Sei sulla pagina 1di 161

THE ROLE OF ANXIETY IN THE RELATIONSHIP BETWEEN BREATHING

EFFORT AND CANCER-RELATED DYSPNEA SENSATION

by

CHIOU-FANG LIOU

Submitted in partial fulfillment of the requirements

For the degree of Doctor of Philosophy

Dissertation chair: Dr. Barbara J. Daly

Frances Payne Bolton School of Nursing

CASE WESTERN RESERVE UNIVERSITY

January 2008
CASE WESTERN UNIVERSITY
SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

Chiou-Fang Liou

Candidate for the P.h.D degree*.

(signed) Barbara J. Daly


(chair of the committee)

M. Linda Workman

Chris Winkelman

Mariana Petrozzi

(date) November 01, 2007

*We also certify that written approval has been obtained for any
Proprietary material contained therein.
Dedication

This dissertation is dedicated to

~ The memories of my late maternal grandmother, Gai Chen-Tsou, who loved me,
believed in me, and inspired me, in so many ways.

~ Patients and their family members who made this study possible.
ii

Table of Content

Abstract……………………………………………………………………………. 1

Chapter One: Background and Significance……………………………………… 3


Statement of Problem…………………………………………………………. 3
Background……………………………………………………………………. 5
Significance of Study…………………………………………………………. 14
Theoretical Work……………………………………………………………… 17
Independent Variable…………………………………………………………. 21
Dependent Variable…………………………………………………………… 22
Moderator Variable…………………………………………………………… 23
Assumption …………………………………………………………………… 24
Hypotheses……………………………………………………………………. 24
Research Questions…………………………………………………………… 25
Summary………………………………………………………………………. 25

Chapter Two: Literature Review………………………………………………… 27


Breathing Effort………………………………………………………………. 28
Dyspnea Perception………………………………………………………....... 33
The Relationship between Anxiety and Dyspnea Sensation…………………. 42
Relationship among Dyspnea, Breathing Effort, and Anxiety……………….. 47
Direction of the relationship between Dyspnea and Anxiety………………… 48
Summary…………………………………………………………………….. 49

Chapter Three: Methodology……………………………………………………. 51


Study Design…………………………………………………………………. 51
Sampling Frame………………………………………………………………. 51
Sampling Plan…………………………………………………………………. 52
Sample Size Determination…………………………………………............ 52
Inclusion Criteria and Rationale……………………………………………. 56
Data Collection Procedure……………………………………………………. 58
Measurement Validation…………………………………………………........ 61
Instrumentation……………………………………………………………….. 69
Data Management…………………………………………………………….. 70
Analyses Plan…………………………………………………………………. 71
Methods of Protecting Human Subjects………………………………………. 78
Summary………………………………………………………………………. 79
iii

Chapter Four: Results……………………………………………………………. 80


Sample………………………………………………………………………… 80
Assessment of the pulmonary function status………………………………… 84
Validation of the anxiety subscale of cancer dyspnea scale……………........... 86
Variable Description…………………………………………………………… 87
Demographic Influences………………………………………………………. 88
Correlation Evaluation………………………………………………………… 90
Research Questions……………………………………………………………. 91
Summary………………………………………………………………………. 98

Chapter Five: Discussion…………………………………………………………… 99


Summary of the Present Study………………………………………………….. 99
Sample Description……………………………………………………………… 100
Variable of Interests……………………………………………………….......... 101
Finding Interpretations………………………………………………………….. 108
Study Model versus Analytical Model…………………………………………. 114
Strength…………………………………………………………………………. 117
Limitations………………………………………………………………………. 118
Implications & Recommendations………………………………………............ 120
Implication for Theory……………………………………………………….. 120
Recommendations for Research……………………………………............... 122
Recommendations for Clinical Practice……………………………………… 126
Recommendations for Nursing Education…………………………………… 126
Subject Burden………………………………………………………………….. 127
Difficulty………………………………………………………………………... 128
Conclusion………………………………………………………………………. 129

Appendix…………………………………………………………………………… 131
References………………………………………………………………………….. 141
iv

List of Tables

Title Page
Table 2-1. Three Physiological Manipulated Experimental
Conditions…................................................................................................ 30

Table 2-2. Physiological Oriented Evidence of Breathing Effort…………. 32

Table 2-3. Operational Perceived Dyspnea……………………………....... 36

Table 2-4. Language of Dyspnea in Non-Malignant and Malignant


Population…………………………………………………………………… 41

Table 2-5. Literature on the Relationship between Anxiety and


Dyspnea……………………………………………………………………… 45

Table 2-6.Literature on the Relationship among Breathing Effort, Dyapnea,


and Anxiety………………………………………………………………….. 48

Table 4-1. Demographic Summaries……………………………………. 83

Table 4-2. Demographic Summaries…………………………………….. 83

Table 4-3. Summaries of Differences on Pulmonary Function Status


between Baseline and Examine Day…………………………………… 86

Table 4-4. Description Summaries on Two Measure of Anxiety………. 87

Table 4-5. Description Summaries of Variables………………………… 88

Table 4-6. The Relationship between Age and Dyspnea Sensation,


Breathing Effort, and Anxiety………………………………………… 89

Table 4-7. Summaries of Differences on Dyspnea Sensation, Self Rating of


Breathing Effort, and Anxiety between Male and Female…. 89

Table 4-8. Correlations among Inspiratory Pressure, Expiratory Pressure,


FEV1%, FVC%, Dyspnea Sensation, Breathing Effort, and
Anxiety……………………………………………………………………. 91

Table 4-9. Summary of Multiple Regression Analysis for Breathing Effort


Predicting Dyspnea Sensation…………………………………… 93

Table 4-10. Summary of Hierarchical Moderator Regression Analysis for


Variables Predicting Dyspnea Sensation……………………………. 97
v

Table 5-1. Respiratory Pressure in this Sample Population versus


Hypothetical Parallel Normal Adult…………………………………….. 105

Table 5-2.
Ventilatory Outputs on Patients with Various Disease Diagnoses…….. 109
vi

List of Figures

Title Page
Figure 1-1.
Physiologocal Conceptual Framework of Cancer-related Dyspnea.................. 18

Figure 3-1. Inclusion Diagram…………………...…………………………....... 60

Figure 3-2. A Portable Respiratory Pressure Meter………………………….. 64

Figure 4-1. Flow Chart of Data Collection……………………………………. 81

Figure 4-2. Scatter Plot of Anxiety against Breathing Effort………………… 94

Figure 5-1. Study Model……………………………………………………….. 115

Figure 5-2. Analytical Model of the Present Study……………………………. 117

Figure 5-3. The Proposed Conceptual Framework of Cancer-related Dyspnea


Sensation………………………………………………………………. 122
vii

Acknowledgements

This long journey won’t end without guidance, supports, or helps. I would like to

express my gratitude to the former dissertation chairperson, my mentor, Dr M. Linda

Workman for guiding me to initiate, organize ideas of conducting this study. Countless

gratitude goes to my other mentor, the present dissertation chairperson, Dr Barbra J. Daly,

for sharing her wisdom as a researcher, as a scholar, in so many ways. I would like to

extend the appreciations to the dissertation committee members, Dr Chris Winkelman for

her helping and supporting and Dr Mariana Petrozzi for her sharing knowledge in

pulmonary medicine. I am grateful for Dr Sara L. Douglas for her sharing knowledge in

the area of data management.

I am indebt to my parents, who endeavored to understand, and then supported my

decision of being 3,000 miles away, studying in another country with such different

cultures. I am grateful for my dearest cousin, my sister, my best friend, Yu-Ching

(Vanessa) Lin, for years being aside me with endless supports; I also thank her two

adorable daughters, 8-year-old Jui-Yun (Rebecca) and 4-year-old Shin-Yun (Cindy), for

their cheerful smiles to make this journey a lot easier.

Special thank goes to my friend, my colleague, my college classmate, Ching-Chuan

(Joanna) Tung, for sending those overseas care packages, supporting phone calls, helping

me to process IRB submission to the hospital in Taiwan. Sincere appreciation goes to my

friend Dr Shih-Tzu Huang, for accompanying me facing difficulties during course works.
viii

In addition, I am grateful for my friends, my colleagues, my classmates, Dr Mary Jo

Prince-Paul, Dr Amany Farag for sharing all the moments of tears, hardships, joys with

me. Sincere thank goes to my friend, my classmate, Tsay-Yi (Amy) Au for her supporting

and praying for me.

I also would like to thank my colleague, Li-Fen Chao, for helping me to communicate

with the hospital in Taiwan while I was in Cleveland. Final sincere appreciations go to Dr

Wen-Cheng (John) Chang, staffs, and members of Chang Guang Memorial Hospital,

Linkuo branch, Taiwan, for their helping with recruitments.


1

The Role of Anxiety in the Relationship between Breathing Effort and Cancer-Related

Dyspnea Sensation

By

CHIOU-FANG LIOU

Abstract

Background: How objective measures of respiratory effort explain dyspnea, and how

psychological distress influences dyspnea sensation requires further investigation in order

to understand the phenomenon of cancer-related dyspnea. Purpose of the study: This

theory derived cross-sectional correlational study sought to examine factors influencing

the intensity of dyspnea in patients with cancer diagnosis. Sample: 28 patients (17

women and 11men), age 41-90, diagnosed with lung cancer, primary or metastastic.

Method: Data collection was carried out at a cancer center in Northern Taiwan. Face-

to-face interview employed; two hand-held devices obtained physiological indicators

(FEV1, FVC, and respiratory pressure), and survey questionnaires assessed for conscious

breathing effort and for anxiety. Major findings: Regression analysis supported the

hypothesis that increasing respiratory muscle work effort is positively related to

individual’s reporting of dyspnea sensation. The results of this regression analysis also

revealed that only self aware breathing effort is a significant factor of predication the

intensity of dyspnea sensation (β = 0.56, p < 0.01) . However, results from the hierarchal

multiple moderate regression analysis did not support the hypothesis that anxiety
2

influences the relationship between respiratory muscle effort and individuals’ perceived

dyspnea. Nevertheless, it revealed that anxiety is another significant factor, together with

self aware breathing effort explaining the intensity of dyspnea sensation (adj.R2 = 0.47, p

= 0.000). The Pearson correlation also verified a moderate significant association

between anxiety and self aware breathing effort, r = 0.62, p < 0.01. Conclusions and

recommendations: The analytical model adds to previous empirical evidence suggesting

a new conceptual framework for explaining cancer-related dyspnea. In addition, further

study is indicated, focusing on testing three targeted mediating effects: decreasing

conscious breathing effort, reducing anxiety, and the aggregation effects of both.
3

Chapter One

Background and Significance

Statement of Problem

The word “Dyspnea” comes from the Greek. The suffix ‘Dys” indicates hard and

difficult; “Ponia” means breathing. The word dyspnea then, refers to a description of

breathing difficulty. American Thoracic Society (1998) defines dyspnea as the subjective

experience of awareness of labored breathing, respiratory distress, and unpleasant or

uncomfortable respiratory sensations.

Although dyspnea is not a disease-specific symptom, it most commonly results

from dysfunction of the cardiac and pulmonary systems. Cancer-related dyspnea, for

example, is a consequence of respiratory dysfunction caused by cancer, including tumor

location and cancer treatment toxicity. In other words, dyspnea is a subjective sensation

with pathological changes, such as hyperpnea and changes in pulmonary function. Those

physiological changes that can be observed are then used as indicators for assessment.

However, most studies exploring dyspnea have used Visual Analogue Scales (VAS) to

assess for the subjective symptom of dyspnea. Using the definition of dyspnea from the

American Thoracic Society, dyspnea should be based on a subjective description of the

sensation along with measurement of any physiological change. Measuring dyspnea by

VAS only captures the subjective sensation. The labored breathing and increased

respiratory effort need more objective and precise measures. Actual physiological
4

measurement would be a good way to proceed. Therefore, a comprehensive examination

of dyspnea sensation should include both objective and subjective measures.

In addition, since cancer is a diagnosis that often is perceived as being

life-threatening, cancer-related dyspnea sometimes clusters with other symptoms,

particular psychological distress. Dyspnea clustering with psychological distress may

complicate the measurement of dyspnea as well as its management. Current knowledge

about dyspnea is only focused on the medical management. Scientific exploration

regarding the nursing management of dyspnea is lacking.

A review of the nursing literature regarding cancer-related dyspnea revealed a

dearth of prospective studies that either used objective measures or addressed the issue of

the possible influence of psychological distress on the sensation of dyspnea. Prospective

studies addressing theses issues are needed to fully characterize the objective and

subjective aspects of cancer-related dyspnea. The results of such studies then could be

used as the basis for the development of more effective nursing strategies to manage

cancer-related dyspnea.

Physiologic experiments supported the concept that dyspnea sensation was related

to respiratory effort. These physiological studies also hypothesized that changes in

breathing effort could be detected by respiratory muscle work and sensations. In addition,

persons experiencing dyspnea verified that psychological distresses could be the result of

bio-behavioral responses to the dyspnea experience. Previous studies of patients with


5

dyspnea identified that the most common psychological distress of dyspnea sensation

associated with cancer-related dyspnea is anxiety. However, gaps remain regarding what

is known about how anxiety influences cancer-related dyspnea. Therefore, the purpose of

this cross-sectional descriptive study was to add to knowledge regarding the mechanisms

of cancer-related dyspnea. This study sought to identify the role of anxiety and to

examine how anxiety influenced cancer-related dyspnea. The findings of this study were

then expected to provide more breadth and depth of evidence for probing cancer-related

dyspnea.

Background

Four sections are presented and discussed in terms of issues related to cancer-related

dyspnea. The first section addresses the prevalence of cancer-related dyspnea. The second

section focuses on describing the nature of pulmonary ventilation because the nature of

pulmonary ventilation forms the basis of understanding the pathophysiological

explanation of dyspnea. The third section explains what is known about the physiological

mechanisms of dyspnea sensation, including the nature of dyspnea sensation and the

Psychophysics Law (Stevens, 1957) of dyspnea sensation. The fourth section illustrates

the possible bio-behavioral component of dyspnea, explaining how an individual might

choose to respond to his or her dyspnea.

Prevalence and the Issue of Cancer Dyspnea

Dyspnea is a common symptom occurring in the advanced cancer population. A


6

large study of the prevalence of dyspnea revealed that among 289 patients with non-small

cell lung cancer, 60% self reported the presence of breathlessness, and about 30% rated

this dyspnea as moderate to severe (Mures & Round, 1993). In another study of 1752

terminal cancer patients, more than 72% of subjects reported the presence of dyspnea

during the last six weeks of their lives (Mor & Reuben, 1986).

Numerous factors related to dyspnea among the cancer population were examined

in variety of studies. Dudgeon and colleagues (2001) explored the prevalence and

associated factors of dyspnea in 923 patients with cancer. They found that only 4% of this

population had a diagnosis of lung cancer; and 5.4% of this sample had other cancer

types with lung metastases. They also found that additional risk factors for dyspnea were

related to persistent pulmonary conditions, such as smoking history, asthma or chronic

pulmonary disease, lung irradiation, and history of exposure to asbestos, coal dust, cotton

dust, or organic dust (Dudgeon, Kristjanson, Solan, Lertzman, & Clement, 2001). This

study did not relate the variable of ethnicity to different causes of cancer-related dyspnea.

A retrospective study conducted in Taiwan examined 125 terminal cancer patients

with dyspnea at admission and 2 days before death. The results from this tudy were

consistent with those conducted in the United States. Among 74 patients, dyspnea was

caused by the physiological changes of cachexia, anemia, pleural effusion, and

lymphangitis. The logistic regression suggested that lung infection, airway obstruction,

acidemia, and pericardial effusion significantly predicted the severity of dyspnea with an
7

odds ratio (OR) of 2.29 to 1.38 (Chiu, Hu, Lue, Yao, Chen, & Wakai, 2004).

The Nature of Pulmonary Ventilation

All body tissues and cells need the oxygen derived from atmospheric air to sustain

critical physiologic functions. Tissues and cells also need to rid themselves of gaseous

metabolic wastes, specifically carbon dioxide. The entrance point for atmospheric oxygen

and the exit site for the majority of carbon dioxide is the respiratory system. Gases are

first exchanged at the alveolar-capillary membrane and are then exchanged again at the

capillary-cell interface at the tissue level because most cells are remote from the

respiratory system.

The purpose of respiratory regulation is to ensure that ventilation, the movement

of atmospheric air into and out from the lungs as a result of intrathoracic pressure

changes through respiratory muscle contraction and relaxation, is sufficient to maintain

appropriate arterial blood levels of carbon dioxide (PaCO2) and oxygen (PaO2). This

regulation stimulates an increase in a person’s rate and depth of ventilation when PaCO2

levels rise or when PaO2 levels decrease. Likewise, respiratory regulation causes a

decrease in a person’s rate and depth of ventilation when PaCO2 levels fall or when PaO2

levels increase to a maximum partial pressure. The neural receptors (central

chemoreceptors) for monitoring changes in PaCO2 levels, located in the central nervous

system, are more sensitive and stimulate faster responses for ventilatory change than do

the peripherally located neural receptors (peripheral chemoreceptors) that monitor and
8

respond to changes in PaO2 levels.

