‘ a 52.35.11. . ahuuﬁrwax .,.. , .
gnu; , , . ma rsmwwﬁgé...”

2*! ,
zotw

I bill.
. . 5...

7...: .

. Y .
cvhuiL. ,
. u! , .
. “

.VU. ’0‘ I :f. Y .i O.“

32.3? , i. . . ,

, , 1 . .
. . if 3 3.
. . é , . .1; 3 .
b...» 9,312
52.5.1»:
§

Hal-Hist
.c

f ,
:13}

.2.
A a.
,:.,.:a.l.;
ﬁﬁuuﬁir
. .1 ,

45‘
s:

x z .2. . . .
2.3 , , ,
. . . . , V 1...:

A . 1.3%.:

V (.393...

x‘
‘ :2 3i).
3 1 5.; n u

.2. 1...
r 1!, A .1.
an. 1..) .11..

3.3.1..
aura}!
A?:F..J

.1 .35.?

1, 4.12:.

y

A \x .7. J
:? 7?!

 

 

 

 

 

3.-.!
xii-v0.52...
73.. 5:11.!
. {(1')}: ‘
.I.f.x).!.(\‘7 1....
Vititv‘iz; vr .

 

‘ . 1.: ... or .
. a... Jihaln
».< :3...
24...}... r

 

                                                                                                                           

3 1293 01051 53 89

This is to certify that the

dissertation entitled

IMPACT OF A DECISION SUPPORT INTERVENTION
FOR MIDLIFE WOMEN ON HEALTH CARE SELF-EFFICACY

presented by
JILL CORRINE KROLL

has been accepted towards fulﬁllment
of the requirements for

PhD Psychology

degree in

 

 

 

'Major professor

 

MS U is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE DI RETURN BOX to remove thi- checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

SC!" 2 5 Wj 1
i ‘3 1

s ‘ .2 .‘ ﬂ "\
1

 

 

 

 

   
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU le An Afﬁrmative Action/Equal Opportunity lnetltuion
Wanna-9.1

IMPACT OF A
DECISION SUPPORT INTERVENTION FOR MIDLIFE WOMEN
ON HEALTH CARE SELF-EFFICACY

By

Jill Corrine Kroll

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Psychology

1993

 

ABSTRACT

IMPACT OF A
DECISION SUPPORT INTERVENTION FOR MIDLIFE WOMEN
ON HEALTH CARE SELF-EFFICACY

By

Jill Corrine Kroll

This research examined the impact of a decision support intervention for
midlife women on self-efﬁcacy and behavior related to active participation in
health care. A model of the relationship between self-efficacy, perceived barriers,
outcome expectations, subjective norms, behavioral intention and participation
in health care decisions was developed and tested. Participants were randomly
assigned to one of three educational interventions addressing menopause and
hormone replacement therapy: a written brochure, a lecture/ discussion-format
intervention or a decision support intervention. Participants completed self-
administered questionnaires prior to the intervention, immediately following the
intervention and two months following the intervention. Self-efﬁcacy increased
significantly among participants in all three conditions following intervention.
There were no significant differences in self-efficacy or self-reported active
patient participation behavior between participants in the decision support
intervention compared to the written brochure. Self-efficacy increased
signiﬁcantly more among participants in the lecture/ discussion-format
intervention than among those in the written brochure condition. Signiﬁcant
predictors of self-reported participation in health care were intention to

participate, perceived barriers to participation and outcome expectations related

to active participation behavior. It was concluded that external factors such as
barriers were more important to active participation behavior than were internal,

individual factors such as perceived skills related to participation.

Copyright by
JILL CORRINE KROLL
1993

Dedicated to my patient family;
John, David and Thomas

who gave up so much

vi

ACKNOWLEDGMENTS

A special thank you goes to the women who participated in the Decision
Making in Menopause Study. They willingly answered my questions and
provided helpful comments and insights. It is my hope that dissemination of
these findings will help their concerns be heard.

I want to thank my advisor, Dr. William Davidson who provided me with
support, advice and helpful feedback during this long period of dissertation
writing and during my years of graduate study.

' I reserve special gratitude for Dr. Marilyn Rothert, principal investigator for
the Decision Making in Menopause Study. I want to thank her for serving on
my dissertation committee, for letting me learn to do research under her
guidance and direction and for providing me with an outstanding role model.

Thank you to Dr. Neal Schmitt and Dr. Thomas Reischl for serving on my
committee, encouraging me in my work, and for guiding and supporting my
learning process.

Thank you to the members of the Decision Making Study research team
who taught me the challenge and fun of good research and collaborative effort.
Thanks to Ellie, who helped me when I was down. Thanks to neighbors,
friends, faculty and fellow students who provided me with help, support and
diversion. Thank you to my parents, Kurt and Laurie, who believed in me even

when I didn’t believe in myself and to my family, John, David and Thomas.

vii

TABLE OF CONTENTS
Page
LIST OF TABLES .............................................................................. xiii
LIST OF FIGURES ............................................................................... xv
CHAPTER
1 INTRODUCTION ................................................................... 1
Statement of the Problem ....................................................... 1
Patient Participation in Decision Making ............................ 2
Psychological Morbidity .................................................. 3
Patient Preference to Participate ..................................... 5
Theoretical Models .................................................................. 6
Theory of Flamed Behavior .................................................. 6
Perceived Behavioral Control .......................................... 8
Similarity Between Social Cognitive Theory and
Theory of Flamed Behavior .................................... 12
Self-Efficacy as a Mediator of Low Desire for Control
of Health Care ............................................................ 15
Can Self-Efﬁcacy be Changed? ..................................... 15
What Factors Impact Self-Efﬁcacy? .................................... 16
Performance accomplishments...........l ................... 16
Vicarious Experience ................................................ 16
Verbal Persuasion ..................................................... 17
Physiological States ................................................... 17
Patient/ Professional Interaction ......................................... 18

Professional as
Decision Maker—Advocacy Models ...................... 18

Joint Decision Making Models ...................................... 19

viii

CHAPTER Page
1 Consumer as Decision Maker—Decision

Support Models ................................................... 20
Critique of Decision Support Models .................... 20
Conclusion ....................................................................... 23
Community Psychology Concepts ..................................... 23
Empowerment ................................................................. 23
Individual Versus System-Level Intervention ............ 24

Individual-Level Intervention for

System-Level Change ......................................... 25
Current Programs to Empower

Health Care Consumers ................................................. 28
The Need for a Program Addressing Menopause ........... 30
Conclusion ............................................................................. 30
The Current Research ........................................................... 32
Hypotheses ............................................................................. 40
Hypothesis 1 .................................................................... 40
Statement of Hypothesis .......................................... 40
Summary of Related Literature ............................... 40
Link Between Intervention and Self-efficacy. 40
Link Between Self-efficacy and Intention ........ 40
Link Between Intention and Behavior ............. 41

Link Between Active Participation
Behavior and Satisfaction ................................... 41
Relation to Model ...................................................... 41
Intervention and Self-efﬁcacy ........................... 41

Self-efficacy and Behavioral Intention ............. 42

ix

CHAPTER Page
1 Behavioral Intention and Behavior ................... 42

Participation and Satisfaction with Decision. 42

Hypothesis 2 .................................................................... 42
Statement of Hypothesis .......................................... 42
Summary of Related Literature ............................... 42

Outcome Expectations ........................................ 43

Subjective Norm .................................................. 44

Relationship Between 3 Variables

and Intention ........................................................ 44
Relation to Model ...................................................... 44

Hypothesis 3 .................................................................... 45
Statement of Hypothesis .......................................... 45
Summary of Related Literature ............................... 45
Relation to Model ...................................................... 46

Hypothesis 4 .......................................................................... 46
Statement of Hypothesis .......................................... 46

Summary of Related Literature ............................... 46

Relation to Model ...................................................... 47

Hypothesis 5 .......................................................................... 47
Statement of Hypothesis .......................................... 47

Summary of Related Literature ............................... 47

Relation to Model ...................................................... 49

2 Method ............................................................................... 52
Setting ......................................................................... 52

Research Participants ............................................... 52

CHAPTER Page
2 Recruitment .......................................................... 52
Random Assignment .......................................... 53
Attrition ................................................................ 53
Characteristics of the Sample ............................ 56
Research Design ........................................................ 59
Power Analysis .......................................................... 59
Procedure ............................................................................... 61
Instructors ....................................................................... 61
Intervention Pilot ............................................................ 62
The Intervention .............................................................. 63
Written Brochure Intervention ................................ 63
Lecture/ Discussion Intervention ............................ 63
Decision Support Intervention ................................ 64
Attrition Prevention ........................................................ 66
Data Collection ................................................................ 67
Measures ............................................................................... 68
Instrument Pilot ............................................................... 68
Outcome Measurement ............ . ...................................... 69
Health Care Self—Efﬁcacy ......................................... 69
Barriers to Participation Measure ........................... 73
Outcome Expectations Scale .................................... 77
Subjective Norms Measure ...................................... 79
Behavioral Intentions Measure ............................... 83
Behavior Self-Report Measure ................................ 86

Satisfaction with Decision ........................................ 89

xi

CHAPTER Page
2 Scale Intercorrelations .............................................. 89
3 Results ............................................................................... 92
Hypotheses ............................................................................. 92
Hypothesis 1 .................................................................... 92
Hypotheses 2, 3 & 5 ........................................................ 98
Hypothesis 4 .................................................................. 104
Additional Analyses ............................................... 104
Summary ........................................................................ 108
4 Discussion ............................................................................ 109
Research Questions and Hypotheses ......................... 109
Summary .................................................................. 121
Limitations ..................................................................... 123
Issues for Future Research ........................................... 124
Conclusions and Recommendations .......................... 125
LIST OF REFERENCES ........................................................................ 133
APPENDICES
APPENDD( A
Recruitment Advertisements ...................................... 145
APPENDD( B
Power Analysis Calculations ....................................... 146
APPENDD( C
Intervention Outline ..................................................... 148
APPENDD( D
Measures ................. ' ........................................................ 168
APPENDD( E

Scale Means and Standard Deviations ....................... 183

xii

Page
APPENDIX F
Calculation of Correction for Unreliability ............... 184
APPENDD( G
Computation for Regression ....................................... 185
APPENDIX H

Significance Test for Correlated r’s ............................ 187

LIST OF TABLES
Table Page(s)
1 Common Constructs in Self-Efficacy Theory and
Theory of Plamed Behavior .......................................... 14
2 Hypotheses ....................................................................... 50—51
3 Attrition by Experimental Group ....................................... 55
4 Demographic Characteristics of Sample ............................ 57
5 Health History and Access to Medical Care ..................... 58
6 Measures Administration Schedule ................................... 60
7 Instructor Schedule ............................................................... 62
8 Barriers to Participation Scale Corrected Item-Total
Correlation ....................................................................... 74
9 Number of barriers Mentioned
on Open-Ended Questions ............................................ 76
10 Outcome Expectations Scale Corrected Item—Total
Correlation ....................................................................... 79
11 Subjective Norms Scale Corrected Item-Total
Correlation ....................................................................... 81
12 Subjective Norms Sources of Inﬂuence on Decision
Open-Ended Questions .................................................. 82
13 Behavioral Intention Scale Corrected Item-Total
Correlation ....................................................................... 85
14 Behavior Self-Report Scale Corrected Item-Total
Correlation ....................................................................... 88
15 Correlation Matrix ................................................................ 91
16 Contrast Coding of Nominal Experimental
Group Variable ................................................................ 94

17 Regression Coefficients and Significance for Regression

of Experimental Group on Self-Efficacy at Time 2
(controlling for Self-Efﬁcacy at Time 1) .................

18 Regression Equations for Self-efﬁcacy

Post-Intervention (Time 2) .......................................

...... 95

...... 97

xiv

Table Page
19 Group Means for Intention, Behavior and
Satisfaction ....................................................................... 97
20 Model as Proposed: t-values ............................................. 100
21 Standardized Residuals for Revised Model .................... 103
22 Revised Model: t-values ..................................................... 103

23 Regression Coefﬁcients and Significance for
Regression of Self-efﬁcacy Time 3 on
Self-efﬁcacy Time 2 and Behavior .............................. 106

XV

LIST OF FIGURES
Figure Page

1 Theory of Reasoned Action (Ajzen 8: Madden, 1986) ....... 7
2 Theory of Plamed Behavior (Ajzen 8: Madden, 1986) ...... 9
3 Self-Efficacy Theory Model (Bandura, 1977) ..................... 11
4 Health Care Self—Efﬁcacy Model ......................................... 39
5 Frequency Distribution

Self-Efficacy Time 1 (Pm-Intervention) ........................ 70
6 Frequency Distribution

Self-Efficacy Time 2 (Post-Intervention) ...................... 71

7 Frequency Distribution
Self-Efficacy Time 3 (2 Months Post-Intervention)... 72

8 Frequency Distribution
Barriers to Participation ................................................. 75
9 Frequency Distribution
Outcome Expectations Scale .......................................... 78
10 Frequency Distribution
Subjective Norm Scale .................................................... 81
11 Frequency Distribution
Behavioral Intention Scale ............................................. 84
12 Frequency Distribution
Behavior Self-Report Scale ............................................. 87
13 Frequency Distribution
Satisfaction With Decision Time 3 ................................ 9O
14 Regression of T1 Self-Efficacy on T2 Self-Efﬁcacy by
Experimental Group ....................................................... 96
15 Path Coefficients for Model as Proposed .......................... 99
16 Path Coefficients for Revised Model ................................ 101

17 Behavior Plotted Against Self-Efficacy at
Time 2 and Time 3 ......................................................... 107

CHAPTER 1
INTRODUCTION
Statement of the Problem

Women face a dilemma as they reach midlife. Hormone Replacement
Therapy [HRT] has been found to signiﬁcantly reduce white women’s risk of
heart disease and osteoporosis and to decrease deaths from all causes
(Cummings, Black 8: Rubin, 1989). It signiﬁcantly reduces hot ﬂashes and vaginal
dryness among most women (Judd, 1987). At the same time it has been found to
signiﬁcantly increase risk of endometrial cancer (Persson, Adami, Lindberg,
Johansson, & Manell, 1989) and has an unknown affect on risk of breast cancer
(Dupont, Page, Rogers 8: Parl, 1989; Bergkvist, Adami, Persson, Hoover 8:
Schairer, 1989). It may also be accompanied by side effects such as vaginal
bleeding, swelling and breast tenderness (Luciano, Turksoy, Carleo 8: Hendrix,
1988). Because the decision whether or not to take hormone replacement therapy
depends on individual values as well as personal risk factors, it has been argued
that the patient should be included in the decision process.

Although involvement of patients in the decision process has been advocated
throughout public health and patient education literature for more than two
decades (Steele, Blackwell, Gutmann 8: Jackson, 1987); studies have found that
many patients prefer to leave medical decisions to professionals (Ende, 1989;
Beisecker, 1988; Strull, Lo 8: Charles, 1984) though few studies have examined
the reasons for this patient preference. The only consistent predictors of patient
desire to participate in medical decisions have been age and education (Cassileth,
Zupkis, Sutton-Smith 8: March, 1980; Blanchard, Labrecque, Ruckdeschel 8:
Blanchard, 1988; Beisecker, 1988). These demographic factors offer little insight

into the reasons behind patient preference. Preference to leave medical decisions

1

2
exclusively to professionals is puzzling when one examines the beneﬁts

associated with active participation and the hazards of lack of involvement.
'nP 'iatin' D ii Makin

There are significant advantages to active patient participation in decisions.
Active patients may be better able to participate in prevention, management and
cure of disease resulting in reduced health care costs and decreased physical and
psychological morbidity. Many illnesses today are directly related to behavior
(Glanz, Lewis, 8: Rimer, 1990) and others require active patient involvement and
cooperation for effective management (Giloth, 1990; McKinlay, McKinlay 8:
Beaglehole, 1989). Informed health care consumers may be able to avoid or
reduce illness through modifying health related behavior.

A second advantage is that informed consumers may reduce the costs
associated with health care. Consumers can beneﬁt by practicing informed use of
limited medical resources (Figge, 1990). Knowing effective self-treatment
methods, how to apply them and when to seek expert advice can. help to
decrease costs.

A third beneﬁt of patient participation in decisions is incorporation of the
patient’s individual needs and values in the treatment choice. Many medical
conditions have several treatment options which are equally effective (Wilson,
Hart 8: Dawes, 1988). In these cases, patient participation in decision making can
ensure that the treatment chosen best meets the patient’s personal needs.

Furthermore, medical technology has provided new choices which demand
decisions based on ethics and values rather than simply medical judgment
(Saxton, 1987). For example, prenatal diagnosis makes it possible to diagnose
and abort defective fetuses (Saxton, 1987). Such decision choices reﬂect ethical
rather than clinical decisions. Society will suffer if decisions resulting from this

technology do not reﬂect the values of health care consumers.

3 .
Further justiﬁcation for involving patients in decision making derives from

research on self-efﬁcacy. Studies of clinical outcome and self-efﬁcacy have
shown that self-efﬁcacy is associated with improved functional status. Allen,
Becker and Swank (1990) found that greater predischarge self-efﬁcacy was the
strongest predictor of functional status at 6 months post-surgery among coronary
bypass patients. ”These ﬁndings support the belief that restoring self-conﬁdence
to perform activities is essential for recovery of the patient with heart disease,”
(Allen, et al., 1990, p. 342). Davis-Berman (1989) found that self-efﬁcacy was a
better predictor of depressive symptoms among a sample of older adults than
was physical illness. ”Results of this study suggest that illness is not the most
salient factor, but rather, that the cognitive interpretation of physical health and
functioning is the more pivotal variable in the prediction of depressive
symptoms,” (Davis-Berman, 1989, p. 213).

A study of an intervention to increase Active Patient Orientation among
hypertension patients found that patients who were afforded a high degree of
Active Patient Orientation were more likely to have their blood pressures under
control and exhibited more positive cognitive and behavioral responses to
illness-management (Schuhnan, 1979). Improved health outcomes related to self-
efficacy suggest that the traditional patient role (clients excluded from
participation in decision making) may adversely affect health outcomes.

E l] .1111.“

Patients who have not had an opportunity to participate in decision making
regarding their treatment may suffer increased depression and anxiety. Among
30 patients with early breast cancer, Morris 8: Royle (1988) found that a
signiﬁcantly higher percentage of the patients who were not offered a choice of
surgery versus excision plus radiotherapy experienced clinical levels of anxiety

and depression pre-operatively and up to 2 months post-operatively compared

4
to patients allowed to choose among the two treatments. Morris and Royle (1988)

concluded that offering a choice of operation may reduce distress. One
explanation for the observed reduction in distress may be that offering a choice
in treatments mitigates the helplessness experienced by individuals who ﬁnd
themselves in the role of patient. The patient role encompasses a loss of control
inherent to illness which is ampliﬁed by denial of the opportunity to participate
in treatment choices and decisions. Repeated exposure to uncontrollable events
has been found to be associated with an increase in depression and an inability to
take corrective actions when events are once again within the individual’s
control (learned helplessness) (Abramson, Seligman 8: Teasdale, 1978).
Individuals who learn that nothing they do in the role of patient will inﬂuence
which treatment they undergo, may generalize this helplessness to other aspects
of the patient role such as adherence to the prescribed treatment or self-care.
Increasing the patient’s sense of helplessness by denying participation in
decision making may result in decreased self-efﬁcacy for carrying out health-
related behaviors and reinforce an external locus of control as it relates to health.
In a review of the literature related to locus of control and health, Strickland
(1978) concluded that patients who perceived that they had little control over
their own health were significantly less likely to change their behavior to
prevent future illness or to improve present health than were patients who
perceived that they had control. Denying patients participation in health care
choices may promote generalized helplessness and depression by affecting locus
of control and self-efficacy.

Further evidence that the loss of control experienced when an individual
assumes the patient role may negatively affect health outcomes can be found in
studies of control. Langer (1983) found increased mortality among nursing home

patients who did not receive an intervention to induce a sense of responsibility.

5
Patients who received a talk emphasizing their responsibility for themselves and

who were given plants to care for themselves had a lower mortality rate than did
patients who received a talk emphasizing the staff’s responsibility for them as
patients and who were given plants that were watered by the staff (Langer,
1983)

Although some studies have concluded that patients do not wish to
participate actively in health care decisions, midlife women may be an exception.
In a study of 262 college-educated midlife women, Duffy (1988) concluded that
highly educated midlife women would not want to leave their health to chance.
Furthermore, although studies have found that many patients prefer to leave
decisions to professionals, a signiﬁcant minority prefer to make decisions
themselves or together with the practitioner (Degner 8: Sloan, 1992; Strull, et al.,
1984).

Patient participation in health care decisions has many advantages for
patients. Patients who are informed and actively involved in decision making are
more likely to have the information and skills needed to maintain their health
and prevent disease. Informed use of limited medical resources can decrease
health care costs. Active patient involvement can ensure incorporation of patient
values in medical decisions and make health care more reﬂective of individual
patient needs and values. Patients with higher self-efﬁcacy and active patient
orientation demonstrate improved clinical outcomes and decreased I
psychological morbidity. Research on learned helplessness and lack of control
suggest that the passive patient role may negatively affect patient health and
well-being. Finally, some patients do wish to participate actively in their health
care.

Given the compelling reasons for active patient involvement in health care,

6
and evidence that some patients do wish to actively participate, why do other

patients prefer to leave decisions to professionals? One explanation is that
individuals believe that they lack the skills necessary for effective participation.
Trinkaus (1991) asked a convenience sample of 779 first-year business students
(91% of whom were age 23 or below) whether they would like their physicians to
provide them with more information about medications prescribed for them and
whether they normally asked for this information. Trinkaus found that although
95% of American respondents (n=431) indicated that they desired more
information about medications only 52% indicated that they normally asked for
this information. Trinkaus speculated that patients might be hesitant to ask for
desired information out of fear of offending, not being given the opportunity, not
knowing how to ask, a feeling of inferiority, a desire to avoid embarrassment or
because the physician behaved in an intimidating, authoritarian manner. Factors
which may inﬂuence patient participation behavior then, include internal control
factors such as conﬁdence or self-efficacy and external control factors such as
system-level barriers to participation (Steele, et al., 1987).

Several models of behavior have been developed which use factors such as
control to predict a variety of behaviors. To better understand the factors
important in determining patient participation behavior an examination of these
models is needed. The next section will describe two theoretical models of
behavior which include the factors Trinkaus and others have suggested may be
important to participation behavior.

IheoneticaLMQdels.
Iheomofﬂannedkhamz

One theoretical model which may help explain patient preferences and
behavior is the ”theory of planned behavior” (Ajzen 8: Madden, 1986). The
theory of planned behavior is an extension of the (Fishbein 8: Ajzen, 1975) theory

7
of reasoned action (Figure 1). According to the theory of reasoned action, any

behavior is preceded by the intention to perform that behavior. The determinants
of intention are 1) attitude toward the behavior and 2) subjective norm. Attitude
toward the behavior is the product of the belief that a given behavior will
produce a specific outcome (behavioral belief) and the subjective value of that
outcome. The subjective norm is the sum of normative beliefs. Normative beliefs
are the product of a) the likelihood that another person or group would approve
or disapprove of the behavior and b) motivation to comply with the other’s

opinion.

w BEHAVIOR

 

 

Consistent with the theory of reasoned action, behavioral intentions have
been found to be highly correlated with volitional behavior, and behavioral
attitudes and subjective norms have been found to predict behavioral intentions
(Ajzen,Timko, 8: White, 1982; Manstead,Profﬁtt, 8: Smart, 1983; Hinsz 8: Nelson,

8
1990). In a study of the relationship between attitudes, intentions and voting

behavior, for example, Ajzen et al., (1982) found that attitude toward voting was
correlated .51 (p <. 01) with intention to vote; subjective norm was correlated .35
(p < .01) with intention to vote, and intention to vote was correlated .70 (p < .01)
with actual voting behavior.

One limitation of the theory of reasoned action however, was that it applied
only to volitional behavior (behavior which is completely under the individual’s
control) (Ajzen 8: Madden, 1986). In order to expand the theory of reasoned
action to include behavior which depends on factors other than simple will (such
as time, money, skills or the cooperation of other people), the theory of planned
behavior added the factor, ”perceived behavioral control” as an additional
predictor of behavioral intention (Ajzen 8: Madden, 1986).

E . 1 B l . l C l

Perceived behavioral control was deﬁned as ”the person’s belief as to how
easy or difﬁcult performance of the behavior is likely to be” (Ajzen 8: Madden,
1986, p. 457). Perceived behavioral control consists of beliefs about two sources of
control; internal factors and external factors. Internal factors include skills,
abilities, knowledge and planning. External factors include time, opportunity,
and dependence on the cooperation of other people (Ajzen 8: Madden, 1986).

According to the theory of planned behavior, perceived behavioral control
will be a significant predictor of behavioral intention beyond attitude toward the
behavior and subjective norms when the behavior in question is not entirely
under volitional control (Ajzen 8: Madden, 1986). Furthermore, perceived
behavioral control will contribute significantly to prediction of the behavior itself
(beyond the variance explained by behavioral intention) when perceived

behavioral control is an accurate reﬂection of actual behavioral control. Perceived

9
behavioral control is considered a partial substitute for a measure of actual

control (Ajzen 8: Madden, 1986).

 

Behavior

  

The theory of planned behavior is shown in Figure 2. According to the
model, perceived behavior is correlated with attitude toward the behavior and
subjective norm as well as with intention. The theory of planned behavior has
been found to predict behavioral intention better than the theory of reasoned
action when the behavior is not completely volitional (Catch 8: Kendzierski,
1990; Netemeyer 8: Burton, 1990; DeVellis,Blalock, 8: Sandler, 1990; Schifter 8:
Ajzen, 1985; Beale 8: Manstead, 1991).

Although the planned behavior model has been found to explain a signiﬁcant
degree of the variance in non-volitional behavior, some studies have found that
the model fails to explain behavior as well as alternative models. Dzewaltowski,

Noble and Shaw (1990) found that perceived behavioral control, subjective norm

10
and intentions failed to account for any unique variance in physical activity

participation over self-efﬁcacy. Multiple R for the theory of planned behavior
model was .52 for behavioral intentions and .32 (p < .05) for prediction of
physical activity behavior. Self-efﬁcacy combined with self-evaluation of
behavior contributed to significant prediction of physical activity behavior (R =
.46; p > .05).

Other studies have found that self-efﬁcacy contributed significantly to
predicting behavior and behavioral intentions when added to the theory of
reasoned action (McCaul,O'Neill, 8: Glasgow, 1988; Tedesco,Keffer, 8: Fleck-
Kandath, 1991; Dzewaltowski, 1989; Brubaker 8: Wickersham, 1990).

It has been noted that the construct of perceived behavioral control (Ajzen 8:
Madden, 1986) is highly related to Bandura’s (1977) construct of self-efficacy
expectations (Ajzen, 1988). Self-efﬁcacy has been defined as ”the conviction that
one can successfully execute the behavior required to produce the outcomes”
(Bandura, 1977). Self-efﬁcacy theory differentiates between efﬁcacy expectations
and outcome expectations. Outcome expectations are a person’s estimate that a
given behavior will lead to certain outcomes (Bandura, 1977). Davis and Yates
( 1982) manipulated self-efﬁcacy and outcome expectations among college
undergraduates using an anagram task and found that performance deﬁcits and
depressive affect occurred only when self-efficacy was low and outcome
expectancy was high, supporting self-efﬁcacy theory. Figure 3 shows a
diagrammatic representation of the difference between efﬁcacy expectations and
outcome expectations. According to self-efﬁcacy theory, efficacy expectations
affect people’s choice of behavioral settings, how much effort they will expend
and how long they will persist in the face of obstacles.

11

-' 1B r1

PERSON —|--> BEHAVIOR —|—> OUTCOME
_ __ _J _ _ _ _ .J _ _

I EFFICACY -| l OUTCOME —|
IE’SEEETE'Q‘E IE’SEEETM'QNS

Many studies have validated the important role of self-efficacy in behavior.
Self—efﬁcacy was found to predict persistence in pain control without medication
(Manning 8: Wright, 1983); post-treatment drinking behavior (Solomon 8: Annis,
1990); and efﬁcient use of analytic strategies in a laboratory management exercise
(Bandura 8: Wood, 1989).

Although self—efﬁcacy theory has been used extensively in predicting
behavior, some aspects of the theory appear to be problematic (Eastman 8:
Marzillier, 1984). For example, although self-efﬁcacy theory makes a distinction
between outcome expectations and efﬁcacy expectations, Maddux,Sherer, 8:
Rogers (1982) found that an experimental manipulation designed to inﬂuence
outcome expectancies, inﬂuenced expectations of self-efﬁcacy as well. Describing
an interpersonal communication technique as effective (manipulation of outcome
expectations) produced an increase in self-efﬁcacy expectations as well as
outcome expectations (Maddux et al., 1982). This raised questions about the
theoretical and practical distinction between the two constructs.

According to self-efﬁcacy theory, when outcomes (defined as the
consequence of behavior) are inherent to a given behavior, or when people
anticipate outcomes based on how well they expect to perform, outcome

expectations can not be differentiated from self-efﬁcacy expectations. In this case,

12
outcome expectations may not add to the prediction of behavior (Bandura, 1986).

Williams and Kinney (1991) found that outcome expectations (anticipated pain)
did not signiﬁcantly add to prediction of pain tolerance behavior when self—
efﬁcacy was held constant.

Although some studies have found that outcome expectations do not add to
the predictability of behavior (Manning 8: Wright, 1983; Solomon 8: Annis, 1990),
theorists have emphasized the value Of the distinction between the two
constructs. ”Both self-efﬁcacy and outcome expectations are likely to be
important and perhaps differentially important as a function of the specific client
problem,” (Kazdin, 1978, p. 180). In the theory of reasoned action and the theory
of planned behavior, ”belief that a given behavior will produce a speciﬁc
outcome” is one of the factors which contributes to attitude toward the behavior
(Ajzen 8: Madden, 1986).

In summary, although there are some questions about the distinction
between outcome expectations and self-efﬁcacy expectations, there is
experimental and theoretical support for maintaining that the two constructs are
distinct and that, at least in some situations, they contribute separately to
prediction of behavior.

D . o. '
1..'l....". .

",..‘ .u :.- l. an M. 94.”. - um . «.10. Im . .,
Several of the constructs which compose the theory of planned behavior are

also part of self-efﬁcacy theory. Table 1 shows the constructs in common and
their deﬁnitions. The ﬁrst similarity between the two theories is that they both
involve an individual’s assessment of personal skills. Self-efﬁcacy involves
organizing cognitive, social and behavioral subskills into action (Bandura, 1986).
In the theory of planned behavior, personal skills, abilities and knowledge are
believed to contribute to perceived behavioral control (Ajzen 8: Madden, 1986).

