Roth, C. P., MacLean, C. H., Saliba, D., Kamberg, C. J .,
et al. (2003). The Quality of Medical Care Provided to Vulnerable Community-
Dwelling Older Patients. Ann Intern Med, 139(9), 740-747.

Wills, C. E., & Holmes-Rovner, M. (2006). Integrating decision-making and mental
health intervention research: research directions. Clinical Psychology, 1 3, 9-25.

Wyatt, G. K., McPhail, J. D., Mullan, P. B., Thomason, C., & Vander Laan, K. J. (2005).
MlllA CPC Research Team Report 2004-2005. Lansing, MI: Capital Area Health
Alliance Advance Care Planning Coalition.

Yamada, R., Galecki, A. T., Goold, S. D., & Hogikyan, R. V. (1999). A multimedia
intervention on cardiopulmonary resuscitation and advance directives. Journal of
General Internal Medicine, 14, 559-563.

215

   

IIIIIIIIIIIIIIIII