The central respiratory center is composed of three groups of neurons located in

the medulla oblongata and pons areas of the brain stem: the dorsal respiratory group, the

ventral respiratory group, and the pneumotaxic center. The function of this integrated

respiratory center is to ensure the proper execution of the respiratory cycle whereby

inspiration, or the movement of air into the lungs (caused by diaphragm and skeletal

muscle contractions that decrease the intrathoracic pressure below atmospheric pressure),

is followed by expiration, or the movement of air out from the lungs (caused by

diaphragm and skeletal muscle relaxation and abnormal muscle contraction that increase

the intrathoracic pressure above atmospheric pressure). The actual work of normal

ventilation or pulmonary respiration starts with the excitation of inspiratory neurons in

the respiratory center sending efferent action potentials to the skeletal muscles of

inspiration. Contractions of these muscles enlarge the thoracic cavity, decreasing the

pressure in the cavity and in the lungs to below atmospheric pressure.

In the healthy person, respiratory regulation of ventilation resulting in changes in

the rate and depth of ventilation occurs quickly and is usually not noticed by the

individual unless blood chemistry changes are abrupt or severe. The normal quiet and

continuous minor adjustments in ventilation resulting from respiratory regulation

matched to PaCO2 and PaO2 levels are termed eupnea and they do not trigger respiratory

effort awareness in the alert individual.


9

Dyspnea as a sensation of increased respiratory effort awareness, on the other hand,

occurs when respiratory regulation results in significantly increased respiratory muscle

work for increased ventilatory effort. At times the respiratory regulation during dyspnea

appears disassociated from actual negative changes in blood chemistry levels (PaCO2 and

PaO2 levels).

Physiological Mechanism of Dyspnea

The Nature of Dyspnea Sensation

Dyspnea, a symptom rather than a discrete disease entity, interferes with normal

breathing. The sensation of dyspnea is a perception resulting from excessive or impaired

respiratory regulation. Dyspnea sensation is a subjective experience and such an abstract

experience results from stimulation of the sensory neuron conduction between peripheral

and central nerve systems. The “length-tension inappropriateness” theory (Comore, 1966)

indicates that the disassociation between the efferent information and neural activity

contributes to the intensity of dyspnea. The stimulation of the afferent intrapulmonary

receptors and stimulation of neural receptors in respiratory muscles and skeletal joints,

increase the sensitivity in the central nervous system through amplification, which then

leads to a reduction of ventilatory capacity or breathing reserve (Carrieri, Janson-Bjerklie,

& Jacobs, 1984). Breathing reserve is the excess breathing capacity that occurs beyond

actual ventilation and is generally expressed as a percentage of maximum breathing

capacity. The greater the need for breathing to maintain appropriate arterial lvels of
10

oxygen and carbon dioxide, such as during increased physical activity, the more breathing

reserve is reduced (Cournand & Richard, 1941). Furthermore, dyspnea sensation occurs

when the breathing reserve decreases to less than 65% to 70% in both normal healthy

individuals and in individuals with pulmonary function impairment (Fishman & Ledlie,

1979).

Therefore, any increasing physical activity requires a corresponding increase in

breathing load. Then, the incoming afferent respiratory information from the receptors in

the airways, lungs, and chest wall structures and the respiratory efferent signals are

integrated in triggering the perception of dyspnea in the central nervous system

(Schwartzstein, Manning, Weiss, & Weinberger, 1990). Dyspnea sensation also occurs

through the process of having receptors in the airways, lungs, and chest walls sensitive to

chemical changes. This incoming afferent information is sent as an attempt to stimulate

the brain to allow respiration regulation. The respiratory center of brain sends the

command to change the respiratory pattern in order to achieve respiratory regulation

(Guyton & Hall, 2006). The receptors in respiratory muscles convert the magnitude of the

stimuli into signals in afferent nerves and the signals are transmitted to the central

nervous system through afferent nerve proprioceptors (Killian & Jones, 1988). The brain

cortex then processes the stimuli and interprets the stimuli into physical sensation, which

results in the subjective sensation of dyspnea (Carla & Eduardo, 1997).

In summary, dyspnea sensation is triggered either from the activation of respiratory


11

muscle afferents or from the corollary discharge within the central nervous system (CNS);

the CNS response is either from cortex or brainstem (Banzett, Im, Lansing, Legedza, &

Thwing, 2000). Dyspnea sensation depends on subjective awareness or perception, which

is not necessarily dependent on blood chemical changes or ventilatory abnormality

(Dudley, Martin, & Holmes, 1968). However, either blood chemical abnormalities or

ventilation abnormalities can also stimulate the sensation of dyspnea. Dyspnea sensation

can be explained by sensory receptors and stimuli. The origin of dyspnea begins with a

stimulus and the stimulus then leads to conscious sensation through neural transmission

(Killian & Jones, 1988; Killian, 1998).

Psychophysics Law of Dyspnea

Stevens (1957) proposed an equation, the psychophysics law, that determines a

relationship between physiological and psychological magnitude of afferent stimuli in

explaining dyspnea. The equation is Ψ=k φn; where Ψ is the psychological magnitude of

the stimulus, φ is the physiological magnitude, and k is a constant that standardizes the

starting point of this equation in a geometrical projection image. The geometrical

projection image of this equation does not show a linear relationship between

physiological magnitude and psychological magnitude with respect to dyspnea sensation.

However, it does indicate that the psychological magnitude and physiological stimuli are

related. This psychophysics equation was adapted by Killian (1998) to explain the unique

sensory function of increased skeletal muscle work contributing to the perception of


12

dyspnea. Killian’s empirical works contributed to the physiologic basis. He also

concluded that sensory physiology and psychophysics both apply to the breathing action

pattern and responses predominant during dyspnea.

The psychophysics law of dyspnea emphasized that the sensation of dyspnea was

the individual’s perception of breathing difficulty; however, other researchers tended to

be more interested in where this sensation occurred in the neurophysiologic explanation.

Duel (1986) stated that the sensation of dyspnea could be related to both cortical and

sensory perception. A physiological oriented study provided further empirical evidence

supporting Duel’s proposition. A functional imaging study with positron emission

tomography indicates that the right side of the posterior cingulate cortex was the site of

neural activation specifically associated with perceived intensity of respiratory

discomfort. The same study also suggests that labored breathing may be processed

through two distinct neural networks: the concomitant processing of the genesis of the

perceived dyspnea and the modulation of perceived respiratory discomfort (Aubier,

Peiffer, Poline, Samson, & Thivard, 2001).

Bio-behavioral component of Dyspnea

Dyspnea is a subjective sensation and it involves the patient’ description and

reaction to this sensation (Comrone, 1966). The bio-behavioral component of dyspnea

indicates how an individual with dyspnea reacts in the presence of this symptom. Steele

and Shaver (1992) presented an ecological model, which used both sensation and
13

affective dimensions in explaining the dyspnea experience. The ecological model of

dyspnea experience assumes that the proportional increase in reported dyspnea magnitude

is consistently accompanied by an increased magnitude of physical stimuli, such as

inspiratory muscle loading. The perceived severity of dyspnea could then act as a

psychological stressor.

Two concerns, in terms of cancer-related dyspnea, need to be addressed. The first

one is that patients with dyspnea have to work hard to breathe, a purely physical

phenomenon. The second concern is that these patients have to face their inability to

perform daily activities due to the increase in breathing difficulty. These limitations can

lead to patients’ fears of not getting sufficient air and may progress to worse feelings such

as not getting well, or dying. The fear of not getting well or dying, then, could raise

patients’ anxiety levels. The increased anxiety could, of itself, increase the respiratory

rate, which then triggers dyspnea sensation as a result of increased respiratory muscle

work.

In addition, cancer has a negative connotation as a life-threatening illness. Thus, in

patients with cancer, dyspnea could be positively correlated with the severity of their

spiritual distress (Altman, Edmonds, Higginson, Mcdonnell, & Sen-Gupta, 2000). Duel

(1986) addressed the issue that subjects with dyspnea often feel anxious. The increase in

respiratory effort leads to worsening feelings of dyspnea due to patients’ uncertainty

about their disease and their fears of death. Both of these factors may increase patients’
14

anxiety, and this anxiety then could lead to breathlessness. Thus, it is reasonable to

consider that a major psychological stressor for patients with cancer is anxiety.

Anxiety is an emotional response, and this psychological factor often varies by

ethnic groups. Therefore, anxiety related to cancer-related dyspnea also was examined

among cancer patients in both western and oriental cultures. The results for both culture

groups were consistent with each other in providing evidence of a positive relationship

between anxiety and cancer-related dyspnea. In the United States, a prospective study

found that lung involvement, anxiety, and maximal inspiratory pressure were factors that

correlated with dyspnea in 135 cancer patients attending pain clinics (Bruera, Pither, &

Schmita, 2000). Smith and colleagues (2001) used two-way ANOVA to investigate the

effect of anxiety on the rating of dyspnea severity and found a significant main effect of

anxiety among low, moderate, and high dyspnea groups. Another study, conducted in

Japan, concluded that the phenomenon of dyspnea was multi-dimensional, including

sense of effort, sense of anxiety, and sense of discomfort (Tanaka, Akechi, Okuyama,

Niskiwaki, & Uchitomi, 2000).

Significance of Study

The empirical works examining dyspnea revealed that this symptom includes at

least two domains: the physiological indicators as well as the psychophysical reaction to

dyspnea. The body of knowledge of effective dyspnea management, however, has

unknown areas, including the relationship between these two domains as well as the role
15

of each domain in explaining dyspnea. Thus, while establishing strategies to help patients

with dyspnea to alleviate their discomfort, it is important to take both the physiological

and the bio-behavioral aspects of dyspnea into consideration. Current cancer-related

dyspnea management is focused on medical strategies, primarily the use of opioids. A

Cochrane collaborative methodology-based systematic review of 18 clinical trials of

dyspnea management included dyspnea from both malignant and non-malignant causes.

The meta-analysis concluded that a significant positive effect of opioid treatment on

dyspnea existed in terms of the standardized mean difference of 0.31 (p≤0.001) (Broadley,

Davies, Higgins, & Jennings, 2002). Nursing specific interventions, on the other hand,

were rarely addressed and were not evidence-based. In order to build knowledge of

effective nursing strategies for cancer-related dyspnea, understanding the mechanism of

dyspnea is important and should be known prior to conducting research involving nursing

interventions. The body of knowledge for cancer-related dyspnea includes factors

contributing to it, as well as how these factors are related to each other in explaining the

phenomenon of dyspnea.

Significance to Nursing Science

The purpose of this study was to assist nurse-scientists, practicing nurses, and

nursing educators to understand the phenomenon of cancer-related dyspnea sensation.

This purpose involved two foci. The first focus was to examine specifically how

breathing effort can explain the severity of dyspnea sensation in cancer patient. The
16

second focus was to explore the possibility of anxiety moderating the nature of the

relationship between breathing effort and the severity of dyspnea sensation. The

significance of this study to the discipline of nursing was to provide evidence to explain

the mechanism of cancer-related dyspnea and to determine the roles of these two factors

in contributing to dyspnea in persons with cancer. Such information then could be used to

guide the development and testing of nursing interventions to manage dyspnea effectively.

The findings were expected to be used as empirical knowledge, in terms of further

nursing intervention on cancer-related dyspnea. The findings also were expected to guide

further research on the mechanism of cancer-related dyspnea.

In addition to building the body knowledge of cancer-related dyspnea, the findings

were also important in nursing education. Roy and Robinson (1992) proposed a paradigm

shift using physiological nursing research in examining dyspnea. They suggested that

integrating physiological-based research into basic and clinical nursing science forms

another stage of developing nursing knowledge. Nursing education is the foundation of

introducing the basic and clinical sciences. Having nursing educators disseminate this

knowledge may allow students to have a more comprehensive understanding of the

mechanism of dyspnea in order to more effectively manage this symptom.


17

Theoretical Work

The theoretical work guiding this study was based on a physiological prospective.

The first part of this framework depicted the relationship between physiological changes

leading to dyspnea. The physiological changes included the neural activity that stimulated

the individual’s perception. The second part of this framework hypothesized a proposition

to explain how anxiety influenced the relationship between the physiological indicators

of dyspnea and individual’s perception of the severity of dyspnea. Figure 1-1 presents the

conceptual framework of cancer-related dyspnea used to guide this present study.


18

Figure 1-1.
Physiological Conceptual Framework of Cancer-related Dyspnea
Moderator:
Anxiety
+

Somatic +
Receptors
Effusion Receive
in Chest and
wall transmit
Individual
Increasing Signal to
Perceive
Respiratory Cerebral Cerebral
the
Muscles’ Cortex
Severity of
Work Effort Interprets
Dyspnea
Tumor Signal
Sensation
related
factors
caused
decreasing
oxygenation +

Note: Direct Influence


Indirect Influence
Anxiety increases the intensity and severity of individual’s perceived
dyspnea sensation
19

Physiological Changes

Tumor location, pleural effusion, and cancer treatment-induced pulmonary toxicity

result in ventilation limitation; such limitations could lead to dyspnea sensation. No

single factor, however, precisely explains cancer-related dyspnea. In addition, factors

contributing to dyspnea are not definitely correlated to each other.

Although cancer-related dyspnea cannot be explained directly by its etiologies,

the effects of these various etiologies influence dyspnea possibly through respiratory

muscle movement. Fishbein, Kearon, and Killian (1989) initially proposed the framework

of the relationship between dyspnea and the sense of respiratory effort, and used this

framework to explain dyspnea in cancer patients. In this framework, These researchers

implied that the intensity of dyspnea could be explained by respiratory muscle effort; for

example, more respiratory muscle effort led to greater intensity of dyspnea sensation. In

addition, based on the psychophysical principles, Fishbein and colleagues demonstrated

that dyspnea sensation, at least in part, was based on proprioceptive information

generated from the mechanical act of breathing. They used both experimental evidence

and clinical observation, inductively assuming that the major specific sensation related to

dyspnea was respiratory effort.

The role of respiratory muscles is to provide force for chest volume changes that

allow lung inflation. The essential inspiratory respiratory muscles are the diaphragm, the

sternocleidomastoideus muscle, and the internal and external intercostal muscles. The
20

abdominal muscles are most important for expiratory movement; the diaphragm is the

most important inspiratory muscle (Levitzky, 1999). Respiratory muscle malfunction,

which results from various mechanisms, often is a factor of breathing difficulty in

patients with cancer (Fishbein, Kearon, & Killian, 1989). Factors contributing to dyspnea

influence respiratory muscles either directly or indirectly. Direct influences among

patients with cancer include pleural effusion and pericardial effusion, creating difficulty

in reducing thoracic cavity pressure. When these conditions occur, the diaphragm

movement is restricted and increased work is needed in order to achieve normal breathing

function. Indirect influences include factors that decrease the oxygenation function of

pulmonary system. Whenever oxygen saturation decreases, the respiratory muscles have

to correspondingly increase their workload, which then results in more oxygen

consumption and a greater imbalance between oxygen need and oxygen delivery.

Respiratory muscles, including the diaphragm, are skeletal muscles and contain

somatosensory receptors. When respiratory movement increases, regardless of the cause,

the somatosensory receptors in respiratory muscles then receive a signal of “hard work”.

This signal of increased breathing effort is transmitted to the cerebral cortex. As soon as

the cerebral cortex is aware of these signals, the cerebral cortex interprets these signals as

an unpleasant feeling (Carla & Eduardo, 1997). The difficulty of breathing then becomes

a sensation perceived by the alert individual. Together, these physical changes and

resulting perceptions can explain the mechanism of dyspnea.


21

Approximately 50% to 70% of dying cancer patients suffered death anxiety

because of their dyspnea (Person, Robert, &Thorne, 1993). The perceived severity of

dyspnea, fear, and anxiety could create a cycle of dyspnea reaction. This reaction cycle

could be seen in the way clients vary in their perceived severity of dyspnea because the

experience of dyspnea is interpreted and rated differently by individual patients

(Widimsky, 1979). In addition, researchers have proposed that terminal cancer patients

may have acute respiratory distress attacks and respiratory panic as a result of rapid

pathological changes from disease progression (Doyle, Hanks, & MacDonald 1998).

Variables

Independent Variable

The independent variable of this study was breathing effort. For the purpose of this

study, breathing effort was defined as increasing intensities of breathing. Increasing

breathing intensity was proposed to lead to exertion to breathe (Lansing, IM, Thwing,

Legedza, & Banzett, 2000). The sense of exertion occurs within the muscle by

stimulating intramuscular sensory receptors. Thus, increasing breathing effort causes

increasing respiratory muscle work load. The sense of muscle work includes the sensation

of effort (which is related to the willed motor command), the sensation of tension (which

is related to tension generated by muscle), and the sensation of displacement (which is

related to muscle contraction’s extent and velocity), and which also provides a vital role

regarding vascular responses in the muscles. In other words, these three sensations occur
22

simultaneously when any muscle, including respiratory muscles, is stimulated.

Killian and Jones (1988) examined studies related to human consciously perceived

sensations and muscular activities. They concluded that the muscular sensations related to

dyspnea were the sensation of effort, the sensation of tension, and the sensation of

displacement. Moreover, high sensation of tension and the sensation of effort were both

found in non-pathologic labored breathing that was artificially stimulated by adding

elastic loads on normal subjects. In Killian and colleague’s (1984) early work, they found

that increasing inspiratory pressures and lung volumes were significantly related to the

magnitude of the individual’s sensing of breathing effort (p≤0.001). In addition to

increasing inspiratory pressure, breathing effort can be explained psychophysically

(Killian, Gandevia, Summers, & Campbell, 1984). In other words, breathing effort also

came from subjective sensation indicating that sensory interpretation occurred in the

brain cortex.