Self-efﬁcacy is measured by asking the individual to indicate conﬁdence in

13
ability to perform a given behavior on a scale from 10 to 100. Perceived

behavioral control has been measured by items such as, ”If I wanted to I could
easily attend this class every session,” (extremely likely/ extremely unlikely)
(Ajzen & Madden, 1986). °

The major difference between self-efﬁcacy and perceived behavioral control
is that in addition to measuring the individual’s assessment of personal skills and
resources, perceived behavioral control measures the individual’s assessment of
external factors which would facilitate or hinder the behavior.

Belief about the relationship Of behavior to outcome is also an important
element of both self-efﬁcacy theory and the theory of planned behavior. Self-
efﬁcacy theory describes outcome expectation as a person’s estimate that a given
behavior will lead to certain outcomes (Bandura, 1977). In the theory of planned
behavior, the underlying elements of attitude toward behavior are 1) behavioral
belief which links the behavior to a certain outcome, and 2) the outcome’s
subjective value (Ajzen 8: Madden, 1986).

In summary, both self-efﬁcacy theory and the theory of planned behavior
include belief about personal skills and belief about the relationship of behavior
to outcome in their models of behavior. The theory of reasoned action differs
from self-efﬁcacy theory in that it posits that the immediate precursor to
behavior is behavioral intention, and it includes subjective norms as a predictor
of behavioral intention.

Although both self-efﬁcacy theory and the theory of planned behavior have
been found to signiﬁcantly predict behavior, is there any evidence that either

theory may help predict patient participation in health care?

Iahlil

14

0 mu 0 u- - ' zlf-Ef' -§Th=x._l010ndTh of lair- 3°h- '

Self—Efﬁcacy Construct

 

Planned Behavior Construct

 

SEW i Remixed Behavior Control

The conviction that one can
successfully execute the behavior
required to produce the outcomes
(Bandura, 1977).

Quicomebspectations
A person’s estimate that a given
behavior will lead to certain

outcomes (Bandura, 1977).

A person’s belief as to how easy or
difﬁcult performance of the behavior
is likely to be. lnternaljagtnm
contributing to perceived control
mcludejkﬂblabﬂimm
and planning. External factors
include time, Opportunity and
dependence on other people (Ajzen
8: Madden, 1986).

l . 1 I l B l .
The degree to which a person has a
favorable or unfavorable evaluation
of the behavior. Consists of
hehaxioraLheliefs whichlimLthe
WW and

l , l . . l

(Ajzen 8: Madden, 1986).

15

l- f' a: . .. u-ol. oro o _I- 'r or u o .f_I_‘-., 4
Woodward and Wallston (1987) found that among older adults desire for
control in health-related situations was mediated by low health care self-efﬁcacy.
One hundred sixteen adults between 20 and 99 years old completed measures of

desire for control of health care, desire for information and health care self-
efﬁcacy. Three age groups (20—39; 40—59; 60+) differed signiﬁcantly on desire for
control of health care and health care self—efﬁcacy. When health care self-efﬁcacy
was used as a covariate for desire for control of health care the main effect for age
was no longer signiﬁcant. Desire to be actively involved in health care decisions
was dependent on self-perception of competence to participate effectively.

The ﬁnding that desire to participate is mediated by self-efﬁcacy is important
because it indicates an avenue through which patient involvement in decision
making can be modiﬁed.

WM

In a critique of self-efﬁcacy theory, Lee (1989) concluded that reliable
interventions to alter efﬁcacy expectations could not be developed until a model
of the relative weighting of various types of self-efﬁcacy information (mastery
experiences, vicarious experience, persuasion and physiological information) has
been developed. Previous research indicating that self-efﬁcacy was changed
through intervention shows that this conclusion was unwarranted. Vallis and
Bucher (1986) increased self-efﬁcacy for pain tolerance during a cold pressor test.
Maddux, Sherer, and Rogers (1982) successfully manipulated self-efﬁcacy for
assertiveness using verbal persuasion and Davis 8: Yates ( 1982) experimentally
induced high and low self-efﬁcacy expectations among a group of 108 male and
female undergraduate students. Sanna (1992) experimentally manipulated self-
efﬁcacy expectancy by providing false feedback on a vigilance task and by
manipulation of task difﬁculty (providing mastery experiences).

16
Although interventions can not be developed based on quantitative formulas

of speciﬁc amounts of persuasion, modeling and mastery, effective interventions

can be developed and tested. Interventions to increase self-efﬁcacy should

provide one or more of the four types of information outlined by Bandura (1982).
t r a1 -E ' ?

Bandura (1977) described four principal sources of efﬁcacy information:
performance accomplishments, vicarious experience, verbal persuasion, and
physiological states.

W. Performance accomplishments are the most
inﬂuential sources of information (Bandura, 1977, Barling 8: Snipelisky, 1983).
Success experiences raise efﬁcacy expectations and repeated successes develop
strong efﬁcacy expectations which reduce the impact of failure experiences.
Performance accomplishments are promoted through skill-building activities
such as participant modeling, performance exposure and self-instructed
performance. Skill-building leads to increased success experiences and
strengthens efﬁcacy expectations. Davis and Yates (1982) found that providing
college students with either mastery or failure experiences in solving anagrams
produced changes in selﬁefﬁcacy. Students who were given difﬁcult anagrams
had lower self-efﬁcacy than those given easy anagrams. Participant modeling of
skills like question-asking and decision making may effectively increase
participant health care self-efﬁcacy.

MW. Self-efﬁcacy may also be inﬂuenced through
observation of live or symbolic models (Bandura, 1977). Schunk and Hanson
(1985) found that children who observed a peer model developed higher self-
efﬁcacy for learning than those who observed a teacher model or no model at all.
Observation of peers modeling communicatiOn and decision making skills in a

group intervention setting may strengthen gains in health care self-efﬁcacy

17
obtained through participant modeling and provide independent efﬁcacy

information.

Managing Efﬁcacy expectations can be increased through persuasion
although expectations induced this way are easily extinguished by disconﬁrming
experiences (Bandura, 1977). Maddux, et al., (1982) manipulated self-efﬁcacy
among introductory psychology students using verbal persuasion. Students who
read essays describing a technique as easy to use had higher self-efﬁcacy for
using the technique than those who read essays describing the technique as
difﬁcult to use. Verbal persuasion can be used in health care settings to introduce
new information which may change expectations and encourage participant
modeling experiences.

W5, Emotional arousal is another source of efﬁcacy
information (Bandura, 1977). Anxiety may be interpreted as a sign of an inability
to perform in a given situation and can operate to inhibit performance of
efﬁcacious behavior.

These four sources of efﬁcacy information can be utilized to modify self-
efﬁcacy. The most effective source of efﬁcacy information has been found to be
performance accomplishments (Bandura, 1982). Following this, vicarious
experience and verbal persuasion have also been found to be effective in
modifying efﬁcacy expectations (Bandura, 1982). But how can self-efﬁcacy
expectations be maintained in instances where the individual must interact with
others? This issue is of particular concern related to health care self-efﬁcacy.
What health care settings are likely to promote health care self-efﬁcacy? How can
patients be trained to interact with the health care system to maximize their
success experiences? What is the impact of cultural role expectations and the
interaction between patients and health care providers on patient participation

behavior and satisfaction?

18
Pti Prf inltrain

The most signiﬁcant aspect of health care settings is the interaction of the
health care professional with the patient. Three basic models have been used to
describe the interaction between patients and medical practitioners; 1)
Professional as Decision Maker, 2) Joint Decision Making and 3) Consumer as
Decision Maker. All three models have potential value in health care. For
example, the professional as decision maker model may be appropriate in
instances in which the patient is incapacitated or otherwise unable to participate
in decision making (Schain, 1980). Below each model is described and critiqued
as to its potential impact on patient health care self-efﬁcacy.
EE'IDHIH—El “I!

In the advocacy model of professional / client interaction, the patient is
regarded as helpless and dependent (Schain, 1980). The professional maintains
an authoritative role and solves problems and makes decisions on the patient’s
behalf (Sharf, 1988; Schain, 1980). In this kind of an interaction it would be
difﬁcult for a patient to develop or maintain conﬁdence in regard to health
matters. The patient would have to remold the nature of the interaction in order
to take an active role in decisions.

In American society, physicians in particular, though to some extent other
health professionals as well, control medical knowledge, medical practice and
the institutions of medicine, (Navarro, 1983). As a consequence, the medical
profession commands high social and economic status (Ehrenreich 8: Ehrenreich,
1978; Navarro, 1983) which makes it difﬁcult and unlikely that a patient could
effectively change the nature of the interaction between herself and the
professional (Roter, 1987).

Modern medical technology further complicates the patient’s efforts at self-
efﬁcacy in health care. As medical technology becomes increasingly more

19
complex and formidable it further removes the patient from an active role.

Expert technicians are required to operate and interpret the information from
sophisticated equipment. Professionals and patients alike may be so fascinated
with medical technology (Figge, 1990) that they treat the machines as decision
makers. Davis-Floyd (1987) reported that obstetricians for example, were likely
to perform a caesarean if a fetal monitor indicated a potential problem although
caesarean rates have not been found to be related to neonatal outcomes.

Patients may ﬁnd the array of machinery and technology so daunting that
they surrender all hope of ever being able to understand the nature of their
health problems, let alone make decisions regarding them. The professional-as-
decision-maker model provides no aid to the patient in attempting to overcome
this intimidation and take an active role.
1.”: .. Ill. 1111

The second model which has described patient/ professional interactions is
the joint decision making model. Schain (1980) and Roter (1987) advocated
models of interaction in which patient and professional shared responsibility for
health care decisions equally. The patient maintained autonomy and dignity
while respecting the skill, opinions and expertise of the professional. Joint
decision making models were designed to mitigate the tendency of professionals
to emphasize medicotechnical concerns and neglect psychosocial issues (Roter,
1987). Psychosocial concerns of the patient are incorporated into decisions
through the patient’s input into the process. Many patients may be unable to
participate equally in decisions with professionals however, due to the social
power differences between themselves and the professional. Joint decision
models lack a strategy for supporting patients in the decision making role and
fail to address the issue of power inequity between clients and professionals.

Because of the difference in power between patients and professionals it is

20
likely that in practice the professional would dominate a joint relationship

between clients and professionals. Joint decision making models would become
professional-as-decision maker models in practice.
' i — ' i M 1

In contrast to the joint decision making models, several articles have outlined
models which specifically address the consumer’s role as decision maker.
Corcoran (1988) delineated the role of an advocate in the decision process.
Corcoran (1988) operationalized Gadow’s 5 steps in assisting a patient in
decision making in the following way: (1) Assure relevant information, (2)
Enable the patient to select information, (3) Encourage patients to seek
information, (4) Offer to develop a decision ﬂow diagram, (5) Disclose nurse’s
view, (6) Share personal values and viewpoints (7) Help patient determine own
values, (8) Discuss the meaning of the experience for the patient.

O’Connor and O’Brien-Pallais (1989) also developed a model for assisting
patient’s in making decisions. O’Connor 8: O’Brien-Pallais established that the
goal of decision counseling should be for the client to make an ”effective”
decision . An effective decision was deﬁned as a decision which was informed,
consistent with personal values and congruent with subsequent behavior
(O’Connor 8: O’Brien-Pallais, 1989). Steps in decision therapy include: (1) Clarify
goals, alternatives and outcomes, (2) Realign unrealistic alternatives, (3) Identify
viable alternatives, (4) Clarify outcome expectations, (5) Realign unrealistic
outcome expectations, (6) Clarify desirability of possible outcomes, (7) Clarify
priority of outcomes, (8) Identify values tradeoffs, (9) Facilitate alternative
selection, (10) Teach self-help skills required for behavioral implementation of
decision.

WW. Effective decision counseling has been

found to be associated with decreased levels of post-decisional regret and

_ 21
increased satisfaction. Cooper, Bledin, Brice, and Mackenzie (1985) found that

women who indicated regret or uncertainty about their decision to be sterilized
were signiﬁcantly more likely to report that their preoperative counseling had
not been adequate than were women who reported that they were deﬁnitely glad
they had been sterilized (p < .005). Parents who recalled being involved in
making choices regarding their newborns with myelomeningocele were
signiﬁcantly more likely to be satisﬁed with the tertiary-care center than were
parents who did not recall involvement in making choices (p < .01) (Charney,
1990).

Decision support models of client—professional interaction have several
advantages over the traditional interaction model and the joint decision making
model. Both the Corcoran (1988) and the O’Connor 8: O’Brien-Pallais (1989)
models deﬁne speciﬁc steps for facilitating client decisions. Unlike the joint
decision making models, the decision support models emphasize that the ﬁnal
decision should rest with the client, with the professional’s role being to serve as
facilitator in the client’s decision process. Because the emphasis is on helping
clients to make decisions themselves, many of the problems associated with
authoritative and shared decision models—such as the power inequity between
professional and client—are circumvented.

Several issues remain to be addressed, however. Control of information by
the health professional has not been adequately addressed in decision support
models. Corcoran (1988) noted that the professional may subtly manipulate a
patient’s decision by the information that is or is not provided. Corcoran’s
model left this problem open because the professional holds the responsibility
for ”assessing” the amount of information the patient desires and ”can cope
with.” The O’ Connor 8: O’Brien-Pallais (1989) model gave no such responsibility

to the professional thus making such an occurrence less likely, though still

22
possible.

Corcoran (1988) indicated that it was important to assess the patient’s desire
to know and provide only the desired level of information. The clinician must
guesshow much information the client desired and withhold other information.
Clinicians have been found to estimate patient preferences for information
inaccurately. Strull, et al. (1984) found that while 55% of patients being treated
for hypertension indicated that they preferred ”Quite a lot” or ”Very Extensive”
discussion about therapy, their clinicians estimated that only 43% desired that
extent of discussion. The professional should communicate to the client that
several levels of information are available so that the client can choose what
information he or she seeks.

A drawback of decision support models has been a lack of operationalized
evaluation criteria. Corcoran (1988) gave no criteria by which to evaluate the
effectiveness of the decision advocacy guidelines but did encourage nurses to test
and reﬁne the model and to develop additional guidelines. O’Connor 8: O’Brien-
Pallais (1989) listed expected outcomes to be achieved by patients through
decision support therapy. Some of the expected outcomes such as ”identiﬁes
viable alternatives” can be measured using existing measures such as decision
analysis. Other outcomes such as ”expresses satisfaction at having made the best
decision under the circumstances” presently lack reliable measures.

Because decision support models place decision making with the client,
interactions based on these models are most likely to promote health care self-
efﬁcacy. Health care self-efﬁcacy is promoted two ways in decision support;
through mastery and persuasion. First, the client experiences competence in the
decision making process. Second, the professional communicates conﬁdence in

the client’s ability to make a decision which is best for him or her.

23
Conclusion

Of the three models of client-professional interaction commonly advocated,
the decision support model is most likely to promote high health care self-
efﬁcacy among patients. Although the model has been developed there have
been few programs which have applied the model in actual practice.

Despite growing demand for active involvement of patients in their health
care, practical guidelines for supporting the client in this role and evaluating
such efforts have been lacking. Also lacking has been a theoretical basis for
evaluating interventions designed to facilitate active patient participation in
medical decision making (Steele, et al., 1987).

MW

Theoretical concepts and tools useful for evaluation of patient participation
interventions may be drawn from community psychology. Community
psychology is concerned with development and evaluation of social programs. It
embraces an approach to these problems which is 1) democratic, 2) innovative
and 3) can be disseminated (Fairweather 8: Davidson, 1986). Inherent to the
concept of democratic change is empowerment (Rappaport, 1987). Full
participation in social change requires providing information and skills to those
who have traditionally been excluded from decision making. Intervention for
social change can occur at the individual, group, organizational or societal levels
but most often requires intervention at all four levels (Fairweather 8: Davidson,
1986). Empowerment and level of change are two principals of community
psychology which apply to evaluation of a patient participation intervention.
Emmment

Empowerment refers to individual determination over one’s own life and
democratic participation in the life of one’s community (Rappaport, 1987).

Empowerment describes both a psychological sense of personal control or

24
inﬂuence and a concern with actual social inﬂuence, political power and legal

rights (Rappaport, 1987). It applies to individuals as well as to organizations
(Rappaport, 1987).

Empowerment has been described as a multi-factor construct which includes
the elements self-efﬁcacy, political efﬁcacy, mastery, desire for control, and locus
of control as well as others (Zimmerman 8: Rappaport, 1988). 1

Health care self-efﬁcacy is one element of psychological empowerment
related to health care. Measurement of this construct can be used to evaluate the
extent to which an intervention serves to empower health care consumers in their
interactions with the health care system. Self-efficacy focuses on the individual
aspect of empowerment while measures of barriers or external perceived control
focus on the health care system. Combined, self-efﬁcacy and perceived barriers
represent the two main elements of psychological empowerment related to
health care.

Fairweather and Davidson, (1986) indicated that while intervention for social
changecould occur at any of four levels, it was most likely to result in successful
and lasting social change if it occurred at multiple levels.

Although intervening at multiple levels is most likely to promote effective
change, it is difﬁcult to do. Attempts to intervene at the system level often meet
with barriers. Primary among the barriers is the tendency of those invested in the
status quo to perpetuate the status quo at all costs (Fairweather 8: Tomatzky,
1977). Gray, Doan and Church (1991) reported that one barrier to empowerment
of persons with cancer was the behavior of health professionals. The health
professionals were invested in the traditional patient / professional model.
Professionals derived social and economic status from their role as decision

makers (Ehrenreich 8: Ehrenreich, 1978; Navarro, 1983). Persons with cancer

25
found it difﬁcult to obtain the needed information from these health

professionals in order to participate in decisions.
- n ' f t - n

One solution to the barriers to intervening at the system level is intervention
at the individual level which stimulates individuals to demand changes at the
system level. An example of this was the alternative birth movement in the
United States. The movement was a reaction by women against the medical
establishment which controlled Obstetrical care (Mathews 8: Zadak, 1991).
Mathews and Zadak (1991) reviewed the history of the alternative birth
movement: The natural childbirth movement gained momentum in the 1960’s
and 1970’s when it converged with the feminist and consumer movements of the
time. The movement emphasized public education, informed choice and self-
help. Educated, middle class Americans began to choose home births and
alternative birthing centers over hospital births. They demanded a more
personalized, family-centered, patient-controlled birthing experience. Hospitals
and obstetricians experienced a loss of control and increased competition for
patients. As a result health professional organizations joined together to endorse
a philosophy Of family-centered birthing and hospitals began to offer in-hospital
combined labor-delivery-recovery birthing rooms. Routine procedures such as
shaving the perineum, giving the laboring woman enemas and conﬁning her to
bed began to fall into disfavor. Hospital rooms were redesigned to minimize the
appearance of technological intervention in birthing. The alternative birth
movement initiated a major transformation in the birthing process in the United
States (Mathews 8: Zadak, 1991). This transformation was a result of
empowering health care consumers rather than directly intervening at the system
level.

Despite outward appearances however, the degree of actual change in

26
Obstetric practice can be questioned. MatheWs and Zadak (1991) pointed out that

alternatives to traditional medical management of childbirth are available only to
highly motivated, well-prepared women who anticipate a completely normal
pregnancy and delivery. The control and prerogatives of the obstetrical
community continue to be maintained.

In spite of continued control by the obstetrical community however,
childbirth in the United States continues to change to meet consumer demands.
As hospitals have changed and consumers have continued their demands for
family-centered care, medical education has slowly begun to adapt. Pregnancy
and childbirth is being redeﬁned as a natural rather than a medical event. The
impact of the alternative birthing movement continues to mold US obstetrical
practices. Providing health care consumers with the information and skills to
challenge the traditional model of care successfully promoted change within the
system. Intervention at the individual level resulted in system-level change
despite resistance from those invested in the current system.

Another way in which system-level change may be effected is through
development, evaluation and dissemination of programs which empower health
care consumers to participate actively in health care decisions. Once an effective
method of empowering consumers is identiﬁed it may be disseminated to the
health care community as an alternative model for interacting with patients.
Dissemination involves four phases; approach, persuasion, activation and
diffusion (Fairweather 8: Davidson, 1986). The approach phase involves
identifying the target population (such as individual health professionals, clinics
or health maintainence organizations) and implementing a plan to inform them
of the innovation (such as advertisements, professional training workshops,
articles in professional journals, etc.). Persuasion involves identifying and

utilizing an effective method of information transmission. It is important to

27
consider the style of persuasion, content of the message and legitimacy of the

advocate (Fairweather 8: Davidson, 1986). The most effective method of
persuasion can be identiﬁed through implementation of multiple methods and
evaluation of their success. Use of respected fellow professionals as advocates
and focusing the content Of the message on time- and cost-efﬁciency as well as
patient well-being are possible aspects of persuasion with health professionals.
The next phase of dissemination is the activation phase. Once the professional or
organization decides to adopt the innovative program it is necessary to activate
the model in the community setting through accurate model replication.
Activating the model in the community while maintaining its integrity requires
transmission of content and process, guidance during implementation and
Ongoing monitoring of implementation (Fairweather 8: Davidson, 1986).
Activating a program to empower health care consumers to participate in health
care decisions will require packaging the program materials in a manner
appropriate to the private ofﬁce or clinic setting, appropriate training of program
personnel and observation and feedback regarding the program implementation.
The ﬁnal phase of dissemination involves spreading the innovation from the
initial adopters to the rest of the target population and society as a whole
(Fairweather 8: Davidson, 1986). During diffusion it is critical to continue
monitoring implementation of the program to maintain its integrity and
effectiveness. With patient empowerment programs it is imperative to avoid
cooptation by the status quo such as partially occurred in the instance of the
natural birth movement. Integrity of the empowerment goal of the program may
be maintained by tracking adoption of the program through copyright of
materials and visible requests to be informed of program adoption. Tracking Of
implementation will allow program developers to closely monitor

implementation and provide feedback regarding congruence with original goals.

28
The speciﬁc method used for each of the four phases of dissemination should be

determined through experimentation (Fairweather 8: Davidson, 1986).
rr r E H a1 ar n r

Two areas in which the patient empowerment approach has been used are
cancer care and diabetes education. Involvement of patients in decision making
is especially important in diabetes education and in cancer care because in
diabetes care the patient is personally responsible for carrying out so much of the
treatment program (Anderson, Funnell, Barr, Dedrick, 8: Davis, 1991) and in
cancer, patients lose much control in their lives (Gray, et al., 1991).

The University of Michigan Diabetes Research and Training Center
developed a training program aimed at teaching diabetes educators to empower
patients. The program included four steps; 1) Help patients determine what part
of their diabetes care is a problem for them, 2) Help patients focus on their
emotions, 3) Help patients clarify their health-related values and establish goals,
and 4) Help patients develop and commit to a speciﬁc plan to achieve their goals
(Anderson, et al., 1991). Participating professionals were required to follow a
simulated diabetes care regimen for 3 days prior to participation in the skills-
based workshop. The goal of the simulation was to introduce the professionals to
the challenges of caring for diabetes on a daily basis. The program was based on
Rogers’ client-centered counseling model and was developed, pilot tested and
offered to diabetes educators. Evaluation of the effectiveness of the program in
empowering patients was not reported.

A second program focused on cancer patients. Because people diagnosed
with cancer lose control in their lives due to illness, side effects and unbalanced
power relations with health professionals, they can beneﬁt from interventions to
empower them in relation to the health care system (Gray, et al., 1991). Gray, et

al. (1991) developed a model to empower people diagnosed with cancer. The

29
model included the following ways Of empowering people: 1) Address symptom

management for persons with cancer, 2) increase funding for outpatient,
community and home-based care, 3) allow patients to determine the amount of
information and participation they receive 4) teach patients stress-management
techniques, 5) teach assertiveness training for patients, 6) develop mutual
support groups for people with cancer, 7) involve patients in setting health care
policies, 8) hire patient advocates, 9) encourage political activism by persons
with cancer, 10) change societal attitudes toward illness and death to decrease
stigmatization of cancer survivors.

Both programs addressed patient empowerment on two levels, change at the
system level and change directed toward patients themselves. At the system
level, the diabetes education program sensitized diabetes educators to the
problems of diabetes patients by having them follow a diabetic care regimen for
3 days. It educated professionals in how to change their behavior in order to
empower patients. The model for empowering persons with cancer
recommended involving patients in setting health care policies and encouraging
political activism.

. Both programs developed to empower medical patients also included
elements directed at intervening with patients themselves. The individual-level
elements of the interventions were similar to those advocated in decision support
models of client-professional interaction.The diabetes program trained educators
to intervene with patients using the four steps described above. The model for
persons with cancer emphasized informing patients and training them in skills
such as stress-management and assertiveness.

The four recommendations shared by the individual-level portions of the
intervention programs and by the decision support models of client-professional

interaction were 1) provide information about the factors important to the health-

30
related decision; 2) help the client clarify his or her values related to the

important factors; 3) involve the client in health care decisions; and 4) help the
client develop and commit to a speciﬁc plan to carry out health care decisions.
These four elements along with system level interventions, form the basis of the
patient participation philosophy.

Although there have been programs aimed at empowering patients with
speciﬁc medical problems, no program has combined the four basic elements of
patient participation into an intervention to aid women in their decision making
related to menopause which is the focus of the research reported in this paper.

Th N f r r A ' M

Menopause is a time of physical changes for women. Recent research
demonstrating beneﬁts and risks associated with hormone replacement therapy
after menopause contribute to the decision making dilemmas women face at
menopause. Because women at midlife face physical changes associated with
menopause as well as new concerns about health changes associated with aging,
the need to be an informed decision maker regarding health issues is especially
important. Additionally, it is important to involve women in health care
decisions at menopause because they are ultimately responsible for carrying out
the care plan selected. The woman herself must carry out many of the activities
which may beneﬁt her health such as exercise programs, diet changes or taking
medicines as directed. Because of the changes and decisions women face at
menopause a program which promotes their participation in their health care is
needed.

mm

Women at midlife face an increased risk of health problems such as

osteoporosis and heart disease. They may choose to respond to these increased

risks through health promotion behaviors such as regular exercise, elimination of

31
smoking and diet modiﬁcations. They may also choose whether or not to take

hormone replacement therapy.

Because the issues related to health maintainence and hormone replacement
therapy are complicated and controversial, midlife women may beneﬁt from
informed, active involvement in decision making related to their health care.
Active consumer participation in health care is important for other reasons as
well. Expanded choices and increased costs of medical care provide medical
practitioners with multiple treatment options. Consumers should help choose the
option which best represents their own values. Furthermore, active participation
and a sense of self-efﬁcacy related to health have been found to result in
improved clinical outcomes and decreased psychological morbidity.

Despite potential beneﬁts related to active participation, health care
consumers frequently decline to participate in health care decisions. One
explanation is they believe they lack skills necessary for effective participation.
Health care consumers have low self-efﬁcacy related to health care.

One solution to this problem is to increase health care self-efﬁcacy. It has
been demonstrated that self-efﬁcacy can be modiﬁed through intervention.
Bandura (1977) described four principal sources of efﬁcacy information through
which intervention could occur; performance accomplishments, vicarious
experience, persuasion and physiological states. An intervention which enhanced
self-efﬁcacy through these sources could promote participation among
individuals for whom low self-efﬁcacy was a barrier.

Health care self-efﬁcacy may also be modiﬁed by the quality of the
interaction between the consumer and the health care professional. Interactions
in which the professional maintains control over decision making may serve to
lower consumer self-efﬁcacy. Alternatively, interactions which promote

autonomy by the consumer may serve to increase self—efﬁcacy. The advocacy

32
and the joint decision making models of interaction may hinder development of

health care self-efﬁcacy by encouraging a passive role for the consumer. In the
advocacy model the professional makes decisions for the patient. In the joint
decision making model the professional may encourage input from the patient
but the status contrast between patient and professional may prevent active
participation by the patient.

The decision support model of interaction is hypothesized to have the most
positive impact on health care self-efﬁcacy. In this model the health care
consumer is assisted to make decisions which are informed, consistent with
values and congruent with subsequent behavior (O’Connor 8: O’Brien-Pallais,
1989)

Few programs to empower health care consumers have been developed and
evaluated. Two programs which have been developed incorporated aspects of
the decision support model of interaction. These programs 1) informed
consumers, 2) helped the consumer clarify values, 3) promoted consumer
decision making, and 4) helped the consumer develop a personal plan to carry
out health care decisions.

Although a few programs have been developed to empower health care
consumers to participate actively in decision making, no program has been
developed and evaluated which addressed the issue of menopause.
Development and evaluation of such a program is needed.

Concepts and methods from community psychology may be useful in
evaluating such a program. The concepts of empowerment and multi-level
intervention may be applied to determine the effectiveness of such a program in
empowering midlife health care consumers.

The Current Research

The purpose of the current research was to determine the impact of three

33
menopause education programs on the health care self-efﬁcacy of midlife women

and to test a model of the impact of self-efﬁcacy on participation in health care
decisions. Because self—efﬁcacy has been found to be related to desire for control
of health care and it has been documented that some patients desired more
information than they actively requested from health care providers, it was
hypothesized that patient participation behavior could be inﬂuenced by
increasing self-efﬁcacy related to active participation behavior. The three
intervention programs to be tested for their impact on self-efﬁcacy were
developed by Rothert (1990) to help women become better decision makers
about their menopausal health and hormone replacement therapy.

The ﬁrst program, (A) provided information in the form of a wriﬁen
brochure. The brochure consisted of three sections addressing 1) the physiology
of menopause and self-care, 2) the pros and cons of hormone replacement
therapy and 3) communication with health care professionals (Rothert, Kroll,
Holmes-Rovner, Rovner, Schmitt, 8: Talarczyk, 1992). Section one of the brochure
described what women could expect at menopause, including information on
what menopause is, when it occurs and the physical changes women may
experience at menopause. Section one discussed the symptoms and increased
risks that may accompany menopause and suggested self-care strategies.

Section two of the brochure described additional risk factors to consider
related to use of hormone replacement therapy and described the two main types
of hormone therapy; estrogen-only therapy and estrogen combined with
progesterone. The beneﬁts and risks of each type of therapy and the increase or
decrease in death rate due to endometrial cancer, heart disease and osteoporosis
were depicted numerically and graphically.

Section three of the brochure was in workbook format with spaces to record

personal and family health history, menstrual history and questions for the

34
health care professional. Examples illustrated how to provide detailed

information to help health professionals care for one’s health and sample
questions were offered to help the reader think of her own questions and
concerns related to menopause. Recommendations for communicating effectively
during the visit included the following: describe symptoms speciﬁcally, let the
professional know about questions at the beginning of the visit (to allow time
during the visit to have them answered), ask questions and ask for clariﬁcation
when necessary, repeat information to ensure understanding and write down
information that will be needed later. 1

' The brochure was designed as a standard treatment control for the decision
support intervention (described below). Although the brochure provided
instruction in effective communication with health care professionals, it was not
anticipated that the brochure alone would be sufﬁcient intervention to increase
self-efﬁcacy related to the health care system.

The second program (B) provided information in a W
format. Program B consisted of three 1 1/ 2 hour sessions using a lecture format
with overhead transparencies as visual aids. Program content was parallel to the
brochure. Session I covered the physiology of menopause and self-care methods
for managing symptoms and decreasing health risks. Session II discussed the
pros and cons of hormone replacement therapy and Session III covered
communicating with health care providers. A 10 to 20 minute question and
answer period was led by a nurse or physician member of the instructor team
following discussion of the pros and cons of HRT. Session III presented a lecture
on communicating with health care providers which paralleled the content of
Section III of the brochure but instructors included anecdotal examples to
illustrate how effective communication could result in more satisfying

interactions with the health care system. For example, an advisory panel of

35
health professionals indicated that it would be helpful for patients to convey that

they have questions at the beginning of a visit in order to allot time to answer
them. The reasoning behind this recommendation was therefore explained in the
lecture, but not in the written brochure.

The lecture/ discussion format program was not anticipated to significantly
impact self-efﬁcacy related to the health care system. Although information on
effective communication with health care providers was presented verbally as
well as in written form, instruction alone was not expected to be an effective
intervention by itself. The lecture / discussion program was designed as a time
and attention control group for the decision support intervention.

The third program (C) was a decision support intervention which provided
information and experience in an MW]; format. The intervention
consisted of a series of exercises to foster active involvement in the decision
process. The program, parallel to Program B in time spent, consisted of three
sessions, each approximately 1 1/ 2 hours long.

The ﬁrst session of the program was identical in content to Programs A and
B. Information on the physiology of menopause and self-care strategies for
managing symptoms and minimizing health risks was presented in a lecture
format.

The second session of Program C presented the information contained in the
brochure and in Program B on the risks and beneﬁts of hormone replacement
therapy. As in Program B, a question and answer period was included following
presentation of the pros and cons of hormone replacement therapy. The women
received information about a previous study by the research team which found
that different women weighed the risk factors related to HRT in differing ways
(Rothert, et al., 1990). This information was presented to help the women

understand that women differed in how they approached the decision about

36
HRT and to introduce an exercise assessing personal risks and values related to

HRT.

In order to better understand their own risks and values related to HRT the
women completed a Personal Risk Assessment, Problem Signiﬁcance Assessment
and Relevance Chart which combined base rate risk, personal risk and personal
values for each of up to 7 factors important to the woman’s decision about HRT.
The Personal Risk Assessment informed the women of their personal risk of
heart disease, osteoporosis and endometrial cancer. The women scored their own
risk assessments and marked base rate risk and their personal risk on their
relevance charts. The purpose of the relevance chart was to help women visually
compare their risks and values across the multiple factors in their decision.

Following the Personal Risk Assessment exercise, the women completed the
Problem Signiﬁcance Assessment on which they rated having each of the
following factors on an 11 point scale with 0=Perfect Health and 100=Death:
fractures from osteoporosis, heart disease, endometrial cancer, hot ﬂashes, cyclic
bleeding, treatment side effects and other. Intermediate points on the scale were
marked in 10-point increments from 10 to 100. The Problem Signiﬁcance
Assessment was designed to help the women clarify their values related to the
tradeoffs in the HRT decision. This information was transferred in graphic form
to the Relevance Chart.

After completion of the Personal Risk Assessment and Problem Significance
Assessment, the women were given a work sheet with a series of six scenarios
describing various combinations of risks and values. One or two of the scenarios
were read in class and sample relevance charts were completed on overhead
transparencies to show how a woman might weigh and combine the information
and come to a decision. The women were then given time to read over the other

scenarios, placing themselves in each situation, and consider how they would

37
weigh the various factors and make a decision related to symptom management

and risk minimization.

Following the scenarios exercise, the women returned to the relevance chart
and assigned a ”relevance” score to each factor based on her risk and signiﬁcance
for that factor (heart disease, fractures, endometrial cancer) or signiﬁcance alone.
The relevance chart visually depicted each woman’s unique combination of risks
and values with base rate information. The women weighed the relevance of
each factor to make a preliminary choice between 1) I choose not to discuss HRT
with my health care provider, 2) I choose to discuss taking HRT with my health
care provider or 3) I choose to continue to take HRT.

During session III of the decision support intervention, participants received
the same lecture on communicating with health care providers presented to
participants in program B. The information was parallel to that provided in the
brochure and included anecdotal examples to support the value of effective
communication and participation in health care. In addition, participants took
part in a role play exercise in which they took turns acting as health care
providers and as active patients. Participants received the following instructions
prior to the role play, ”For those of you who are playing health care providers,
remember you have a schedule to keep and patients in the waiting room but you
are seeing one of your regular patients. For those of you who are playing
patients, you have called ahead and informed the receptionist you want to make
an appointment to discuss menopause.” The purpose of this exercise was to
provide a participant modeling experience in which the women could gain
conﬁdence actively participating in a simulated health care encounter. The
women separated into pairs and took turns acting as the health professional and
as the patient. In addition to gaining experience and conﬁdence in active patient

behavior, this allowed participants to gain understanding of how a health care

38
provider might feel when facing an active patient. Following the role play there

was a discussion period in which the instructor asked the group the how the role
play went and how it worked. One comment included that when playing the
health care provider it felt uncomfortable not to know the answers to the I
patient’s questions. This allowed a brief discussion about whether health care
providers may know the answers to all questions. Sometimes discussion focused
on past experiences in actual health care encounters and how the concepts
learned in the class could be applied to the situation in the future.

The purpose of the role play was to increase self-efﬁcacy related to active
participation in health care, which was expected to lead to increased active
behavior in health care encounters. Although mastery experiences have been
found to be the most effective in increasing self-efﬁcacy, practical limitations
required that the intervention use a simulated rather than actual health care
encounter.

The hypothesized impact of the intervention on health care self-efﬁcacy and
participation behavior is depicted in Figure 4. Because mastery experiences have
been found to be the strongest sources of self-efﬁcacy information, it was
hypothesized that intervention C, which included experience in a simulated
health care encounter, would signiﬁcantly increase self-efﬁcacy. The brochure
alone and the lecture/ discussion format were not expected to signiﬁcantly
increase self-efﬁcacy related to the health care system. The relationship between

the hypotheses and the proposed model are described below.

 

 
     
 

39

 
  

   
    
   
   
 
 

EiguLeA.
Attitude Toward the Behavior
E: . . Intention
E Subjective to
a) Norms Partici ate
2 (31)
22
.5-
Behavior Outcome
_4

 

 

 

Perceived Behavioral

Control (7)

40
Hypotheses

Hypothesis 1

W

Post-intervention, participants in Program C will have greater conﬁdence in
their ability to participate actively in health care decisions which will result in
greater intention to participate, increased participation behavior and greater
satisfaction with decision compared to participants in programs A and B.
SummaryoLBolafedLiteratoro

WM. In the Health Care Self-Efﬁcacy
model (ﬁgure 4), Intervention C (1) was hypothesized to impact the self-efficacy
aspect of perceived behavioral control (2) through participant modeling and
persuasion. Since participant modeling has previously been found to increase
self-efﬁcacy among snake phobics (Bandura, Adams, 8: Beyer 1977), it was
hypothesized that the role play exercise would increase participants’ health care
self-efﬁcacy in a similar way. The brochure alone and the lecture/ discussion
format intervention were not expected in signiﬁcantly inﬂuence self-efﬁcacy.

WW. It was hypothesized that the
increase in self-efﬁcacy (2) induced by Intervention C (1) would result in
increased intention to actively participate in health care decisions (3). The impact
of self-efﬁcacy on behavioral intention has been well documented. For example,
self-efﬁcacy was found to be predictive of intention to perform dental hygiene
behaviors (McCaul, O’Neill 8: Glasgow, 1988; Tedesco, Keffer 8: Fleck-
Kandath,1991) and intention to participate in physical activity (Dzewaltowski,
Noble 8: Shaw, 1990). In previous studies, the correlation between self-efﬁcacy
and behavioral intention was found to be substantial; .89 for teeth-brushing
(Tedesco, Keffer 8: Fleck-Kandath,1991) and .81 for physical activity
(Dzewaltowski, Noble 8: Shaw, 1990).

41
Wm. Because previous studies have found

that behavioral intention was a signiﬁcant predictor of behavior, it was
hypothesized that intention to participate actively in health care (3) would
predict self-reported participation in health care (4). Intention has been found to
predict class attendance among college students (Ajzen 8: Madden, 1986),
brushing and ﬂossing among dental patients (Tedesco, Keffer 8: Fleck-Kandath,
1991), physical activity participation among undergraduate college students
(Dzewaltowski, Noble 8: Shaw, 1990), and testicular self-examination among
college students (Brubaker 8: Wickersham, 1990).

ii fa a o a . The outcome of

 

active participation in health care (4) was hypothesized to be satisfaction with
decision (5). Parents who recalled being involved in the decision making
regarding their hospitalized children were signiﬁcantly more likely to be
satisﬁed with their care than were parents who did not recall involvement
(Chamey, 1990). It was therefore anticipated that patients who actively
participated in their health care would be more satisﬁed with the health care
decision made.
RolaﬁonJoModel

Hypothesis 1 regarded the relationship between intervention C (1), self-
efﬁcacy (2), intention to participate (3), participation in health care (behavior) (4)
and satisfaction with health care decision (outcome) (5).

W. The health care self-efﬁcacy model shown in
Figure 4 shows the hypothesized relationship between the intervention (1) and
self-efﬁcacy (2) suggested by the literature. The intervention was hypothesized to
increase self-efficacy through participant modeling and persuasion. In the model,
the factors self-efﬁcacy (2) and perceived barriers (6) paralleled Shifter 8: Ajzen’s
(1985) construct ”perceived behavioral control (7).” Self-efﬁcacy expectations

42
represented the internally controlled aspects of perceived behavioral control and

perceived barriers represented the externally controlled aspects.

W. The model also showed the expected
relationship between self-efﬁcacy expectations (2) and behavioral intention (3). It
was hypothesized that health care self-efﬁcacy would be positively related to
intention to participate in health care and that intention would predict
participation behavior.

WW. As shown in Figure 4, it was anticipated
that the decision support intervention ( 1) would signiﬁcantly increase health care
self-efﬁcacy expectations (2) and that increased self-efﬁcacy would result in
increased intention to participate in health care (3). Intention to participate was
hypothesized to predict self-report of actual participation in health care (4) which
would result in increased satisfaction with the health care decision (5).

W. In the health care self-efficacy
model, satisfaction with decision (5) was the desired outcome stemming from
patient participation behavior. It was hypothesized that self-reported
participation in health care (4) would be signiﬁcantly associated with increased
satisfaction with health care decision (5).

Hypothesis 2
Statemmmﬁﬂmolhosia

Outcome Expectations, Subjective Norms and Health Care Self-Efﬁcacy will
signiﬁcantly predict Intention to Participate in health care decisions.
SummanLoLRelatedliteramre

According to the theory of planned behavior, behavioral intention was
determined by three factors, Attitude Toward a Behavior, Subjective Norm and
Perceived Behavioral Control (Shifter 8: Ajzen, 1985). Attitude Toward a

Behavior had two determinants, behavioral beliefs and outcome evaluation.

43
Behavioral beliefs was deﬁned as the likelihood that performing a behavior

would result in a given outcome and outcome evaluation was deﬁned as the
positive or negative value placed on a given outcome (Ajzen 8: Fishbein, 1980).
These two concepts were identical to the concepts Outcome Expectation
(Bandura, 1977) and Outcome Value (Teasdale, 1978) in self-efﬁcacy theory
(Tedesco, Keffer 8: Fleck-Kandath, 1991).

Wong. Outcome expectations were deﬁned as the belief that
a given behavior would lead to a speciﬁc outcome (Bandura, 1977). According to
the Theory of Planned Behavior, the determinants of Attitude Toward the
Behavior were behavioral beliefs (subjective probability that performing the
behavior will lead to a given outcome) and the outcome’s subjective value (Ajzen
8: Madden, 1986). Each behavioral belief and outcome value were multiplied
together and the products summed to produce Attitude Toward the Behavior
(Ajzen 8: Madden, 1986). Behavioral beliefs were identical to Bandura’s (1977)
Outcome Expectations. Teasdale (1978) clariﬁed the nature of outcome
expectation as the sum of the probability of each outcome times the value of that
outcome. Solomon and Annis (1990) used Teasdale’s (1978) and Ajzen 8:
Madden’s (1986) method of multiplying outcome expectations by their value
when developing a measure of outcome expectancy related to drinking behavior.

Although outcome expectations, alone or in combination with self-efﬁcacy
expectations, have not been found to predict behavior as well as self-efﬁcacy
expectations alone (Solomon 8: Annis, 1990; Manning 8: Wright, 1983); they have
been found to contribute signiﬁcantly to prediction of behavioral intention
(Ajzen 8: Madden, 1986; Manstead, Profﬁtt 8: Smart, 1983; Maddux, Sherer 8:
Rogers, 1982; Crawley, 1990; Netemeyer 8: Burton, 1990; Shifter 8: Ajzen, 71985)‘
Maddux, Sherer and Rogers (1982) found that outcome expectancy was

significantly related to behavioral intention to use a novel communication

44
technique and that outcome expectancy was correlated with self-efﬁcacy.

thjggﬁxeﬂonn. Subjective norm had two determinants; normative beliefs
and strength of normative beliefs. A normative belief was the perceived
likelihood that other people would approve or disapprove of a given behavior.
Subjective norm consisted of the product of each normative belief and its
strength (motivation to comply with the other’s belief) (Ajzen 8: Madden, 1986).
Subjective norm has been found to signiﬁcantly contribute to prediction of
behavioral intention for some behaviors but not for others (Ajzen 8: Madden,
1986; Beale 8: Manstead, 1991 ; Brubaker 8: Wickersham, 1990;McCau1, O’Neill 8:
Glasgow, 1988).

W. Attitude toward behavior,
subjective norm and self-efﬁcacy have been found to signiﬁcantly predict dental
hygiene behavior intentions (R2 = .31) (McCaul, O’Neill 8: Glasgow, 1988),
intention to practice testicular self-examination (Brubaker 8: Wickersham, 1990),
intention to exercise (Gatch 8: Kendzierski, 1990), intention to lose weight (Shifter
8: Ajzen, 1985), and intention to attend college class lectures (Ajzen 8: Madden,
1986). ‘

RolaﬁonfoModol

Hypothesis 2 regarded the prediction of intention to participate in health care
(3) from the factors health care self-efﬁcacy (2), outcome expectations (8) and
subjective norm (9). As shown in the model of patient participation behavior,
intention to participate was hypothesized to be determined by self-efﬁcacy
expectations, outcome expectations and subjective norms. It was hypothesized
that all three factors would contribute signiﬁcantly to prediction of intention. The
participant modeling exercise in intervention C (1) targeted self-efﬁcacy (2) as
shown in the model, but was not part of the model itself. The model described

the hypothesized relationships among self-efﬁcacy, outcome expectations,

45
subjective norms, perceived barriers and intention, behavior and satisfaction.

Hypothesis 3

W

Health Care Self-Efﬁcacy and Intention to Participate will signiﬁcantly
predict Participation in Health Care.
WWW

Self-efﬁcacy represented the internal control aspect of perceived behavioral
control (Shifter 8: Ajzen, 1985) and has been found to signiﬁcantly improve
prediction of behavior beyond behavioral intention (Ajzen 8: Madden, 1986;
Shifter 8: Ajzen, 1985; Tedesco, Keffer 8: Fleck-Kandath, 1991). Although self-
efﬁcacy may inﬂuence behavior through its impact on behavioral intention, it has
been found to be more highly correlated with some behaviors than behavioral
intention suggesting a direct, unmediated relationship to behavior as well. For
example, self-efﬁcacy was found to correlate .41 with exercise behavior, while
intention was correlated .32 (Dzewaltowski, Noble 8: Shaw, 1990). The degree to
which self-efﬁcacy (perceived behavioral control) has a direct inﬂuence on
behavior has been hypothesized to depend on the degree of actual control
present in a given situation. Perceived behavioral control has been found to
improve prediction of behavior in situations in which individuals lack complete
behavioral control and have experience with the behavior in question (Beale 8:
Manstead, 1991 ; Ajzen 8: Madden, 1986). Since active participation in health care
has been found to be a situation in which some individuals perceive a lack of
internal control (self-efﬁcacy) (Woodward 8: Wallston, 1987), it was anticipated
that perceived behavioral control would exert a direct inﬂuence on participation
behavior as well as impact intention to participate. According to the theory of
planned behavior, perceived control represents actual behavioral control and

experience results in a more accurate prediction of actual control (Ajzen 8:

46
Madden, 1986). Since the health care encounter was a situation with which most

midlife women had experience, it was expected that their prediction of actual
control would be highly accurate and therefore perceived control would be
representative of the actual control available in that situation and directly related
to behavior.
WM

As shown in Figure 4, it was hypothesized that there would be a direct
relationship between self-efﬁcacy (2) and participation behavior (4) not
accounted for by behavioral intention (3) alone. It was therefore hypothesized
that intention and self-efﬁcacy together would signiﬁcantly predict participation
in health care decisions (4). Intervention C was hypothesized to exert an

inﬂuence on behavior only via self-efﬁcacy.

Hypothesis 4
Siafcmentofﬂmoﬂiesis
Self-Efﬁcacy Expectations and Outcome Expectations will be signiﬁcantly
correlated.
SummancofRelamiLm

According to theory, attitude toward the behavior and perceived behavioral
control would be expected to be correlated (Ajzen 8: Madden, 1986) however,
some studies have found a correlation between the two factors (Dzewaltowski,
Noble 8: Shaw, 1990), while others have not (Shifter 8: Ajzen, 1985). Social
cognitive theory maintained that while self-efﬁcacy expectations and outcome
expectations were distinct constructs, self-efﬁcacy expectations and outcome
expectations would be correlated except when no level of competence could
produce desired outcomes (Bandura, 1988). Self-efficacy has been found to be
signiﬁcantly correlated with outcome expectations in several studies (Manning 8:
Wright, 1983; Dzewaltowski, Noble 8: Shaw, 1990) supporting the theory. It has

47
also been shown that self-efﬁcacy expectations and outcome expectations can be

manipulated independently (Davis 8: Yates, 1982) and that outcome expectations
contributed separately to self-efﬁcacy’s prediction of behavior (Dzewaltowski,
Noble 8: Shaw, 1990; Manning 8: Wright, 1983) supporting the distinction
between the factors.
R ' t l

The patient participation model (figure 4) paralleled the theory of planned
behavior and health care self-efﬁcacy (2) was therefore expected to be correlated
with outcome expectations (8) since it was anticipated that patients would
perceive a relationship (positive or negative) between patient participation
behavior and quality of medical care. Intervention C was expected to inﬂuence
self-efﬁcacy expectations directly.

Hypothesis 5

Siatemontotﬂxoothoaii

Perceived Barriers and Health Care Self-Efﬁcacy will contribute signiﬁcantly
to prediction of Intention to Participate in Health Care Decisions and
participation behavior.
SummancoLRelatedLiteramre

According to the theory of planned behavior, perceived behavioral control
included the elements of internal, personal control (skills and abilities) and
control by external sources (time, opportunity, and other people) (Ajzen 8:
Madden, 1986). The internal control factor was conceptually identical to self-
efﬁcacy (Ajzen 8: Madden, 1986; Dzewaltowski, Noble 8: Shaw, 1990; Catch 8:
Kendzierski, 1990; Brubaker 8: Wickersham, 1990) and the external control factor
was similar to the concept ”perceived barriers” in the Health Belief Model (Beale
8: Manstead, 1991). Combined, self-efﬁcacy and perceived barriers included both

the internal and external control aspects of Perceived Behavioral Control.

48
In the Theory of Planned Behavior, both internal and external aspects of

perceived behavioral control contributed to prediction of intention and behavior.
According to social cognitive theory, perceived self-efﬁcacy was the determining

factor in intention and behavior:

”Human behavior is, of course governed largely by perceptions Of
personal efﬁcacy and social environments rather than simply by their
Objective properties. Thus, individuals who believe themselves to be
inefﬁcacious are likely to effect limited change even in environments
that provide many opportunities. Conversely, those who have a ﬁrm
belief in their efﬁcacy, through ingenuity and perseverance, ﬁgure
out ways of exercising some measure of control in environments
containing limited opportunities and many constraints,” (Bandura 8:
Wood, 1989; p. 806).

Recent research has suggested that self-efﬁcacy alone was a better predictor
of intention and behavior than perceived behavioral control which incorporated
both internal and external control factors. Dzewaltowski, Noble and Shaw (1990)
found that self-efﬁcacy was more highly correlated with behavioral intention (r =
.81) and exercise behavior (r = .41) than was perceived behavioral control (r = .40
and .10, respectively).

Perceived behavioral control has however, been found to signiﬁcantly
contribute to prediction of behavioral intention and behavior in several studies
(Dzewaltowski, Noble 8: Shaw, 1990; Beale 8: Manstead, 1991 ; Gatch 8:
Kendzierski, 1990; Netemeyer 8: Burton, 1990; Shifter 8: Ajzen, 1985; Ajzen 8:
Madden, 1986). According to the theory of planned behavior, perceived control
predicted behavior because it was a substitute for a measure of actual control
(Ajzen 8: Madden, 1986). Since it was actual behavioral control which was
believed to predict behavior in the theory of planned behavior (Ajzen 8: Madden,

1986), external control factors may add to prediction of behavior.

49

BolationjoModol

As shown in the model, it was hypothesized that perceived barriers to
participation (6) would contribute signiﬁcantly to prediction of intention to
participate (3) and participation behavior (4). Intervention C was expected to
directly inﬂuence self-efﬁcacy (2) but not perceived barriers (6) since external, or
system-level barriers were outside the realm of the intervention.

Table 2 shows the relationship between the model, hypotheses, instruments

and analyses. The numbers in parentheses refer to Figure 4.

50

 

 

 

 

1:1th
protheees
Moots mm: 1 ___axs_sAn 1 i
1. Intervention C (1) 0Health Care Self- 1 0Multiple Regression
will increase self- Efﬁcacy Scale 1 0Path Analysis
efﬁcacy (2) which OBehavioral
will inﬂuence Intentions Scale
Intention (3) 8: 0Behavior Self-
Participation (4) Report Scale
which will impact OSatisfaction with
Satisfaction (5) Decision Scale
2. Outcome OOutcome OPath Analysis
Expectations (8), Expectations Scale
Self-Efﬁcacy 0Health Care Self-
Expectations (2) 8: Efﬁcacy Scale
Subjective Norms OSubjective Norms
(9) will predict Scale
Intention (3) OBehavioral
Intentions Scale
3. Self-Efﬁcacy 0Health Care Self- OPath Analysis
Expectations (1) 8: Efﬁcacy Scale
Intention (3) will OBehavioral
predict ' Intentions Scale
Participation OBehavior Self-
Behavior (4) Report Scale

 

 

51

 

W
Hmothegs
Hypothesis Regiments Analgesic
4. Self-Efﬁcacy (2) 8: 0Health Care Self- OPearson Correlation
Outcome Efﬁcacy Scale Coefﬁcient
Expectations (8) OOutcome
will be Expectations Scale
signiﬁcantly
correlated
5. Perceived Barriers OPerceived Barriers 0 Path Analysis

(6) 8: Health Care
Self-Efﬁcacy (2)
will predict
Intention to
Participate (3) and
Participation
Behavior (4)

 

 

Scale
0Health Care Self-
Efﬁcacy Scale
CBehavioral
Intentions Scale
OBehavior Self-
Report Scale

 

 

 

CHAPTER 2
Method

ms

The experimental interventions were conducted by the Decision Making in
Menopause Study in classrooms at Michigan State University. Classrooms were
equipped with desks and overhead projectors.

ar P ' i

Went. Three hundred seventy nine women were recruited by the
Decision Making in Menopause Study through advertisements in the city
newspaper, stories in community newspapers and radio and television
announcements. Advertisements were directed at women over 40 who wanted to
learn more about menopause and hormone replacement therapy. Most responses
came from advertisements in the city newspaper. The newspaper in which the
advertisements appeared had a circulation of 322,100 adults, 167,300 of whom
were women. By age, the newspaper circulation was 90,700 for age 35—49; 51,200
for age 50—64; and 39,900 for age 65 and older. Since approximately 52% of the
circulation was women, it was estimated that the advertisements potentially
reached about 42,000 women between the ages of 45 and 64. The total population
of women age 45 and over in the tri-county area was 60,348 in 1990. Interested
persons were directed to call the study for information. See Appendix A for
advertisement. Women who called were told that the purpose of the study was
to test three educational programs to help women in their decision making
related to menopause and hormone replacement therapy. They were informed
that participation required attendance at up to three class sessions, completion of
monthly calendars and attendance at follow-up sessions to complete

questionnaires. Random assignment was described and explained as similar to a

52

53
lottery to ensure the women understood that they would have an equal chance of

being in one of the two classroom groups or the brochure group. The three
programs were described and participants who agreed to random assignment
were registered for the study.

W. Random assignment was accomplished by the author
or assistants following a written protocol and using a table of random numbers.
Registration forms were placed in alphabetical order or order of registration and
a one-digit number was read from the table for each registration form.
Participants were assigned to Group A if the number on the table was a 1, 2 or 3;
Group B if the number was a 4, 5 or 6 and Group C if the number was a 7, 8 or 9.
If the number read was a zero the next number in the column was used.
Participants were informed of their group assignment at, not prior to, the ﬁrst
data collection session to prevent selective attrition based on group assignment.

Am. Three hundred seventy nine (379) women registered for the study.
Three hundred (300) women attended session 1. Of those who registered but
failed to attend session I, approximately half canceled by phone prior to the
program. The most frequent reasons given for cancellation were scheduling
conﬂicts or illness. Of the women who attended session I, 252 (84%) completed
all sessions and provided Time 1 and Time 2 data. Four participants were
eliminated from analyses because they were not randomly assigned to group,
resulting in a total sample size of 248 women at the conclusion of the
intervention program. Distribution across experimental groups was as follows:
Group A=87; Group B=78; Group C=83. Two hundred and two participants
returned two-month follow-up questionnaires by mail (53% of those who
registered; 67% of those who attended session I; 80% of those who attended
session III). Distribution across experimental groups for the 2-month follow-up

was: Group A=76; Group B=62; Group C=64. Of the 202 who completed the

54
follow-up questionnaires, 67 had visited their health care providers and

provided behavior self-report data. Distribution across experimental groups was:
Group A=22; Group B=24; Group C=21. Table 3 shows attrition.

Attrition was monitored by comparing the demographic and other data from
participants who terminated participation in the study with data from
participants who continued in the study. Chi-square’s or Pearson correlation
coefﬁcients were computed comparing participants who dropped between
session I, session III and follow-up. There were no differences in class
membership, history of hysterectomy or oophorectomy, time since last menstrual
period (menopausal status), education, or self-efﬁcacy at time 1. There was
selective attrition by experimental group, with more Group A participants than
Group B or Group C participants continuing participation. It is hypothesized that
this difference was due in part to less demand made on Group A participants.
Group A participants needed only to attend sessions I and III to complete
questionnaires, while Group B and Group C participants had to attend session II
as well. Group B and Group C participants who attended sessions I and III but
failed to attend the second session were counted as attrition since they did not
receive the full treatment. Of those who remained in the study at session III,
there was no difference in attrition by experimental group at follow-up. There
was no signiﬁcant attrition by experimental group between session III and
follow-up.

There was selective attrition by race, with more non-white participants
dropping from the study between session I and session 111 than would be
expected due to chance alone. The intervention may have been less useful for
non-white women because of the small number participating. Attrition by race
was constant across experimental groups and attrition between session 1]] and

follow-up was not signiﬁcantly different by race.

55

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 3
A a r
Group A Group B Group C Can. TOTAL
Registered Wave I 51 52 53 156
Wave II 62 64 75 22 223
Total 113 116 128 22 379
Session I Wave I 41 (80%) 41 (79%) 37 (70%) 119 (76%)
(0/0 Of 0 o o o
Registr.) Wave II 55 (89 A.) 57 (89 A.) 69 (92 A.) 181 (81 A.)
Total 96 (85%) 98 (84%) 106 (83%) 300 (79%)
Session II Wave I 38 (93%) 33 (89%) 71 (91%)
(‘70 0f 0 o 0
Session I) Wave II 50 (88 A.) 56 (81 A.) 106 (84 A.)
Total 88 (90%) 89 (84%) 177 (87%)
Session 11] Wave I 39 (95%) 34 (83%) 29 (78%) 102 (86%)
(0/0 Of 0 o o 0
Session 1) Wave II 49 (89 A.) 47 (82 A.) 54 (78 A.) 150 (83 A.)
Total 88 (92%) 81 (83%) 83 (78%) 252 (84%)
Follow-up Wave I 33 (80%) 29 (71%) 26 (70%) 88 (74%)
(2 mos '1 Wave II 43 (78%) 33 (58%) 38 (55%) 114 (63%)
Total 76 (67%) 62 (53%) 64 (50%) 202 (53%)
(% Reg)
(% of (79%) (63%) (60%) (67%)
Sess. 1)
(% of (86%) (77%) (77%) (80%)

 

 

Sess. 3)

 

 

 

 

 

 

 

56
Experience with menopausal symptoms affected attrition during the

intervention programs. Participants who had never experienced symptoms were
more likely to complete the intervention. Those who were not currently
experiencing symptoms but had in the past were less likely to complete the
intervention. There was no difference in return of follow-up questionnaires by
experience with symptoms although participants who had never experienced
menopausal symptoms were less likely to provide behavior self-report data (i.e.
they were less likely to have visited a health care provider during the follow-up
period). Difference in attrition during the intervention may have been due to the
decreased relevance the intervention might have for women whose symptoms
were in the past rather than current or anticipated. There was no difference in
experience with menopausal symptoms at time 1 across experimental groups.

In order to ensure adequate statistical power at follow-up (Time 2 and Time

3), twice as many participants were recruited at Time 1 as were needed at Time 3.

See W515 below.
WW. Participants were two hundred forty eight

women who were at least 40 years old, who were participating in a study of
menopause education programs. The women were recruited by the Michigan
State University Decision Making in Menopause Study through media requests
for participants (Rothert, 1990). Selected demographic characteristics of the
sample and the Tri-County Area from which the sample was drawn are shown in
Table 4. Comparable data for the Tri-County Area was not available for
employment status. Sixty-three percent of the sample was employed full-time;
20% part-time, 5% were retired; 11% were not employed and 2% listed ”other" '
for employment. The sample was primarily White, employed full- or part-time,
with incomes of at least $50,000 per year. Health history and access to care are

shown in Table 5.