Dependent Variable

The dependent variable was the severity of dyspnea sensation and was defined as

the cognitive response and the perceptional sensitivity of an individual rating his or her

severity of dyspnea. The cognitive response of dyspnea was a subjective perception and

depended on personal experience. Studies of human sensory perception often reported the

problem of subjectivity in which the sensation was experienced as a highly personal

event. In addition to the physiological explanation of sensory perception in the central


23

nerve system, previous personal experience with dyspnea may have influenced the

individual to believe he or she was having a dyspnea episode even when the physical

sensory stimuli was not present (Dudel, 1986). Detecting the signals of dyspnea sensation

relied on the individual’s awareness. Perception was a result of a sensation accompanied

by its interpretation (Dudel, 1986). The interpretation could be based on personal

experience and therefore became a unique sensation impression.

Moderator

In this study, anxiety was considered a moderator that influenced the relationship

between breathing effort and the perceived severity of dyspnea. The definition of anxiety

was a psychological reaction that occurred along with dyspnea sensation. Human

sensation and its interpretation resulted in perception and the perception was influenced

by the emotional response to the particular sensation experienced. The perceived severity

of dyspnea along with fear and anxiety created a cycle of dyspnea reaction. This reaction

cycle may be seen in the way clients varied in their perceived severity of dyspnea,

because the experience of dyspnea was interpreted and rated differently by individual

patients (Widimsky, 1979). A study of six home-based palliative care teams found that the

presence of dyspnea positively correlated with the severity of patients’ spiritual distress

(Altman, Edmonds, Higginson, Mcdonnell, & Sen-Gupta, 2000). In a cross-sectional

study of 75 oncology outpatients with dyspnea, the regression analysis suggested that

anxiety significantly predicted the interpretation of shortness of breath as dyspnea


24

(β=0.244, p≤0.05) (Dugeon, Lertzman, & Askew, 2001).

Assumption

The premise of the proposed conceptual framework (figure 1-1) was that

cancer-related dyspnea was the symptom of pulmonary dysfunction caused by cancer and

cancer treatment; the severity of cancer-related dyspnea was influenced by breathing

effort and the psychological reaction (anxiety). Two assumptions were drawn in this

present study. The first assumption was the mechanism of cancer-related dyspnea:

regardless of cause, dyspnea occurred whenever the respiratory muscle movement was

increased above a normal level, with neural signals transmitted to the cerebral cortex; the

cerebral cortex then interpreted this signal as labored breathing. The individual, then,

perceived and reported dyspnea sensation. The second assumption was that the

psychological reaction of dyspnea sensation, anxiety, triggered hyperpnea. The increased

breathing rate then increased respiratory muscle work load. Therefore, the dyspnea

sensation then was perceived as worse.

Hypotheses

The physiological and pathological processes of dyspnea sensation were

overridden by the cerebral cortex and were influenced by any alteration of respiratory

muscle movement (West, 1985). Two hypotheses were drawn from the principle

assumption: (a) increasing respiratory muscle work effort would be positively related to

individual’s reporting of dyspnea sensation; and (b) anxiety would influence the
25

relationship between respiratory muscle effort and individuals’ perceived dyspnea.

Research Questions

Three questions guided this research:

(1) Does breathing effort significantly predict individuals’ perceived severity of dyspnea?

(2) Is anxiety positively related to individuals’ rating their severity of dyspnea sensation?

(3) Is anxiety a moderator influencing the relationship between respiratory work effort

and individuals’ perceived the severity of dyspnea, in a positive direction?

Summary

A review of the nursing literature regarding cancer-related dyspnea revealed a

dearth of prospective studies that either used objective measures or addressed the issue of

the possible influence of psychological distress on the sensation of dyspnea. Prospective

studies addressing theses issues are needed to fully characterize the objective and

subjective aspects of cancer-related dyspnea. The results of such studies could be used as

the basis for the development of more effective nursing strategies to manage

cancer-related dyspnea.

Previous studies of patients with dyspnea identified that the most common psychological

distress of dyspnea sensation associated with cancer-related dyspnea is anxiety. However,

gaps remain regarding what is known about how anxiety influences cancer-related

dyspnea. Therefore, the purpose of this cross-sectional descriptive study was to identify

the role of anxiety and to examine how anxiety influenced cancer-related dyspnea. The
26

findings of this study were then expected to provide more breadth and depth evidence for

probing cancer-related dyspnea.


27

Chapter Two

Literature Review

Chapter one introduces the theoretical work of cancer-related dyspnea and

proposed conceptual framework of the relationship between breathing effort and dyspnea

sensation along with a possible moderator, anxiety, in this relationship. This chapter

continues addressing these variables by adding and integrating current relevant empirical

works.

The operational definitions of breathing effort and dyspnea sensation were based

on the empirical works of physiological-oriented studies. This present study sought to add

evidence for explaining the mechanism or mechanisms of cancer related dyspnea,

particularly focusing on the relationship among breathing effort, anxiety, and the severity

of dyspnea sensation. Thus, the following review of relevant literature emphasized those

works concerned with the relationship between anxiety and dyspnea. In addition to

critiquing and integrating this literature, a possible direction between anxiety and dyspnea

was discussed. The summary, along with the limitation of the empirical works examined

in this chapter, addressed any knowledge gaps of the phenomenon regarding factors

contributing to cancer-related dyspnea.


28

Breathing Effort

Several physiological-oriented studies indicate that breathing effort is an added

workload on respiratory muscles for performing respiratory movement. Further empirical

evidence from these studies suggest that without invasive procedures, mesuring breathing

effort can be achienved through measuring lung volume changes, and inspiratory pressure

changes through the mouth.

Operational Breathing Effort

Examining breathing effort in relation to physiological changes was difficult to

conduct in persons with illness. Thus, physiologists created several experimental

conditions to mimic pulmonary impairment that could which lead to increased breathing

effort. Three relevant experimental conditions had been introduced since 1980s. Those

experimental conditions then were tested on healthy subjects aged between 20 and 40

years. One experimental condition was a series of air tight drums, to produce target

pressures at functional residual capacity (FRC) to match the pressures of breathing effort,

which then led to increased lung volume (Killian, Gandevia, Summers, and Campbell,

1984). Another experimental condition used to increase breathing effort was the

purposeful induction of respiratory muscle fatigue (Yan, 1999; Chen & Yan, 1999).

Those experimental conditions then were used to examine how changes in lung

volume and inspiratory muscle fatigue correlated with the sensation of breathing effort /

breathing difficulty. Results from these studies revealed three important findings
29

regarding the relationship between breathing effort and changes in lung volume under

manipulated conditions. The subjects’ perceived breathing efforts were associated with

increased inspiratory threshold/pressure, lung volume increases, and changes in end-

expiratory lung volume (Killian, Gandevia, Summers, & Campbell, 1984). Additionally,

changes lung volume were significantly associated with muscle fatigue (Yan, 1999).

Another conclusion was that the imposed inspiratory threshold had a greater impact on

the perceived breathing difficulty than on the respiratory muscles (Chan & Yan, 1999).

The limitations of these three studies mainly results from fairly small sample sizes. The

variability regarding demographic characteristics (men and young age) of sample in these

three studies was also small.


30

Table 2-1.
Three Physiological Manipulated Experimental Conditions.

Author / Sample Experimental Measurement Results


Year Description Conditions
Killian, N=5 healthy A series of air Phyiological The
Gandevia, sujects aged tight drums, to measures and perceieved
Summers, 26-34. produce target psychophysical breathing
& pressures at techniques effort was
Campell, functional were employed associated
1984 residual capacity to collect data. with lung
inducing volume
breathing effort. change,
p≤0.05.

Yan, 1999 N=7 healthy Fatiguing Physiological Increasing


male inspiratory indicators and the
subjects in muscle: subjects the Borg sclae inspiratory
their 30s. were asked to threshold
breathe through a was
mouth piece and associated
wore a nose clip. with
breathing
difficulty,
p≤0.001.
Chen & N= 8 healthy Same as Yan’s End-expiratory Imposed
Yan, 1999 laboratory conditions lung volume inspiratory
workers (1999). threshold
aged 31- 41. may explain
the perceived
breathing
difficulty

Breathing effort is the result of respiratory muscle movement (contraction),

which then causes lung volume changes. Thus, it seemed logical that evaluating

respiratory pressure through the mouth could be a practical and less invasive method to
31

measure breathing effort. The rationale for use of this measure was supported by the fact

that inspiratory pressure against an occluded airway was associated with integrated

mouth pressure (p≤0.05) (Jones, Killian, Summers, & Jones, 1985), and increasing

breathing effort was highly correlated to perceived respiratory difficulty (n = 24, r = 0.86,

p < 0.001) (O'Donnell, Bertley, Chau, & Webb, 1997). In addition to the use of

physiological measures, the perceived sensation of breathing effort and dyspnea sensation

also could be considered or “measured” psychophysically (Killian, Gandevia, Summers,

& Campbell, 1985), assessed as a subjective-rating of breathing exertion.


32

Table 2-2.
Physioloical Oriented Evidences of Breathing Effort

Author / Study Design / Sample Measurement Results


Year Purpose Description
Jones, Cross-sectional; N=10 healthy Delta VO2 (for Inspiratory
Killian, To investigate subjects oxygen pressure was
Summers, how respiratory uptake), associated
& Jones, muscle work integrated with both
1985 influences the mouth oxygen uptake
increased pressure and integrated
inpsiratory mouth
threshold. pressure,
p≤0.05.

O’Dnonell, Cross-sectional N=24 The Borg A high


Bertley, Compared the consisted scale, correlation
Chau, qualitative aspects n= 12 respiratory (r= 0.78)
&Webb, of breathing patients with pressure, between
1997 difficulty severe breathing ventilation
chronic pattern, and output and
airflow lung volume. perceieved
limitation; breathing
n=12 effort.
age-matched
healthy
subjects

In summary, the physiological-oriented studies discussed above provides evidence

supporting how healthy individuals sensed changed in breathing effort induced by

manipulation of respiratory muscle work load. The current study thus hypothesized that

cancer-related dyspnea occurs when respiratory muscle work load increases, regardless of

etiology. It would be difficult to directly test muscle effort on subjects with cancer using
33

physiological manipulation. However, the hypothesis in the current study only required

examination of the respiratory muscle effort; thus, the above studies provided sufficient

evidence in terms of explaining the relationship between sense of breathing effort and

respiratory muscle movement.

Dyspnea Perception

Dyspnea sensation is an accentuated report of breathing difficulty. Such sensation

is a subjective experience and results from respiratory muscle hard work with or without

a chemical change. However, a study about the interrelationships between the extent of

thoracic muscle movements and respiratory chemical drive on16 healthy subjects verified

the sensed dyspnea occurred before a chemical change (as PCO2 level increased)

(Chonan, Mulholland, Cherniack, & Altose, 1987).

Operational Dyspnea Perception

Although dyspnea sensation often followed from pulmonary impairment, the

sensation might not occur synchronously with the impairment. For example, one study

had demonstrated that subjects with an FEV1 of less than 50% of the predicted normal

did not consistently sense airway obstruction. Thus, measuring breathing difficulty might

be more based on perception than on indicators of lung function impairment (Rubinfield

& Pain 1976). A psychophysical test could be used to measure perception.

The psychophysical test is a type of physiological measurement that is a perceptual

estimate and is based on ratio-scaling methods. An example of using this type of


34

measurement in nursing research is asking subjects’ to rate the severity of their symptoms.

The reseacher generally asks subjects to give a number to rate their symptom as it

normally occurs. This rating number will become a reference scale and allows subjects to

estimate the severity of their symptoms as to whether the symptom is less or worse than

usual.

Psychophysical testing could be useful in exploring unknown phenoemenon and

building knowledge (Borg, 1982; Burns & Grove, 2001). During early the 1980s,

researchers asked subjects to rate their breathing efforts before and after dyspnea

episodes to explore the mechanism of perceived breathing difficulty in healthy subjects as

well as in subjects with cardio-pulmonary disease. The modified Borg scale was included

in psychophysical testing and often employed to measure breathing exertion in clinical

practice and reseach.

Another frequently used measure of dyspnea sensation often appearing in the

literature was the Visual Analogous Scale (VAS). Correlation studies of the VAS and

modified Borg scale indicated that both measures tapped dyspnea sensation (Lush,

Janson-Bjerklie, Carrieri, & Lovejoy 1988). The difference between the psychophysical

test and VAS was that the VAS was able to assess a wider range of dyspnea sensation than

the modified Borg scale. (Wilson & Jones, 1989).

Gottfried used two different physiological measures to examine factors influencing

dyspnea sensation in both healthy subjects and those with COPD. These two studies
35

verifed that patients with COPD had less sensitivity to breathing difficulty compare with

healthy subjects (Gottfired, Altose, Kelsen, & Cherniack 1981; Gottfried, Redline, &

Altose 1985). Thus, because patients with COPD might have different sensitivity to a

change in breathing status compared with healthy subjects, measuring objective

indicators of dyspnea would be preferred to more subjective measures, such as the VAS.

In addition, although dyspnea sensation was noted for its unique and individualized

characteristics, studies showed that the sensation of dyspnea may be or may not be

affected by personal or demographic characteristics, such as age or gender (Adams,

Chronos, Lane, & Guz, 1986). Two additional studies found that perception of dyspnea

was different among individuals with different demographic characteristics. Subjects of

younger ages and female gender (Brand, Rijcken, Schouten, Koeter, Weiss, & Postma,
H H H H H H H H H H H H

1992; Janson-Bjerklie, Carrieri, & Hudes, 1986) tended to report more severe dyspnea

sensation among those patients with emphysema-bronchitis, asthma, cardio-vascular

disease, and restrictive pulmonary disease. Dyspnea sensation also was influenced by the

presence of other symptoms, such as pain and anxiety in patient with lung cancer (Smith,

Hann, Ahles, Furstenberg, Mitchell, Meyer, Maurer, Rigas, & Hammond, 2001).
36

Table 2-3
Operational Perceieved Dyspnea

Author / Study Design / Sample Measurement Results


Year Purpose Description
Rubin & Descriptive with N=82 subjects Subjective 15% of
Pain, 1976 physioloical with measure of subject could
manipulation methacholine-in perceived not perceieve
environment duced asthma dyspnea and dyspnea with
To identify the objective FEV1 less
realtionship measure of than 50% of
between forced FEV1 the predicted
expire volume normal value.
(FEV) and These 15%
perceieved were not
dyspnea demographic
ally different
from the rest
of subjects.

Gottfried, Correlational N=50 subjects, FEV The


Altose, Compare the n=29 healthy Magnitude perception of
Kelsen, & perception subjects; estimate of change in air
Cherniack, ofchange in air n= 14 COPD breathing flow is
1981. flow among three patients; difficulty impaired in
groups n= 7 asthma COPD
patients.
37

Table 2-3(continued)
Operational Perceieved Dyspnea

Author / Study Design / Sample Measurement Results


Year Purpose Description
Adams, Correlational N=23 Forced expire Subejcts had
Chornos, To determine the N=19 healthy volume (FEV), wide
Lane, & measurement of subjects n=4 of functional variability of
Guz, 1986 dyspnea them were volume reporting
sensation regular smoker capcacity dsypena and no
(FVC), demographic
magnitude of differences
estimate of the were noted.
intensity of
sensation.

Janson- Descriptive N=68 subjects Visual Reports of the


Bjerklie, Correlational with analogue scale intensity of
Carrieri, & Compare non-malignant (VAS) dyspnea on
Hudes, physical and pulmonary VAS was
1986 emotional illness different by
response of disease group,
experiencing gender,
breathing p≤0.05
difficulty

Chonan, Descriptive N=16 healthy Tidal volume, Dsypnea


Mulhollan Correlational subjects, n=14 minute sensation was
d, To determine the males and n=2 ventilation, related to
Cherniack, realtionship females VAS both tidal
& Altose, between thoracic volume and
1987 movement and minute
the intensity of ventilation,
dyspnea p≤0.05
sensation
38

Table 2-3(continued)
Operational Perceieved Dyspnea

Author / Study Design / Sample Measurement Results


Year Purpose Description

Lush, Descriptive N=5 subjects Borg scale The Borg


Janson- To investigate with VAS scale and
Bjerklie, factors related to pulmonary VAS are
Carrieri, & dyspnea disease and highly
Lovejoy, receiving correlated,
1988 mechanical r=0.92,
ventilator. p≤0.001

Wilson & Descriptive N=10 healthy VAS High


Jones, 1989 Testing the subjects Borg Scale correlation
repeatability of (r=0.7)
both VAS and between
Borg scale borg scale
and VAS,
but VAS
has a wider
range.

Brand, Descriptive N= 412 Borg score Reports of


Rijcken, To determine middle-aged Force severe
Schouten, characestics subjects expiratory dsypnea
Koeter, differences on volume (FEV) sensation
Weiss, & reports of were
Postma, dsypnea associated
1992 sensation with
younger
ages,
female, and
severity of
illness.
39

Table 2-3(continued)
Operational Perceieved Dyspnea

Author / Study Design / Sample Measurement Results


Year Purpose Description

Smith, Descriptive N=120 patietns Survey Dyspnea


Hann, To determine the with stage one questionnaires was worse
Ahles, relationship to stage four in female
Fursteberg, among QOL and lung cancer group,
Mitchell, anxiety, dsypnea, p≤0.05.
Meyer, and body Dyspnea
Maurer, conciousness was related
Rigas, & to anxiety,
Hammond, p≤0. 05.
2001

The language of dyspnea in non-Malignant Population verus Maligant population

Patients with lung cancer also experience dyspnea during the trajectory of their

illness; however, few empirical works specific to this population have been done

previously. Researchers hypothesize that examining breathlessness descriptions from

patients with cardiopulmonary diseases might help determine the degree of breathlessness

in patients with lung cancer. The current literature focuses on the description of dyspnea

sensation in various disease diagnosis groups which includes cardio-pulmonary illness,

pulmonary illness, and cancer.