 

 

Table 4
D m 1 . C1 . l' E S 1
Variable N % Tri-County Area
Age
40-45 92 37.1%
46-50 114 46.0%
51-55 34 13.7%
56-60 6 2.4%
61-65 2 .8%
Race (Females age 40—64)
African-American 9 3.6% 5.3%
Hispanic 4 1.6% (2.2%)
American Indian 0 0.0% 0.8%
White 234 94.40/0 91 .50/0
Asian / Paciﬁc Islander 0 0.0% 1.4%
Other 1 .4% 1.0%
Household Income
Under $14,999 10 4.0% 21.0%
$15,000-$49,999 100 40.3% 53.0%
$50,000-$99,999 113 45.6% 23.0%
$100,000 and over 24 9.7% 3.0%
Missing 1 .4% --
Education (Persons age 25+)
Less than 12 years 2 .8% 16.0%
High School Graduate 24 9.7% 28.0%
Greater than 12 years but no degree 64 25.8% 24.0%
Technical / Community College degree 31 12.5% 8.0%
Bachelors Degree 59 23.8% 14.0%
Graduate/ Professional Degree 63 25.4% 10.0%
Other 4 1.6% --
Missing 1 .4% --
Marital Status (Females age 45+)
Ever Married 231 93.1% 96.0%
Never Married 17 6.9% 4.0%

 

58

 

 

Table 5
HeelthﬂstomandAoesaJsLMedisalLare
variable N 0/0
Last Natural Menstrual Period
Still Have Regular Periods 143 57.7%
Less than 3 months ago 20 8.1%
3 to 12 Months ago 21 8.5%
More than 12 months ago 59 23.8%
Not Sure 5 2.0%
Experience With Menopausal Symptoms
Have Never Experienced Symptoms 43 17.3%
Currently Experiencing Symptoms 123 49.6%
Have Experienced Symptoms in the Past 22 8.9%
Not Sure 60 24.2%
Hysterectomy
YES 36 14.50/0
N o 212 85.5%
Oophorectomy
Both Ovaries Removed 17 6.9%
One Ovary Removed 8 3.2%
No 223 89.9%
History of Cancer
Breast Cancer 4 1.6%
Endometrial Cancer 4 1.6%
Other Cancer 5 2.0%
None 235 94.8%
Source of Payment for Medications
Completely Out-of-Pocket 31 12.5%
Partly Out-of-Pocket/ Partly Other Source 187 75.4%
Completely Paid by Other Source 30 12.1%

 

59

Beecerehllesigo

This was a longitudinal experimental study with two experimental groups
and a standard treatment control group. Participants were randomly assigned to
one of three educational programs on menopause. Program A served as the
control group with a written brochure as the standard intervention. Program B,
1 controlling for time and attention as well as information used a lecture and
discussion format. Program C utilized an innovative decision support
intervention developed by Rothert (1990) which consisted of exercises and role
play to communicate the information. The independent variable was education
method and the major dependent variables related to consumer participation in
health care decisions were Self-efficacy; Perceived Barriers; Subjective Norm;
Outcome Expectations; Behavioral Intention and Participation Behavior.
Satisfaction with Decision was an outcome variable. Table 6 shows the schedule
of administration for each measure.
W

Power was calculated for a two-tail F-test because although the major interest
was in a positive relationship, a negative relationship between the intervention
and self-efﬁcacy was potentially possible. For an analysis of variance, 21
participants per group would produce a statistical power of .80 for a large effect
magnitude of .40 (f = .40) at a2 = .05, requiring a ﬁnal total N of 63 (Cohen, 1988).
It was also determined that an N of 63 would provide a power of 80 or greater
for a multiple regression analysis with fewer than 16 independent variables.
Power was also calculated for a medium effect size. It was found that an N of 156
(53 per group) would be required to detect a medium effect size (f = .25) for
analysis of variance and a multiple regression analysis with fewer than 13
independent variables. Power ([3) for detecting a small effect size (f = .02) with

=248 was less than 29. See Appendix B.

Health Care Self-
Efﬁcacy Scale

 

 

60

ul

Iimel

Health Care Self-
Efﬁcacy Scale

Perceived Barriers
Scale

Subjective Norms
and Motivation to
Comply Instrument

Outcome
Expectations and
Outcome Values
Scale

Intention to
Participate in Health
Care Decisions Item

 

 

Iimel

Health Care Self-
Efﬁcacy Scale

Behavior Self-Report
Instrument

Satisfaction with
Decision Scale

61
Proceﬂre

1413mm

Program instructOrs were members of the Decision Making in Menopause
Study research team. Two instructors team-taught each intervention session for
programs B and C and attended the data collection sessions for program A.
Instructors were organized into clinician / non-clinician teams. Each team
member practiced the appropriate portion of the intervention in front of an
audience prior to implementation of the intervention. Timing, content and
delivery were tested in an intervention pilot (see below) which was attended by
all instructors to ensure consistency across instructors. There was minor ’
disruption of instructor teams due to absenteeism. Absent members of an
instructor team were replaced by a practiced instructor from another team.
Clinicians were replaced by team members trained as clinicians and non-
clinicians were replaced by non-clinicians. One member of the research team was
trained in both clinician and non-clinician aspects of the intervention and served
as substitute where possible. There were 2 instances of replaced instructors for
Programs B or C out of 24 sessions. Table 7 shows the balancing of instructor
team across group and wave.

As expected, one-way Analysis of Variance failed to indicate differences on
outcome variables by instructor team, supporting the assumption that the
intervention was implemented consistently across instructor teams. Co-
instructors served as manipulation checks for each other, assuring that each

intervention session was implemented as designed.

62

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 2
r h 1
Instructor Teams
Experimental Group
Wave Class
A B C
1 Y Z X
1
2 Z X Y
3 X Y Z
2
4 Y Z X
I | I' 2.! |

Prior to implementation of the intervention study, outlines of the content and
format for each session of Program B and Program C were developed. The
handouts and intervention instruments for Program C (see Inelntemenﬂnn,
below) were prepared and pilot-tested on a convenience sample of 21 women
age 40 and over recruited from among the faculty and staff of a large rnidwestern
university (see InstnnnenLBith section for information on recruitment).
Participants were contacted by phone regarding participation in the intervention
pilot. Those who agreed to attend all three sessions were registered. Pilot study
participants received a buffet luncheon and a draft copy of the brochure
developed for the intervention.

The intervention pilot for Program C was held during the lunch hour for
three consecutive days in a private dining room at a convenient campus location.
All three intervention sessions were video recorded. The video recordings were

transcribed and used in modifying the intervention outlines. Questions

63
generated by pilot participants during the question and answer period in the

intervention were compiled and the answers were either incorporated into the
intervention or anticipated in future question and answer periods. Answers were
generated so that they would be consistent across instructors.

Feedback from the intervention pilot indicated that participants liked the
question and answer period and the informational content. Participants felt that
adequate time for the program and clarity of the intervention instruments could
’ be improved. It was also suggested that audio-visual aids be used during the
lecture portions of the program.

Based on the intervention pilot, the program content was modified and
reorganized. Some aspects of the intervention, such as inclusion of a bar chart
showing women’s decisions related to hormone replacement therapy, were
eliminated. Instead this information was conveyed brieﬂy in a descriptive
format based on feedback that the bar chart was confusing. Other content was
reordered to improve understandability and use of time. Over-head
transparencies were added to the lecture portions of the programs and the
intervention instruments were modiﬁed for understandability.

Ihelntemenﬁon

mnenjmehmelntemenﬁen. The written brochure, described in the section
W provided information on menopause, hormone
replacement therapy and communication with health care professionals (Rothert,
et al., 1992). The brochure comprised the entire intervention for participants in
Program A, and provided a written adjunct to the lectures for Programs B and C.
See Appendix C for outline.

W. The second program (Program B) consisted
of three 1 1/ 2 hour sessions usinga lecture format with overhead transparencies

as visual aids. Participants in Program B received copies of the brochure to use

64
during the lectures which paralleled the brochure in content. In addition,

participants were provided with 3X5 cards on which they could write questions
to be read and answered by the instructor during the group question and answer
period. To help participants think about the HRT decision, data was presented
which depicted how different women may weigh various factors related to the
decision. This helped to introduce discussion of personal risk factors and to show
that women vary in their values related to those factors. Information about the
impact of personal risk factors on heart disease, osteoporosis and endometrial
cancer was presented and participants were encouraged to think about their
personal values related to each of the factors important to the HRT decision. The
format and content of Program B was discussed in detail in the sectionlhe
W. See Appendix C for an outline of the content and format of the
three sessions.

W. The third program (Program C) was a
decision support intervention developed to aid women in decision making
related to their menopausal health. See Appendix C for outline.

The ﬁrst session of the program was identical in content to Programs A and
B. Session I of Intervention C was identical in format as well as content to Session
I of Program B. For the initial wave of participants, the lecture on the pros and
cons of HRT was also included in Session 1. Time constraints resulted in
modiﬁcation of the program so that this information was presented in Session H
for Wave II programs.

As in Program B, a question and answer period using 3X5 cards was
included following presentation of the pros and cons of hormone replacement
therapy. As in Program B, data was presented which depicted how different
women may weigh various factors related to the decision. The women then

completed the Personal Risk Assessment, Problem Signiﬁcance Assessment and

65
Relevance Chart. As the women marked risk on their relevance charts, they were

ﬁrst instructed to mark one box for base rate risk for each of the factors to help
them understand the relative risk for each factor. For example, since heart
disease is much more common than endometrial cancer, the heart disease risk
”arm” of the relevance chart was marked comparatively longer than the
endometrial cancer risk ”arm.” Women then marked a speciﬁed additional
number of boxes for their personal risk if it was above population base rate.
Women who had few additional risk factors did not mark any additional boxes
on the chart beyond that for base rate risk. The purpose of the relevance chart
was to help women visually compare their risks and values across the multiple
factors in their decision.

Following the Personal Risk Assessment exercise, the women completed the
Problem Signiﬁcance Assessment. The women scored their Problem Signiﬁcance
Assessments in the following way: A factor rated 30 or below was of low
signiﬁcance, a factor rated greater than 30 but less than 70 was of moderate
signiﬁcance and a factor rated 70 or greater was of high signiﬁcance. This
information was transferred in graphic form to the Relevance Chart. The three
factors for which risk information was relevant (heart disease, osteoporosis and
endometrial cancer) had two ”arms” on the relevance chart; one for risk and one
for signiﬁcance. For these factors the women marked one box on the signiﬁcance
arm for factors they rated as low signiﬁcance, two boxes for factors moderate in
signiﬁcance and all three boxes for factors rated high in signiﬁcance for their
decision making. For the other factors only the signiﬁcance arm was present and
it was marked in the same way.

After completion of the Personal Risk Assessment and Problem Signiﬁcance
Assessment, the women participated in the scenarios exercise to help them think

about how to combine their personal risks, values and base rate risk.

66
Following the scenarios exercise, the women completed the Relevance Chart

and made a decision about whether to consult their health care providers about
taking HRT. Session III included the lecture and role play on communicating
with health professionals. See the section, W for a detailed
description of the intervention.

Women in all three intervention programs completed a personal plan form
during week 3 of the intervention. The women chose up to three new activities to
prevent disease, promote health or manage symptoms. The women were asked
to choose activities in one of four categories; 1) exercise, 2) calcium intake, 3)
HRT or 4) other. They recorded their planned activities on the personal plan
form along with how often they planned to engage in the activity. The women
were asked to keep a monthly calendar indicating when they engaged in each
planned activity.

9 II 'I' E |°

To decrease the likelihood of attrition from the study, a rafﬂe for cash prizes
was offered to participants in each of the three experimental groups. Participants
in Programs B and C were expected to return for the Time 2 data collection
because it was incorporated into the third session of their intervention programs.
Therefore, participants in Programs B and C were only offered a rafﬂe at the
follow-up data collections (6 and 12 months post-intervention for the main
study) and eligibility for the rafﬂe was dependent on having returned all interim
data. Participants in Program A (the brochure only group) were expected to be
less likely to return at Time 2 since they had already received the complete
intervention (the brochure) at Time 1. Therefore, participants in Program A were
told that they would be eligible to participate in a rafﬂe for one of three cash
prizes ($25, $25, or $50) when they returned to complete questionnaires at Time

2. Participants in Program A also received a reminder post-card to attend the

67
Time 2 data collection since there was a one-week gap between Time 1 and Time

2 for this group. Participants in Programs B and C attended intervention sessions
weekly decreasing the likelihood that they would forget to attend the third
session of the program.

Participants received monthly mailings from the study which was expected
to help minimize attrition. Participants who failed to return follow-up
questionnaires received at least one reminder telephone call. Participants who
indicated that they had discarded the questionnaires (usually through
misunderstanding) were mailed replacement instruments. At least three attempts
were made to reach participants by phone if they had not returned their
questionnaires.

Datamation

Prior to commencement of the intervention program, participants completed
written consent forms, results request forms and a series of written
questionnaires. Data collection took approximately 45 minutes during sessions I
and III of the intervention program period. Following data collection,
participants were informed regarding whether they had been randomly assigned
to the brochure-only program or one of the class programs. Participants in
program A then received brochures and were free to return home. Participants in
programs B and C received brochures and the Session I lecture. Participants were
informed that participation in the study was completely voluntary and that they
could terminate participation at any time without penalty. The instructor team
answered questions about the study and what would be expected of the women
during the study period. Participants in program A were informed that they
would be able to participant in a rafﬂe for one of three cash prizes when they
attended the follow-up data collection at Time 2 (see Won). Prior
to data collection, the purpose of the study and instructions for completing the

68
instruments were brieﬂy presented. The Health Care Self-Efﬁcacy Scale was

completed at Time 1 and Time 2. Other instruments completed at Time 2
included the Perceived Barriers Measure, Subjective Norms Scale, Motivation to
Comply scale, Outcome Expectations Scale, Outcome Values scale and Intention
to Participate scale. Participants answered the questionnaires on a computer-
scored answer sheet.

Two months following the intervention, participants were mailed the follow-
up instruments and instructed to complete and return them. The instruments
included the Health Care Self-Efficacy scale, Behavior Self-Report Measure and
Satisfaction with Decision Scale. Participants answered directly on the
questionnaires rather than on a computer-scored answer sheet.

Measms
Instrumentmot

All measurement instruments were pilot tested using a convenience sample
of 120 women age 40 and over recruited from among the faculty and staff of a
large rnidwestem university. A letter explaining the study was sent to all female
faculty and academic staff (n=approximately 1,300) and non-academic staff age
40 and over (n=1,611). Two hundred ﬁfteen women interested in more
information returned a registration form by mail or called the study ofﬁce. Of
these, 185 were scheduled to attend one of the data collection sessions. Thirty
women indicated interest in participating but were unable to attend any of the
sessions and 54 women were excluded because the sessions they registered for
were canceled. Several women canceled prior to the data collection session. The
remaining women were scheduled to attend a session to complete questionnaires
with 120 completing instruments. Instrument pilot participants received
refreshments and a draft copy of the brochure developed for the intervention.
Instrument pilot participants completed a consent form and draft copies of the

69
instruments and provided written and verbal feedback regarding the clarity of

the instructions. The feasibility of using computer-scored answer sheets was also
tested during the instrument pilot. All scales were found to have adequate
internal consistency (a _>_ .70) and computer-scored answer sheets were found to
be feasible with this population. Most participant comments were related to
clarity of items and instructions. Confusing items and instructions were revised
for clarity.

Wm

Healm Care $lf-Efﬁeaey. A modiﬁed version of a measure developed by
Woodward 8: Wallston (1987) was used (Woodward, 1984). The modiﬁed scale

was pilot-tested for clarity and found to have adequate internal consistency (a =
.91; M =63.82; SD = 13.12). The scale (Appendix D) consisted of 5 items
measuring self-efﬁcacy related to desire for control over health care and 3 items
measuring self-efﬁcacy related to desire for health information. The modiﬁed
scale correlated .66 with the original scale. Responses were on a scale from 1 (Not
At All Conﬁdent) to 10 (Extremely Conﬁdent). Responses were coded from 1 to
10. The scale score was calculated by taking the mean of all non-missing items on
the scale resulting in a scale score with a range of 1 to 10.

The Health Care Self-Efﬁcacy scale was conﬁrmed to have adequate internal
consistency with the study sample at Time 1 (or = .91) and Time 2 (a = .93). The
scale means and standard deviations at Time 1, Time 2 and Time 3 are shown in
Appendix E. Figure 5, Figure 6 and Figure 7 show the frequency distributions for
Health Care Self-Efﬁcacy at Time 1, Time 2 and Time 3. A higher scale score
indicated greater conﬁdence in ability to participate actively

in health care decisions.

7O

l-Efﬁa Tim 1Pr intrvn'n

Di

n

 

 

1|]
0
4

35—:

__.___.4|—JA<

O 5
3 2
a»

d — H a) d

0 5
2 1
285$on

._._

mg
mg
new
new
was t
mg .m.
moo m
mN.© NJ
mum coo
m~.m a
mi. m
as am
mom 4...
mum &
m2
m~.~
m5
m3

71

O O
.59! I.',,- ..._!l‘

‘nt '1"-

0
5

5
4

0
4.

5
3

 

0 5 0
3 2 2
55:35

Ra
Ra
m3
m3
max t
w A.
mus M
£8 cows
mg m
m2. m
mg m
m2 m
mom on
m2
m3.
m5
m2

72

_U01-_L

_ -11 °

m D. n I

 

 

mum
mud
m3
mg
m? t
mg m.
mks We
mN.© NJ
mum coo
m
mad a
man. m
. 0
mm a &
m5. m
m3 ax
m2
mum
m5
m3

 

5:25on

 

73
W. The Perceived Barriers scale (AppendixD)

was a 5—item scale which assessed the impact of potential barriers to participation
in decisions such as health care practitioner, payment options, health state, lack
of access to information and conﬂicting information. Responses were on a 5-
point scale with 3 anchors: 1 = Not at All, 3 = Somewhat and 5 = A Great Deal.
The measure also included 2 open-ended items to assess other perceived barriers.
Cronbach’s alpha coefﬁcient was .62. Appendix E shows the scale mean and
standard deviation. A higher scale score indicated greater perceived barriers to
participation in health care decisions. Table 8 shows the corrected item-total
correlation for the scale items. Figure 8 shows the scale frequency distribution.
Table 9 shows the frequency of responses to two open-ended items asking about
additional barriers to participation. Responses to the two items over-lapped and
were combined in the table. In cases where a participant wrote in the same
answer for both questions, only one was counted. Where multiple barriers were
mentioned by a single respondent, all were counted and listed under the heading
’mentioned.’ The heading ’N’ indicates the number of participants who listed a

barrier in that category.

 

Item Corrected Item-Total
Correl
How much does your doctor (or regular health
care practitioner) limit your participation in medical
decisions? .22

How much do your payment Options (such as health
insurance) limit your participation in medical decisions? .29

How much does your health state or personal risk
factors limit your participation in medical decisions? .38

How much does lack of access to accurate information limit
your participation in medical decisions? .57

How much does conflicting medical advice limit your
participation in medical decisions? .45

 

75

Fi

1

Di tributi nBarri r t Parti i ai n

Fr

 

 

76
Table 2

L “1.!" 0

‘ 101‘. 0 '10- - (1‘0 .1 ‘

 

Category

Mentioned

 

HealthLareEroxidmu
0Hard to ﬁnd good provider
0Hard to get appointment with provider
OProvider too busy / in a hurry
GProvider doesn't listen
OProvider arrogance/ personality
OProvider doesn't inform patient
OProvider discourages patient participation
OGeneral distrust of health care providers
0Poor communication among

multiple providers / fragmented care

33

 

0Limitations on provider choices
0Limitations on treatment choices
OLimitations imposed by HMO's
0Lack of insurance/ cost

OGeneral cost

18

16

 

E 11' 'l |'

0Personal lack of time

OPersonal lack of knowledge

OPersonal health problems

0Lack of personal support

OFear/ Embarrassment/ Problem Making Decisions

26

 

OMedicine doesn't have all the answers
0Lack of research
0Conflicting advice

12

11

 

System-Rem

OSystem not prevention-oriented
0Lack of sexual equality

0Lack of access to medical records
OReferral system

 

Miseellanmua

OMakes own decisions
0Doesn't want to participate
OFamily input

OJOb

OUnspeciﬁed

 

 

18

 

18

 

 

77
W. The Outcome Expectations scale was a 7-item

measure which assessed the extent to which the individual expected that active
patient participation behavior would lead to speciﬁc outcomes. An example item
was, ”Patient participation in medical decisions and choices results in: the best
medical care.” The response scale was a 5 point Likert-type scale with 1 =
Extremely Unlikely and 5 = Extremely Likely with a midpoint of 3 = Neither
Likely Nor Unlikely. For each outcome in the scale above, the respondent
indicated its value on a 5 point scale, with 1=Extremely Bad and 5 = Extremely
Good with a midpoint of 3 = Neither Bad Nor Good. The Outcome Expectations
scale was computed by multiplying each expected outcome by its value and
taking the mean of the non-missing values. The scale was found to have
adequate internal consistency (a = .87). A higher scale score reﬂected an
expectation that patient participation in decisions would result in the better
medical care. A lower scale score reﬂected an expectation that patient
participation would result in worse medical care. Table 10 shows the corrected
item-total correlation for the scale items. The corrected item-total correlations
shown are for the product of the expectation and value. Figure 9 shows the scale
frequency distribution. See Appendix D for the instrument. Appendix E shows

the scale mean and standard deviation.

78

 

 

 

- .qdq qqqd qd-q «an. dq-q d-<-
_ _ _ _ H —

0 5 0 5 0
2 1 1

Gama—moi

79

.4- 03‘, 9‘ 1.11., 11‘ I" "Wu-1.0.. 01‘4”!

 

Item Corrected Item-Total
Correlatibn

 

Patient participation in medical decisions and choices results in:

0Medical procedures that are unnecessary for the patient .55
0Medical procedures that are harmful for the patient .56
0The wrong medications for the patient .56
0The right medical tests and examinations for the patient .67
0The right treatment for the patient .79
0The right medical decisions for the patient .79
0The best medical care .72

 

W. The Subjective Norms measure was developed to

assess the perceived opinion of others about patient participation in health care
decisions. The items paralleled the desire to participate item developed by Strull,
et al. (1984). The response scale was 1 = Clinician, 2 = Clinician after considering
patient’s opinion, 3 = clinician and patient together equally, 4 = patient after
considering clinician’s opinion, 5 = patient. The instructions for the scale were,
”Some people think that the decision to take hormone replacement therapy
should be made by the clinician while other people think the decision should be
made by the patient. Indicate how each of the following people think the
decision should be made: Medical practitioner, spouse, relatives, friends, other
(open-ended)” For each of the people listed, the individual answered the

question, ”how much do you want to do what [this person] wants you to?” The

80
5-point response scale had anchors at 1 = Not At All and 5 = Very Much. For

each item, the perceived norm was multiplied by the motivation to comply and
the scale was computed by taking the mean of the non-missing items. Since each
score was weighted by desire to comply, a higher scale score indicated a belief
that signiﬁcant others thought medical decisions should be made with greater
patient participation and that there was high motivation to comply with their
opinions. A lower score indicated a belief that others thought decisions should be
made with less patient participation and that there was lower motivation to
comply. See Appendix D for the measure. Cronbach’s alpha coefﬁcient of
internal consistency was .75. Appendix E shows the scale mean and standard
deviation. Figure 10 shows the scale frequency distribution. Table 11 shows the
corrected item-total correlation for the 4 scale items. The scale item indicated is
the product of the subjective norm and motivation to comply. Table 12 shows
responses to the open-ended item on the Subjective Norms measure. The table
lists signiﬁcant others who inﬂuenced the decision regarding whether or not to

take hormone replacement therapy.

81

 

Item Corrected Item-Total
Correlat_ibn

 

Your doctor (or regular health care practitioner) thinks the
the decision should be made by: .31

Your spouse, partner or signiﬁcant other thinks the decision
should be made by: .63

Your parents, relatives and children think the decision should
be made by: .71

Your friends, peers and classmates think the decision should
be made by: .53

 

 

 

mmmmmmmmmmmmmmmmmmmmmmmm

H H [\ H
Normssweeasezesaeeaaaaa

Scale Score Category Midpoint

Table 12

10‘ aV‘. \0 1-1 10- ‘ f

82

n ‘ n D i ion he‘ll-En- ‘0 .1

 

Source

N

 

Meia

0Media in general
OFeminist authors

0 Health writers
OAuthors in general

 

Self.

 

RespeﬂecLthmgei
OReligion

0Counselor

0 Parents

 

Mmenﬂholﬂeﬁxoeﬁmee.
OFriends

OMother

0 Sisters

OOther women in general

 

IglowIedgeablehers.
0 Researchers
ODoctor

02nd Opinion

OFreinds / Relatives in Health Occupation

 

IIS 't'i

 

 

 

 

83
Wm. This measure listed 8 behaviors related to the

next visit to the health care provider (or = .85). The individual was asked to
indicate the extent to which she intended to do each of the activities. The 5 point
response scale had anchors at 1 = Not At All and 5 = Completely. The scale was
computed by taking the mean of the non-missing items. See Appendix D for the
measure. Appendix B shows the scale mean and standard deviation. A higher
scale score indicated greater intention to participate actively in health care
decisions. Figure 11 shows the scale frequency distribution. Table 13 shows the

corrected item-total correlation for the 8 scale items.

 

wad.
mod
wmﬁ
mad

8

mmd
wmd
mﬁm

‘3. 6'8
N N
ore Category Midpoint

co
cq
N
SC

e

H 28%
we.“
m3
,.wme
..meH

 

120 -

1001

_
0
8

4 1 E — a

w
55853.5

 

85

 

 

Table 13
311.. '0 -. .1 ‘19.: lo. ‘ 01‘ ‘9 ‘ 1- 0 ._ o '11“
Item Corrected Item-Total
Correlation
Intention to:

Carefully consider and understand my lanes related to the
tradeoffs of hormone replacement therapy or other care .44

Gather the information I need to make an informed decision
about hormone replacement therapy or other care .57

Carefully consider my mm related to hormone
replacement therapy or other care. .61

Ask for an extended appointment to discuss my concerns
about my health care .52

Prepare for my visit by thinking about the questions I want to
ask, smiteihemﬁnm and bring the list with me to the visit .67

At the beginningof the visit, tell my doctor / practitioner that
I have questions I would like to discuss. .67

During my visit, ask the questions I had prepared, repeat
the answers to be sure I understand them and aelsjm:

clanﬁeahonineeessam .67

After my visit, review the visit, follow the treatments or
recommendations agreed upon and call my
doctor / practitioner if I have additional questions or
unexpected side effects. .67

 

86
W. The behavior self-report measure included 6

items regarding participation behavior during the last visit to a medical
practitioner (or = .86). The 5-point response scale had anchors at 1 = Not At All
and 5 = Completely. Behaviors assessed included: considered values, personal
risks and gathered information prior. to visit, asked questions during the visit,
and followed recommendations and reported side effects after the visit. A ﬁnal
question was adapted from an item developed by Strull, et al. (1984). The item
asked, ”Which Of the following choices best describes how the decision about
whether or not you would take hormone replacement therapy mentally
made?” Response choices ranged from 1) The clinician made the decision, using
all that’s known about hormone replacement therapy to 5) I made the decision
using all I know and learned about hormone replacement therapy. Response
choices parallel the item on desire to participate in decisions. See Appendix D for
the measure. Appendix E shows the scale mean and standard deviation. A higher
scale score indicated more active participation in health care decisions. Figure 12
shows the scale frequency distribution. Table 14 shows the corrected item-total

correlation for the 8 scale items.

87

I12

Di tri ' nB havi r lf-R al

Fr

 

 

Dame—moi

88

5‘ 1 .0 1‘1-',",001‘t1-1' .1. .. .1!--." . ‘I-Q.

 

Item Corrected Item-Total
Correlation

Carefully considered and understood my 1311.128. related to the
tradeoffs of hormone replacement therapy or other care. .71

Gathered the information I needed to make an informed decision
about hormone replacement therapy or other care. .66

Carefully considered my nemnaLrislga related to hormone
replacement therapy or other care. .51

Asked for an extended appointment to discuss my concerns
about my health care. .41

Prepared for my visit by thinking about the questions I wanted to
ask, W and bringing the list with me

to the visit. .71
At themginnjng of the visit, told my doctor/ practitioner that
I had questions I would like to discuss. .67

During my visit, asked the questions I had prepared, repeated
the answers to be sure I understood them and askedfor

elanﬁcahonifnmessam. .75

After my visit, reviewed the visit, followed the treatments or
recommendations agreed upon and called my
doctor/ practitioner if I had additional questions or
unexpected side effects. .65

 

89
Wen. This 6-item measure, developed by Rothert (1990)

was based on the O’Connor 8: O’Brien-Pallais (1990) model of an effective
decision. Example items included, ”The decision I made was the best decision
possible for me personally,” and ” I am satisﬁed with my decision.” The S-point
Likert-type response scale ranged from 1 = Strongly Disagree to 5 = Strongly
Agree. The scale was computed by taking the mean of the non-missing items (a =
.91). See Appendix D for measure. The Satisfaction with Decision scale was
conﬁrmed to have adequate internal consistency with the study sample at Time 3
(a = .91) Appendix B shows the scale mean and standard deviation. Figure 13
shows the frequency distribution for the scale at time 3.
mlelntercorreleﬁons

Table 15 shows the intercorrelations among the scales. The alpha coefﬁcients
of internal consistency are shown on the diagonal. Correlations corrected for
unreliability are shown above the diagonal (Edwards, 1954). See Appendix F for
calculation formula. Correlations which were signiﬁcant at p < .05 are indicated

by an asterisk.

90

 

 

uu-d «ddwqﬁﬁqd—uuqq—uu-

wmm

kudoswﬁm

8.8
8.8
wow
2.... t
wen .m.
mod N
wm.m NJ
mam coo
m
wwd a
no.“ m
mom &
an m
we; &
«.3
w?
m:

91
1321.115.

Carrelatien Manix

Beh. Beh. Sat Self-E Outc. Subj. Barr. SE SE
Int. /Dec. T2 Exp. Nrm T1 T3

Beh. .85 .44 .15 .17 .06 .07 -.04 .17 .13
Intent. (248)

Beh. .38* .86 .50 .05 —.17 —.02 —.37 .12 .38
(67) (67)

Sat. .13 .44 .91 .45 .22 .11 -.42 .38 .59
w/dec. (184) (67) (184)

Self-Bf .15” .05 .41” .93 .35 —.06 —.60 .73 .74
T 2 (248) (67) (184) (248)

Outcm .05 -.15 .20* .31* .87 .04 —.18 .21 .28
Expect. (248) (67) (184) (248) (248)

Subj. .05 —.02 .09 —.05 .04 .75 .00 .03 .07
Norm (243) (66) (180) (243) (243) (243)

Barr. —.03 -.27* —.32* —.45* -.13 .03 .62 —.43 —.62
(247) (67) (184) (247), (247) (242) (247)

Self-Bf .15" .10 .34" .68" .19" .03 -.32* .91 .67
T 1 (244) (65) (182) (244) (244) (239) (243) (244)

Self-Bf .12 .34" .54" .69" .26" .06 -.47* .61" .93
T3 (184) (67) (180) (184) (184) (180) (184) (182) (184)

"p<.05

CHAPTER 3
Results
The purpose of this research was to determine the impact of three
menopause education programs on health care self-efﬁcacy of midlife women
and to test a model of the impact of self-efﬁcacy on participation in health care
decisions. Data on Self-Efﬁcacy, Outcome Expectations, Subjective Norms,
Intention to Participate and Participation Behavior were collected from 248
participants in a menopause education study and used to explore 5 hypotheses.
The hypotheses regarded the effect of the interventions on self-efﬁcacy and

patient participation behavior related to medical decisions.

Moses.
Emmet

Hypothesis 1 predicted that participants in the Decision Support Intervention
(Program C) would have greater conﬁdence in their ability to participate actively
in health care decisions which would result in greater intention to participate,
increased participation behavior and greater satisfaction with decision. Table 15
shows the correlations between self—efﬁcacy (Time 1 and Time 2), perceived
barriers, subjective norms, outcome expectations, behavioral intention and
behavior self-report.

A multiple regression analysis was conducted with Self-Efﬁcacy at Time 2 as
the dependent variable. Self-Efﬁcacy at Time 1 was entered on the ﬁrst step to
control for preexisting between-participant differences. Experimental group was
contrast-coded (Cohen 8: Cohen, 1983) and the two contrast-coded group
variables were entered on the second step. Contrast 1 compared Experimental
Group C (Decision Support Intervention) with Groups B (Lecture / Discussion)
and A (Brochure Only). Contrast 2 Compared Group B with A. Table 16 shows

92

93
the coding scheme for experimental group. Two variables representing the

interaction between Self-Efﬁcacy at Time 1 and experimental group were entered
on a third step. The nominal variable experimental group was contrast-coded
orthogonally so that the sum of the products of the coding coefﬁcients was zero.
See Appendix G for computation. Table 17 shows the regression coefﬁcients (B)
and their standard errors (s.e.), standardized regression coefﬁcients ([3) and their
standard errors (s.e.), partial correlation coefﬁcients (pri ), Multiple R, and R2-
change for each step in the analysis. The constant represents the average Y
intercept and is used to calculate the intercepts for each of the three regression
equations. An asterisk indicates the F values which were statistically signiﬁcant
at p < .05. Self-Efﬁcacy at Time 1 signiﬁcantly predicted Self-Efﬁcacy at Time 2
(Rz-change = .46, p < .05). The two contrast-coded variables representing the
variance due to experimental group did not signiﬁcantly increase prediction of
Self-Efﬁcacy at Time 2 (Rz-change = .01, p > .05). The interaction between
experimental group and self-efficacy at Time 1 was carried by two variables
(interaction 1 and interaction 2) computed from the products of Time 1 Self-
Efﬁcacy and the two contrast-coded group variables respectively. The interaction
between Time 1 Self-Efﬁcacy and Experimental Group contributed signiﬁcantly
to prediction of Self-Efﬁcacy at Time 2 (R2-change = .02, p < .05). Figure 14 shows
Self-Efficacy Post-Intervention (Time 2) plotted against Self-Efﬁcacy Pre-
Intervention (Time 1) for each experimental group. Table 18 shows the regression
equations for Self-Efﬁcacy post-intervention (Time 2). See appendix G for
calculations. The hypothesis that participants in Program C would have greater
conﬁdence in ability to participate actively in health care decisions than groups A
and B was not supported (Bcontmt 1 = .15; p > .05). There was however, a
signiﬁcant interaction effect between experimental group and self-efﬁcacy at

Time 1.

94
The experimental treatment differentially impacted those for whom self-
efﬁcacy was low pre-intervention compared to those for whom self-efﬁcacy was
high pre-intervention (Rz-chg = .02; p < .05). The signiﬁcant difference was due
to the contrast between Experimental Group B and Group A (B -.13;

p < .05).

interaction 2 =

One-way Analysis Of Variance indicated that Behavioral Intention, Behavior
Self-Report and Satisfaction with Decision were no greater for Experimental
Group C than for Groups A and B. (F = .39, .04, .10, respectively; p > .05). Table 19

shows the means and standard deviations of each variable by experimental

group.

121219.16.
C C 1. ill . 1E . 15 V .1]

 

Contrast Contrast
Experimental Group Variable Variable

 

1 2
Brochure (A) —.5 -1
Lecture/ Discussion (B) —.5 +1
Decision Support (C) +1 0

Total 0 0

 

(-.5)(-1) + (--5)(1) + (1)(0) = 0

 

 

 

95

mo. v 8..

 

 

................................................................................
.IOUUICIIIIIOIIIIII..0...- OIUIOO 0.0... COCO-II. ’0‘. ll... .ICOII IIOICIOOIII: ...............................................
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

.................. *Ow WNH m:.v mh.m AnguchUv

 

 

2.. $3 em.) 63 2... 268883 m

 

 

m. 3.- a3 «7 338.1 H 88885 a

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N m~nw€m>
.2. me an. .2. 63 N6. 83 NS 868:8 m
H Giants!
u.u.”.”.n.umwn ”Hung “mm” mm ”unwwuﬂwﬂnw “mnwwuw”HHHHNNMNNHWH ”w””Huuuuuwwmwmmw 2. .NC. 30¢ no. Amvo mm. ummbﬂoU N
H ugh.
.388 6+. .638 me. we. .63.: mo. 63 me. 33%. 889328 H
mew... We as: “w... m m E 3.8 a 3.8 m 6363, new
. H. . t u N . “4,. .. . .: - 41:1-..
.3. ~. .EE..xmw. .M .m H... . . .1. 3.1“. 1:-.. 3.... .4. .. . H

W

96

W1 lf- fﬁ a T lf- f'c x ' ntal r

Self-Efﬁcacy Post-Intervention (Time 2)

10

 

 

 

Decision Support (C)

Lecture / Discussion (B) /

   

Brochure Only (A)

 

I l

I l r I l I l

3 4 5 6 7 8 9
Self-Efﬁcacy Pre-Intervention (Time 1)

 

97

 

31213.18.
{'14: In a. 1!. 0, 1° -E' on ,0 ,.:“o -41:
For Brochure Only Experimental Group (A): YA = .715(SET1) + 2.595

A

For Lecture/ Discussion Experimental Group (B): YB = .455(SET1) + 4.835

A

For Decision Support Experimental Group (C): YC = .540(SET1) + 3.940

 

 

 

 

 

 

 

 

Experimental Group Mean (s.d.)
. (n)
Vanable
Brochure Lecture / Discuss Decision TOTAL
(A) (B) Suvport (C)

. 4.4 (.64) 4.5 (.63) 4.4 (.55) 4.4 (.60)

“mm“ (87) (78) (83) (248)
. 3.8 (.97) 3.7 (.65) 3.8 (.87) 3.8 (.83)

Bahama (22) (24) (21) (67)
. . 4.0 (.72) 4.0 (.62) 4.1 (.57) 4.1 (.64)

sansfam‘m (68) (54) (60) (182)

 

 

 

 

 

 

98
HypothesesLiandﬁ

Hypotheses 2, 3 and 5 related to testing the model of active patient
participation behavior shown in Figure 4. Hypothesis 2 stated that Self-Efﬁcacy,
Outcome Expectations and Subjective Norms would signiﬁcantly predict
Intention to Participate in health care decisions. Hypothesis 3 stated that Health
Care Self-Efﬁcacy and Intention to Participate would signiﬁcantly predict
Participation Behavior. Hypothesis 5 stated that Perceived Barriers and Health
Care Self-Efﬁcacy would contribute signiﬁcantly to prediction of Intention to
Participate in Health Care Decisions and Participation Behavior.

A path analysis using LISREL VII in SPSSX was conducted to analyze the
extent to which the observed relationships among the variables ﬁt the theoretical
model. The intercorrelations among the variables were corrected for attenuation
due to unreliability of the measures (See Appendix F for calculations) and the
corrected correlation matrix was analyzed by LISREL VII. Table 15 shows the
correlations among the variables analyzed. Missing data were handled by
pairwise deletion. Figure 15 shows the path coefﬁcients for the model as
proposed. An asterisk indicates the paths which were signiﬁcant. x2 with 7
degrees of freedom was 32.71 (p < .000) indicating a poor ﬁt. Goodness of ﬁt
index was .90, adjusted goodness of ﬁt was .58 and root mean square residual
was .171. Signiﬁcant residuals (indicating needed paths) were 3.74 between Self-
Efficacy and Satisfaction with Decision and —2.16 between Barriers and
Satisfaction with Decision. Table 20 shows the t-values for the paths in the model

as proposed.

99

  
 

Outcome
Expectations (x2)

 

 

 

d»
g.

Subjective
Norms (x3)

 

 

 

 

 

(:5? d3 Satisfaction
. Behavior (y2) _> with gapision
3

Behavioral
Intention (y1)

 
  

 

 

 

 

 

 

 

 

  
 

Self-Efficacy (x1)

 

 

714 = '05

 

Barriers (x4)

 

 

*p < .05

100

 

 

 

 

181116.11
Model as Erbpbsed: t—valuee
Intent Behav. Sat Self-Eff Outcm. Subj. Barr.
w / Dec. T2 Expect. Norm
Intent. 0.000 0.000 0.000 1.344 —0.030 0.655 .369
Behav. 4455* 0.000 0.000 -2.066* 0.000 0.000 -4.014*
Sat 0.000 4.509” 0.000 0.000 0.000 0.000 0.000
w / Dec

 

 

 

 

 

 

 

 

 

 

Based on the standardized residuals, t-values and modification indices the
model was modiﬁed for better prediction of behavior and outcome. Figure 16
shows the path coefﬁcients for the revised model. x2 with 10 degrees of freedom
was 14.63 (p = .146) indicating a good ﬁt. Goodness of ﬁt index was .945, adjusted
goodness of ﬁt was .845 and root mean square residual was .107. No non-
signiﬁcant paths were included in the model. Table 21 shows the standardized
residuals for the revised model. Table 22 shows the t-values fOr the revised
model. Hypothesis 2 was not supported. Intention to Participate in Health Care
Decisions was not predicted by Outcome expectations, Subjective Norms and
Self-Efficacy. Although Self-Efﬁcacy was signiﬁcantly correlated with Intention

101

Egocolo
'i eri M 1

 

Outcome
Expectations (x2)

 

 

Behavioral
Intention (y1)

 

   

 

r = .31"

 

 
 

 

 

 

 

Barriers (x4)

 

 

Self-Efﬁcacy (x1)

 

*p < .05

Satisfaction with
Decision (y3)

 

 

102
(SeeTable 15), it was not a significant predictor in the multivariate path analysis.

Neither Outcome Expectations nor Subjective Norms were signiﬁcantly
correlated with Intention. Hypothesis 3 was partially supported. As shown in
Figure 16, Behavioral Intention signiﬁcantly predicted Participation Behavior,
but Self-Efﬁcacy did not. Self-Efﬁcacy was not signiﬁcantly correlated with
Behavior (See Table 15). Hypothesis 5 was partially supported. Although neither
Self-Efficacy nor Barriers signiﬁcantly predicted Behavioral Intention, Barriers
was a signiﬁcant predictor of Participation Behavior (see Figure 16). Self-Efﬁcacy
was not a signiﬁcant predictor of Participation Behavior.

It was expected that Self-Efﬁcacy and Barriers would be highly correlated
and have similar patterns of correlation with other variables because they were
believed to be opposite poles of the same construct, Perceived Behavioral Control
(after Azjen 8: Madden, 1986). This assumption was found to be unsupported.
Although Self-Efﬁcacy and Barriers were moderately correlated (r = -.55;
corrected for attenuation), Barriers but not Self-Efﬁcacy was signiﬁcantly
correlated with Behavior (r = -.38 and .05 respectively—corrected for
attenuation). See Appendix F for calculation of corrected correlation coefﬁcient.

According to the proposed model, outcome of Active Patient Behavior was
satisfaction with the health care decision. The Path analysis indicated that Active
Patient Behavior signiﬁcantly predicted Satisfaction with Decision (see
Figure 16). Unexpectedly, Self-Efﬁcacy was also a predictor of Satisfaction with
Decision (‘Y =.43; p < .05). The indirect effect of Behavioral Intention on
Satisfaction with Decision was .213 and the total effect of Intention on
Satisfaction was .213, indicating that Behavioral Intention had no direct effect on
satisfaction. The variance in Satisfaction with Decision due to Behavioral

Intention was due to the indirect effect of Intention via Behavior.

103

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

12121121.
diz i l f r R i l
Intent. Behav Sat Self- Outcm Subj. Barr.
w/Dec EffT2 Exp Norm
Intent. 0.000
Behav. —0.001 —0.001
Sat. w/ Dec —0.558 —0.280 —0.254
Self-Eff. 1.328 —0.345 -0.190 0.000
Outc. Exp 0.469 0.316 1.459 0.000 0.000
Subj Nrm 0.597 —0.104 1.076 0.000 0.000 0.000
Barriers -0.312 -0.192 -0.473 0.000 0.000 0.000 0.000
IobJelZ
Intent Behav. Sat Self-Eff Outcm. Subj. Barr.
w/ Dec. T2 Expect. Norm
Intent. 0.000 0.000 0.000 0.000 -0.000 0.000 0.000
Behav. 4.392“ 0.000 0.000 0.000 -2.636 0.000 -3.934*
Sat 0.000 4.937" 0.000 4.394" 0.000 0.000 0.000,
w/ Dec

 

*p < .05

 

 

104
Hypbtheeie 4
Hypothesis 4 stated that Self-efﬁcacy expectations and Outcome Expectations

would be correlated. A Pearson Correlation coefﬁcient indicated that Self-
Efﬁcacy at Time 2 was correlated signiﬁcantly (p < .05) with Outcome
Expectations at Time 2 (r = .31) supporting the hypothesis.
AddiﬁoualAnalxoee

It was observed that behavior was sigrniﬁcantly related to self-efﬁcacy at
Time 3 but not at Times 1 or 2. The correlations between behavior and self-
efﬁcacy were signiﬁcantly different for Time 1 and Time 3 (t = -2.0; df = 62; p <
.05); Time 2 and Time 3 (t = -2.5; df = 64; p < .05, 2-tail) but nOt for Time 1 and
Time 2 (t = .48; 62 df; ns). See Appendix H for sample calculations. I

An analysis of variance was conducted to determine whether the correlation
was due to selection bias. Were those who chose to visit their health care
providers higher in self-efﬁcacy at Time 2 than those who chose not to? There
was no signiﬁcant difference (F=.6506) in self-efﬁcacy at Time 2 between women
who visited their health care providers (n=66; m =8.0; s.d.=1.45) and women who
did not visit their health care providers (n=182; m=8.2; s.d.=1.45) during the
follow-up period. This ﬁnding conﬁrmed that the observed relationship between
self-efﬁcacy at time 3 and behavior was not due to selection bias.

The correlation between participation behavior and self-efﬁcacy was
signiﬁcantly geater at Time 3 than at Time 2 and was not due to selection bias.
This suggested that behavior affected self-efﬁcacy. To determine the extent to
which post-behavior self-efﬁcacy was inﬂuenced by experience (behavior), a
multiple regession was performed which controlled for self-efﬁcacy at Times 1
and 2. Table 23 shows the regession coefficients and signiﬁcance values for the
regession equation. The multiple regession analysis showed that participation

behavior explained 9% of the variance in self-efﬁcacy at Time 3 when self-

105
efﬁcacy at Times 1 and 2 were controlled. This supported the speculation that

active participation behavior increased self-efﬁcacy when pre-behavior levels of
self-efﬁcacy were controlled for.

Figure 17 gaphically depicts the relationship between self-efﬁcacy and
behavior. Participation behavior is shown on the vertical y-axis (height) with self-
efﬁcacy at Time 3 plotted on the horizontal x-axis (width) and self-efﬁcacy at
Time 2 plotted on the z—axis represented by depth. The ﬁgure is rotated in space
so that the moderate correlation between behavior and Time 3 self-efﬁcacy is
apparent. Note the elliptical pattern of the points plotted between the x and y
axes. The relationship between behavior and self-efﬁcacy at Time 2 can be seen
by viewing the plot from front to back (depth). As can be seen, there is no pattern
to the depth at which each point is plotted and its height on the vertical axis. At
Time 2, most self-efﬁcacy scores cluster near the high end of the scale, regardless

of behavior.

8. v a...

 

3:36:08

 

 

and 8. .23 mm. . an. m. o3 H. a: B. 8323 m

N 95.
new 8. 83H 8. 6m. .88 Hm. 8H3»... G: on. 885928 N

 

 

 

 

 

 

H 68:.
.3me mm. .3: _E. am. new _3. Tm: mm T38. 883928 H

cwmi «W... m m _HR no.3 n

 

Eula 8.3 m ozaea> 68m

 

 

 

 

 

 

 

 

 

 

 

 

107

.14...

>nm<m wwmma @3339.
M 4 5 3

5
3 2 2

 

10

5

4

Self-Efficacy T3

108
Mum;

Of the 5 hypotheses, 1 was completely supported, 2 were partially supported
and 2 were unsupported by the ﬁndings. Hypothesis 1, that the Decision Support
Intervention would result in greater self-efﬁcacy than the other two
interventions, was not supported. The multiple regression analysis indicated
that, among women low in self-efﬁcacy pre-intervention, the Lecture / Discussion
intervention, not the Decision Support Intervention, resulted in the highest post-
intervention self-efﬁcacy.

Hypothesis 2, that Self-Efficacy, Outcome Expectations and Subjective Norms
would significantly predict Intention to Participation in Health Care, was not
supported. The path analysis indicated that none of the variables measured
significantly predicted Intention.

Hypothesis 3, that Self-Efficacy and Intention would signiﬁcantly predict
Participation Behavior, was partially supported. Behavioral Intention did predict
Participation Behavior, but Self-Efﬁcacy did not.

Hypothesis 4, that Self-Efﬁcacy expectations and Outcome Expectations
would be significantly correlated was completely supported. Self-Efﬁcacy and
Outcome Expectations were significantly, mildly correlated, indicating that they
were related but distinct constructs.

Hypothesis 5, that Self-Efﬁcacy and Perceived Barriers would both contribute
to significantly to prediction of Intention to Participate and Participation
Behavior, was partially supported. Neither Self-Efficacy nor Barriers predicted
Intention, but Barriers significantly predicted Behavior.

In addition to the hypothesized findings, it was found that active
participation experience in the health care setting, as measured by self-reported
participation behavior, significantly predicted post-behavior self-efficacy.

CHAPTER 4
Discussion

The purpose of this study was to examine the impact of a decision support
intervention on the self-efficacy and behavior of health care consumers related to
active involvement in health care. Past research suggested that higher self-
efficacy and an active patient orientation resulted in improved clinical outcomes
and decreased psychological morbidity (Allen, et al., 1990; Davis-Berman, 1989;
Schuhnan, 1979; Morris 8: Royle, 1988). Self-Efficacy was targeted as an avenue
for intervention because it was found that although health care consumers
desired information about their health care they did not ask for more information
(Trinkaus, 1991). Trinkaus (1991) hypothesized that fear, lack of skills and lack of
confidence might explain this contradiction. The importance of skills in
determining patient participation was supported in a study by Woodward &
Wallston (1987) who found that self-efﬁcacy mediated low desire for control of
health care.

Written questionnaires were used to measure self-efficacy, perceived barriers,
outcome expectations, subjective norms, intention, behavior and satisfaction with
decision among midlife women attending an educational program on
menopause and HRT. Data was collected at three points in time; prior to the ﬁrst
class period of the three session program, following the final class and two
months following the final class. The impact of the intervention on self-efﬁcacy
was examined using multiple regression and the impact of self-efﬁcacy, barriers,
outcome expectations and subjective norms on behavioral intention, self-
reported behavior and satisfaction was assessed using path analysis.

Researchﬂnesﬁonundﬁxpothm
Did the Decision Support Intervention increase self-efﬁcacy related to active

patient participation behavior significantly more than the other two programs?

109

1 10
The first hypothesis stated that the Decision Support Intervention would

result in greater conﬁdence in ability to participate actively in health care
decisions, greater intention to participate, increased participation behavior and
greater satisfaction with decision. This study found no signiﬁcant difference in
post-intervention self-efﬁcacy between the Decision Support Intervention
(Program C) and the other two programs. When pre-intervention self—efﬁcacy
was controlled for, self-efﬁcacy at time 2 (post-intervention) increased most in
the Lecture / Discussion condition and least in the Brochure condition with the
increase in Program C between the other two programs. There were no
signiﬁcant differences in behavioral intention, behavior or satisfaction with
decision between the three experimental conditions.

It had been hypothesized that the decision support intervention would
increase self-efﬁcacy significantly more than either the lecture/ discussion or
brochure because it included a participant modeling exercise designed to provide
a mastery experience related to active participation behavior. According to
Bandura (1982) enactive attainments were the most inﬂuential sources of efﬁcacy
information because they were based on authentic mastery experiences. The
participant modeling exercise was designed as an enactive experience in which
participants would gain conﬁdence by carrying out the target behavior in a role-
play situation. The lack of a signiﬁcant difference between the decision support
group and the lecture/ discussion group suggested that verbal persuasion was as
effective in increasing self-efﬁcacy as participant modeling. This conclusion was
counter to previous studies which found that participant modeling signiﬁcantly
increased self-efficacy compared to control groups whereas persuasion resulted
in little actual improvement in behavior (Bandura, 1977).

The role-play situation may not have been similar enough to the naturalistic

situation to have a greater impact on self-efﬁcacy than verbal persuasion. Since,

111
(as discussed below) self-efﬁcacy was related to actual behavior in the natural

situation a more realistic mastery experience (role-play with actual health
professionals, for example) may have had a greater impact on self-efﬁcacy than
the participant modeling exercise used.

In fact, both the lecture/ discussion program and the decision support
intervention included a question and answer period in which the women had an
opportunity to ask questions of a healthlcare provider. This opportunity may
have had an impact on the women’s self-efficacy. In the group setting the women
were able to ask questions of either a nurse or physician and observe other
women asking questions as well. They also received support from each other for
their concerns and questions. Anecdotal observation suggested that the question
and answer opportunity was highly valued by the women. Frequently the
question period had to be terminated due to time limitations before the women
had exhausted their questions. Because of the more individualized nature of the
exercises in the decision support intervention the women had less time to
participate in discussion among themselves and with the health care provider.
This may partially explain the observed impact of the interventions on self-
efﬁcacy.

Among women for whom pre-intervention self-efﬁcacy was low, the increase
in self-efﬁcacy was signiﬁcantly greater for those in the lecture/ discussion
intervention than in the brochure-only group. As noted above, this may have
been due to the greater opportunity for interaction with the health care provider
offered by the question and answer period during the lecture/ discussion
intervention. It was also possible that self-efficacy was inﬂuenced similar to
information acquisition. Written information alone has been previously found to
be a less effective method of communication than verbal information combined

with written information (Miller & Shank, 1986).

112
Counter to hypothesis, there was no significant difference between the

lecture/ discussion intervention and the decision support intervention in post-
intervention self-efﬁcacy. This ﬁnding was counter to previous studies which
have found that modeling increased self-efﬁcacy significantly more than
information alone (Maibach 8: Flora, 1993). There are several possible
explanations for the failure of the role-play intervention to differentially
inﬂuence self-efficacy. The role-play intervention may not have continued long
enough for the women to significantly increase their confidence in their ability to
execute active participation behavior. Since the role—play was only a brief portion
of the class session, it lasted only 15 minutes. A more extended role-play session
may have produced more signiﬁcant effects.

Another explanation for the lack of difference between the decision support
intervention and the other two programs may be that the study sample was an
exceptionally conﬁdent and motivated group and the intervention was not
needed for the majority of the women. They were already highly conﬁdent in
their skills related to interacting with health care providers. This possibility was
supported by the distribution on the self-efﬁcacy scale. As noted previously,
mean pre-intervention self-efﬁcacy was 7.6 (s.d. 1.67) on a scale from 1 to 10.

It may also be that the scale was not sensitive enough to detect differences
between those who were conﬁdent in their ability to actively participate and
those who were not. Woodward and Wallston ( 1987) found mean self-efﬁcacy
subscale scores of 38.72 and 21.81 among individuals age 40 to 59; 16% of whom
were female and 15% of whom had less than a high school education. When
these subscale scale scores (r = .86) were summed and divided by 8 to place them
in units comparable to those used in this study, the mean scale score for all 8
items was 7.57, identical to that found at Time 1 in this study. This suggested that
the lack of ﬁndings using this scale could be due to the restricted distribution

113
found with this scale in all populations.

Did the proposed model adequately describe the relationship between self-
efﬁcacy, barriers, outcome expectations, subjective norms, intention and self-
reported behavior?

Hypothesis 2 stated that Self-Efﬁcacy, Outcome Expectations and Subjective
Norms would significantly predict Intention to Participate in health care
decisions. Of the three variables expected to predict behavioral intention, only
self-efﬁcacy was signiﬁcantly (mildly) correlated with behavioral intention and
in the multivariate analysis the path from self-efﬁcacy to intention was not
statistically signiﬁcant. None of the factors measured in this study signiﬁcantly
predicted behavioral intention. This was counter to recent studies which have
found a signiﬁcant relationship between at least one of these factors and
behavioral intention (Basen-Engquist 8: Parcel, 1992; Tedesco, Keffer 8: Fleck-
Kandath, 1991 ; Manstead, Proffitt 8: Smart, 1983; Netemeyer 8: Burton, 1990;
Catch 8: Kendzierski, 1990; Brubaker 8: Wickersham, 1990; Schifter 8: Ajzen,
1985; Vallerand, Deshaies, Cuerrier, Pelletier 8: Mongeau, 1992; Geringer, Marks,
Allen 8: Armstrong, 1993; Nash, Edwards 8: Nebauer, 1993; Fishbein, et al., 1992;
Rosén, 1992; Boldero, Moore 8: Rosenthal, 1992; Cochran, Mays, Ciarletta, Caruso
8: Mallon, 1992; Gallois, Kashima, Terry, McCamish, Timmins 8: Chauvin, 1992;
Powell-Cope, Lierman, Kasprzyk, Young 8: Benoliel, 1991).

One explanation for the failure of the variables to predict intention may have
been the impact of previous experience. Iepson and Rimer (1993) found that only
two of nineteen variables (income and knowledge of correct mammography
frequency) contributed signiﬁcantly to prediction of mammography intentions
among women who had previously had mammograms. None of the other
variables drawn from the health belief model and the theory of reasoned action

contributed significantly to intention. It was speculated that a previous aversive

114
experience with mammograms may have helped to explain the variance in

mammography intention among women with prior experience (Jepson 8: Rimer,
1993). Since interacting with health care providers was a nearly universal
experience, previous success or failure in active participation may be an
important factor in intention to actively participate in future health care.

Another explanation for these ﬁndings may have been the difference between
active participation behavior and other health-related behaviors. Active
participation behavior involved interaction with the health care provider in the
form of question-asking and participation in decision making. The behavior was
interactive in nature rather than solely dependent on the skills and motivation of
the individual. The model may have failed to predict intention because intention
to execute interactive behavior may have different antecedents than other health-
related behavior such as condom use.

Hypothesis 3 stated that Health Care Self-Efﬁcacy and Intention to
Participate would signiﬁcantly predict Participation Behavior. Self-efﬁcacy was
expected to contribute directly to behavior as well as via intention because it was
recognized that actual control over participation in health care might be limited
by external factors. Although behavioral intention signiﬁcantly predicted
behavior, self-efﬁcacy did not contribute signiﬁcantly to this prediction. The
signiﬁcant relationship between intention and behavior suggested that active
participation in health care was due at least partly to intention to execute specific
behaviors. This supported Ajzen 8: Fishbein (1980) who argued that people
considered the implications of their actions beme they decided to engage in a
given behavior. These ﬁndings indicated that the midlife women in this study
decided whether or not to engage in active participation behavior prior to taking
action in the health care setting. Active patient behavior was found to be

intentional behavior and the ﬁndings supported previous studies which have

115
found strong correlations between intentions and behavior (Ajzen 8: Fishbein,

1980). The lower mean on the behavior self-report scale (3.76) compared to the
behavioral intention scale (4.40) indicated that the women intended to participate
more actively than they actually did. The discrepancy between behavioral
intention and actual behavior was partially explained by other variables in the
model, but self-efﬁcacy was not a factor.

As noted above, the path between self-efﬁcacy and intention was not
statistically signiﬁcant. Unexpectedly, self-efﬁcacy measured pre- and
immediately post-intervention also failed to predict active participation behavior
directly, but self-reported behavior was signiﬁcantly correlated with self—efﬁcacy
at follow-up. Women who reported active patient participation behavior during
a visit to their health care providers were more likely to have higher self-efﬁcacy
at follow-up than women who reported less active patient behavior. This
supported the speculation of Jepson 8: Rimer ( 1993) that, among individuals with
previous experience, behavioral intention may be less dependent on variables
such as attitudes and more dependent on the quality of the past experience.
Bentler and Speckart (1979) found that previous behavior was a signiﬁcant
determinant of intention to repeat the behavior. The finding that self-efﬁcacy
increased signiﬁcantly between Time 1 and Time 2 in all three groups but
remained stable—neither increased nor decreased—between Time 2 and Time 3
suggested that behavior inﬂuenced self-efﬁcacy and intention rather than the
reverse. Since mean self-efﬁcacy did not change between Time 2 and Time 3, self-
efficacy may have increased for some women and decreased for others following
the encounter with the health care system. Actively participating in their health
care may have increased some women’s self-efﬁcacy related to the health care
system while the experience of system-level barriers may have decreased it for

other women. The relationship between intention and behavior may have been

116
mediated by perceived barriers, with those who experienced more barriers

participating less actively than those who experienced fewer barriers.

Bandura (1977) warned that persuasion could result in decreased self-efﬁcacy
if performance expectations were increased through persuasion while the
conditions to facilitate effective performance remained unchanged. Persuasively
increasing performance expectations in the presence of situational barriers could
lead to decreased self-efﬁcacy. This circumstance may have occurred in this
study. The impact of experience with the health care system may have inﬂuenced
self-efﬁcacy in a negative direction for women who encountered situational
barriers to active participation while women who had success experiences in
participating in their health care may have increased their self-efﬁcacy.