Results from descriptive studies provided similarity as well as differences in

language used to describe dyspnea between patients with non-malignant and malignant
40

diagnoses. Subjects with non-malignant disease and subjects with malignant diagnoses

both used the phrase “cannot get enough air” when describing their dyspnea sensation.

The cancer group also used “feel out of the breath” to describe their breathing difficulty

(Corcoran, Fielding, Grosby, Hughes, Tattersfied, & Wilcock, 2002).

The descriptions of dyspnea consisted of physical sensation, affective and

evaluative, and low energy (Skevington, Pilaar, Routh, & Macleod, 1997). For example,

patients with chronic air limitation chose "work/effort" and "heaviness" of breathing to

describe their breathing difficulty, and chose "increased inspiratory difficulty"

"unsatisfied inspiratory effort", and "shallow breathing" to describe their breathing

discomfort (O'Donnell, Bertley, Chau, & Webb 1997).

In terms of dyspnea experience in patients with cancer, the content analysis of

breathlessness assessment notes revealed the descriptions of dyspnea, including the

feeling of being unable to get enough breath, or of panic, or impending death (O'Driscoll,

Corner, & Bertley,1999).


41

Table 2-4.
Language of Dyspnea in Non-Malignant and Malignant Populations

Author / Study Design / Sample Measurement Results


Year Purpose Description
O’Dnonell, Descriptive N=24 subects, Borg scale Only those
Bertley, Comaprsion n=12 patietns (perceived with CAL
Chau, Compare the with severe inspiratory described
&Webb, qualitative chronic air difficulty), their
1997 aspects of limitation respiratory dyspnea as
breathing (CAL) and n=12 pressure, “ increased
difficulty age-matched breathing inspiratory
healthy subjects pattern, and difficulty”,
lung volume and :
unsatisfied
inspiratory
effort”.

Skevington, Descriptive N=48 dyspneic Descriptions Hierarchal


Pilaar, Assess the subjects, 31% breathlessness cluster
Routh, & description of with cancer from analysis
Macleod, dyspnea diagnosis and interviewing confirmed 4
1997 69% with 48 subjects classes of
non-malignant breathless-
cardiac ness.
pulmonary
diagnosis.

O’Donell, Qualitative N=52 lung Content Both


Corner, approach cancer patietns analysis of physical and
&Bailey, To investigate breathlessness emotional
1999 the subjective assessment sensations
descriptions of form were
breathlessness associated
with
dsypnea
descriptions
42

Table 2-4 (continued).


Language of Dyspnea in Non-Malignant and Malignant Populations

Author / Study Design / Sample Measurement Results


Year Purpose Description
Corcoran, Descriptive N=261, n=131 Survey Cluster
Fielding, To examine patients with questionnaire analysis
Grosby, dyspnea lung cancer; found
Hughes, description n=130 patietns both
Tattersfied, between cancer with cardiac groups use
& Wilcock, and non-cancer pulmonary “can not
2002. groups disease get
enough
air”.
Cancer
group also
uses "feel
out of the
breath”.

The Relationship between Anxiety and Dyspnea

Loop Feedback Mechanism

The body of knowledge of the relationship between anxiety and dyspnea indicated

that anxiety also could be the result of dypsnea, as opposed to its cause (Rice, 1950; Lum,

1981; Rosser & Guz, 1981). The loop feedback mechanism was proposed by Dudly,

Martin, and Holmes (1968) and the concept of this mechanism was that dyspnea becomes

a continuous process by a feedback loop mediated through anxiety. This mechansim was

based on their early study on a total of 36 subjects including healthy subjects and those

who had diffuse obstructive pulmonary syndrome (DOPS). Dudly and colleages designed
43

a protocol to create a certain degree of various discomforts and to examine whether

subjects’ dyspnea experiences were worse under the presence of those created discomfort

circumstances. This experiment showed that both psychological and physiological

changes are associated with dyspnea. The limitation of this early study is that Dudly used

the findings from individual analysis so that the conclusion was less reliable and then

lacked generalizability. However, research regarding exploring the phenomenon of

dyspnea on both non-malignant and malignant populations tended to assume that

emotional responses will worsen individual’s dyspnea and empirical evidences from these

studies had supported a significant relationship between anxiety and dyspnea sensation.

The following two sections address the empirical evidence of supporting the loop control

mechanism.

Anxiety and Dyspnea in the Non-malignant population

Empirical evidence regarding the relationship between anxiety and dyspnea mainly

have been focused on subjects with non- malignant disease such as chronic obstructive

pulmonary disease (COPD) and subjects with asthma. One study examined relationships

between anxiety and behavioral styles, anxiety, and dependency in 12 subjects with

asthma and 12 healthy subjects. The research findings concluded that those subjects who

were anxious or dependent had significantly greater threshold values for inspiration and

expiration, compare with those who were adaptive or rigidly independent (Hudgel,

Cooperson, & Kinsman, 1982). Another study of 52 patients with COPD also revealed
44

that the intensity of shortness of breath (SOB) correlated significantly to perceived

breathing effort, thus concluding that a measure of dyspnea’s affective component would

be important for outcome measures of dyspnea (Carrieri-Kohlman, Gormley, Douglas,

Paul, & Stulbarg 1996).

Anxiety and Dyspnea in Malignant Population

Cancer-related dyspnea is a complex symptom that varies by gender, age groups,

and may cluster with other symptoms, particular with anxiety (Smith et al., 2001). A

study of 135 patients with cancer also concluded there was a significant relationship

between anxiety and dyspnea sensation (Bruera, Schmitz, Neumann, & Hanson, 2000).

Another study also had the same conclusion, yet the correlation coefficient between

anxiety and dyspnea sensation was low (r = 0.29). The level of significance in this study

was a result of its a large sample size and measure (VAS) with narrow variability

(Dugeon & Lertzman, 1998). Anxiety was found to be an emotional response; therefore,

using the VAS alone might not be sufficient to capture such emotional response.

Three non-nursing studies conducted in Asia, including two in the Chinese

population and one in the Japanese population, also found a significant correlation

between dyspnea and anxiety (Chiu, Hu, Lue, Yao, Chen, and Wakai , & 2004; Chan,

Richardson, and Rishardson, 2005; Tanaka, Akechi, Okuyama, Nishiwaki, &Uchitomi,

2002). However, no nursing studies have been conducted examining cancer-related

dyspnea in Taiwan.
45

The results of two qualitative studies indicated that the acute dyspnea experience

was associated with anxiety as well as emotional functioning (Bailey, 2004). Such

dyspnea experiences were observed with patient issues such as “panic beckons,” loss of

control, and dying (Tarizan, 2000).

Table 2-5.

Literature on the Relationships between Anxiety and Dyspnea

Author / Study Design / Sample Measurement Results


Year Purpose Description
Hudgel, Correlational N=24, Psychophysical Anxious
Cooperon, Comparison n=12 healthy test dependent
& Kinsman, Study subjects; subjects had
1982. hypothsize that n=12 asthma a greater
dyspnea in subjects inspiratory
asthma patients threshold.
is partially
dependent on
psychological
factors.

Carrieri- Descriptive N= 52 COPD VAS The


Kohlman, To investigate if subjects and intensity of
Gormley, COPD subects have dyspnea shortness of
Douglas, are able to breath was
Paul, & differientiate related to
Stulbarg, distress and breathing
1996 anxiety effort, and
associated with distress was
their dyspnea. related to
anxiety, both
with
p≤0.001
46

Table 2-5 (Continued).

Literature on the Relationships between Anxiety and Dyspnea

Author / Study Design / Sample Measurement Results


Year Purpose Description
Dudgeon, & Descriptive N= 100 VAS-Anxiety Anxiety is
Lertzman, To evaluate the terminally ill and Dyspnea related to
1998 relationship patients, Physiological shortness of
among dyspnea including 49 indicators breath,
sensation, patients with p≤0.001.
anxiety, and lung cancer.
respiratory
labor.

Bruera, Prospective 135 cancer VAS Anxiety was


Schmitz, study patients associated with
Pither, To investigate attending pain dyspnea,
Neumann, factors related clinics p≤0.001.
& Hanson, to dyspnea.
2000.
Traxian, Qualitative N=10 nurses In depth One of the
2000. To explore the caring dying interview thematic
lived experience patients summaries
of caring stated patient’s
terminal ill look and panic
subjects with beckons.
dyspnea.

Tanaka, To investigate N= 171 lung HAD- Multiple


Akechi, factors cancer patietns psychological regression
Okuyama, contributing to distress indicated that
Nishiwaki, cancer related VAS-Dyspnea anxiety was a
& Uchitomi, dyspnea. significant
2002. predictor of
dyspnea.
47

Table 2-5 (continued).

Literature on the Relationships between Anxiety and Dyspnea

Author / Study Design / Sample Measurement Results


Year Purpose Description
Bailey, 2004 Qualitative N= 10 COPD In–depth Described that
approach patient-family interview dyspnea
To explare the groups experience
lived experience was related to
of both family anxiety.
and patietns
with acute
dyspnea episode

Chiu, Hu, Retrospective N=125 hopice Chart review, Anxiety was


Lue, Yao, study and palliative Borg scale associated
Chen, & To investgate care patients with dyspnea,
Makai, 2004 factors r=0.21,
associated with p≤0.05.
dyspnea in
terminal ill
patients

Relationship among Dyspnea, Breathing Effort, and Anxiety

Dyspnea has both subjective and objective perspectives. The relationship between

these two perspectives and dyspnea is still unclear with respect to how these two

perspectives explain dyspnea sensation. One study suggests that dyspnea is associated

with objective perspectives, such as diaphragm excursion, and smoking history; Dudgeon,

Lertzma, & Askew, 2001).


48

Direction of the relationship between Dyspnea and Anxiety

The above discussed empirical evidence indicates the existence of a moderate to

high correlation between anxiety and dyspnea. The issue related to direction, whether

dyspnea explains anxiety or whether anxiety explains dyspnea in persons with cancer,

remains controversial. Mystakidou, Tsilika, Parpa, Katsouda, Galanos, and Vlahos (2005)

suggest that dyspnea significantly explains anxiety. Contrary to this finding, Tanaka,

Akechi, Okuyama, Nishiwaki, and Uchitomi, (2002) suggest that psychological distress,

including anxiety, is a significant predictor of dyspnea.

Table 2-6.
Literature on the Relationships among Breathing Effort, Dyspnea, and Anxiety

Author / Study Design / Sample Measurement Results


Year Purpose Description
Dudgeon, Cross-sectional N=75 cancer VAS-Anxiety Anxiety and
Lertzman, Correlational patients with VAS-Dsypnea dyspnea
& Askew, Examine both moderate to Physiological were related
2001 subjective and severe dyspnea indicators for to each other,
objective respiratory r=0.26,
factors muscle strength p≤0.05.
associated with
dsypnea
Tanaka, To investigate N= 171 lung HAD- Multiple
Akechi, factors cancer patietns psychological regression
Okuyama, contributing to distress indicated that
Nishiwaki, cancer related VAS-Dyspnea anxiety was
& Uchitomi, dyspnea a significant
2002. predictor of
dyspnea.
49

Table 2-6(continued).
Literature on the Relationships among Breathing Effort, Dyspnea, and Anxiety

Author / Study Design / Sample Measurement Results


Year Purpose Description
Mystakidou, Correlational N= 120 HAD-T for Multiple
Tslilika, comparsion advanced psychological regression
Parapa, study cancer subjects distress analysis
Katsouda, To investigate indicated that
Galanos, & anxiety and dyspnea
Vlahos, depresssion in significantly
2005 cancer patients explained
anxiety,
beta=0.025,
p≤0.05.

Summary

Dyspnea is the consequence of increasing breathing effort. Based on the findings

from several physiologically-oriented studies, breathing effort is associated with changes

in lung volume that can be detected by respiratory pressure measurement without

invasive procedures. Perceived dyspnea is a complex mechanism involving both physical

and psychological aspects, thus using psychophysical rule is the most appropriate method

to explore and explain it. The sensation of dyspnea is an individualized experience but

age and gender tend to influence behavioral responses as well as self report of the

severity of sensation. The main behavioral response of dyspnea, based on empirical

works, is anxiety. Empirical works have supported the existence of a significant moderate

to high correlation between anxiety and dyspnea.


50

The limitation of the current literature regarding anxiety and dyspnea is a lack of

evidence regarding the direction of the relationship between anxiety and dyspnea

sensation. Although the loop feedback mechanism indicates that dyspnea is moderated

through anxiety, current literature is limited on methodology and analysis plans. The

empirical evidence clearly suggest a significant relationship between anxiety and dyspnea

in both the non-malignant and the malignant population. However, little evidence

supporting the loop feedback mechanism provides a hypothesis on the causality of

dyspnea. Advanced empirical evidence is needed to support this hypothesis. Thus, the

current study design is focused on examining the associated between anxiety and dyspnea

sensation, and also on identifying the direction of this association.


51

Chapter Three

Methodology

This chapter presents the methodological considerations and the decision-making

choices used to conduct this study. The content of this chapter includes: study design,

sampling frame, measurement and instrumentation, data collecting procedures, data

management, analysis plan and threats to internal and external validity. The methods used

to protect human subjects are also addressed in this chapter.

Study Design

This study searched for the factors associated with the phenomenon of dyspnea in

patients with cancer. The design of this non-experimental study was cross-sectional

correlational.

Sampling Frame

Rationale for the elements of the sampling frame (descriptions about the setting

and sampling plan, determination of sample size, and inclusion criteria) for this study is

presented in this section.

Setting and Rationale

In Taiwan, cancer has been the leading cause of death since 1981. Based on the

latest official statistical report in 2001, the incidence of lung cancer is 6555/100,000,

making it the leading cancer occurrence in Taiwan

(http://crs.cph.ntu.edu.tw/crs_c/annual.html, 2005). Although a national cancer registry


52

exists in this country, this database could not be accessed for the purposes of recruitment.

The alternate effective method of accessing potential participants through a specific

health care system was used. This study was conducted in the cancer center of Chung

Gung memorial hospital (CGMH) in Northern Taiwan. This center was selected because

it was a regional referral center for the four major counties of northern Taiwan.

Sampling Plan

The potential sample population in this study was identified as patients with lung

cancer, considering the feasibility regarding contacting this patient population. Therefore,

this study used a non-random convenience sampling frame. The potential participants

from which the convenience sample was drawn were referred by CGMH clinicians.

Sample Size Determination

Power analysis was used to determine the probability of detecting the effect of

interest, which accurately reflected the existence of the effect in the larger population.

The calculation of power for this study was consistent with the theoretical foundation for

establishment of power. The four parameters which were critical to power analysis

included the type of statistical significance test, significance criteria, effect size, and

sample size. In addition, these four parameters of power analysis were relevant to one

another. Statistical testing of null hypotheses, consideration of the risk for making type I

and type II errors, and literature review on statistical findings from relevant research,

were used to guide the selection of statistical significance test type, significance criteria,
53

and effect size. These parameters then were employed to calculate an estimated sample

size for this study.

Type of Statistical Significance Test

The statistical significance test was used as a statistical approach in terms of

rejecting the null hypothesis for a specified value and replacing the null with an

alternative hypothesis. The purpose of regression analysis for this study was to examine

the amount of variance of the dependent variable (Y, the perceived severity of dyspnea)

that can be explained by all the predictors (Xs, breathing effort, anxiety, and the joint

effect of both). The concept that the variance of Y could be accounted for by all the Xs

was based on correlation analysis. The statistical significant test, F distribution, was

dependent on the concept of the correlation coefficient reaching a specified critical value,

which then was used to indicate whether the null hypothesis should be rejected.

Effect Size

Calculating the effect size aimed to quantify the strength of the probability with

regard to breathing effort, anxiety, and the joint effect of both in explaining the perceived

severity of dyspnea. Thus, the effects from both the simple main effect and the interaction

effect were expected to influence the determination of one artificial number as

representing the effect size of this study. The effect size of the simple main effect was

associated with the parameter of r2Y.M and the parameter of r2Y.MI was associated with both

the simple main effect and the interaction effect. The determination of the association
54

between r2Y.M and r2Y.MI was that when there was a larger effect size of the simple main

effect, there should also be a corresponding large effect size of the interaction effect

(Jaccard & Turrisi, 2003).

Previous physiological empirical works provided the evidence to support the

concept that respiratory muscle work effort has a close relationship with dyspnea induced

artificially in healthy subjects by adding elastic loading (Killian, Gandevia, Summers, &

Campell, 1984; Killian & Jones, 1988; Killian, 1998). In addition, in some studies,

anxiety significantly predicted shortness of breath (R2=0.13) (Dugeon, Lertzman, &

Askew, 2001). Based on this study and the principle of calculating the effect size on a

regression model, R2/1-R2, the effect size of the simple main effect was set as 0.15, a

moderate effect size. Because the simple main effect was expected to be moderate, one

could logically assume the interaction effect would also be moderate in this study.