Hypothesis 4 postulated that self-efﬁcacy expectations and outcome
expectations would be signiﬁcantly correlated. The ﬁnding in this study that self-
efﬁcacy and outcome expectations were signiﬁcantly, though moderately,
correlated supported previous studies which found there to be a relationship
between these constructs (Manning 8: Wright, 1983). The modest correlation
supported the theoretical distinction between the two concepts however,
suggesting that conﬁdence in ability to participate actively could be
distinguished from a belief that participating actively would result in speciﬁc
outcomes (Bandura, 1977; Maddux, Sherer 8: Rogers, 1982; Grembowski, et al.,
1993). This ﬁnding also partially supported the theory of planned behavior in
which attitude toward the behavior was signiﬁcantly correlated with perceived
behavioral control (Ajzen 8: Madden, 1986). The lack of a signiﬁcant correlation
between barriers and outcome expectations however, indicated that attitude
toward the behavior (outcome expectations) was correlated only with the
internal (self-efﬁcacy) aspect of perceived behavioral control. Although the
theory of planned behavior treated perceived behavioral control as a

117
unidirnensional construct with internal and external aspects on a continuum, the

ﬁndings in this study suggested that perceived behavioral control was a
multidimensional construct. The ﬁnding that self-efﬁcacy and barriers were
distinct dimensions of perceived control simply reconﬁrmed previous
understanding regarding the multidimensionality of control (Wallston, Wallston
8: DeVellis, 1978; Levenson, 1974; Marshall, 1991 ; Abrahamson, Schludermann 8:
Schludermann, 1973; Coan, Fairchild 8: Dobyns, 1973; Annandale, 1987).

Hypothesis 5 stated that Perceived Barriers and Health Care Self-Efﬁcacy
would contribute signiﬁcantly to prediction of Intention to Participate in Health
Care Decisions and Participation Behavior. Neither self-efﬁcacy nor barriers
contributed signiﬁcantly to prediction of behavioral intention. As noted above,
this may have been due to the inﬂuence of past experience on intention, making
psychological variables such as perceived barriers and self-efﬁcacy less
important in determining future intentions.

Although Time 2 self-efﬁcacy failed to contribute to prediction of active
participation behavior, perceived barriers was a significant factor. The direct
effect of barriers on behavior supported the supposition that active patient
behavior was not completely under volitional control of the respondents. Bentler
and Speckart (1979) found that attitudes and norms predicted behavior directly,
rather than via intention when the behavior in question was not completely
under volitional control. Ajzen and Madden (1986) deﬁned a behavior as being
under complete volitional control when the individual could decide at will
whether or not to perform the behavior. The more that performance of the
behavior was contingent on the presence of opportunities and adequate
resources, the less the behavior was under volitional control (Ajzen 8: Madden,
1986). The unique contribution of barriers to prediction of behavior suggested

that participation behavior was not completely under volitional control.

118
Unsolicited comments on the self-efﬁcacy instrument reinforced the conception

that for some women participation was not completely under their own control.
For example, next to an item asking about conﬁdence in ability to ask questions
one woman wrote, ”I can certainly ask. There's no guarantee that the health
professional will be willing to answer my questions—or even able to
communicate clearly.” Next to an item asking whether she would know how to
help her health care provider decide what medical procedure she should get
another woman wrote, ”But that doesn’t mean he or she’ll listen to/ agree with
me! OR make a referral.” Next to an item asking whether she was conﬁdent she
would be able to understand information about what a medical procedure would
do to her, a respondent wrote, ”If they TELL YOU!” These comments suggested
that self-efﬁcacy failed to predict behavior because the critical factors in
determining active participation behavior were intention and external barriers
rather than internal factors such as conﬁdence.

In addition to the four general categories of barriers to participation included
in the barriers scale, women wrote in barriers on an open—ended question. Six
general categories of barriers were identiﬁed; aspects of health care providers,
payment-related barriers, personal limitations, knowledge-related limitations,
system-related limitations and miscellaneous. A general picture of an
unresponsive medical system arose from these comments suggesting that the
women did not perceive participation in their health care to be under their own
control.

As noted above, self-efﬁcacy was signiﬁcantly related to behavior at follow-
up. Rather than self-efﬁcacy inﬂuencing behavior, it appeared that experience
with the health care system inﬂuenced self-efﬁcacy. The impact of success and
failure experiences on self-efﬁcacy has been well-documented with successful

experiences increasing self-efﬁcacy and failure experiences resulting in a

119
decrease in self-efﬁcacy (Davis 8: Yates, 1982; Bandura 8: Cervone, 1986). The

relationship between barriers and self-efﬁcacy suggested that some women who
desired to actively participate in their health care and felt conﬁdent in their skills
failed to do so because of the presence of external barriers to participation. These
women may have suffered a decrease in self-efﬁcacy as a result. Other studies
have found a discrepancy between desired participation in health care and actual
behavior. For example, Trinkaus (1991) found that college undergraduates
desired more information from their health care providers than they actively
requested. In this study, the need for a more active role for consumers in health
care was conﬁrmed by the finding that active participation behavior and greater
self-efﬁcacy contributed signiﬁcantly to prediction of satisfaction with the health
care decision.

The path model which ﬁt the data best (shown in Figure 16) indicated that
behavior was predicted by behavioral intentions, perceived barriers and outcome
' expectations. Satisfaction with decision was predicted by behavior and self-
efﬁcacy. Women who intended to participate actively and who perceived few
barriers to active participation were significantly more likely to report actively
participating in their health care. Women who participated more actively and
had greater conﬁdence in their ability to participate were signiﬁcantly more
satisﬁed with their decision regarding their health care. The relationship between
outcome expectations and behavior was spurious and probably due to the
relationship between self-efficacy and barriers but improved the ﬁt of the model
to the data.

The relationship of self-efﬁcacy to behavior and satisfaction with decision
supported three ﬁndings. First, active participation in health care among midlife
women was dependent on the woman’s intentions to participate actively and on

barriers such as lack of access to information and conﬂicting medical advice

120
rather than on the woman’s conﬁdence in her ability to participate effectively.

Second, the woman’s success or failure in effectively overcoming perceived
barriers to actively participate through such behaviors as asking questions, and
helping to make choices, inﬂuenced her conﬁdence in her ability to participate
actively in the future. Third, women who actively participated in their health
care and were more conﬁdent in their ability to participate actively were more
satisﬁed with their decisions regarding their health care.

According to the ﬁndings in this study, intention was a significant predictor
of participation behavior but the factors which inﬂuenced intention were not
identiﬁed in this study. None of the variables measured signiﬁcantly predicted
intention. Knowledge of behavioral intention does not, by itself, provide any
insight into the factors inﬂuencing participation behavior. Intentions may
depend on an individual’s assessment of personal skills compared to difﬁculty of
the task, expectations that a given behavior will produce a desired outcome and
perception that the behavior is the social norm. All of these factors were
measured in this study and failed to predict intention. Understanding of the
ﬁndings in this study must therefore focus on the other factors which contributed
to prediction of behavior such as barriers.

Previous reports in the literature supported the ﬁndings in this study that
barriers to active participation were an important factor in patient / health care
provider interaction. Barriers to active patient participation cited in the literature
included lack of courtesy and a patronizing attitude by physicians, physician
lack of time, ﬁnancial barriers, confusing explanations by physicians, and
unwillingness of physicians to inform patients (Jones 8: Phillips, 1988).
Additional barriers included patient loss of personal control, lack of privacy and
social isolation for patients, lack of an integrated process for development of new

patient skills, education of physicians and nurses, and a focus on cognitive rather

121
than behavioral or psychosocial educational interventions for patients (Giloth,

1990). The emotionally charged nature of the health care setting, particularly
hospitals, may generate fear and anxiety and hinder patient participation in
decisions (Dawson, Youngner 8: Connors, 1990).

In conclusion, although the purpose of this research was to increase active
patient participation in health care by increasing participants’ perceived self-
efﬁcacy, the ﬁndings suggested that barriers to active patient participation were
primarily due to factors external to the individual rather than internal. Lack of
participation was not due to lack of skills, but to lack of available information,
conﬂicting medical advice and payment restrictions.

Smmarx

This research identiﬁed the following main ﬁndings:

1. Self-efﬁcacy related to active participation in health care signiﬁcantly
increased in all three educational conditions; Written Brochure Only,
Lecture/ Discussion and Decision Support.

2. The Lecture/ Discussion intervention was signiﬁcantly more effective in

. increasing self-efficacy than was the written brochure alone.

3. Self-efﬁcacy at 2-month follow-up was signiﬁcantly greater than at pre-
intervention indicating that the observed increase following all interventions
remained stable over time.

4. Intention to participate in health care was not signiﬁcantly predicted by any
of the measured variables. Self-efﬁcacy, outcome expectations and subjective
norms failed to signiﬁcantly predict intention to participate.

5. Active participation behavior was significantly predicted by intention to
participate, perceived barriers to participation and outcome expectations.

6. Outcome expectations were moderately, positively correlated with self-

efﬁcacy supporting both the association between these variables and the

122
distinction between them.

7. Perceived barriers to participation were signiﬁcantly negatively correlated
with self-efﬁcacy . The pattern of correlation of these variables with other
variables differed supporting previous understanding of control as a
multidimensional construct.

8. Active participation behavior and self-efﬁcacy predicted greater satisfaction
with decisions.

This study was designed to determine the impact of an educational
intervention on the self-efﬁcacy and behavior of midlife women regarding active
participation in their health care. The major ﬁnding for this study was that
women who intended to participate actively in their health care and who
perceived fewer barriers to participation were more likely to report actively
participating during a visit to their health care providers. Self-efﬁcacy did not
predict active patient behavior when measured immediately post-intervention,
but was signiﬁcantly correlated with behavior two months following the
intervention. Self-efﬁcacy did increase signiﬁcantly from pre-intervention to
post-intervention, however, and increased most among women in the
Lecture/ Discussion condition.

The relationship between self-efﬁcacy, barriers, intention and behavior
suggested that experience with the health care system was an important factor in
intention and participation behavior. The correlation between self-efﬁcacy at
follow-up and behavior suggested that successful participation behavior
increased self-efﬁcacy rather than vice versa and the independent inﬂuence of
barriers on behavior and lack of predictors of intention suggested that intention
may have been inﬂuenced by experience rather than attitudes or expectations.

The ﬁndings in this study suggested that active patient participation in health

care was an intentional behavior inﬂuenced by barriers to participation and that

123
those who participated more actively were more likely to be satisﬁed with their

decisions.
Limitations

Limitations of the current research included a potential selection effect for the
behavior self-report data since only 66 of the 248 women in the study reported
visiting a health care provider during the follow-up period.

An alternative explanation for the observed relationship between self-efﬁcacy
and behavior at follow-up may have been that those who were higher in self-
efficacy immediately following the intervention were more likely to visit their
health care providers. An analysis of variance however, showed that this was not
the case. There was no signiﬁcant difference in Time 2 self-efﬁcacy between
women who reported visiting their health care providers and women who did
not visit a health care provider.

Another possible limitation of the study was the lack of a no-treatment
control group. Although the increase in self-efﬁcacy between pre- and post-
intervention suggested that the Decision Support Intervention increased self-
efﬁcacy above baseline this ﬁnding would have been stronger if it could have
been compared to a no-treatment control group which showed no increase in
self-efﬁcacy due to testing, history or regression toward the mean. A no-
treatment group would have substantially increased costs however, and would
have been of little practical value since the goal was to develop an intervention
which was more effective than the standard treatment.

The study sample also limited the generalizability of the ﬁndings.
Participants for this study were recruited primarily by newspaper
advertisements and were self-selected volunteers. As noted in the description of
the sample, study participants were more likely to be white, college-educated
and have higher incomes than the population from which they were recruited.

124
Additionally, study participants were drawn from an urban and suburban

Midwestern population which may limit generalization to rural populations and
non-Midwestern populations. In spite of these limitations however, it was
believed that the sample adequately represented the population of women who
would be most likely to seek participation in educational programs on
menopause.

Issues for Future Research

The ﬁndings from this study suggested that barriers limited midlife women’s
ability to actively participate in their health care. These factors limited the
women’s behavior in spite of their conﬁdence in their skills and an intention to
act. The barriers assessed included lack of access to information, conﬂicting
medical advice, personal health state, payment options and their health care
providers. Future research should include further assessment of the factors
patients perceive as barriers to participation. Research is also needed to
determine how women can be empowered to overcome identiﬁed barriers to
effective participation in health care decisions.

As noted previously, the role-play intervention may have been less effective
in changing self-efﬁcacy than a more realistic modeling or participant modeling
experience. Future studies should strive for realistic models and mastery
experiences which involve health professionals in real or simulated health care
encounters. Future efforts to increase self-efﬁcacy should realistically describe
potential barriers within the health care system and avoid excessively increasing
expectations in the face of systemic barriers to participation.

It has been argued that extent of control in health care should vary according
to the needs of the individual patient (Corcoran, 1988) and that it may be harmful
to offer choices to individuals with low desire for control since they may

experience more distress under choice conditions (Rokke, Absi, Lall 8: Oswald,

125
1991). Others have argued that choice may be necessary even when individual

patients preferred less participation because society may demand consumeristic
accountability (Beisecker 8: Beisecker, 1993). Furthermore, studies have found
increased distress among individuals with a high desire for control and low
feeling of control (Logan, Baron, Keeley, Law 8: Stein, 1991). Further study is
needed to understand the complex relationship between desire for control of
health care, participation in health care, clinical outcomes and patient
satisfaction. Future research should study the circumstances under which
patients choose to retain decision making autonomy and the circumstances
under which they choose to delegate responsibility to professionals or intimate
others. Special emphasis should be placed on identifying the antecedents to
delegation of responsibility and the impact of relinquishment of control on
clinical outcomes and satisfaction.
C l . 1 B l |°

This dissertation discussed the dilemma of midlife women regarding the
decision whether or not to take hormone replacement therapy and the relevance
of patient participation issues to this problem. It was suggested that participation
in health care may have an important impact on physical and psychological
health and that empowering medical consumers to participate actively may have
many advantages for patients. The theory of planned behavior and self-efﬁcacy
theory were offered as theoretical models for explaining observed differences in
patient participation behavior. Three models of patient-health care professional
interaction were described and critiqued for their potential impact on patient
self-efﬁcacy related to health care participation. It was suggested that decision
support models, in which decision making rested with the consumer and the
professional served as facilitator were most likely to promote self-efﬁcacy

regarding participation in health care.

126
A need for a decision support intervention to empower midlife women in

their decision making regarding hormone replacement therapy was identified. It
was hypothesized that women who participated in a decision support
intervention and active participation role-play exercise would experience
increased self-efﬁcacy related to participation in health care and that this would
lead to greater intention to participate actively, more active participation
behavior and greater satisfaction with decision. A study was designed which
compared three programs designed to aid women in their decision making
regarding hormone replacement therapy. The three interventions studied were a
decision support intervention, a lecture/ discussion program and a written
brochure. The variables measured included self-efﬁcacy related to active
participation in health care, perceived barriers to participation, outcome
expectations regarding active participation behavior, subjective norms related to
active participation in health care, intention to participate, participation behavior
and satisfaction with decision. All scales used were found to have adequate
internal consistency.

Self-efﬁcacy was found to increase significantly following intervention in all
three experimental conditions with the greatest increase occurring in the
lecture/ discussion condition. It was conjectured that the decision support
intervention and role play failed to increase significantly more than the
lecture/ discussion intervention because the role play was not similar enough to
the naturalistic situation and because both the lecture/ discussion program and
the decision support intervention included a question and answer period with a
health professional which may have inﬂuenced self-efﬁcacy.

Active participation in health care was found to be an intentional behavior
inﬂuenced by perceived barriers. Active participation behavior was found to
correlate signiﬁcantly with self-efﬁcacy at follow-up suggesting that behavior

127
inﬂuenced self-efficacy rather than vice versa. Active participation behavior and

self-efﬁcacy were signiﬁcant predictors of satisfaction with decision. The ﬁndings
in this study suggested that self-efﬁcacy related to participation in health care
could be increased through intervention but that active participation in health
care was inﬂuenced by the presence of external barriers to participation. The
importance of the situational context in determining participation behavior was
concluded to be critical to understanding and modifying participation behavior.
Future interventions should therefore be directed at removing barriers to patient
participation in health care. This will require intervention with the health care
system itself, with health care professionals, with patients and with society
overall. Some changes in the system, with health care professionals and with
patients have already been advocated in the literature. These will be discussed
below along with recommendations for intervention at the societal level.

It is recommended that interventions at the system level address barriers
associated with institutional policies and procedures such as hospital rules and
reimbursement restrictions. Giloth (1990) advocated the following system-level
strategies to support patient involvement in care: 1) Change the physical
environment of the health care setting to support patient education and
involvement, 2) Make policy changes which support patient access to
infOrmation 3) Change hospital rules to accommodate individual patient needs,
4) Develop programs to decrease health care staff stress, 5) Make patient
involvement in decision making an integral part of the health care delivery
process, 6) Increase patient and family responsibility for health care delivery, and
7) Help patients adapt health care regimens to their own lifestyles. Furthermore,
it is recommended that changes be made in the current system of payment for
health care services. Beisecker and Beisecker (1993) noted that regulations
imposed by third-party payers may result in barriers in the patient-professional

128
relationship and may serve as a form of paternalism in which neither the patient

nor the professional has complete decision making authority.

It is recommended that barriers which prevent health professionals from
encouraging active patient participation also be addressed. Many barriers related
to health care professionals have been identiﬁed in the literature (Anderson 8:
Sharpe, 1991). These include risk of malpractice suits (Green, 1988), ethical
considerations (Mower 8: Baraff, 1993), bureaucratic structure of the practice
setting (Stevens, Diederiks, 8: Philipsen, 1992) and patient behavior (Levinson,
Stiles, Inui, 8: Engle, 1993). Suggestions for addressing these barriers have
included educating health care providers in communication skills (Marteau,
Humphrey, Matoon, Kidd, Lloyd, 8: Horder, 1991) and teaching empathy skills
(Brock 8: Salinsky, 1993) and reducing time pressures on health care providers
(DiMatteo, Shugars, 8: Hays, 1993). It is recommended that future research study
identify additional ways in which these barriers can be minimized and that
intervention strategies be evaluated for effectiveness in promoting increased
patient participation and satisfaction.

Barriers to active patient participation in health care have also been identiﬁed
related to patients themselves. In a study of the factors inﬂuencing patient-
physician communication, physicians most often attributed communication
problems to the patient rather than to their own limitations (Levinson, et al.,
1993). Communication problems identiﬁed related to the patient included lack of
patient trust in the health care provider, patients with too many problems,
patient lack of compliance with treatment, patient communication difﬁculties
and demanding behavior by patients (Levinson, et al., 1993). Other factors which
may limit patients in actively participating in the health care encounter include
relatively less social power compared to health professionals (Navarro, 1983;

Duran-Arenas 8: Kennedy, 1991), loss of control associated with illness and the

129
patient role (Chipperﬁeld, 1993; Logan, et al., 1991), emotional distress such as

anxiety associated with illness, diagnosis and treatment (Weijts, Widdershoven 8:
Kok, 1991), and sex discrimination (Bullough 8: Bullough, 1975; Wilcox, 1992). It
is recommended that interventions to decrease patient-related barriers to
participation be developed and disseminated.

The ﬁndings in this study suggested that active patient participation was
related to intention to participate actively and perceived barriers to participation.
The literature suggested that there have been extensive attempts to intervene
with health care professionals and patients to improve communication and
promote active patient participation and dissemination of these interventions
may address some of the difﬁculties identiﬁed. In spite of these efforts however,
both patients and health care professionals remained frustrated. by poor
communication in the health care setting (Anderson 8: Sharpe, 1991 ; Levinson, et
al., 1993). Interventions aimed at society as a whole rather than at individual
patients or health care professionals and consideration of the barriers facing
health care providers and patients will be required to develop effective social
interventions to promote active patient participation.

In order to address the societal barriers to patient participation it is
recommended that future research develop interventions to modify the basic
assumptions and expectations held by patients and health care professionals
related to the nature of health, illness and social roles. One aspect of participation
in health care which needs examination is the nature of health-related decisions.
Like many health-related behaviors, hormone replacement therapy is
prevention-oriented rather than treatment-oriented. Taking HRT has been found
to have long-term health beneﬁts as well providing short-term symptom-relief.
These beneﬁts must be weighed against known and unknown long-term risks

such as endometrial cancer, and short-term discomforts such as side effects and

130
resumption of cyclic bleeding. Furthermore, the HRT decision is a dilemma for

health care providers as well as consumers since lack of research on women’s
health issues has resulted in a dearth of information from which to make
decisions. Changing society to a more public health orientation rather than a
disease or treatment orientation would encourage more research directed at
understanding universal transitions such as menopause and promotion of
wellness throughout the life-span. Increased understanding and knowledge
related to healthy functioning would provide both health care providers and
consumers with information needed to make informed decisions and would
empower patients in maintaining their health.

It is also recommended that the inﬂuence of social role expectations for
patients and health care professionals be redeﬁned. Redeﬁnition of these roles
may occur through adoption of models of human behavior consistent with the
active patient concept. Theoretical approaches likely to encourage development
of active patient role expectations include cognitive-behavioral models (Beck,
1976), theories associated with feminist therapy (Greenspan, 1993) and the theory
of rnicropolitics (W aitzkin, 1991). In cognitive-behavioral theory for example,
affect and behavior were explained by the thoughts which preceded observed
behaviors (Beck, 1976). Because this model of human behavior emphasized the
rational, cognitive aspects of behavior it may promote recognition of patients’
capacity for participation and encourage inclusion of the patient’s stated
concerns, values and preferences in decision making. Active patient behavior
may also be facilitated by minimizing the social status differences between health
care professionals and patients. Recognition of the similarity in needs and
capacities of patients and professionals may promote a more equal relationship
while acknowledging the special skills of professionals (Higgs, 1991).

Adoption of the principles of feminist therapy, which closely paralleled the

131
decision support model of patient participation may contribute to greater patient

participation and control in health care. In feminist therapy the patient was
encouraged to actively question the professional and the professional’s role was
to nurture this strength in the patient (Greenspan, 1993). Application of this
philosophy to the health care encounter may help to modify the social role
expectations for patients and health care professionals. It is also recommended»
that the principles of micropolitics be applied to the health care encounter. In
micropolitics the patient’s social context was considered critical to understanding
and healing. Health professionals were encouraged to examine their role in social
control and empower patients through social change and modiﬁcations in
medical discourse (Waitzkin, 1991).

Finally, although theory may provide a basis for future research, this study
indicated that little was currently understood about the factors inﬂuencing
patient participation in health care. Recommendations for identiﬁcation and
modiﬁcation of factors inﬂuencing health care should begin with further
research. Although both health professionals and patients have been held
responsible for problems in the health care encounter, the inﬂuence of social
factors limited the ability of individuals to modify the situation. Soliciting input
and suggestions from patients and health care professionals regarding the
strengths and weaknesses of the current social context is the ﬁrst step toward
understanding and promoting active patient participation in health care.

In summary, in order to address the difficulties present in the current health
care system it will be necessary to modify the way in which health care
professionals and patients view their roles in relation to the health care
encounter. Accomplishing these goals will require consideration of the nature of
health-related decisions and the barriers facing health care providers and

patients, including system-level barriers, barriers speciﬁc to health professionals

132
and patients and societal expectations. Recommendations included

disseminating interventions already developed, developing interventions to
address societal barriers and soliciting input from patients and health care
providers regarding needed changes. Suggestions for modifying societal barriers
included promoting a public health orientation among professionals and patients
and encouraging adoption of models of human behavior which promoted

respect for the concerns, values and preferences of the patient.

LIST OF REFERENCES

133

LIST OF REFERENCES

Abrahamson, D., Schludermann, S, 8: Schludermann, E. (1973). Replication of

dimensions of locus of control. WW
Iimhnlogx, AL 320

Abramson, L. Y., Seligman, M. E. P., 8: Teasdale, I. D. (1978). Learned
Helplessness in Humans: Critique and Reformulation. MW

Radium 87. 49-74.
Ajzen, I. (1988). Wm. Milton Keynes: Open

University.

Ajzen, I & Fishbein, M. (1980)- Understandmgattitudmndnnedicnngmml
behavior. Englewood Cliffs, NJ: Prentice-Hall.

Ajzen, I 8: Madden, T. I. (1986). Prediction of goal-directed behavior:
Attitudes, intentions and perceived behavioral control. Imam

ExpeﬁmentalmciaLEsxchologL 22, 453-474.

Ajzen, I. ,Timko, C., 8: White,]. B. (1982). Self-monitoring and the attitude-
behavior relation Wag): 42. 426-435

Allen, J. K., Becker, D. M., 8: Swank, R. T. (1990). Factors related to functional
status after coronary artery bypass surgery. Wing, 12, 337—343.

Anderson, L. A, 8: Sharpe, P. A. (1991). Improving patient and provider
communication: A synthesis and review of communication interventions.

W12, 99-134.

‘ Anderson, R. M., Funnell, M. M., Barr, P. A., Dedrick, R. F., & Davis, Wayne K.
(1991). Learning to empower patients: Results of professional education
program for diabetes educators. W _14, 584—590.

Annandale, E. C. (1987). Dimensions of patient control m a free-standing birth
Center W 25, 1235-1248

Bandura, A. (1977). Self-efﬁcacy: Toward a unifying theory of behavioral
change W84. 191-215

134

Bandura, A. (1982). Self-efﬁcacy mechanism in human agency. man
Paxchologist 31. 122-147-

Bandura, A- (1986). WWW
them. Englewood Cliffs, NJ: Prentice-Hall.

Bandura, A. (1988). Self-regulation of motivation and action through goal
systems. InV. Hamilton, G. H. Bower 8: N. H. Frijda (Eds.,) Cognitive

me 37—61). Boston. Kluwer

Academic.

Bandura, A. Adams, N. E., 8: Beyer, I. (1977). Cognitive processes mediating
behavioral change We;
125—139.

Bandura, A. 8: Cervone, D. (1986). Differential engagement of self-reactive

inﬂuences m cognitive motivation WWW
DecisionJchesses 38, 92—1 13

Bandura, A., 8: Wood, R. (1989). Effect of perceived controllability and
performance standards on self-regulation of complex decision making.

Wﬁ, 805-814

Barling, 1., 8: Snipelisky, B. (1983). Assessing the determinants of children’s

academic self-efficacy beliefs: A replication. W
Research, Z, 371-376.

Basen-Engquist, K. 8: Parcel, G. S. (1992). Attitudes, norms, and self—efﬁcacy: A
model of adolescents’ HIV-related sexual risk behavior. Healthﬁﬁmﬁgn

W19. 263-277

Beale, D. A., 8: Manstead, A. S. R. (1991). Predicting mothers' intentions to limit
frequency of infants' sugar intake. Testing the theory of planned behavior.

MW 21. 409-431.

Beck, A. T. (1976)..Cogmnmtherap1and_the_enmionaLdmrders New York:
Penguin.

Beisecker, A. E. (1988). Aging and the desire for information and input in
medical decisions: Patient consumerism in medical encounters. 111g

ﬁemtologist. 330—335.

Beisecker, A. E., 8: Beisecker, T. D. (1993). Using metaphors to characterize
doctor-patient relationships: Patemalism versus consumerism. Health

Communication, 5, 41-58.

135

Bentler, P. M. 8: Speckart, G. (1979). Models of attitude-behavior relations.
W86. 452—464.

Bergkvist, L., Adami, H. O., Persson, 1., Hoover, R., 8: Schairer, C. (1989). The .
risk of breast cancer after estrogen and estrogen-progestin replacement. The

Mglandjoumalnmedidneﬁnﬁ). 293-297.

Blanchard, C. G, Labreque, M. S, Ruckdeschel, 1. C., 8: Blanchard, E. B. (1988).
Information and decision making preferences of hospitalized adult cancer

patients WE 1139-1145

Boldero, 1., Moore, S, 8: Rosenthal, D. (1992). Intention, context, and safe sex.

Australian adolescents’ responses to AIDS. W1
3831321191938 22, 1374-1396.

Brock, C. D., 8: Salinsky, 1. V. (1993). Empathy: An essential skill for
understanding the physician-patient relationship in clinical practice. Family
Medicine, 25, 245-248.

Brubaker, R. G., 8: Wickersham, D. (1990). Encouraging the practice of testicular
self-examination: A ﬁeld application of the theory of reasoned action.

mums 154—163.

Bullough, B., 8: Bullough, V. L. (1975). Sex discrimination 1n health care,
Nursingﬂuﬂmkml) 40-45

Cassileth, B. R., Zupkis, R. V., 8: Sutton-Smith, K. (1980). Information and
participation preferences among cancer patients. W
Medicine. 22. 832-836.

Chamey, E. B. (.1990) Parental attitudes toward management of newborns with

Chipperﬁeld, 1. G. (1993). Perceived barriers 1n coping with health problems.
JournaLoLAgingandHealth, 5, 123—139.

Coan, R. W., Fairchild, M. T., 8: Dobyns, Z. P. (1973). Dimensions of
experienced contrOI Ihﬂoumalnfﬁomalﬁsxchology, 21. 53-60

Cochran, S. D., Mays, V. M., Ciarletta, 1., Caruso, C, 8: Mallon, D. (1992).
Efficacy of the theory of reasoned action in predicting AIDS-related sexual

risk reduction among gay men loumaLQLApleLEmholggL 22,
1481-1501.

136

Cohen, 1. (1988). Statistical power analysis for the behavioral sciences (2nd ed.).
Hillsdale, N1 : Lawrence Erlbaum.

CohenJ &Cohen P (1983) W
Wed) Hillsdale NJ Lawrence Erlbaum

Cooper, 1. E., Bledin, K. D. ,Brice, B., 8: Mackenzie, S. (1985). Effects of female
sterilization: One year follow-up m a prospective controlled study of

psychological and psychiatric outcome.
22, 13-22.

Corcoran, S. (1988). Toward operationalizing an advocacy role. Imam
Professionallﬂursing, 4(4) 242-248

Crawley, III, F. E. (1990). Intentions of science teachers to use investigative
teaching methods. A test of the theory of planned behavior. 183.32.8le.

W22, 685—697.

Cummings, S. R., Black, D. M., 8: Rubin, S. M. (1989). Lifetime risks of hip,
colles', or vertebral fracture and coronary heart disease among white

postmenopausal women. AcclneesoﬂntemaLMedicine. 142, 2445-2448.

Davis, F. W., 8: Yates, B. T. (1982). Self-efﬁcacy expectancies versus outcome
expectancies as determinants of performance deﬁcits and depressive affect.

We, 23-35

Davis-Berman, 1. (1989). Physical self-efﬁcacy, perceived physical status, and

depressive symptomatology in older adults. W 124,
207—215.

Davis-Floyd, R. E. (1987). The technology model of birth. IgmmaLQfAmgﬁgan
More. 19.0. 479-495.