Detection of an interaction effect in this study, as in any regression model, required at

least a moderate effect size (Cohen, 1988). Thus, the effect size of the interaction effect in

this study was determined as 0.13, a moderate effect size. Based on Cohen’s (1988)

suggestion in which the moderate effect size was approximately r2YI.M=0.13 for the

interaction effect and was equivalent to f2=0.15 in multiple regression, the parameter of

effect size used for calculating sample size in F distrubution was 0.15.

Significance Criteria

The null hypothesis of this study was that the joint effect of breathing effort and
55

anxiety does not explain the perceived severity of dyspnea. Two significant criteria were

related to the standards of assessing the risks about making errors regarding acceptance

or rejection of this null hypothesis. The first criterion was the parameter of alpha and it

represented the probability of making a type I error, falsely rejecting the null hypothesis

when a true difference did not exist. The second significance criterion was the parameter

of power such that one minus the parameter of power represented the probability of

making a type II error, mistakenly accepting the null hypothesis when a true difference

did exist.

The decision of determining these two criteria was dependent on what type of error

the researcher in this study was willing to make when testing the null hypotheses. The

purpose of this study was to identify the role of anxiety in explaining the perceived

severity of dyspnea in cancer patients. The findings were expected to be able to provide

an in-depth understanding of the mechanism of cancer-related dyspnea and further

helping clinicians and researchers to identify effective management strategies for this

symptom. For the purpose of this study, investigating the relationship among breathing

effort, anxiety, and the severity of dyspnea, avoiding making type I errors and type II

errors were equally important. Thus, determination of these two critical criteria was based

on Cohen’s (1988) suggestion with regard to conducting standard behavioral research.

The significance criterion of setting an alpha was dependent on determining the existence

of the phenomenon being studied, as rejecting the null hypothesis. The small value of α
56

indicated more strength to determine the existence of this phenomenon. The parameter of

alpha was set at 0.05 and the power was set at 0.8, the suggested minimal parameter for

prevention of making a type II error in behavioral research (Cohen, 1988).

Sample Size

Based on the above parameters, F distribution, the combination of large effect size

for simple main effect and moderate effect size for interaction effect, with an alpha of

0.05 and a power of 0.8, and using by Cohen’s (1988) power table, the sample size

needed for this study was estimated to be 55.

Inclusion Criteria and Rationale

In order to have enough variation for regression analysis and to control for

confounding variables in this study, the following inclusion criteria and rationale were

used.

(1) The age of each subject must be at least 21 years old. This study was focused on the

adult.

(2) A histological diagnosis of lung cancer, regardless of stage. Dyspnea is a common

symptom in lung cancer subjects. Primary lung cancer or other types of cancer with lung

metastasis cause lung tissue to be diffused with cancer cells, leading to airway

obstruction and poor oxygenation. If there is a massive obstruction in the lung, alveolar

atelectasis also occurs. Atelectasis is a pathological change in the lung characterized by

alveolar collapse, which leads to gas exchange difficulty (Guyton, & Hall, 2006). These
57

pathological changes increase the respiratory muscle work load needed for breathing

leading to the dyspnea sensation.

(3) Self-report of moderate to severe dyspnea by indicating the severity of dyspnea

sensation between 40-80 on a 0-100 Visual Analogous Scale (VAS). Regression analysis

is based on correlations among variables and thus it is important to have enough variation.

In order to have enough variation of the data for regression analysis, subjects were those

who rated their intensity of dyspnea between at least 4 and 8 on a 0 to 10 visual analogue

scale, which represented moderate to severe dyspnea.

(4) Subjects could not have any indication of cognitive impairment. Dyspnea sensation is

a subjective experience and is dependent on subject self report. Thus, it is important to

make sure that no indication of cognitive impairment exists for each subject. Cognitive

function was determined by the Katzman’s short memory concentration test. This test was

a 6-item Orientation-Memory-Concentration Test that had been validated as a measure of

cognitive impairment. The score was based on counting the numbers of errors made by

the person taking this test. If the errors were greater than 6, the subject was considered to

be cognitively impaired and not eligible to participate in the study.

(5) Subjects needed to be able to speak and understand Chinese or Taiwanese. Because

this study was conducted in Taiwan and Chinese was the official language, subjects

needed to be able to speak and understand Chinese. Taiwanese was another language

commonly used among elderly Taiwanese.


58

Data Collection Procedure

This study was approval by the Institute Review Board (IRB) from both Chung

Gung Memorial Hospital, Linkuo branch in November, 2006 and from University

Hospitals of Cleveland in January, 2007. Other requirements, in addition those of the IRB,

were satisfied and approval from CGMH was obtained in January, 2007. After approval,

screening, consenting for the potential participants, and enrolling the eligible participants

for this study were carried out at the CGMH. Screening of potential participants relied on

both the medical chart review for confirmation of the histological diagnosis of lung

cancer and on clinical reports from the patients’ physicians. Because the investigator did

not have access to the medical charts or clinical reports, screening potential participants

mainly relied on the clinicians’ referrals.

Clinician assistance was needed in the process of screening potential participants.

The investigator first sought those clinicians who provided medical care as well as those

who provided nursing care to patients with lung diagnosis. The clinicians for potential

subjects included physicians from both the oncology and pulmonary medicine

departments, and staff nurses from the oncology and pulmonary medicine units.

In order to start the screening process, the investigator first held two formal

meetings: one meeting with the nurse-managers and staff nurses from cancer and

pulmonary medicine units; and another meeting with oncologists, and pulmonologists.

Clinicians from both the nursing and medical disciplines agreed to refer potential
59

participants to the investigator.

Screening Potential Subjects and Enrolling Subjects

After identifying subjects, the investigator asked them for permission to evaluate

cognitive function using the Katzman’s short-memory –concentration test to determine

eligibility. Those subjects who were found to be cognitively intact and who met all other

inclusion criteria were then invited to participate in the study. Reasons for refusal to

participate among eligible potential subjects were documented. Figure 2 illustrates the

inclusion diagram of this study.


60

Figure 3-1. Inclusion Diagram

All Cancer Treatment Patients age≥ 21

*A.
Histological diagnosis of Lung Cancer
(Medical Chart & Clinician report)

Report of their dyspnea rated between 40 and


80 on a 0 to 100 mm VAS (Clinicians
Referral)

Lung Cancer Patients


**B. (Age≥ 21 and Self-reported Dyspnea as
Moderate to Severe)

Refused Cognitive Status


(Katzman’s Test
Score < 6)

Document
Consented
Reasons

Enrolled Subjects

Note: * A. illustrates the process of screening the potential subjects.

** B. illustrates the processes of consenting and enrolling subjects.


61

Data Collection

Data collection occurred during two face-to-face interviews by the investigator. At

the first interview, measurements of respiratory rate/depth and ventilation output, as

measured by FEV1 (forced expiratory volume exhaled in the first 1 second) and FVC

(force vital capacity), were collected as the baseline line of subjects’ breathing status. The

second interview was held on a separate day from the first interview. Measurements

respiratory rate/depth and ventilation output (FEV and FVC) were collected for the

second time. After the observable measures were obtained, physiological indicators of

respiratory pressure were measured for data collection. These physiological measures

were assessed before any survey questionnaires were administered to prevent tiring of

subjects, which could affect their breathing status. In addition, to reduce subjects’ burden

with regard to dyspnea, vertical visual analogs instruments for breathing effort, dyspnea

sensation were used. Only the subscale of sense of anxiety of cancer dyspnea scale

(which had four items), and the 6-item short-form of the State anxiety inventory were

used with these subjects.

Measurement Validation

Katzman’s Short-Memory-Concentration Test

Katzman’s short-memory-concentration test was used to screen subjects regarding

their cognitive status, to ensure compliance with the inclusion criteria for this study. This

instrument was a 6-item scale to assess a person’s orientation, short task concentration,
62

and learning and recalling simple information. This scale was initially derived from a

26-item mental status survey by Katzman, Brown, Fuld, Peck, Schechter, and Schimmel

(1983). This derivation involved using the data from subjects (N=321) who completed

the 26-item mental survey. A “least square discrimination analysis” was used to compress

the survey into 6 items. Katzman and colleagues reported that this 6-item test was able to

predict the scores on a validated 26-item mental status survey. Then, testing this 6-item

scale on 170 additional patients in two skilled nursing homes, a health-related facility,

and a senior center was performed. Among those 170 patients completing 26-item and

6-item scales, 38 patients were identified to have plaque counts in their brain lobes

(obtained after death). Plaque formation is an indication of the presence of dementia from

the processes of Alzheimer’s disease. The Pearson correlation verified that both 26-item

and 6-item scales were significantly correlated with the plaque, (r=0.595 and 0.542,

respectively; p ≤ 0.001) (Katzman, Brown, Fuld, Peck, Schechter, & Schimmel, 1983).

Independent Variable

The independent variable was breathing effort, which was defined as increasing

intensity of breathing that causes exertion to breathe through the stimulation of

intramuscular sensory receptors. Based on the literature review of several physiologic

studies, the exertion to breathe was proposed to be measured by the respiratory muscle

work effort. However, breathing effort data could also be from an individual’s subjective

description as a personal discomfort experience associated with the act of breathing.


63

Because breathing effort measurement could be obtained from several sources, both

objective (the physiological indicator) and subject measures (vertical visual analogs scale,

VVAS) were carried out to assess breathing effort. The subjective measure of breathing

effort was obtained using a 0-100 vertical visual analogous scale.

The physiological measurement of breathing effort was assessed using respiratory

pressures. To measure the contractile forces of respiratory muscle, the pressures

generated from the respiratory muscle contraction were evaluated. This selection of this

measure was based on the work of Fishbein, Kearon, and Killian (1989), who suggested

that despite larger variability in the general population, using static maximum inspiratory

and expiratory pressures was reasonable as indicators of respiratory muscle work effort in

cancer subjects with dyspnea. Their rationale was that severe dyspnea in cancer patients

was not likely to occur unless the static pressures fell below 40 cmH2O. Furthermore, a

significant positive correlation between perceived magnitude of effort and inspiratory

pressure as well as lung volume was found in among healthy subjects, p≤0.05 (Killian,

Gandevia, Summers, & Campell, 1984).

The non-invasive method selected to measure inspiratory and expiratory pressures

for the purposes of this study involved measuring the pressure from air flow through the

mouth (Rochester, 1988) using a portable respiratory pressure meter. Figure 3 illustrates

this device. Smith and Royall (1992) reported that a bedside measurement of maximum

static respiratory pressures was reliable to represent monitoring respiratory muscle


64

weakness in both acute and chronic conditions. In addition, Hamnegard and colleagues

(1994) recommended that the mouth pressure meter was able to accurately record

respiratory pressure.

Figure 3-2.

A Portable Respiratory Pressure Meter

The validation of any physiological measure relies on accuracy, precision,

sensitivity, and error. The accuracy, as equivalent to validity, was the extent to which this

instrument measured breathing effort. Precision, as the equivalent to reliability, was

assessed by examining the reproducibility, repeatability, or consistency among

measurements of the same concept. Sensitivity was the determination of whether or not

this physiological measure detected the true phenomenon. Hamnegard, Wragg, Kyoussis,

Aquilina, Moxham, and Green (1994) compared the results of PImax and PEmax from

the mouth pressure meter to results from laboratory standard pressure recording

equipment. The paired t-test showed an insignificant difference between these two

methods of measuring respiratory pressure. The insignificance was evidence of accuracy

of this physiological indicator.

Thus, measuring respiratory pressure through the mouth was a non-invasive and
65

reliable indicator of the respiratory muscle strength. The device MicroRPM (Respiratory

Pressure Meter) was used to measure the respiratory muscle effort in this study. This

device not only measured the maximum inspiratory and expiratory pressures but also

provided a data-based software package to calculate the Sniff Nasal Inspiratory Pressure

(SNIP). The MicroRPM was designed to monitor the strength of respiratory muscles

primarily in the COPD population as well as in patient populations who have respiratory

muscle function affected by other diseases. This device was housed as a hand-held item

and was light weight with battery operation. The result was read as a digital number

displayed in units of centimeters of water (cmH2O) for comparison with normal adult

maximum static mouth pressures based on the following equations: (1) inspiratory

pressure for males: 142-(1.03 x age) , (2) inspiratory pressure for females:(0.71 x height

in cm) – 43, (3) expiratory pressure for males: 180-(0.91 x age), and (4) expiratory

pressure for females: 3.5+ (0.55 x height in cm) (Wilson, Cooke, Edwards, & Spiro,

1984). The company also reported the validation of the device including its specificity

and accuracy. The specificity of this device included an operating pressure in the range of

±300 cmH2O, burst pressure in the range of ±700 cmH2O, and a resolution of one

cmH2O. The accuracy was in the range of ± 3 %. This device also was considered

accurate at room temperature between 32 and 104 Fahrenheit and room humidity from

30% to 90% (http://www.micro-direct.com/).


66

Moderator Variable

The moderator variable was anxiety, defined as a psychological reaction that occurs

along with dyspnea sensation. This type of anxiety was classified as “state” anxiety. The

Spielberger state-trait anxiety inventory (STAI) with two subscales, the trait scale and the

state subscale, is considered a gold standard of measuring anxiety in psychology.

However, the state scale of anxiety included a total of 20 items and administrating this

scale could increase subjects’ burden. Thus, the 6 item short-form state scale of the STAI

was used to measure the state anxiety in this study. This version of the scale was found to

have a Cronbach’s α of 0.82, indicative of good internal consistency (Marteau & Bekker,

1992).

In addition, because anxiety in this study was specifically accompanied by dyspnea,

the second measure was the “sense of anxiety” subscale of the cancer dyspnea scale

(CDS). The CDS, developed by a group of Japanese researchers, is a 12-item self-rating

scale composed of three subscales: sense of effort, sense of anxiety, and sense of

discomfort. The subscale ‘sense of anxiety’ had four items. The responses of all items,

including those in the subscale ‘sense of anxiety’, were Likert-type scales. The response

was categorized by 1 as not at all, 2 as a little, 3 as somewhat, 4 as considerably, and 5 as

very much. Based on the scoring of responses, higher scores indicated higher intensity of

each item, which means a more severe symptom. In addition, the sample population in

this study is Taiwanese; Taiwanese and Japanese have some degree of similarity in terms
67

of oriental culture. Thus, it was reasonable to use this instrument with this study’s sample

population.

The psychometric reports regarding this scale were based on 166 cancer patients in

a National Cancer Center in Japan. The scale had well established construct validity and

convergent validity. Each subscale correlated with VAS reports of dyspnea sensation (r

=0.57; p < 0.001) as well with as the modified Borg scale (r =0.52; p < 0.001). The Borg

scale is a tool for assessing individual rating of one’s perceived exertion. The score of

Borg scale ranged from 6 (no exertion at all) to 20 (maximal exertion)

(http://www.cdc.gov/nccdphp/dnpa/physical/measuring/perceived_exertion.htm). The

original Borg scale was previously modified to increase the ease of administration. The

modified scale ranged between 0, no exertion at all, 0.5, just noticeable exertion, to 10,

maximum exertion. In addition, the subscale “sense of anxiety” correlated with

State-Trait Anxiety Inventory (STI) with r=0.28 and 0.33, p < 0.001, respectively. The

STI was initially developed for adults with anxiety and this measure was broadly

employed in measuring anxiety across populations.

The reports for the anxiety subscale’s reliability included both Cronbach’s alpha

indicating internal consistency and test-retest reliability representing the stability of CDS.

Cronbach’s alpha coefficiencies were 0.83 of the subscale “sense of effort,” 0.81of the

subscale “sense of anxiety,” and 0.91 of the subscale “sense of discomfort.” These

Cronbach’s alpha coefficiencies indicated good internal consistency. The test-retest


68

reliability was based on 37 subjects with two administrations of the CDS an average of

6.9 days apart. The correlation coefficiencies between the two administrations among the

three subscales were 0.71, 0.69, and 0.58, respectively (p≤0.01), meaning that CDS was a

stable instrument (Tanka, Akechi, Okuyama, Niskiwak, & Uchitomi, 2000).

Dependent Variables

The dependent variable was the perceived severity of dyspnea. The definition of

the perceived severity of dyspnea was the cognitive response and the perceptional

sensitivity of an individual rating his or her severity of dyspnea. The VVAS (Vertical

Visual Analogous Scale) was used to measure perceived severity of dyspnea because this

scale used a self report to measure a subjective experience. This self report scale was a

100 mm vertical line with two anchors indicating minimal and maximal extremes of

breathlessness on each end. The bottom anchor indicated no dyspnea at all and the top

anchor indicated dyspnea as severe as it can be. The person marked a point on this line to

rate his or her perceived severity of dyspnea. The vertical format was used because it has

been found to be easier for subjects in a sitting position to use.

Gift (1989) validated the use of the VVAS as a measurement of clinical dyspnea

among asthmatic and COPD subjects. The concurrent validity was established by

comparing the VVAS to peak expiratory flow rates in asthmatic subjects. The results

showed that the correlation coefficiency between the VVAS and the PEFR was good ( r =

-.85), supporting the validation of VVAS. In addition, construct validity of VVAS then
69

was established by using the tool under two circumstances, severe obstruction and small

airway obstruction, among both asthmatic and COPD subjects. The results showed that

the ratings of dyspnea were different in these two extremes; this means that the VVAS

was able to tap the domain of the concept of dyspnea sensation under different

pathophysiological changes resulting from different causes. Thus, the investigator

concluded that for comfort, a vertical format was easier to use for those who need to sit

due to serve dyspnea.