Dawson, N. V. ,Youngner, S. T., Connors, A. F. (1990). Phase II: Inﬂuencing

decision making 1n Sopport lounmofﬂimcallipidemiologx, AﬂSuPpl )
1035—1085.

Degner, L. F., 8: Sloan, 1. A. (1992). Decision making during serious illness.

What role do patients really want to play? JonmaLQLClimELEpidmniongL
_45, 941-950.

DeVellis, B. M. ,,Blalock S. 1., 8: Sandler, R. S. (1990). Predicting participation in
cancer screening. The role of perceived behavioral control. 1mm

War. 29, 639-660.

137

DiMatteo, M. R., Shugars, D. A., 8: Hays, R. D. (1993). Occupational stress, life

stress and mental health among dentists. Wm
QrgamaﬁonaLBsxchologL 86, 153-162.

Duffy, M. E. (1988). Determinants of health promotion in midlife women.
31, 358—361.
Dupont, W. D., 8: Page, D. L. (1991). Menopausal estrogen replacement therapy

and breast cancer. mixesoﬂntememmcide 151., 67-72.

Duran-Arenas, L., 8: Kennedy, M. (1991). The constitution of physicians’ power:
A theoretical framework for comparative analysis. W
Medicine, 32, 643—648.

Dzewaltowski, D. A. (1989). Toward a model of exercise motivation. magi
MW 11. 251-269

Dzewaltowski, D. A. ,N,oble 1. M, 8: Shaw, 1. M. (1990). Physical activity
participation: Social cognitive theory versus the theories of reasoned action

and planned behavior mmﬁpdrﬂerciseﬂsycmm 12. 388-405

Eastman, C, 8: Marzillier,]. S. (1984). Theoretical and methodological

difficulties 1n Bandura’ s self-efﬁcacy theory. Cogmﬁyﬂherapmd
Research, 8. 213-229

Edwards, A. L. (1954) mmmethedsfeuhehehmoralmces New
York: Holt, Rinehart and Winston.

Ehrenreich, B, 8: Ehrenreich, 1. (1978). Medicine and social control. In B.

Ehrenreich 8:] Ehrenreich (EdS) WW
(pp. 39-79). New York: Monthly Review.

Ende,1., Kazis, L., 8: Moskowitz, M. A. (1990). Preferences for autonomy when
patients are physician-WWW 5. 506—509.

Fairweather, G. W, 8: Davidson, W. S. (1986). Anintmducnoanmmumty,
W New York: McGraw-Hill

Fairweather, G W & Tomatzky, L G (1977) Experimentalmethodsforsecial
W. New York: Pergamon.

Fishbein, M., Chan, D. K-S., O’Reilly, K., Schnell, D., Wood, R., Beeker, C.,
Cohn, D. (1992). Attitudinal and normative factors as determinants of gay
men’ s intentions to perform AIDS-related sexual behaviors: A multisite

analysis humaLoprpliedSociaLEsxclmlogLZZ. 999-1011

138

Figge, D. C. (1990). The tyranny of technology. WE
andﬁmecology, 16; 1365-1369.

FiSthinI M., & Ajzen, 1- (1975). WW1;
mtreducnentdtheomndresearch Reading, MA: Addison-Wesley.

Fishbein, M., 8: Stasson, M. (1990). The role of desires, self-predictions, and
perceived control 1n the prediction of training session attendance. Journalof

Apphedﬁndalﬁsyrhnham 173—198.

Gallois, C., Kashirna, Y., Terry, D., McCamish, M., Timmins, P., 8: Chauvin, A.
(1992). Safe and unsafe sexual intentions and behavior: The effects of norms

and atﬁtudeS-lmamdeLApphedmcialﬁsygmeuzz, 1521-1545.

Gatch, C. L., 8: Kendzierski, D. (1990). Predicting exercise intentions: The theory
of planned behavior. Reseamhgianmlyferimiseandjport 6.1. 100-102.

Geringer, W. M., Marks, 8., Allen, W. 1., Armstrong, K. A. (1993). Knowledge,
attitudes, and behavior related to condom use and STDs in a high risk

population. W30, 75—83.

Giloth, B. E. (1990). Promoting patient involvement: Educational,
organizational, and environmental strategies _EaﬂenLEdngaﬁQnand
Counseling .15. 29-38.

Glanz, K., Lewis, P. M., 8: Rimer, B. K. (1990). The scope of health education:
Parameters of a maturing ﬁeld. In K. Glanz, F. M. Lewis, 8: B. K. Rimer
(EdSw)-.l1-1V0 -._11 ‘I ‘1- -10 11:10; _‘.‘- -1I.110- 1‘
(pp. 3—16). San Francisco. Iossey-Bass.

Gray, R. E., Doan, B. D., 8: Church, K. (1991). Empowerment issues in cancer.
32311111811128, .11-2(4) 22—28.

Green, 1. A. (1988). Minimizing malpractice risks by role clariﬁcation: The

confusing transition from tort to contract. AnnalseﬂntemalMegLicine, 1112,
234-241.

Greenspan, M. (1993). Anemppmachjmmenkjhmpetznded)

Bradenton, FL: Human Services Institute.

Grembowski, D., Patrick, D. Diehr, P., Durham, M., Beresford, 8., Kay, E., 8:
Hecht, 1. (1993). Self-efﬁcacy and health behavior among older adults.

MW :2. 89—104.

139

Higgs, R. (1991). Looking after yourself. In R. Colrney, (ed. )Qexagping
11.:London
Tavistock / Routledge.

Hinsz, V. B, 8: Nelson, L. C. (1990). Testing models of turnover intentions with

university faculty IonmaleprphedmoaLEsychnlngL 251, 68-84

1epson, C., 8: Rimer, B. K. (1993). Determinants of mammography intentions
among prior screenees and nonscreenees.

was. 40—51.

Jones, 1- A., 8: Phillips, G. M. (1988)-WW.
Carbondale, IL: Southern Illinois University.

1udd, H. (1987, May). Efﬁcacy of transderrnal estradiol. W
thfetricundﬂynemlogLMS), 1326—1331.

Kazdin, A. E. (1978). Conceptual and assessment issues raised by self-efﬁcacy

theory WWW]. 177-185.
Langer, E. 1. (1983). MW. Beverly Hills: Sage.

Lee, C. (1989). Theoretical weaknesses lead to practical problems: The example
of self-efﬁcacy theory.

W211, 115—123.

Levenson, H. (1974). Activism and powerful others: Distinctions within the

concept of internal-extemal control. MW 18,
377-383.

Levinson, W., Stiles, W. B., Inui, T. S., 8: Engle, R (1993). Physician frustration
in communicating with patients. Madiglﬁara, 31, 285—295.

Logan, H. L., Baron, R. S., Keeley, K., Law, A., Stein, S. (1991). Desired control
and felt control as mediators of stress in a dental setting. ﬁgauhfsyghglggy,
1Q, 352-359.

Luciano, A. A., Turksoy, R. N., Carleo,1., 8: Hendrix, 1. W. (1988). Clinical and
metabolic responses of menopausal women to sequential versus continuous

estrogen and progestin replacement therapy. W51, Z],
39-43.

Maddux,]. E., Sherer, M, 8: Rogers, R. W. (1982). Self-efﬁcacy expectations and
outcome expectancy: Their relationship and their effects on behavioral

intentions Cogniﬁxelhmpundﬁesearch 6. 207-211-

140
Maibach, E. 8: Flora, 1. A. (1993). Symbolic modeling and cognitive rehearsal.
Using video to promote AIDS prevention self-efﬁcacy. Comm
Research. .29. 517-545.

Manning, M. M., 8: Wright, T. L. (1983). Self-efﬁcacy expectancies, outcome
expectancies, and the persistence of pain control 1n childbirth. IgumaLQf

PersonaliﬂandioeiaLpsxchdegLﬂQL 421-431

Manstead, A. S. R. ,,Profﬁtt C, 8: Smart, 1. L. (1983). Predicting and
understanding mothers' infant-feeding intentions and behavior. Testing the

theory of reasoned action We 44
657-671. _

Marshall, G. N. (1991). A multidimensional analysis of internal health locus of
control beliefs: Separating the wheat from the chaff? Imnnalgﬁmsgnahm

W61. 483-491.

Marteau, T. M., Humphrey, C., Matoon, G., Kidd, 1., Lloyd, M., 8: Horder, 1.
(1991). Factors inﬂuencing the communication skills of ﬁrst-year clinical

medical students. Madjgalﬁdagaﬁgn, _25, 127-134.

Mathews, 1. 1., 8: Zadak, K. (1991). The alternative birth movement in the
United States: History and current status. Womenﬁﬂealﬂl, 12(1), 39—36.

McCaul, K. D. ,O,'Neill H. K. 8: Glasgow, R. E. (1988). Predicting the
performance of dental hygiene behaviors: An examination of the Fishbein
and Ajzen model and self-efficacy expectations. loumaLoLAppliedmal
W18: 114-128.

McKinlay,1. B. ,McKinLay, S. M., 8: Beaglehole, R. (1989). A Review of the
Evidence Concerning the Impact of Medical Measures on Recent Mortality

and Morbidity m the United States Jntemanenallenrnathealthkmees.
12, 181-208.

Miller, G, 8: Shank, 1. C. (1986). Patient education: Comparative effectiveness
by means of presentation. W 22, 178—181.

Morris, 1., 8: Royle, G. T. (1988). Offering patients a choice of surgery for early
breast cancer: A reduction in anxiety and depression in patients and their

husbands Wenceandlylﬂme. 26(6). 583-585-

Mower, W. R., 8: Baraff, L. 1. (1993). Advance directives: Effect of type of
directive on physicians’ therapeutic decisions. W1
Medicinelﬁﬁ. 375—381.

141

Nash, R., Edwards, H., 8: Nebauer, M. (1993). Effect of attitudes, subjective
norms and perceived control on nurses’ intention to assess patients’ pain.

InumaLofAdxancedNnrsinglﬁ. 941-947.

Navarro, V. (1983). Radicalism, Marxism, and medicine. mammal
efﬂealﬂiﬁemices. 13. 179-203.

Netemeyer, R. G. 8: Burton, S. (1990). Examining the relationships between
voting behavior, intention, perceived behavioral control, and expectation.

hnmaLQprpmaoemEsmielegan, 661-680.

O’Connor, A. 8: O’Brien-Pallais, L. L. (1989). Decisional conﬂict. In G. K.

McFarland 8: E. A. McFarlane, (Eds), Eli—5W
(pp. 573-578). Toronto: C. V. Mosby.

Persson, I., Adami, H. O. Lindberg, B. S. ,,10hansson E. D. B, 8: Manell, P.
(1983). Practice and patterns of estrogen treatment in climacteric women in a

Swedish population Aetaﬂhsfetnmetﬁycelegmaﬁandmam 62. 289-
296.

Powell—Cope, G. M., Lierman, L. M., Kasprzyk, D., Young, H. M., Benoliel, 1. Q.
(1991). The theory of reasoned action in prediction of breast self-
examination: A comparison of two studies. W
InternationaL .12. 51-61.

Rappaport, 1. (1987). Terms of empowerment / exemplars of prevention: Toward

a theory for community psychology AmericanjeumaLQLCmnmnnint
Psychologioll 121-144.

Rokke, P. d. A,bsi, M. a., Lall, R, 8: Oswald, K. (1991). When does a choice of
coping strategies help? The interaction of choice and locus of control.
_14, 491—504.

Rosén, A. (1992). Beliefs, attitudes, and intention in the context of abortion.

WW2; 1464-1480.

Roter, D. (1987). An exploration of health education’s responsibility for a
partnership model of client-provider relations. EaﬁenLEdagaﬁonand
Cennsdingﬂ. 25—31.

Rothert,..M(1990) 1011‘1 ' '
(Research Grant No. NR01245-04A2). Washington DC. U. S. Department of
Health and Human Services, National Center for Nursing Research.

 

142

Rothert, M., Kroll,1., Holmes-Rovner, M., Rovner, D., Schmitt, N ., 8: Talarczyk,
G. (1992). Menopause: Taking care of your health. East Lansing, Michigan
State University.

Rothert, M., Rovner, D., Hohnes, M., Schmitt, N ., Talarczyk, G., Kroll,1., 8:
Gogate, 1. (1990). Women’s use of information regarding hormone

replacement therapy. WM 13, 355-366.

Sanna, L. 1. (1992). Self-efﬁcacy theory: Implications for social facilitation and
social loafingI Idumalctﬁerserisrlincandmglﬁsychclogio 62, 774-786.

Saxton, M. (1987). Prenatal screening and discriminatory attitudes about

disability. Womendcﬂealthldﬂ & 2), 217-224-
Schain, W. S. (1980). Patients' rights in decision making. _Cancar, _46, 1035-1041.

Schifter, D. E., 8: Ajzen, I. (1985). Intention, perceived control, and weight loss:

An application of the theory of planned behavior. MW
Sociallisychclegw .49, 843-851.

Schulrnan, B. A. ( 1979). Active patient orientation and outcomes in
hypertensive treatment. Madicalﬁaza, 12, 267—280.

Schunk, D. H, 8: Hanson, A. R. (1985). Peer models: Inﬂuence on children’ s

self-efficacy and achievement ImmaletEdncaticnaLEsxchclch ZZ.
313—322.

Sharf, B. F. (1988). Teaching patients to speak up: Past and future trends.
.EatienLEdncaticnendﬁcnnseling. 11. 95—108.

Solomon, K. E, 8: Annis, H. M. (1990). Outcome and Efﬁcacy Expectancy 1n the
Prediction of Post-Treatment Drinking Behaviour. W
M85, 659-665.

Steele, D. 1., Blackwell, B. G,utmann, M. C., 8: Jackson, T. C. (1987). The
activated patient. Dogma, dream, or desideratum? BaﬁenLEdngaﬁQnand

Counselinglﬂ. 3—23.

Stevens, R, Diederiks,1., 8: Philipsen, H. (1992). Physician satisfaction,
professional characteristics and behavior formalization 1n hospitals. ﬁnal

.‘ic1ence_&_Medicine. 35, 295—303.

Strickland, B. S. (1978). Internal-External Expectancies and Health-Related
Behaviors W46. 1192-1211

143

Strull, W. M., Lo, B, 8: Charles, G. (1984). Do patients want to participate in

medical decision making IcnmLQLtheArnencanL/ledicalessdcaunnﬁz.
2990—2994.

Tedesco, L. A.,Keffer, M. A., 8: Fleck-Kandath, C. (1991). Self-efﬁcacy, reasoned
action, and oral health behavior reports: A social cognitive approach to

compliance Immalntﬁehaxicralbdedidme, 14, 341-355.

Teasdale, 1. D. (1978). Self-efﬁcacy: Toward a unifying theory of behavioural
change?Ad1ances1n_Beha)eniesearch_andlheapLL 211-215.

Trinkaus, 1. (1991). Medications and information for patients: A quick look.
PaxchclcgicLlReans. 68, 911-914.

Vallerand, R. 1., Deshaies, P, Cuerrier, 1., Pelletier, L. G., Mongeau, C. (1992).
Ajzen’ s and Fishbein’ s theory of reasoned action as applied to moral

behavior. A conﬁrmatory analysis. W1
Psechclcgy. 62. 98-109

Vallis, T. M., 8: Bucher, B. (1986). Self-efﬁcacy as a predictor of behavior
changes: Interactions with type of training for pain tolerance. Cognitive

IherapLandRegarch. 10. 79-94.

Waitzkin, H. (1991).
' . New Haven: Yale.

Wallston, K. A., Wallston, B. S., 8: DeVellis, R. (1978). Development of the
multidimensional health locus of control (MHLC) scales. Hgalmﬁdagaﬁgn

Queens-moo, 160—170.

Weijts, W., Widdershoven, G., 8: Kok, G. (1991). Anxiety-scenarios in
communication during gynecological consultations. WM

Counselinglﬁ. 149—163-

Wilcox, V. L. (1992). Effects of patients’ age, gender, and depression on medical
students; beliefs, attitudes, intentions, and behavior. W

ﬁcciaLEsxchelchR 1093—1110.

Williams. S. L. 8: Kinney, P. K. (1991). Performance and nonperformance
strategies for coping with acute pain: The role of perceived self-efﬁcacy,

expected outcomes and attention. MW 15, 1-19.

Wilson, R. G., Hart, A., 8: Dawes, P. 1. D. K. (1988). Mastectomy or conservation:
The patient's choice. W 227, 1167—1169.

144
Woodward. N J- (1984)I Agﬂﬂrencesmﬁnhsdesmucntm.

I‘ VI-.‘ 1I-I1110 .I'I I-eIi‘ .1 1.1-1.L1-J.-l
- .Unpublished master’s thesis, Vanderbﬂt University,
Nashville, TN.

Woodward, N. 1., 8: Wallston, B. S. (1987). Age and health care beliefs: Self-

efﬁcacy as a mediator of low desire for control. EMMA-81118: 2,
3—8.

Zimmerman, M. A., 8: Rappaport, 1. (1988). Citizen participation, perceived

control and psychological empowerment. WW1);
Psychology. 16 725—750.

APPENDICES

APPENDD( A

145

APPENDIX A
RECRUITMENT ADVERTISEMENTS

Women over 40

 

Learn more about
Menopause and Hormone
Replacement Therapy

Michigan State University is looking for women to participate in
a study which will inform them about menopause and hormone
replacement therapy. Women will attend three eduwtional
sessions at MSU in January on either Thursday evenings or
Saturday mornings. Women will also provide additional data at
intervals during the following 12 months. .

Dates: Thmday. Januaryluamitom 7-8:30 pm
SaMday,Januay9.16&23komlo-ll:30an

Call now for more information (517) 355-6744
mum! investigator Marilyn Rothert RN. PhD. mm
MIchigan State University
College of Nursing

 

 

 

APPENDD( B

146

APPENDD( B
POWER ANALYSIS CALCULATIONS

ANALYSIS OF VARIANCE
f = Effect Size Index

a2 = Two-Tailed Signiﬁcance Level
11 = Numerator Degrees of Freedom; 11 = k — 1; where k = number of groups
N = Number of participants

B = Power
Large effect size: f = .40; Medium effect size: f = .25 (p. 355, Cohen, 1988)

k=3;u=3—1=2;a2=.05
From Table 8.3.13 (p. 314) Cohen (1988):
For medium effect size: B = .80; N = 156

For large effect size: B = .80; N = 63

Multiple Regression Analysis
:2 = Effect Size Index

a2 = Two-Tailed Significance Level
N = Number of Participants

7. = The noncentrality parameter of the noncentral F distribution = f 2 (u + v + 1)

u = Numerator Degrees of Freedom; 11 = number of independent variables

v = Degrees of Freedom for Error Variance = N - u — 1

B = Power

147

Cases 0 8: 1: Test of R2 (See p. 415 Cohen, 1988). For 13 independent variables:

MW

{2 = 3311213.,“ = f2/ 1 + r2 = .35/21+.35 = .26; n = .51
N=63;a2=.05;u=13;v=63—13-1=49

1. = .35 (13 + 49 + 1) = 22.05

Forl=.35;B=77

For 7L = .22; B = 81.5

For it = .34; B = 86

For f2 = .5; 2. = 9.45 K = (.15)(63) = 9.45; B = 41

E H 1' m 5' ,
£2: .15; (See p. 413 Cohen, 1988)=R= .36

N=156;a2=.05;u=13;v=156—13—1=142

1:.15 (13+142+ 1)=23.40
From table 9.3.2 (pp. 420 — 423) Cohen (1988):

For5lewithN=156;94<B>98
For13IszithN=156;81<B>92

For15lewithN=156;78<B>91

EQLSmallﬁffectﬁze:
r2 = .02; (See p. 413 Cohen, 1988) = n = .02

N=248;a2=.05;u=13;v=248-13-1=234

2. = .02 (13 + 234 + 1) = 4.96; 19 < [3 > 29 (See p. 421 Cohen, 1988)

APPENDIX C

p—
O

148

APPENDIX C
INTERVENTION OUTLINES

Outline of Intervention A
Session I

Introduction and informed consent forms (5 minutes)
a. Importance of study and data

i. "Participants are an important part of this study and we need your

help“

b. Discuss incentives to continue participation
Data Collection (30 minutes)
a Problem Signiﬁcance Assessment
b Sociodemographic
c. Perceptions (Menopause Problem Scale and Control Scale)
(1. Symptoms Instrument
e. Symptom Management/Self-Care Instrument
f. Knowledge Instrument
g Satisfaction with Health Care Provider Instrument (Part I)
h Satisfaction with Decision Instrument
i. Health Care Self-efﬁcacy Instrument
Distribute Brochures

Session 111

Personal Plan Form and Activity Record Calendar (15 Minutes) (Clinician)

Data Collection (25 Minutes)

a. Knowledge

b Satisfaction with Decision

c Health Care Self-Efﬁcacy

d. Barriers to Decision Scale

e. Outcome Expectations Scale

f. Subjective Norms Scale

g. Behavioral Intentions Scale

Emphasize the importance of data collection and returning in six months

a. Discuss incentives to return calendars and return in six months and one
year

149

Outline of Intervention B
Session I

Bold denotes topics in knowledge instrument

1. IntroduCtion and informed consent forms (5 minutes)
2. Data Collection (30 minutes)

'5‘“? 2'7.“ 9‘9 57!”

1.

Problem Signiﬁcance Assessment

Sociodemographic

Perceptions (Menopause Problem Scale and Control Scale)
Symptoms Instrument

Symptom Management/Self-Care Instrument

Knowledge Instrument

Satisfaction with Health Care Provider Instrument (Part I)
Satisfaction with Decision Instrument

Health Care Self-efﬁcacy Instrument

3. Physical changes of menopause (20 Minutes) (Clinician)
[Participants may follow along in the brochure]

What 13 menopause?
i. Deﬁnition of menopause
ii. Menopause is a normal transition
iii. Changes at midlife
iv. Surgical menopause
When does menopause occur?
1. Average age of menopause is 50
What happens to my body at menopause?
i. Childbearing years
(1) Ovary contains follicles which decrease in number over
life-span
(2) Two glands 1n the body to know about
(a) Pituitary
(b) Ovary
(i) Responds to the signals from the pituitary
(ii) Produces two primary hormones
l) estrogen
2) progesterone
(3) Two hormones that drive the ovary to do its job
(a) FSH
(b) LH
(4) Feedback cycle of hormonal changes during menstrual
cycle

150

(5) Hormonal changes at menopause
(6) Fwdback cycle of childbearing years interrupted
(7) Levels of pituitary hormones rise
(8) Decrease in production of estrogen and progesterone
ii. Menstrual periods change
(1) Pattern of change varies between women
iii. Ovulation can occur and pregnancy can result even without a
menstrual period
(1) Should use birth control for one full year

Will I experience symptoms at menopause? (Clinician)

a.
b.

gar-spam

Sheehy quote
Hot flashes one of most common symptoms
i. Related to estrogen deficiency
80 % of women experience symptoms
20% of women seek help for symptoms
Description of experience of hot ﬂashes
Estrogen reduces hot ﬂashes in 95 -96% of cases
Other changes due to estrogen decrease
i. breast
ii. vaginal lining
(1) lining becomes less thick
(2) dryer, thinner, shorter
(3) Can result in burning or stinging with urination
(a) difﬁculty or pain with intercourse
(4) Can result in increased chance of vaginal infections
Other things which occur around the time of the menopause

i. anxiety
ii. weight gain
1ii. bloating

iv. Not shown to be directly due to estrogen deﬁciency

Increased Risks at Menopause (Clinician)

a.
b.

Two diseases that increase: Osteoporosis and heart disease

Osteoporosis

i. Deﬁnition ‘

ii. Around 50,000 women die per year as a result of osteoporotic
fractures

iii. Fracture may result in nursing home placement
iv. Osteoporosis a calcium disease and protein disease
v. Osteoporosis accelerates rate of development at menopause
vi. Women most at risk for fractures:
(1) Thin, white women who smoke and do not exercise
regularly

151

c. Heart Disease

i. Major cause of death for men and women
ii. 390,000 women per year die of heart disease -
iii. Heart disease in women tends to increase in rate of

development around the time of the menopause
6. Self-Care Strategies (10 Minutes) (Clinician)
a. For hot ﬂashes:
i. Eat a good, healthy diet
' (1) - To help hot ﬂashes decrease stimulatory types of foods
such as spices
ii. Wear layered clothing
(1) Can take of layers if get hot
b. Skin (like the bones) gets thinner and dryer around the menopause
i. . Natural skin lubricants like lanolin or other non-perfumed
lubricants may help
c. Vaginal dryness
i. Use water-soluble lubricant
d. For many people these self-eare strategies are enough to take care of the
symptoms and signs that occur at the time of the menopause but don’t do
anything do anything about the long-term, killing complications of
osteoporosis and coronary heart disease.

152

Intervention Outline B
Session 11

Pros and Cons of Hormone Replacement Therapy (30 Minutes) (Clinician)
Five areas to think about related to HRT (SCOBES)

S--Symptoms (menopausal symptoms)

C--Coronary heart disease

O—- -Osteoporosis

B--Breast cancer

E--Endometrial cancer

S--Side effects of HRT

Deﬁnition of base rates

Frequency of occurrence

Important because it addresses "double what?" "What are these
relative to each other?"

Two main regimens

a.

i.
ii.

iii.
iv.

v.
vi.
i.

ii.

1.

ii.

Estrogen- Only
(1) Typical regimen
(2) Whether it would be prescribed for you would depend on
your personal history and medical examination
(3) Side effects include:
(a) headache
(b) bloating
(0) weight gain
(d) breast tenderness
(4) No resumption of cyclic bleeding with estrogen-only
(5) No chance of a woman getting pregnant
Estrogen combined with Progesterone
(1) Typical regimens
(2) Side effects

(a) More women experience side effects with
combination therapy than with estrogen-only
(b) Side effects similar to those experienced with

estrogen:
(i) headache
(ii) bloating

(iii) weight gain
(iv) breast tenderness

d.

153

Impact of Hormone Replacement Therapy on factors
Symptoms

i.

(1)

(2)

Recap of Base Rate Risk

(a) Hot ﬂashes 80% of women experience them but
only 20% seek help for them

(b) Vaginal atrophy-becomes dryer and more easily
injured and may cause difficulty with intercourse

Effect of HRT

(a) Both ERT and PERT relieve hot ﬂashes by about
90-95%

(h) Both regimens relieve vaginal dryness

(c) Clinical data has not found that ERT or PERT
affect other symptoms such as mood swings

Coronary Heart Disease

(1)

(2)

Recap of Base Rate Risk

(a) Most frequent killer of both men and women.
394,000 women die each year

(b) Heart disease occurs somewhat later in women than
in men, however.

Effect of HRT

(a) Estrogen-only cuts risk in half

(b) Addition of progestogen has been questioned.

(c) There are less data about the effect of progestogen.
Most research has been done on .625 of estrogen

Osteoporotic Fractures

(1)

(2)

Recap of Base Rate Risk

(a) 50,000 women die each year from complications
of f ractures.

(b) Fractures occur predominantly among women and
increase greatly with age. By age 80, 1 out of every
3 women will have had a hip fracture. Of those hip
fractures, 20% will result in death

Effect of HRT

(a) Risk '5 cut in half with ERT and PERT

iv.

154

Breast Cancer

(1)

(2)

(3)

Base Rate risk

(a) 41,000 women die each year from breast cancer

(b) Every woman over age 50 ought to have a
mammogram every year

(c) Breast cancer is usually treated with surgery, and/or
chemotherapy

Effect of HRT

(a) Do not have the bottom line on this

(b) Some studies indicate an increased risk, some
decreased risk, some no change in risk
(i) Increase is usually 1.3 to 1.8 increased risk

Important to know the following when you read reports of

studies:

(a) What kind of estrogen

(b) Dose

(c) Duration of use

Endometrial Cancer

(1) Base Rate Risk
(a) Cancer of the uterus
(b) 3,000 women die each year from endometrial
cancer
(c) Is related to hormone replacement therapy
(d) Mortality related to endometrial cancer is smaller
(e) Symptoms of endometrial cancer is bleeding
(0 A woman who experiences bleeding after
menopause should see her health care
professional
(g) Lower base rate
(2) Effect of HRT
(a) ERT increases risk 6-fold
(b) PERT negates the increased risk. No increased
risk ‘
(c) Diagnosed earlier
(d) Occurrence increases but mortality has not
Side effects
(1) Experienced by more women more severely with PERT
(2) Sequential PERT usually causes resumption of cyclic
bleeding
(3) Neither will allow you to become pregnant again

10.

155

Question and Answer Period (10 Minutes) (Clinician)
Clusters Narrative (10 Minutes) (Social Scientist)
a. Factors included in the study

(1) Hot ﬂashes

(2) Osteoporosis

(3) Endometrial Cancer

(4) ERT vs PERT
b. Women look at the pieces of information differently
c. Three major groups of women

1. Group I: 120 Women highly inﬂuenced by hot ﬂashes

ii. Group II: Were interested in hot ﬂashes and osteoporosis about
equally and were concerned about endometrial cancer.

iii. Group HI: (40) Hot ﬂashes extremely important to them. Would
not want to resume monthly blwding.

d. The reason for presenting this information is that you may see yourself in
one of these groups (give or take a little). It gives you an idea of how
some other women have weighed these factors

Risk Discussion (20 Minutes) (Social Scientist)

a. Factors which can affect base rate risk
1. Fracture

(1) Weight-bearing exercise (at least 3 times per week for 20
minutes or more)

(2) Diabetes Mellitus

(3) Age

(4) White! N on-white

(5) Smoking

(6) Hypertension

(7) Calcium intake

(8) Over/Underweight

(9) Oophorectomy

ii. Heart Disease

(1) Family History

(2) Age (over 50)

(3) Smoking ,

(4) Exercise (At least 3 times per week for 20 minutes or
more)

(5) Diabetes Mellitus

(6) Hypertension

iii.

156

Endometrial Cancer Risk

(1) Overweight (15 pounds or more)
(2) Never pregnant

(3) Previous difﬁculty getting pregnant
(4) Irregular periods

(5) Excessive menstrual blwding

11. Assess personal values (10 Minutes) (Social Scientist)
Factors to consider

a.

1.
ii.

iii.

iv.
v.
vi.

Menopausal symptoms
Fractures due to osteoporosis
Heart disease

Endometrial cancer

Cyclic bleeding

Possible side effects

Other factors

157

Intervention Program B
Session III

1. Personal Plan Form and Activity Record Calendar (15 Minutes) (Clinician)

Communicating with Health Care Professionals (10 Minutes) (Social Scientist)

a. Two main aspects of effective communication:

i.

Effective listening allows:
(1) Carry out health care provider’s recommendations
(2) Help you understand information
(a) Example
Effective information-giving and question-asking allows:
(1) Give info health care provider needs to give best care
(2) Get back answers that you understand

b. Preparing for the visit

i.
ii.

iii.

iv.