Instrumentation

Although it was important to ensure all the methods of measurement used in this

study were reliable, only the CDS had been validated previously in the cancer

population. The VVAS was validated by subjects who had asthma and COPD. No

updated study had been conducted using the respiratory pressure meter through the

mouth to assess dyspnea in cancer patients. The physiological indicator in the studies

conducted by Killian and colleagues indicated a close relationship between sensation of

effort and dyspnea. Thus, the physiological indicator was acceptable.

All the survey questions, including Katzman’s short-memory-concentration scale,

6-item short-form of the STAI, and CDS, were double translated. Double translation was

first to translate instruments into Chinese and then translate the Chinese version back to

English by two persons working independently. The investigator translated these

surveys from English version to the Chinese version. Then, the second person translated
70

the Chinese version back to an English version. The English version translated from the

Chinese version was compared with the original version of each scale. The investigator

and the second person discussed the disagreement of items regarding language use and

meaning of each item. Mutual agreement between the investigator and the second

person was achieved; the final Chinese versions of each scale were then completed.

In addition to Chinese, Taiwanese was the other language commonly used among

the elderly in Taiwan. However, the Taiwanese language did not have character changes

in writing; only the pronunciation differed from Chinese using the same characters.

Thus, interviews with those subjected who only spoke Taiwanese was carried by the

investigator. The investigator asked the questions (written in Chinese) in spoken

Taiwanese. The investigator then recorded the subject’s verbal response.

Data Management

Raw data were entered by investigator. Data cleaning included checking errors by

comparing data collection sheet and data base, and examining the description of

demographic data to determine whether any outliers existed was employed after all raw

data were entered in the data base. SPSS (version 15.0) software was used to carry out the

analyses.
71

Analyses Plan

Descriptive Statistics

Descriptive data analysis was used first to examine the normality of the data set

reported in chapter four.

Inferential Statistics

Inferential statistical analysis was used to answer the research questions. Research

questions 1: Does breathing effort significantly predict individuals’ perceived severity of

dyspnea? This question was answered by linear regression analysis in which the overall

variance, significance, and the strength and significant level of each β were evaluated.

Research question 2: Is anxiety positively related to individuals’ self-rating of their

intensity of dyspnea sensation? This question was answered by Pearson correlation which

represented by its strength, p value, and direction.

Research question 3: Is anxiety a moderator influencing the relationship between

respiratory work effort and individuals’ perceived intensity of dyspnea, in the positive

direction? This question was answered by the moderated multiple regression analysis.

The evaluation of the regression analysis was discussed in the following which included

statistical concerns of performing MMR analysis as well as the interpretation of results.

Moderator Multiple Regression Analysis

The focal statistical analysis for examining whether anxiety was a moderator was

the moderator multiple regression (MMR). The statistical approach of MMR analysis was
72

first introduced by Sanders (1959) and is aimed at examining how one variable

(moderator) influences the natural relationship between two other variables. The term

moderation with regard to statistical perspective is the interaction effect, as the jointing

effect of breathing effort and anxiety in explaining the severity of dyspnea sensation.

In addition, the MMR analysis, just as multiple regression analysis, was

conceptualized by multivariate analysis. The benefit of using a multivariate statistical

approach was to merge the variance and co-variances of the independent variables into

one vector. The variance was the correlation within each variable and the covariance was

the correlation between variables. Thus, the effects of breathing effort and anxiety on the

perceived severity of dyspnea were taken into account as one, as opposed to being

examined individually. This approach removed the overlapping influence of breathing

effort and anxiety, the noise. Therefore, the explanation of the causal relationship of

breathing effort, anxiety, and the perceived severity of dyspnea would be more credible.

Predictors

The identification of independent variables and moderator variables when

constructing the MMR model requires strong theoretical support (Jaccard & Turrisi,

2003). Chapters I and II have identified that breathing effort is the independent variable

and the presence of anxiety is the probable moderator in the theoretical framework

guiding this study examining why there is variability in the perceived severity of dyspnea

sensation among patients with cancer. In summary, the predictors of the MMR model in
73

this study are identified as breathing effort, anxiety, and the joint effect of breathing effort

and anxiety.

The Effect of Predictors

In the moderator regression model with respect to mathematical technique, two

types of effects, the simple main effect and the interaction effect, need to be examined

(Aiken & West, 1991). Statisticians have addressed that the interaction effect would not

exist unless the main effect exists; thus, this section addresses both the main effect and

the interaction effect in the statistical model. The simple main effect in this study is the

degree to which breathing effort and anxiety explain the perceived severity of dyspnea.

The interaction effect is the joint effect of breathing effort and anxiety together in

explaining individuals’ perceived severity of dyspnea. In terms of statistical concern, the

benefit of using interaction effect is that this method partitions the independent variable

and the moderator variable into subgroups. Using partitioning maximizes the predictor’s

effects on the perceived severity of dyspnea and this maximum effect can be either in a

positive direction or a negative direction (Aiken & West, 1991; Jaccard & Turrisi, 2003).

In this study, the moderation effect assumes a positive direction; in other words, the

presence of anxiety will cause the individual to experience a more severe dyspnea

sensation. Statisticians have claimed that it is of no value to interpret the interaction

effects if the overall regression is insignificant (Bideian & Mossholder, 1994). Thus, the

MMR model must be able to allow the investigator to assess the simple main effect and
74

the interaction effect separately and then in a hierarchical model, using more than one

step to enter the variables (Aiken & West, 1991; Cohen, Cohen, West, & Aiken, 2003).

The SPSS statistical package contains the program for hierarchical regression

analysis, which allowed the researcher to enter the predictors into sequences of blocks.

The sequence of entering predictors in this study was entering the main effect, including

both breathing effort and anxiety, into the first block and then entering the interaction

effect into the second block. The presence of an interaction effect was determined by the

product of breathing effort and anxiety and represented by an artificial number of

breathing effort * anxiety.

The significance of the simple main effect was based on the p value of its F

distribution from the first block. Examining interaction effect in this study was based on

the concepts of the Omnibus interaction significant test, globally examining the entirety

of the interaction effect between breathing effort and anxiety (Jaccard & Turrisi, 2003).

Thus, the significance of the interaction effect also was based on the p value of its F

distribution from the second block.

The parameter of R2 was the amount of the variance in the tolerance being

accounted for by the predictors. The incremental R2 indicated how much more variance

arose from the second block and then, determined the influence of the joint effect on the

perceived severity of dyspnea. Thus, the strength of the interaction effect was reported by

the incremental R2 (Bideian & Mossholder, 1994; Jaccard & Turrisi, 2003).
75

Regression Diagnosis

One of the important considerations of the regression model was its stability, as this

model was able to capture all the data obtained from subjects. Because MMR was based

on ordinary least square (OLS), this meant the relationship between the independent

variables and the dependent variable must be linear. A linear model was based on the

concept of assuming that data obtained from subjects (observational value) would

aggregate and form an ideal straight line. Because the observational value cannot be on

the exact dot (referring to true value) of this ideal straight line, the presence of a distance

indicated the difference between observational value and true value. This distance was

the residual. The perfect linear model is one in which the observational value and true

value are on the same dot; therefore, the closer these two values are, the smaller the

residual, indicating closeness to a perfect linear model. Thus, the specific statistical

assumptions of the linear regression model were based on their residual assessment.

Four residual assumptions (mean of zero, no autocorrelation, normal distribution,

and homoskedasticity) of the regression analysis were examined. Violating these residual

assumptions would reduce the statistical validity of this study. The violation of the

assumptions of normal distribution and homoskedasticity could be overcome and made

more robust by the F distribution (Cohen, Cohen, West & Aiken, 2003). However,

violating the assumptions of mean of zero and no autocorrelation would lead to specific

errors in any linear regression model. In this study, the regression model was set up as an
76

intercept, which did not violate the assumption of mean of zero. Autocorrelation occurred

when “nesting” data was present (referring to data nesting to organization) and time

series data (referring to data collection occurring more than one time point). The data in

this study were collected at the same hospital and at one time point; thus, providing

freedom from violation of the assumption of no autocorrelation.

The multicollinearity was the inter-correlation among all the predictors and a high

correlation coefficiency indicated that different predictors were, in fact, measuring the

same latent variables. Multicollinearity of predictors could lead to bias R square of the

overall regression model. This means that the results might indicate an insignificant

model, yet, in fact, the model should be significant in reality. Thus, the presence of

multicollinearity would bias the regression model and then could cause a type II error.

Two considerations of multicollinearity in this study were examined before the analysis

and were addressed to examine and deal with multicollinearity theoretically as well as

statistically.

The first consideration examined whether there was an inter-correlation between

breathing effort and anxiety. In terms of statistical approach, the parameter of tolerance

and variation-inflation factor (VIF) could be used to assess the potential presence of this

multicollinearity. These two parameters were, in fact, the same artificial number because

VIF is the reciprocal of tolerance. Tolerance only examined the predictors and was

defined as the proportion of variance in one independent variable that was not accounted
77

for by other independent variables. The parameter of tolerance was the number between 0

and 1. Thus, if the tolerance was close to 0, the inter-correlation would be high,

indicating multicollinearity. Although research reports stated that a correlation between

anxiety and breathing effort should exist; no empirical evidence supported a high

correlation. Thus, the nature of anxiety and breathing effort could correlate with each

other to some degree, but they were not measured using the same latent construct.

Therefore, the risk of having this type of multicollinearity was relatively small.

The second consideration of multicollinearity was specific to the MMR model.

Multicollinearity would indicate the presence of high correlations: the correlation

between the independent variable and the interaction effect, the correlation between the

moderator variable and the interaction effect. This multicollinearity would increase the

difficulty of detecting the interaction effect (Aiken & West, 1991).

In this study, detecting this type of multicollinearity relied on examining the

correlation coefficiency of the breathing effort and the interaction effect, and the

correlation coefficiency of the anxiety and the interaction effect. If both correlation

coefficiencies were low, no multicollinearity would exist. In this situation, a statistical

manipulation, centering the independent variable and the moderator variable, could be

used before performing the regression analysis. After centering the predicators, the

regression coefficiency of the independent variable and the moderator variable, which

indicate the simple main effect, would then remain the same. However, it would change
78

the regression coefficiency of the interaction effect and reduce the difficulty of detecting

the interaction effect (Aiken & West, 1991; Tabachick & Fidel, 2001). Thus, the

procedure of centering predictors could reduce the difficulty of detecting the interaction

effect (Neter, Wasserman, & Kutner, 1989).

Methods of Protecting Human Subjects

When consenting for participants, the investigator provided each potential subject

with an explanation regarding the purpose of this study and its requirements. Potential

subjects also were informed that there was no risk or benefit anticipated from this study.

However, the physiological measurement could cause minor discomfort in some cases

and resting could ease this discomfort. Potential subjects also were informed that their

participation in this study was voluntary and they could drop out from this study anytime.

In addition, their participation or attrition would not affect their medical relationship with

CGMH.

All information was entered by study identification number. Personal identification,

name and medical number, which matched with the study ID number, was stored

separately in a different place from the data collection sheet. All the information was

stored in locked file cabinets and only the investigator had access. All data were reported

as group data. No individual subject could be identified in any reporting of the data from

this study.
79

Summary

This non-random cross-sectional correlation study was conducted at the cancer

center of a medical center in Northern Taiwan. All required documents, including the IRB

application, were reviewed and approved. Power analysis with α of 0.05, moderate effect

size of 0.15 on regression analysis, and power of 0.8 was employed to estimate sample

size as 55. Patients aged 21 or older, with primary or metastastic lung cancer diagnosis,

with self-reported dyspnea sensation between 40-80 on 0-100 VVAS, who were

cognitively intact, and had the ability to understand and speak Chinese or Taiwanese were

eligible for participation in this study. Several validated survey questionnaires along with

hand-held device for measuring physiological respiratory indicators were used to collect

data. All the potential participants were referred from clinicians; and two independent

face-to-face interviews were employed to collect data by the investigator.

Simple linear regression analysis was employed to examine whether breathing

effort explained the sensation of dyspnea. The Pearson Correlation was employed to

examine the association between anxiety and dyspnea sensation. Moderated multiple

regression (MMR) analysis was used to examine the extent to which anxiety influenced

the relationship between breathing effort and the severity of dyspnea sensation.
80

Chapter Four

Results

This chapter summarizes results from the linear and hierarchical moderator

regression analyses. All participants had interviews at two time points separated by at

least one day. This chapter first introduces the sample’s demographic characteristics.

Then, the validation of instruments and the assessments of pulmonary function status are

presented. Third part of this chapter presents descriptions of the variables used in all

analyses. Finally, the three research questions proposed in chapter one are addressed by

statistical analysis.

Sample

Figure 4-1 depicts a flow chart of data collection procedures, which includes the

screening of potential participants and consenting subjects. Between January 12 and

March 16 2007, a total of 473 patients with various cancer diagnoses from three oncology

clinics, one pulmonary medicine clinic, and four in-patient units were screened for study

eligibility at the cancer center of a teaching hospital in Northern Taiwan. Eighty-seven of

these potential subjects were diagnosed with either primary or metastastic lung cancer. A

separate list was used to record all 87 patients to avoid the possibility of screening and

recruiting the same potential participant more than once.

Out of these 87 potential participants, 51 (58.9%) did not meet the inclusion

criteria of self-rating of dyspnea, leaving 36 potential participants (41.1%) eligible for


81

this study. Of these 36, 4 did not keep their clinic appointments. Thirty-two eligible

potential participants were then referred to the investigator by physician clinicians.

Twenty-eight (32.1%) were successfully consented and enrolled into this study and four

other potential subjects refused to participant in this study.

Figure 4-1.

Flow chart of data collection.

N= 473 patients with various cancer diagnosis

N = 87 patients with primary or metastatic lung cancer

N= 51
patients did
not meet the
N= 36 potential participants
inclusion
(With self report of dyspnea and
criteria
cognitively intact)

N= 28 consented N= 4 were
N= 4 and enrolled missed
refused
82

All 28 subjects stayed in the study and were followed at 2 time points, ranging

from 3 days to 7 days apart. Among those 28 participants enrolled into this study, only 20

participants were able to complete the physiological measure and survey questions while

8 participants (1 male and 7 females) completed the survey questions only.

For those subjects who could not complete physiological measures, 3 subjects had

severe muscle wasting, with body mass indices (BMI) of 13.7, 14.6, and 20. One subject

had treatment-related abdominal pain, and one subject stopped participating as a result of

fear that the dyspnea would become worse. The remaining three subjects had a variety of

personal reasons for not completing all parts of the study.

Data were analyzed from the final sample of 28 subjects, 20 of whom completed

all parts of the study measures and 8 of whom completed only the survey questionnaires.

Table 4-1 presents the demographic characteristics of the subjects. The mean age of

subjects in the study was 64.61 (SD 13.55) with a range of 41 to 90 years. The majority

were female (60.7%), and unemployed (50.0% not working and 35.7% retired). Most of

the subjects (n=22, 78.6%) had less than a high school education. Most had a diagnosis of

primary lung cancer (82.1%), with advanced stage (39.3% for stage three and 57.1% for

stage four), and no history of COPD (89.3%). The mean hemoglobin level was 11.45 (SD

2.02) with range of 8.1 to 15.5 mg/dl. Mean body height (in centimeters) was 163.9 (SD

3.78) for the men and 155.6(SD 5.6) for the women. All subjects in this study were found

to be cognitively intact as assessed by the Katzman’s scale. The mean score of


83

Katazman’s scale is 0.61 and the standard deviation is 0.96.

Table 4-1.
Demographic Summaries
(N=28)

Variable Mean(range) Median SD


64.61 68.5 13.55
Age (41-90)
Hemoglobin 11.36(8.1-15.5) 11.2 2.02
Katzaman 0.61(0-3) 0 0.96
Body Height (in cm)
-Male 163.9 164 3.78
(160-172)
-Female 155.6 153 5.60
(145-165)

Table 4-2.
Demographic Summaries
(N=28)

Frequency Percentage (%)


Gender
-Female 17 60.7
-Male 11 39.3

Lung cancer
-Primary 23 82.1
-Metastasis 5 17.9

Stage
- Stage one 1 3.6
-Stage three 11 39.3
-Stage four 16 57.1

With History of COPD


-Yes 3 10.7
-No 25 89.3
84

Table 4-2 (continued).


Demographic Summaries
(N=28)

Frequency Percentage (%)


Employment
-working 4 14.3
- retired 10 35.7
- not employed 14 50.0

Education
-No education / home 8 28.6
school*
-Elementary School 11 39.3
(6 years)
-Middle school (9 3 10.7
years)
-High School (12 1 3.6
Years)
-Associate Degree 3 10.7
-Bachelor Degree 1 3.6
-Graduate school 1 3.6

Assessment of the pulmonary function status

The objective indicator of breathing effort for this study was respiratory pressure,

including both inspiratory and expiratory pressures, measured by a portable, handheld

device. The assurance of no change in pulmonary function status during the measurement

of respiratory pressure was based on each subject’s pulmonary function of FEV1 (forced

expiratory volume in 1 second), FVC (forced vital capacity), and shallow breathing

counts. All subjects were asked to take the first test of FEV1 and FVC using the

hand-held device at the first interview and take the second test of FEV1 and FVC in
85

addition to the respiratory pressure at the second interview. These two interviews were

from 3 days to 7 days apart. All raw scores of FEV1 and FVC from both time points

were transformed to the percent of predicted value based on each participant’s age,

gender, and body height (in centimeter). The calculation was made using a web-site

calculator (www.dmac.adelaide.edu/au/copd/index.html).
H H

Three separate repeated measure analysis of variance (RMANOVA) were carried

out to examine whether the pulmonary function status (FEV1%, FVC%, and shallow

breathing count) changed on the second interview compared with the first interview. The

insignificant result of the Muchly’s test indicated that the assumption of compound

symmetry was met in each of these three analyses. The F-test on each three separate

results yielded no difference between baseline and the second examining day for FEV1%,

FVC%, and shallow breathing, with p > 0.05 for all three measures (table 4-4).
86

Table 4-3
Summaries of Differences on Pulmonary Function Status between
Baseline and Examine Day. (N=20 ~28)

Variable Baseline Examine day Repeated


(n=20~28) (n=20~28) ANOVA (p)

Mean(SD) Mean(SD)

FEV (%) 56.71(23.66) 58.55 1.15(0.3)


(19.78)
FVC (%) 55.63 (21.32) 56.15 (20.8) 0.41(0.53)
SB* 25.64 (3.4) 26.46(3.75) 3.75 (0.06)

Note: SB*= shallow breathing; all p values of the F-test were insignificant.

Validation of the anxiety subscale of cancer dyspnea scale

Anxiety was defined in relation to cancer; thus, the Anxiety subscale of the

Cancer Dyspnea Scale (CDS-anxiety) was used to assess subjects’ anxiety while

experiencing dyspnea. The short form State Anxiety Inventory (STAI) was completed

by all subjects to validate the CDS-anxiety. The scoring of STAI and CDS-anxiety both

indicated that higher score means more anxious. Out of a possible scale range of 0-16,

the mean score of CDS-anxiety was 4.0 (SD 3.97), which indicated an overall low

anxious status. The mean score of SATI was 13.75 (SD 3.76) with a possible scale

range of 6-30 (table 4-4). Pearson correlation was calculated to examine the

relationship between these two scales and the results indicated a significant moderate

correlation, r=0.49, p ≤ 0.05.


87

Table 4-4.

Description Summaries on Two Measures of Anxiety.

Variable Mean(range) Median SD

CDS-anxiety 4.00 ( 0 -16) 4.0 3.97


(with scale range of 0-16)

Short form state anxiety


inventory 13.75 (8 - 21) 13.5 3.76
(with scale range of 6-30)

Variable Descriptions

Table 4-5 presents description summaries of each variable proposed in chapter one.

The dependent variable, dyspnea sensation defined as a person’s subjective rating or

rating of the intensity of dyspnea, was measured on 0-100 vertical visual analogous scale

(VAS). The mean dyspnea sensation of all participants, was 51.52 (SD 14.31) with a

range of 40 to 80m with higher scores indicating more severe dyspnea.

The independent variable, breathing effort, was examined by objective measures

(inspiratory pressure and expiratory pressure) as well as a subjective self-rating of

breathing effort on a 0-100 vertical VAS. The mean inspiratory pressure was 49.6 (SD

14.31) and mean expiratory pressure was of 41.42 (SD 22.54).

The moderator variable was anxiety defined in relation to cancer. Thus, the Anxiety

subscale of the Cancer Dyspnea Scale (CDS-anxiety) was applied. Subjects’ mean score

of CDS-anxiety was 4 (SD 3.97).


88

Table 4-5.

Description Summaries of Variables

Variable Mean(range) Median SD

Dependent variable:
Dyspnea Sensation 51.52(40-80) 60.0 14.31
(0-100 VAS)

Independent variable:
Inspiratory Pressure 49.60(18-108) 49.0 21.27
Expiratory Pressure 41.20 (16-147) 35.0 22.54
Self rating breathing 34.61 ( 0 - 80) 35.0 22.54
effort (on 0-100 VAS)

Moderator variable:
Anxiety (CDS-anxiety ) 4.00 ( 0 4.0 3.97
(with scale range of 0-16) -16)

Demographic Influences

Age was one variable to be examined for its influence on dyspnea sensation, self-

rating of breathing effort, and anxiety. Pearson Correlation analysis indicated that

dyspnea sensation, or the subjective perception of breathing effort, and anxiety were not

correlated with age, r = 0.12, 0.12, and 0.21 respectively, all p > 0.05 (table 4-6). Thus,

age did not function as covariate variable in this study.


89

Table 4-6.
The Relationship between Age and Dyspnea sensation, Breathing Effort,
and Anxiety.

Dyspnea sensation Breathing effort CDS-anxiety


Age -0.12 -0.12 -0.21

Note: all p values were insignificant.

In addition to age, gender was another demographic characterestics examined for

its possible influence on dyspnea sensation, self-rating of breathing effort, and anxiety.

Three separate Analysis of Variance (ANOVA) tests were performed to examine

differences of dyspnea sensation, breathing effort, and anxiety between males and

females. These three analyses each met the assumption of homogeneity with insignificant

results in Levine’s test. No significance difference in dyspnea sensation, breathing effort,

or anxiety was noted between genders. However, men had higher scores on both dyspnea

sensation and breathing effort than did women, while women showed higher anxiety

scores than did the men.

Table 4-7.
Summaries of Differences on Dyspnea Sensation, Self Rating of
Breathing Effort, and Anxiety between Male and Female.
N=28

Female (SD) Male (SD)


ANOVA (p)
n = 17 n = 11

Dyspnea 57.38 (14.02) 62.82(14.79) 0.96(0.34)


sensation
Self rating of 31.18 (21.06) 39.91(24.74) 1.00(0.33)
breathing effort
CDS-anxiety 4.53 (4.12) 3.45(3.64) 0.50 (0.49)
90

Correlation Evaluation

The regression analysis is the statistical method based on correlation among

variables. Pearson correlation was used to examine the relationship among dyspnea

sensation, respiratory function status, including inspiratory pressure, expiratory pressure,

FEV1%, and FVC%, self-rated breathing effort, and anxiety. Self-rating of breathing

effort and anxiety significantly correlated with dyspnea sensation, with r = 0.64, p < 0.01.

Breathing effort also had a significant correlation with anxiety, r = 0.62, p < 0.01(table

4-8).
91

Table 4-8.
Correlations among Inspiratory pressure, Expiratory Pressure, FEV1%, FVC%
Dyspnea sensation, Breathing effort and Anxiety (N= 28).

1 2 3 4 5 6 7

1. Inspiratory
Pressure 1.00

2. Expiratory 0.85** 1.00


Pressure

3. FEV1 (%) 0.30 0.18 1.00

4. FVC (%) 0.37 0.19 0.90** 1.00

5. Dyspnea 0.05 -0.001 -0.17 -0.05 1.00


sensation
6. Self rating -0.13 0.01 -0.21 -0.13 0.64** 1.00
of breathing
effort

7. -0.16 -0.01 -0.42 -0.30 0.64** 0.62** 1.00


CDS-anxiety

Note. **p < 0.01(2-tailed).

Research Questions

The three research questions proposed in chapter one was answered by following

inferential statistical approached.

(1) Does breathing effort significantly predict individuals’ dyspnea sensation?

The variable, breathing effort, was measured by physiological indicators

(inspiratory pressure and expiratory pressure) and subjective ratings of breathing

effort. A high correlation between inspiratory pressure and expiratory pressure was
92

found, using Pearson correlation analysis, with r= 0.85, P < 0.01. To avoid

multicollinearity between these two variables in the regression analysis, inspiratory

pressure was selected to be one predictor. The reason expiratory pressure was dropped

as a predictor was that inspiratory pressure had a higher correlation coefficient (r =

0.05) than expiratory pressure (r = 0.001). The subjective rating of breathing effort

was positively correlated with dyspnea sensation, with r = 0.64, p < 0.01. Thus,

another predictor for this regression analysis was the subjective rating of breathing

effort.

Both inspiratory pressure and the subjective rating of breathing effort were

entered in the same box to regress the dependent variable, dyspnea sensation.

Regression diagnosis, including multicollinearity and residual assumptions were

carried out. As explaining in chapter three, tolerance and VIF were the two

parameters which examined whether or not multicollinearity exists. In this regression

analysis, the value of tolerance of 0.98 and VIF of 1.02 indicated no multicollinearity

in this regression model.

Four residual assumptions (mean of zero, independence, normal distribution,

and homoskedasticity) were examined in addition to multicollinearity. Results

indicated no violation of the any of the residual assumptions. This regression used a

constant; thus, the assumption of mean of zero was not violated. The Durbin-Watson

value of 2.4 supported no violation of independence. The P-P plot of residual was
93

close to a straight line, indicating normal distribution of residuals. The scatter plot of

standard residuals against the standard dependent variable was random; thus, the

assumption of homoscadasticty was not violated.

To answer question 1, the regression model generated a total variance of 23.0%

of the dyspnea explained by inspiratory pressure and subjective rating of breathing

effort, adj. R2 = 0.23, P <0.05. Only subjective rating of breathing effort was a

significant predictor of dyspnea sensation, β = 0.56, 95% CI = 0.10~0.73, p= 0.01

(table 4-9).

Table 4-9

Summary of Multiple Regression Analysis for Breathing Effort Predicting


Dyspnea Sensation (N=20).
Variable B SE B Β
(95%CI)

Inspiratory pressure 0.08 0.14 0.12


(-0.21~0.37)
Self rating breathing effort 0.42 0.15 0.56**
(0.10~0.73)

Note. adj. R 2 = 0.23 (p< 0.05). ** p ≤ 0.01

(2) Is anxiety positively related to individuals’ rating their intensity of dyspnea sensation?

Pearson correlation analysis was used to answer this question. The intensity of

dyspnea sensation was measured on a 0-100 Vertical Visual Analogous Scale. Anxiety

was indicated by the subscale of cancer dyspnea scale (CDS). The result of the Pearson

correlation (figure 4-2) verified a significant correlation between dyspnea sensation and
94

anxiety, in a positive direction, with r = 0.64, p < 0.01. This means that increasing anxiety

was associated with individuals’ increasing their rating of dyspnea sensation.

Figure 4-2.

Scatter Plot of Anxiety against Breathing Effort.


90

80

70
dyspnea sensation (0-100)

60

50

40

30
-10 0 10 20

CDS_ANX

(3) Is anxiety a moderator influencing the relationship between respiratory work effort

and individuals’ perceived intensity of dyspnea, in a positive direction?

Moderation Regression analysis was used to answer this question. Independent

variables were entered in hierarchical order to examine both the main effect and the

interaction effect. Based on the statistical conclusion from research question one, the

subjective dimension, self-rating of breathing effort, had a stronger and positive influence

on dyspnea sensation than did the objective dimension of breathing effort, inspiratory

pressure. Self-rating of breathing effort, then, was selected as one main effect entered in

the first block as step one. Another predictor entered in step one was anxiety measured

with CDS-anxiety.
95

The second step was to enter the interaction effect between breathing effort and

anxiety. The interaction term was generated by multiplying these two main effects, as

breathing effort * anxiety. However, the Pearson correlation analysis demonstrated a

significantly high correlation between the created interaction term and breathing effort,

and anxiety, with r = 0.76 and 0.92 respectively, and all p < 0.01. The high correlation

was expected to lead to multicollinearity which could then increase the difficulty of

assessing the interaction effect. The statistical technique of centering predictors was used

to avoid multicollinearity in this regression analysis. Mean scores of both breathing effort

and anxiety were subtracted from raw score; that is to say, breathing effort (centered) =

breathing effort – mean of breathing effort, and anxiety (centered) = anxiety – mean of

anxiety, were generated. Then, these two new variable were used to create a new

interaction term = breathing effort (centered) * anxiety (centered).

To summarize, the raw scores of breathing effort and anxiety were entered into the

first block and the interaction term after centering the predictors was entered into the

second block. Regression diagnosis proceeded, which included assessing

multicollinearity and residual assumptions. The value of tolerance of 0.61, VIF of 1.63

for step one and tolerance of 0.44 t0 0.61 and VIF 0f 1.6-2.3, indicated no

multicollinearity. Four residual assumptions (mean of zero, independence, normal

distribution, and homoskedasticity) were examined in addition to multicollinearity.

Results indicated no violation of the any of the residual assumptions. This regression
96

used a constant; thus, the assumption of mean of zero was not violated. The

Durbin-Watson value of 2.2 supported the result of no violation of independence. The P-P

plot of residual was close to a straight line indicating normal distribution of residuals. The

scatter plot of standard residuals against the standard dependent variable was random;

thus, the assumption of homoscadasticty was not violated.

The results were significant for both step one and step two regression analyses.

In step one, a total of 47% of the variance of dyspnea sensation was explained by the

predictors, adj. R2 = 0.47, p= 0.000. Both self-rating of breathing effort and anxiety were

significant predictors, β = 0.40, 0.39 respectively, all p < 0.05. The strengths of these two

predictors were closed to each other. This analysis demonstrated strong and significant

main effects from both self-rating of breathing effort and anxiety on dyspnea sensation.

Analysis of the second step examined whether anxiety was a moderator that

could influence the relationship between breathing effort and dyspnea sensation. The

final model from the regression analysis generated a significant model although no

interaction could be detected. A total of 45% of the variance in dyspnea sensation was

significantly explained by self-rating of breathing effort, anxiety, and the interaction

between breathing effort and anxiety, adj. R2 = 0.45, p= 0.001. However, the R square

change from step one to step two was undetectable (ΔR2 = 0.000, p =0.98). In the final

analysis, only breathing effort was a significant predicator of dyspnea sensation, β = 0.40,

p < 0.05. Therefore, in answer to research question three, anxiety was not supported as a
97

moderator influencing the relationship between breathing effort and dyspnea sensation,

although it can be considered a predictor for explaining dyspnea sensation.

Table 4-10.
Summary of Hierarchical Moderator Regression Analysis for
Variables Predicting Dyspnea Sensation
Dependent variable = Dyspnea Sensation, (N=28).

Variable B SE B Β
(95%CI)

Step one
Self rating breathing effort 0.25 0.11 0.40*
(0.02~0.49)
Anxiety (CDS-anxiety) 1.44 0.66 0.39*
(0.09~2.79)

Step two
Self rating breathing effort 0.25 0.12 0.40*
(0.12~0.49)
Anxiety (CDS-anxiety) 1.45 0.79 0.40
(-0.19~3.10)
Interaction of self rating -0.001 0.02 -0.04
breathing effort and (-0.050.05)
anxiety (after centering
predictor)
Note. adj. R2 = 0.47 (p = 0.000) for step 1; Adj. R2 = 0.45 (p = 0.001) for step 2;
ΔR2 = 0.000(isp).
*p < 0.05.
98

Summary

Dyspnea sensation, self-rating of breathing effort, and anxiety were not

influenced by age and gender. In patients with lung caner, dyspnea sensation was

significantly correlated with both self-rating of breathing effort and anxiety, both r = 0.64,

p < 0.01. Breathing effort and anxiety also had a significant correlation, r = 0.62, p < 0.01.

The regression analyses generated results indicating that both self-rating of breathing

effort and anxiety were equally significant in explaining dyspnea sensation (Adj. R2 = 0.

47, p=0.000). Anxiety, however, did not function as moderator in the relationship

between breathing effort and dyspnea sensation in this sample population.


99

Chapter Five

Discussion

This chapter discusses the statistical description of each variable and the statistical

conclusions answering the research questions. The variables of interest with regard to the

role of anxiety in the relationship between breathing effort and cancer-related dyspnea

focus on the congruence of the descriptive results compared with the conceptual and

operational definitions. Major findings from inferential statistical analyses integrate not

only the similarities but also the contradictions, with the results from previous research.

This integration then inductively allows construction of an analytical model of the results

of this completed study. Derivations from the analytical model added to previous

empirical works facilitate the construction of a new conceptual framework for explaining

cancer-related dyspnea sensation. Recommendations for clinical practice and nursing

education are addressed accordingly. In addition, this section addresses the issues of

subject burden and the difficulties of employing physiological measures in a clinical

setting.

Summary of the Present Study

A physiological model of cancer-related dyspnea guided this cross-sectional

correlational study. Two hypotheses were derived from the study model: (1) increasing

respiratory muscle work effort is positively related to individuals’ reporting of dyspnea

sensation; and (2) anxiety influences the relationship between respiratory muscle effort
100

and individuals’ perceived severity or intensity of dyspnea. This study then sought to

examine: the extent to which both subjective and objective aspects of breathing effort

explained dyspnea; and whether anxiety moderated the relationship between breathing

effort and dyspnea sensation in cancer patients.

To summarize the results from the statistical analyses, the influences of

demographic characteristics, age, and gender, on dyspnea sensation, self-awareness of

breathing effort, and anxiety, were insignificance. In patients with lung caner, dyspnea

sensation had significant moderate correlations with self-awareness of breathing effort

and with anxiety. Self-awareness of breathing effort was significantly and moderately

associated with anxiety. The hierarchical multiple moderate regression analysis indicated

that both self-awareness of breathing effort and anxiety were equally significant in

explaining 47% of the total variance of dyspnea sensation. Anxiety, however, did not

function as moderator in the relationship between breathing effort and dyspnea sensation

in this sample population.

Sample Description

Based on the most recent reports from the Bureau of Health Promotion,

Department of Health in Taiwan (2003), 7,415 persons were reported to have a lung

cancer diagnosis. These statistics were obtained from various health care facilities, such

as hospitals and cancer treatment centers. The incidence of lung cancer in the male

population was approximately twice the incidence in female population. The median age
101

of this larger Taiwanese patient population was 71 years for men and 68 years for women.