Allows effective use of appointment time

Allows you to assemble the information your health care provider
will need

Indicate if extra time is needed for a visit when the appointment is
made

(1) Example

See brochure for list of activities to prepare for visit

c. During the visit

i.
ii.

1V.

vi.

vii.

viii.

At beginning say have questions
Describe your symptoms speciﬁcally
(1) Example
(2) Describe:
(a) duration of the symptom
(b) ’ when it occurs
Say what you think may be the cause of your problem
Listen carefully to questions
Answer the questions completely and directly
Ask the questions you have prepared ahead of time
(1) Example
Write down the answers to questions and information you will
need later
(1) Example
If your health care provider makes a recommendations and you
know right off that this 1s something that you won t be able to do.
(1) Examples

 

158

d. After the visit

1. Review the visit after you leave the ofﬁce
(1) When you get home, ask yourself if your questions were
answered

(a) If yes, carry out the recommendations
(b) If your goals weren’t met you can:
(i) Call your health care provider and ask for
further information or clariﬁcation
(ii) If you decide that you need to see a
specialist, you can call up and ask for a
referral
(iii) Ask your friends for some suggestions of
referrals to other health care providers.
Data Collection (25 Minutes)
Knowledge
Satisfaction with Decision
Health Care Self-Efﬁcacy
Barriers to Decision Scale
Outcome Expectations Scale
Subjective Norms Scale
Behavioral Intentions Scale

rhea-9.61»

159

Outline of Intervention C
Session I

Bold denotes topics in knowledge instrument

y...

Introduction and informed consent forms (5 minutes)
Data Collection (30 minutes)

53‘2” rm 99 9‘?

i.

Problem Signiﬁcance Assessment

Sociodemographic

Perceptions (Menopause Problem Scale and Control Scale)
Symptoms Instrument

Symptom Management/Self-Care Instrument

Knowledge Instrument

Satisfaction with Health Care Provider Instrument (Part I)
Satisfaction with Decision Instrument

Health Care Self-efﬁcacy Instrument

Physical changes of menopause (20 Minutes) (Clinician)
[Participants may follow along in the brochure]

a.

What is menopause?
1. Deﬁnition of menopause
ii. Menopause is a normal transition
iii. Changes at midlife
iv. Surgical menopause
When does menopause occur?
1. Average age of menopause is 50 ,
What happens to my body at menopause?
i. Childbearing years
(1) Ovary contains follicles which decrease in number over

life-span ~'

(2) Two glands in the body to know about
(a) Pituitary
(b) Ovary

(i) Responds to the signals from the pituitary
(ii) Produces two primary hormones
1) estrogen
2) progesterone
(3) Two hormones that drive the ovary to do its job
(a) FSH
01) LB
(4) Fwdback cycle of hormonal changes during menstrual
cycle

160

(5) Hormonal changes at menopause
(6) Feedback cycle of childbearing years interrupted
(7) Levels of pituitary hormones rise
(8) Decrease in production of estrogen and progesterone
ii. Menstrual periods change
(1) Pattern of change varies between women
iii. Ovulation can occur and pregnancy can result even without a
menstrual period
(1) Should use birth control for one full year

Will I experience symptoms at menopause? (Clinician)

a.
b.

Orr-map

Sheehy quote
Hot flashes one of most common symptoms
i. Related to estrogen deficiency
80% of women experience symptoms
20% of women seek help for symptoms
Description of experience of hot ﬂashes
Estrogen reduces hot ﬂashes in 95-96% of cases
Other changes due to estrogen decrease
1. breast
ii. vaginal lining
(1) lining becomes less thick
(2) dryer, thinner, shorter
(3) Can result in burning or stinging with urination
(a) difﬁculty or pain with intercourse
(4) Can result in increased chance of vaginal infections
Other things which occur around the time of the menopause

i. anxiety
ii. weight gain
iii. bloating

iv. Not shown to be directly due to estrogen deﬁciency

Increased Risks at Menopause (Clinician)

a.
b.

Two diseases that increase: Osteoporosis and heart disease

Osteoporosis

i. Deﬁnition

ii. Around 50,000 women die per year as a result of osteoporotic
fractures

iii. Fracture may result in nursing home placement
iv. Osteoporosis a calcium disease and protein disease
v. Osteoporosis accelerates rate of development at menopause
vi. Women most at risk for fractures:
(1) Thin, white women who smoke and do not exercise

regularly

6.

7.

161

Heart Disease

i. Maior cause of death for men and women

ii. 390,000 women per year die of heart disease

iii. Heart disease in women tends to increase in rate of
development around the time of the menopause

Self-Care Strategies (10 Minutes) (Clinician)

a.

For hot ﬂashes:
i. Eat a good, healthy diet
(1) To help hot ﬂashes decrease stimulatory types of foods
such as spices
ii. Wear layered clothing
( 1) Can take of layers if get hot
Skin (like the bones) gets thinner and dryer around the menopause
i. Natural skin lubricants like lanolin or other non-perfumed
lubricants may help
Vaginal dryness
i. Use water-soluble lubricant
For many people these self-care strategies are enough to take care of the
symptoms and signs that occur at the time of the menopause but don’t do
anything do anything about the long-term, killing complications of
osteoporosis and coronary heart disease.

Pros and Cons of Hormone Replacement Therapy (30 Minutes) (Clinician)

a.

Five areas to think about related to HRT (SCOBES)

i. S--Symptoms (menopausal symptoms)

ii. C--Coronary heart disease

iii. O--Osteoporosis

iv. B--Breast cancer

v. E--Endometrial cancer

vi. S--Side effects of HRT

Deﬁnition of base rates

i. Frequency of occurrence

ii. Important because it addresses ”double what?" ”What are these
relative to each other?"

162

c. Two main regimens
i. Estrogen-Only

(1)
(2)

(3)

(4)
(5)

Typical regimen

Whether it would be prescribed for you would depend on
your personal history and medical examination

Side effects include:

(a) headache

(b) bloating

' (c) weight gain

(d) breast tenderness
No resumption , of cyclic blwding with estrogen-only
No chance of a woman getting pregnant

ii. Estrogen combined with Progesterone

(1)
(2)

Typical regimens

Side effects

(a) More women experience side effects with
combination therapy than with estrogen-only

(b) Side effects similar to those experienced with

estrogen:
(i) headache
(ii) bloating

(iii) weight gain
(iv) breast tenderness

(1. Impact of Hormone Replacement Therapy on factors
i. Symptoms

(1)

(2)

Recap of Base Rate Risk

(a) Hot flashes 80% of women experience them but
only 20% seek help for them

(b) Vaginal atrophy—becomes dryer and more easily
iniured and may cause difficulty with intercourse

Effect of HRT

(a) Both ERT and PERT relieve hot flashes by about
90-95%

(h) Both regimens relieve vaginal dryness

(c) Clinical data has not found that ERT or PERT
affect other symptoms such as mood swings

iii.

iv.

163

Coronary Heart Disease
(1) Recap of Base Rate Risk
(a) Most frequent killer of both men and women.
394,000 women die each year
(b) Heart disease occurs somewhat later in women than
in men, however.
(2) Effect of HRT
(a) Estrogen-only cuts risk in half
(b) Addition of progestogen has been questioned.
(c) There are less data about the effect of progestogen.
Most research has been done on .625 of estrogen
Osteoporotic Fractures
(1) Recap of Base Rate Risk
(a) 50,000 women die each year from complications
of fractures.
(b) Fractures occur predominantly among women and
increase greatly with age. By age 80, 1 out of every
3 women will have had a hip fracture. Of those hip
fractures, 20% will result in death
(2) Effect of HRT
(a) Risk is cut in half with ERT and PERT
Breast Cancer
(1) Base Rate risk
(a) 41,000 women die each year from breast cancer
(b) Every woman over age 50 ought to have a
mammogram every year
(c) Breast cancer is usually treated with surgery, and/or
chemotherapy
(2) Effect of HRT
(a) Do not have the bottom line on this
(b) Some studies indicate an increased risk, some
decreased risk, some no change in risk
(i) Increase is usually 1.3 to 1.8 increased risk
(3) Important to know the following when you read reports of

studies:
(a) What kind of estrogen
(b) Dose

(c) Duration of use

Ve

vi.

164

Endometrial Cancer

(1) Base Rate Risk
(a) Cancer of the uterus
(b) 3,000 women die each year from endometrial
cancer
(c) Is related to hormone replacement therapy
((1) Mortality related to endometrial cancer is smaller
(e) Symptoms of endometrial cancer is bleeding
(1') A woman who experiences bleeding after
menopause should see her health care
professional
(g) Lower base rate
(2) Effect of HRT
(a) ERT increases risk (Hold
(1)") PERT negates the increased risk. No increased
risk
(c) Diagnosed earlier
(d) Occurrence increases but mortality has not
Side effects
(1) Experienced by more women more severely with PERT
(2) Sequential PERT usually causes resumption of cyclic
bleeding '
(3) Neither will allow you to become pregnant again

p—s
0

PM“?

165

Intervention Outline C
Session II

Question and Answer Period (10 Minutes) (Clinician)

Clusters Narrative (10 Minutes) (Social Scientist)

a. Aspects to combine into decision
i. Your risk
ii. What’s important to you

b. Factors included in the study

(1) Hot ﬂashes

(2) Osteoporosis

(3) Endometrial Cancer
(4) ERT vs PERT

c.‘ Women look at the pieces of information differently

(1. Three major groups of women
i. Group 1:120 Women highly inﬂuenced by hot ﬂashes
ii. Group 11: Were interested in hot ﬂashes and osteoporosis about

equally and were concerned about endometrial cancer.
iii. Group III: (40) Hot ﬂashes extremely important to them. Would
not want to resume monthly bleeding.

e. The reason for presenting this information is that you may see yourself in
one of these groups (give or take a little). It gives you an idea of how
some other women have weighed these factors

Personal Risk Assessment (20 Minutes) (Social Scientist)

a. Complete forms
i. Bar length represents chance of dying W
ii. Bar length shows the chance of this being your cause of death (not

necessarily soon)
iii. Questions on risk assessment represent factors which can modify

base rate
iv. The risk assessment is simply a guide; estimates
b. Transfer to the relevance chart

Problem Signiﬁcance Assessment (10 Minutes) (Social Scientist)
Relevance Chart (15 Minutes) (Social Scientist)

Scenarios (30 Minutes) (Clinician)

a. Case 1

b. Case 2

Decision (Clinician)

166

Session III

1. Personal Plan Form and Activity Record Calendar (15 Minutes) (Clinician)

Communicating with Health Care Professionals (10 Minutes) (Social Scientist)

a. Two main aspects of effective communication:

1.

ii.

Effective listening allows:
(1) Carry out health care provider’s recommendations
(2) Help you understand information
(a) Example
Effective information-giving and question-asking allows:
(1) Give info health care provider needs to give best care
(2) Get back answers that you understand

b. Preparing for the visit

1.
ii.

iii.

iv.

Allows effective use of appointment time

Allows you to assemble the information your health care provider
will need

Indicate if extra time is nwded for a visit when the appointment is
made

(1) Example

See brochure for list of activities to prepare for visit

c. During the visit

1.
ii.

iii.
iv.

vi.

vii.

viii.

At beginning say have questions
Describe your symptoms speciﬁcally
(1) Example
(2) Describe:
(a) duration of the symptom
(b) when it occurs
Say what you think may be the cause of your problem
Listen carefully to questions
Answer the questions completely and directly
Ask the questions you have prepared ahead of time
(1) Example
Write down the answers to questions and information you
will need later
(1) Example
If your health care provider makes a recommendations and you
know right off that this is something that you won’t be able to do.
(1) Examples

 

 

167

d. After the visit

i. Review the visit after you leave the ofﬁce
(1) When you get home, ask yourself if your questions were
answered

(a) If yes, carry out the recommendations
(b) If your goals weren’t met you can:
(i) Call your health care provider and ask for
further information or clariﬁcation
(ii) If you decide that you need to see a
specialist, you can call up and ask for a
referral
(iii) Ask your friends for some suggestions of
referrals to other health care providers.
Communicating with Health Care Provider Role Play ( 15 Minutes) (Social

Scientist)
a. Instructions:
i. For those of you who are playing health care providers, remember

you have a schedule to keep and patients in the waiting room but
you are seeing one of your regular patients.
ii. For those of you who are playing patients, you have called ahead
- and informed the receptionist you want to make an appointment to
discuss menopause
b. Afterwards:
i. For those who were playing health care providers, how did it go?
How did it work to have this patient in here asking you questions?
ii. For those who were patients, how did it go?
Data Collection (25 Minutes)
Knowledge
Satisfaction with Decision
Health Care Self—Efﬁcacy
Barriers to Decision Scale
Outcome Expectations Scale
Subjective Norms Scale
Behavioral Intentions Scale

erase-.0571»

 

APPENDD( D

 

168

APPENDIX D
MEASURES

ID 1'
DECISION MAKING IN MENOPAUSE STUDY

 

Mark your answer for each question on the answer sheet provided.
These sheets will be scored by machine so it is very important that
you completely darken in your answer choice in the proper location
on the answer sheet with the pencil provided. Do not leave any stray
pencil marks on the sheet outside of your answer space and makesure

. ‘1]... ° 1 1. "11 .-H-. .".\ -. e . U'I°I
W.

For example, consider this question:
The ﬁrst month of the year is:

l = March
2 = July

3 = January
4 = October

To choose January, you should darken in 3 on your answer sheet as
is done below.

2 assaeaaaba
At the top of each page is a reminder to check to be sure you are

ﬁlling in the correct row on your answer sheet for the questions.

This publication is part of 'Women's Judgments of Estrogen Replseemelu Therapy." which is supported by Grant 0R0! NR0124S-04A2
from the Nations] Carter for Nursing Resesrweh. Nelsons! last'sues of Health. its com are solely the responsibility oflher ashore and
bumﬂywtheofﬁcal vaewsoflheNstiealeeuerfor Nuniagllesareh.

169

DECISION MAKING IN MENOPAUSE STUDY
BARRIERS TO PARTICIPATION IN HEALTH CARE DECISIONS

Some people want to help make decisions about their health are but an not. Other factors affect how
much input they an have in decisions. These questions are about how much you feel that these factors

WWW.
How much do these factors limit your participation in medial decisions? Use the scale below:

 

l 2 3 4 5
Not at All Somewhat A Great Deal 6 = NOT APPLICABLE

24. How much does your doctor (or regular health are practitioner) limit your participation in
medial decisions?

25. How much do your payment options (such as health insurance) limit your participation in medial
decisions?

26. How much does your health state or personal risk factors limit your participation in medial
decisions? .

27. How much doeslack ofaccess to accurate information limit your participation in medial
decisions?

28. How much does conﬂicting medial advice limit your participation in medial decisions?

29. How much do other aspects of the health are system limit your participation in medial
decisions? Plase specify:

 

30. How much do other factors limit your participation in medial decisions? Please
specify:

 

° 1992 Michigan State University
College of Nursing

170

DECISION MAKING IN MENOPAUSE STUDY
OUTCOME EXPECTATIONS SCALE

These statements are about patient involvement in health are decisions in general. W
mfonmunnmionmthcunxnlxcmcnuimbuhmem Then statements are about the outcome
of patient participation, that is, W that patient participation will result in the outcomes
described? Please answer questions 31-37 using the following response sale:

1 2 3 4 5
Extremely Unlikely Neither Likely Likely Extremely
Unlikely Nor Unlikely Likely

Patient participation in medial decisions and choices results in:
31. Medial procedures that are unnecessary for the patient
32. Medial procedures that are harmful for the patient
33. The wrong mediations for the paient
34. The right medial tests and examinations for the patient
35. The right treatment for the patient
36. The right medial decisions for the patient
37. The best medial are

Outcome Values Sale

Pleaseusethefollowingsaletoanswerquestions38-44. Foreachoutcome,indiatehowgoodorhad
itwouldbeform:

1 2 3 4 5
Extremely Bad Neither Bad Good Extremely
Bad Nor Good Good

38. Medial procedures that are unnecessary for me
39. Medial procedures that are harmful for me

40. The wrong mediations for me

° 1992 Michigan State University
College of Nursing

171

Check your answer sheet. You should now be ﬁlling in row 41

Please use the following sale to answer questions 4144. For each outcome, indieate how good or bad
it would be for you:

1 2 3 4 5
Extremely Bad Neither Bad Good Extremely
Bad Nor Good . Good

41. 'I‘herightmedialtestsandexaminationsforme
42. The right treatment for me
43. The right medial decisions for me

44. The best medical are

° 1992 Michigan State University
CollegeofNursing

172

DECISION MAKING IN MENOPAUSE STUDY
SUBJECI'IVE NORMS QUESTIONNAIRE

Some people think that the decision to take hormone replacement therapy should be made by the
clinician while other people think that the decision should be made by the patient. Indiate how you
believe each of the following people think the decision should be made. Use the scale below to answer
questions 45-49. (Clinician means your doctor or regular health are practitioner).

1 2 3 4 5
Clinician Clinician After Clinician Patient After Patient
Considering and Patient Considering

5 = NOT APPUCABLE Patient's Opinion Together Equally Clinician’s Opinion
45. Your doctor (or regular health are practitioner) thinks the decision should be made by:
46. Your spouse, partner or signiﬁant other thinks the decision should be made by:
47. Your parents, relatives and children think the decision should be made by:
48. Your friends, peers and classmates think the decision should be made by:
49. Other people you think inﬂuence your decision (Please specify: ) think
the decision should be made by:
MOTIVATION TO COMPLY
Please use the following scale to answer questions 50-54 below.

L J 3 4 5 5 = NOT APPLICABLE
Not at All Very Much

50. Howmuchdoyouwanttodowhatyourdoctor(orregularhealtharepractitioner)wantsyouto.
51. How much do you want to do what your spouse, partner or signifant other wants you to do?

52. How much do you want to do what your parents, relatives and children want you to do?

53. How much do you want to do what your friends, peers or classmates want you to do?

54. How much do you want to do what the other persons you listed above want you to do? (Please
specify: )

 

° 1992 Michigan State University
College of Nursing

173

DECISION MAKING IN MENOPAUSE STUDY
BEHAVIORAL INTENTIONS QUESTIONNAIRE

To what extent do you man or plan to do the following activities related to your next visit to your
doctor (or regular health are practitioner)? Please use the following sale to answer questions 55-62.

 

2 3 4 5

Not At All Completely

55. Carefully consider and understand my mum related to the tradeoffs of hormone replacement
therapy or other are.

56. Gather the information I need to make an informed decision about hormone replacement therapy
or other are.

57. Carefully consider my W513: related to hormone replacement therapy or other are.

58. Ask for an extended appointment to discuss my concerns about my health are.

59. Prepareformyvisitbythinkingaboutthequestionslwanttoask,Wandbring
the list with me to the visit.

60. At the beginning of the visit, tell my doctor/practitioner that 1 have questions I would like to
discuss.

61. Duringmyvisit,askthequestionslhadprepared,repeattheanswerstobesurelunderstand

62. Aftermy visit,reviewthevisit,followthetratmentsorrecommendationsagreeduponandall
my doctor/practitioner if I have additional questions or unexpected side effects.

63. Regarding the decision whether or not to take hormone replacement therapy, how do you

Worplanthatthedecisionwillbemade?chmmne. (Clinician meansyourdoctoror
regular health are practitioner).

1- Iintendtohavemyelinicianmakethedecision,usingallthat'slmownabouthormone
replacement therapy.

2 - I intend to have my clinician make the decision but strongly consider my opinion.

3 a lintend that myclinicianandlwillmakethedecision together, onanequal basis.

4 - I intend to make the decision myself but stongly consider the clinicians’opinion.

5 - I intend to make the decision myself, using all I know or learn about hormone replacement

therapy

° 1992 Michigan State University
College of Nursing

174

DECISION MAKING IN MENOPAUSE STUDY
HEALTH CARE SELF-EFFICACY MEASURE

Beloware8sinrationsyoumightexperienceasapatient. Onasaleof1t010,circlethe
numbewhichbeﬂdamibahowanﬁdatyouwouﬂfedmyomabﬂitywhandlemesimadon
mm.

 

1 2 3 4 5 6 7 8 9 10

Not at all Conﬁdent Extremely Conﬁdent
HOW CONFIDENT ARE YOU-
1. —1hatwhayouneedmedialareyouanpmvideimportantinputaboutwhatwinbe
doneto you?

 

l 2 3 4 5 6 7 8 9 10
Not at all Conﬁdent Extremely Conﬁdent

2. --In your ability to understand any medial procedures which might be done to you?

 

l 2 3 4 5 6 7 8 9 10
Not at all Conﬁdent Extremely Conﬁdent

3. -That when you need medial are, your input would help you get the best are?

 

l 2 3 4 5 6 7 8 9 10
Not at all Conﬁdent Extremely Conﬁdent

4. -That once a health professional has begun doing a medial procedure with you, you can
ask important questions about the procedure?

 

l 2 3 4 5 6 7 8 9 10
Not at all Conﬁdent Extremely Conﬁdent

175

-That you would know how to help your health are provider decide what medial
procedures you should get?

 

1 . 2 3 4 5 6 7 8 9 10
Not at all Conﬁdent Extremely Conﬁdent

«Thatyouanaskyourhalthareprovider important questions about your healthafter
amedialexambeforeyouaretold?

 

1 2 3 4 5 6 7 8 9 10
Not at all Conﬁdent . Extremely Conﬁdent

-‘1'hatyouwouldbeabletounderstand information about whatamedial procedure
woulddotoyou?

 

l 2 3 4 5 6 7 8 9 10
Not at all Conﬁdent Extremely Conﬁdent

-That, given many choices about what would be good for your health, you could make
the best choice?

 

l 2 3 4 5 6 7 8 9 10
Not at all Conﬁdent Extremely Conﬁdent

176

DECISION MAKING IN MENOPAUSE STUDY
SATISFACTION WITH DECISION MAKING PROCESS INSTRUMENT

You have been considering whether to consult your health care provider
about hormone replacement therapy. Answer the following questions about
your decision. Please indicate to what extent each statement is true for you
AT THIS TIME by circling your answer.

1. I am satisﬁed that I am adequately informed about the issues important
to my decision.

 

Strongly Disagree Neither Agree ' Agree Strongly
Disagree . Nor Disagree Agree

2. The decision I made was the best decision possible for me personally.

Strongly Disagree Neither Agree Agree Strongly
Disagree ' Nor Disagree Agree

3. I am satisﬁed that my decision was consistent with my personal values.

Strongly Disagree Neither Agree Agree Strongly
Disagree Nor Disagree Agree

4. I expect to successfully carry out (or continue to carry out) the decision
I made.

Strongly Disagree Neither Agree Agree Strongly
Disagree Nor Disagree Agree
5. I am satisﬁed that this was my decision to make.

Strongly Disagree Neither Agree Agree Strongly
Disagree Nor Disagree Agree

° 1992 Michigan State University
College of Nursing

10.

177

I am satisﬁed with my decision.

Strongly Disagree Neither Agree Agree . Strongly

Disagree Nor Disagree Agree
UNDERSTANDING MY DECISION

This decision is hard for meto make.

Strongly Disagree Neither Agree Agree Strongly

Disagree Nor Disagree Agree

It’s clear what choice is best for me.

Strongly Disagree Neither Agree Agree Strongly
Disagree Nor Disagree Agree

I’m unsure what to do in this decision.

Strongly Disagree Neither Agree Agree Strongly
Disagree Nor Disagree Agree -

I know all the possible choices open to me in protecting my health after

menopause .

Strongly Disagree Neither Agree Agree Strongly
Disagree Nor Disagree ' Agree

° 1992 Michigan State University L
College of Nursing

ll.

12.

13.

14.

178

I understand the risks and beneﬁts of taking hormone replacement
therapy.

Strongly Disagree Neither Agree Agree Strongly
Disagree Nor Disagree Agree

RESTRICTIONS

My health state restricted my decision related to hormone replacement
therapy.

Strongly Disagree Neither Agree Agree Strongly
Disagree Nor Disagree Agree
My health state restricted my decision related to calcium intake.
Strongly Disagree Neither Agree Agree Strongly
Disagree Nor Disagree Agree
My health state restricted my decision related to exercise.

Strongly Disagree , Neither Agree . Agree Strongly
Disagree Nor Disagree Agree

° 1992 Michigan State University
College of Nursing

179

15. On a scale of 1 to 10 with 1 = No conﬁdence at all and 10 = complete
conﬁdence, circle the number which indicates your conﬁdence in the
correctness of your decision. 1 signiﬁes no conﬁdence at all. 10
signiﬁes complete conﬁdence; 2 through 9 indicate some middle level
of conﬁdence.

 

12 3 4 5 6 7 8 910

No Conﬁdence Complete
at All Conﬁdence

° 1992 Michigan State University
College of Nursing

180

Behavior Self-Report

. Please answer the following questions WW

brochure. If you have not visited your health care provider since completing the
program, please check here: and skip these questions.

 

To what extent did you do the following activities related to your last visit to
your doctor (or regular health care practitioner)?

1. I carefully considered and understood my nines related to- the tradeoffs
of hormone replacement therapy or other care.

1 2 3 4 5
Not At All _ Completely

2. I gathered the information I needed to make an informed decision about

hormone replacement therapy or other care.

1 2 ' 3 4 5
Not At All - Completely

3. I carefully considered my mum related to hormone replacement
therapy or other care.

1 2 3 4 5
Not At All Completely

181

I asked for an extended appointment to discuss my concerns about my
health care.

1 2 3 4 5
Not At All Completely

I prepared for my visit by thinking about the questions I wanted to ask,
W and bringing the list with me to the visit.

1 2 . 3 - 4 5
Not At All , Completely

At the beginning of the visit, I told my doctor/practitioner that I had
questions I would like to discuss.

l 2 3 4 - . 5
Not At All Completely

During my visit, I asked the questions I had prepared, repeated the
answers to be sure I understood them and WW

necessary-

1 2 3 4 5
Not At All Completely

After the visit, I reviewed the visit, followed the treatments or
recommendations agreed upon and called my practitioner if 'I had
additional questions or experienced unexpected side effects.

I‘J
U
h

1 5
Not At All Completely

182

Which of the following choices best describes how the decision about
whether or not you would take hormone replacement therapy mug
made? (Clinician means your doomr or other health care practitioner).
(Circle One). -

, 1 I- The clinician made the decision, using all that’s known about
hormone replacement therapy. -

The clinician made the decision but strongly considered my
opinion.

The clinician and I made the decision together, on an equal basis.

I made the decision, but strongly considered the clinician’s opinion.
I made the decision using all I know and learned about hormone
replacement therapy.

Not.Applicable. I: am not yet experiencing menopause.

2

MAN
ll llll II

a
ll

 

APPENDD( E

183 '

Appendix E

Scale Means and Standard Deviations

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Scale It Mean SD.

Self-Efﬁcacy T1 244 7.6 1.67
Self-Efficacy T2 248 8.1 1.45
Self-Efﬁcacy T3 184 8.1 1.31
Barriers 247 2.32 .77
Outcome Expectations 248 18.5 3.60
Subjective Norm 243 10.5 3.67
Behavioral Intention 248 4.4 .60
Behavior Self-Report 67 3.8 .83
‘ Satisfaction with Decision T3 184 4.0 .63

 

 

 

187 ‘

APPENDIX F

184’

APPENDIX F
Calculation of Correction for Unreliability

1' = 1'
XYtrue xy obt

an a”

 

r = True correlation
xy true

r
xy obt
«xx = Reliability of measure x

= Observed correlation

‘Eyy = Reliability of measure y

From : Edwards, A- L. (1954). Statisﬁcalmeﬂmdaionthehehaxioralsciences
New York: Holt, Rinehart and Winston.

185 '

APPENDD( G

185 ‘

APPENDIX G
COMPUTATION S FOR REGRESSION

From Cohen 8: Cohen (1983); p. 329:

Y=wa+ Bcc+ Bee+ chwc+ Bwewe+ A

Where:
w = A quantitative variable (Self-Efﬁcacy at Time 1)

c, e = A set of contrast-coded variables representing the variance
in a nominal variable with 3 experimental groups (Contrast 1,
Contrast 2)

wc, we = A set of variables representing the variance due to the interaction
between the contrast-coded experimental group and the
quantitative variable (Interaction 1, Interaction 2)

A = Constant

Calculations:

Equationl

Y = .57(w) + .15(c) + 1.12(e) - .03(wc) — .13(we) + 3.79
where:

Y = Self-Efficacy at Time 2

w = Self-Efﬁcacy at Time 1

c = Contrast 1 [Group C (+1) against Groups A (—.5) and B (-.5) pooled]

e = Contrast 2 [Group B (+1) against Group A (-1)]

wc = Interaction 1 [Product of Self-Efficacy at Time 1 and Contrast Variable 1]
we = Interaction 2 [Product of Self-Efﬁcacy at Time 1 and Contrast Variable 2]
A = Constant

186 ‘
Equation 1 Reordered:
Y = .57(w) - .03(wc) — .13(we) + .15(c) + 1.12(e) + 3.79

= [.57 — .03(c) — .13(e)]w + .15(c) + 1.12(e) + 3.79

Substituting coding for Experimental Group:

A

YA = [.57 — .03(—.5) — .13(—1)]w + .15(—.5) + 1.12(—1) + 3.79
= .715(SET1) + 2.595

9, = [.57 — .03(—.5) — .13(+1)]w + .15(—.5) + 1.12(-1) + 3.79
= .455(SET1) + 4.835

YC = [.57 - .03(+1 )- .13(O)]w + .15(+1) + 1.12(0) + 3.79

= .540(SET1) + 3.940

 

APPENDD( H

 

187

APPENDIX H
Significance Test for Correlated r’ s

 

 

 

Correlationatnl .
Self-Eff 1 Self-Eff 2 Self-Eff 3

Self-Eff 2 .66

(65)
Self-Eff 3 .50 .52

(65) (67)
Behavior .10 .05 .34

(65) (67) (67)

 

Correlation: Behavior & Self-Efﬁcacy Time 2 = rxy = .05; n=67
Correlation: Behavior & Self-Efﬁcacy Time 3 = rvy = .34; n=67

Correlation: Self-Efficacy Time 2 8: Self-Efﬁcacy Time 3 = rxv = .52; n=66
IRI = 1—r2xy—r2‘fy-r2xv+2rxyrvyrxv

 

E: r :r t=(rxv-rw)‘l(n-1)(1+r)

\/2C- 191111.; (1 1.3)

I R l = 1 — (.05)2— (.34)2— (.52)2+ 2(.05)(.34)(.52) = .63

 

?=.05+.34 t=(.—05 .-34)~J(67 1)(1+.5'2'

¢2 (667-31) .63+(.2) (1- .52)3
7..

From: Cohen,]., 8: Cohen, P. (1983).

 

(2nd ed.). Hillsdale, NJ:
Lawrence Erlbaum. Pp. 56—57.