The major lung cancer type (88% of males and 94% of females) for the population was

non-small cell lung cancer (NSCLC) (http://crs.cph.ntu.edu.tw/crs_c/annual.html). In this


H H

present study, the majority of subjects were female (60.7%) with a median age of 68.5

years. Eighty-two percent of subjects had a diagnosis of NSCLC. Although the

demographic characteristics of this sample population differed somewhat from the Parent

population with regard to gender, the characteristics for age and lung cancer type were

very similar. Therefore, this sample was comparable to the population in Taiwan.

Variables of Interests

As introduced in chapter one, three variables of interest were addressed in this

study: the severity of dyspnea sensation (dependent variable), breathing effort

(independent variable), and anxiety (moderator). The following discussion focuses on

whether the subjects’ descriptions of these three variables of interest were consistent with

the conceptual and operational definitions used by the researcher to guide this study.

Dependent Variable

The dependent variable was the severity of dyspnea sensation and was defined for

the purposes of this study as the cognitive response and the perceptual sensitivity of an

individual when rating his or her severity of dyspnea. In this study, most participants

rated their dyspnea sensation as moderate on a 0-100 mm visual analogue scale (VAS).

Previous studies indicated that the personal variables of gender and age could influence
102

self-report of the intensity of dyspnea sensation, with younger patients and women

reporting greater intensities of dyspnea sensation. The results of the current study did not

demonstrate any influence of either gender or age on dyspnea sensation.

In this study, male participants reported a somewhat higher mean score (62.82) for

dyspnea sensation than female participants (57.38), although this difference did not reach

statistical significance. This result was contradictory to some other studies in which

women reported greater intensity of dyspnea sensation than men (Brand, Rijcken,

Schouten, Koeter, Weiss, &Postma, 1992). The findings of this study, however, were

consistent with those of a large descriptive study by Smith et al (2001), in which men

with lung cancer reported more severe dyspnea than did women with lung cancer. These

findings were also comparable to those of Adams, Chormos, Lane, and Guz (1986), in

which no gender differences on dyspnea sensation were observed.

The small sample size of the completed study contributed to the inconclusive result

of whether or not gender influences the intensity of dyspnea sensation. Increasing the

sample size in future studies could be beneficial in determining whether or not a true

gender difference exists for the intensity of dyspnea sensation among patients with lung

cancer.

In addition to the small sample size, restriction of the inclusion criteria (enrolling

only subjects with moderate to severe dyspnea) may also have influenced this finding.

Fifty-one of the 87 available patients with lung cancer were excluded from this study due
103

to either the absence of dyspnea or the presence of only mild dyspnea sensation (figure

4-1 in Chapter four). The majority of those who did not meet the inclusion criteria for

presence of moderate to severe dyspnea were men.

Previous studies suggested age was a possible factor influencing self-rating of the

intensity of dyspnea sensation. For example, Brad and colleagues (1992) reported that

greater intensity of dyspnea sensation was associated with younger ages. In contrast, the

current study determined that dyspnea sensation was not associated with age. However,

the subjects comprising this study’s population were drawn from a narrow age pool. Lung

cancer incidence was more common among older adults, and the majority of subjects in

this study were older than 65 years of age. Although the narrow age range led to a skewed

distribution from normal when considering the entire population of Taiwan, it was

consistent with the population of persons with lung cancer. In Taiwan, as in the rest of the

world, most patients with a lung cancer diagnosis were older than 65 years of age. Thus,

the determination of the influence of age on intensity of dyspnea sensation among

patients with lung cancer would not be likely to be clarified in future studies regardless of

sample size. Future studies examining the influence of age on dyspnea sensation would

require the inclusion of subjects whose dyspnea was caused by conditions that occur

across a wider age spectrum.


104

Independent Variable

Breathing effort was conceptualized as an increase in muscle workload and the

individuals’ awareness of the need to use extra effort to breathe. However, results from

this study verified only that the subjective awareness was valid in assessing breathing

effort in patients with lung cancer. The findings from this study determined that

subjective awareness of breathing exertion, as measured using a 0-100 mm VAS, was

significantly and moderately associated with the severity of dyspnea sensation. This

finding was comparable to a study of 52 patients with COPD (Carrier-kohlman, Gormley,

Douglas, Paul, & Stulbwr, 1996). In addition, using a subjective rating for breathing

exertion to assess breathing effort was consistent with the empirical work form Rubin and

Pain (1976). Rubin and Pain determined that measuring the sensations of airway

obstruction and airflow restriction with a perceptual estimate was preferred to more

objective measures of lung function impairment.

The objective measures used to assess breathing effort (described in chapter three)

were the static maximum respiratory inspiratory and expiratory pressures as evaluated

with a non-invasive, hand-held portable respiratory meter measuring airflow through the

mouth. Because the inspiratory pressures and expiratory pressures were generated as a

result of respiratory muscle contractions, this mouth pressure measure could be

extrapolated to represent respiratory skeletal muscle strength, with narrow pressure

changes indicative of muscle weakness.


105

Although this objective measure of static maximum inspiratory pressure and

expiratory pressure had not been employed in any of the previous studies of dyspnea, this

present study measured respiratory pressures in 10 male and 10 female subjects. Thus, a

hypothetic parallel of normal adults’ maximum static mouth pressure was first calculated

(Wilson, Cooke, Edwards, & Spiro, 1984) and then compared to the sample population in

this study. This comparison, shown in Table 5-1, indicated that respiratory pressures in

the sample population were lower than the calculated hypothetic static pressures.

Table 5-1.

Respiratory Pressures in this Sample Population versus


Hypothetical Parallel Normal Adult

Sample Population (SD) Normal adult maximum


static mouth pressure (SD)

IP* 49.60 (21.27) 68.44 (9.03)


EP** 41.20 (22.54) 99.78 (15.62)

Note: IP*: inspiratory; EP**: expiratory pressure

The Pearson correlation analysis, however, did not support that either maximum

inspiratory pressure or maximum expiratory pressure was associated with self-awareness

of breathing effort or dyspnea sensation. Thus, there was only weak, at best, empirical

evidence for supporting the hypothesis explaining that cancer-related dyspnea sensation

results from decreases in respiratory pressures corresponding to lung volume changes,

which then stimulate an individual’s awareness of breathing exertion (Fishbein, Kearon,

& Killian, 1989). This result also was contradictory to other physiologist’s empirical
106

evidence, which indicated that the dyspnea sensation was associated with pulmonary

impairment in healthy persons and persons with pulmonary dysfunction, including COPD

and asthma (Killian, Gandevia, Summers, & Campbell, 1984; Jones, Killian, Summers, &

Jones, 1985; Yan, 1999, Chen & Yan, 1999).

In summary, in patients with lung cancer, breathing exertion was associated with an

individual’s subjective awareness, but not associated with the decrement in respiratory

pressures that could indicate a worse pulmonary performance. Thus, the measurement of

breathing exertion was more a factor of subjective awareness of breathing effort than

objective physiological changes. It is possible, howeve, that the physiologic measure

employed in this study was not sensitive enough or accurate enough to capture

respiratory muscle-induced changes in static pulmonary pressures.

Anxiety as a Moderator

The moderator in this study was conceptualized as anxiety, a psychological

reaction in response to dyspnea episodes in persons with cancer. The diagnosis of cancer

was viewed as a chronic illness and a potentially life-threatening situation. Both

perceptions were considered to be continuous stressors to patients who had such a

diagnosis. Thus, this study employed two survey questionnaires, the anxiety subscale of

the Cancer Dyspnea Scale (CDS), and the short form of the state anxiety inventory

(STAI), to measure the variable of anxiety. The anxiety subscale of CDS was used to

assess for anxiety specific to the dyspnea experience, and the short form STI was used to
107

assess for general anxiety.

Results verified that the anxiety subscale of CDS correlated to the intensity of

dyspnea sensation and self-awareness of breathing effort. In addition, the correlation

between the anxiety subscale of CDS and short form STI was 0.49, p < 0.01. Thus, the

anxiety subscale of CDS was valid in assessing for anxiety within the dyspnea experience

and also general anxiety.

This present study also examined whether age influenced anxiety, and whether

anxiety was different between the two genders. Results revealed that anxiety was not

influenced by age in this sample population. However, whether or not gender influenced

anxiety in patients with lung cancer was not clear. The analysis only revealed that female

subjects reported higher anxiety related to dyspnea (5.36) than did male subjects (3.29).

The small sample size was a barrier in determining whether the difference reached

statistical significance.

Nevertheless, field notes from data collection indicated that worries and concerns

from the experience of dyspnea differed between male and female subjects. On the one

hand, most female subjects worried about the care family members would receive now

that the subjects were unable to participate in their usual care activities. Male participants,

on the other hand, tended to be concerned about who would provide care for them (the

subjects) during their illness. In Taiwan, female roles regarding family responsibilities

include raising children, doing chores, and taking care of elder/ill family members. Thus
108

the female family roles were mother, daughter, and daughter-in-law. Male roles, such as

father, son, and son-in-law, share only a small amount of those traditional female family

responsibilities. The field notes were consistent with anecdotal observations from clinical

practice and conversations with female patients. These anecdotal observations and

conversations revealed that the disease diagnosis and symptoms caused female patients

with cancer to worry about changes in their roles, being mother, wife, and

daughter-in-law. The level of worry corresponded to their illness severity.

No research conducted in Taiwan has explored gender difference in anxiety in

patients diagnosed with cancer. However, two European studies regarding cancer

patients’ emotional status found that women had more distress and more anxiety than

men. One study was conducted in Italy among 60 patients with genitourinary cancer Italy

(Rispoli, Pavone , Bongini , Di Loro , Ponchietti , & Rizzo, 2005), and the other study

was conducted in the United Kingdom among 426 patients with melanoma

(Newton-Bishop, Nolan, Turner, McCabe, Boxer, Thomas, Coombes, A'Hern, & Barrett,

2004).

Finding Interpretations

Dyspnea in Patients with Various Disease Diagnoses

Previous empirical studies focused on the language used in describing the dyspnea

experience among patients with various pulmonary diagnoses, including COPD, asthma,

and cancer. The descriptions of their dyspnea experience were somewhat comparable
109

among the different disease groups (Skevington, Pilaar, Routh, & Macelod, 1997;

Corcoran, et al, 2002). However, to date, little evidence exists about the relationship

between physiological changes and dyspnea sensation among patients with various

diagnoses. Adding the evidence comparing physiological changes in relation to dyspnea

among various disease diagnoses may help to understand the phenomenon of

cancer-related dyspnea. FEV1 (the maximum amount of air that can be exhaled in the

first second of expiration) and FVC (the maximum amount of air that can be exhaled as

quickly as possible after maximum inspiration) were two physiological indicators

employed in studies with patients with COPD and asthma. In order to be able to compare

findings between studies, this present study compared the results of FEV1 and FVC to

those of another research study that evaluated the physical sensation of breathing effort

and its affective response in patients with other pulmonary diagnosis (Meek, Suzanne,

Lareau, & Hu, 2003). Table 5-2 presents the comparison of the ventilatory outputs.

Table 5-2.
Ventilatory Outputs on Patients with Various Disease Diagnoses

Lung Cancer a COPD b Asthma b Healthy b


n = 20 n = 32 N = 27 n = 33

FEV1 58.55±19.78 39.3 ±16.0 83.4±22.0 103.5±15.0


FVC 56.15±20.8 52.8 ±17.0 86.6±20.0 98.3±16.0
Note : a. Lung cancer patients were from this present study; all 20 participants
were with lung cancer diagnosis, primary or metastases.
b. Patients with COPD, Asthma, and healthy participants were from
Meek and colleagues’ research to evaluate breathing effort, including
its physical sensationand affective response.
110

As presented in table 5-2, patients with lung cancer had a worse performance on

FEV1 and FVC than did patients with asthma and healthy persons. Comparison regarding

these two ventilation outputs between patients with lung cancer and patients with COPD

was rather different. Patients with lung cancer had slightly better FVC than did patients

with COPD while the performance of FEV1 in patients with lung cancer greatly exceeded

that of patients with COPD. The probable cause of this difference is related to the basic

pathologies of the two disease processes. COPD involves air-trapping and lung

over-inflation, whereas ling cancer alone does not result in these changes (to the same

degree). Air-trapping and over-inflation greatly increase the work of expiration. In

addition, both studies measured breathing effort on a 0-100 mm VAS, but used different

instruments assessing for anxiety. Meek et al (2003) conceptualized and measured

distress as a reaction to dyspnea on a 0-100 mm VAS. The investigator of this present

study conceptualized anxiety as a psychological reaction to dyspnea as measure by the

anxiety subscale of CDS. In addition, both studies examined relationships between

breathing effort and ventilation outputs, and the relationship between breathing effort and

anxiety/distress. However, breathing effort associated with ventilator outputs was only

established in patients with COPD (r = 0.56 in FEV1% and 0.54 in FVC, both p < 0.01),

not in patients with lung cancer. Breathing effort associated with anxiety/distress was

determined in patient with lung cancer (N= 28, r = 0.62, p < 0.01) as well as in patients

with COPD (N= 30, r =0.45, p < 0.01).


111

To summarize this comparison, dyspnea was associated with severely decreased

ventilation performance in patients with COPD while the experience of dyspnea in

patients with lung cancer was different. In patients with lung cancer and patients with

COPD, dyspnea was associated with having anxiety/distress. However, the patients with

lung cancer did not have the severity of decline in ventilation performance that the

patients with COPD had. Therefore, adapting the mechanism of dyspnea in patients with

COPD cannot be applied to cancer-related dyspnea without more evidence to support and

justify.

Language of Describing Dyspnea Sensation

In spite of the similarities in description of the dyspnea sensation, culture could

have an influence on the meanings of these descriptions of dyspnea. Language chosen to

describe dyspnea by patients from the United Stated and patients from China and Taiwan

seemed to be consistent. In two studies from the United States, physical sensation,

affective and evaluative, and low energy were three categories of dyspnea description; the

descriptions of dyspnea included the feeling of being unable to get enough breath, or of

panic, or impending death (Skevington, Pilaar, Routh, & Macleod, 1997; O'Driscoll,

Corner, & Bertley, 1999). Chinese patients experiencing dyspnea described their

breathing difficulty as mixed with an uncomfortable feeling of breathing movement,

affecting activity, and symptoms rather than by using the terms dyspnea, or breathing

difficulty. The most common term used to describe dyspnea was fatigue. Other
112

descriptions, such as “singing”, “whistling while breathing”, “coughing”, “expectoration”,

“cannot walk on level or up the stair”, “irritated”, and “out of control and feel crazy”

constantly appeared from groups of patients in China (Han, Zhu, Li, Chen, Put, Van de

Woestijne, & Van de Bergh, 2005). The present study also observed that Taiwanese

patients expressed their dyspnea with various descriptions; some descriptions appeared to

link directly with dyspnea while some sounded ambiguous to the investigator. Those

ambiguous descriptions sometimes could not be clarified and thus contributed to the

difficulty of evaluating dyspnea sensation. The examples of less ambiguous descriptions

were “having shortness of breath”, “unable to catch breath” while the ambiguous

description was “had to squat down and sit on floor until symptoms became less”. In

addition to using words to describe the dyspnea experience, one subject expressed his

experience only with tears.

Dyspnea and Breathing Effort

Previous research employed either objective or subjective measures to assess

dyspnea sensation in patients with pulmonary impairments (Lush, Jason-Bjerklie, Carrieri,

& Lovejoy, 1988), or in healthy persons under experimental conditions to mimic

pulmonary function impairment that would induce breathing difficulty (Chonan,

Mullholland, Cherniack & Altose, 1987; Wilson & Jones, 1989). The present study

proposed to determine the extent to which breathing effort (measured with both objective

physiological changes and subjective awareness) explained dyspnea sensation in patients


113

with lung cancer. Results revealed that despite pulmonary impairment in patients with

lung cancer, only the individuals’ subjective awareness of breathing effort influenced

their dyspnea sensation (β = 0.56, p≤ 0.01). Thus, the intensity of the dyspnea sensation

was derived from individual’s perception of ventilatory exertion, and was not dependent

on any measured change in ventilatory activity from normal.

Anxiety and Dyspnea

The results of this study verified that anxiety was significantly associated with

dyspnea. These results were comparable to one previous study also conducted in Taiwan

(Chiu, Hu, Lue, Yao, Chen, & Makai, 2004) and another study conducted in a Chinese

population from Hong Kong (Chan, Richardson, and Richardson 2005). The findings

were consistent with those of studies conducted in United States (Dugeon, & Lertzman,

1998; Beruera, Schmitz, Pither, Neuman, & Hanson, 2000; Dugeon, Lertzman, & Askew,

2001; Bailey, 2004). Thus, the association between anxiety and dyspnea was consistent

across western and oriental cultures.

Breathing Effort, Anxiety, and Dyspnea Sensation

This present study proposed and examined a novel hypothesis, anxiety as a

moderator influencing the relationship between breathing effort and cancer dyspnea

sensation. However, results from the hierarchal multiple moderate regression analysis did

not support this hypothesis. As opposed to being a moderator, anxiety was a strong

predicator, together with breathing effort, explaining dyspnea sensation. A similar


114

conclusion, that anxiety was a significant factor influencing dyspnea sensation in cancer

patients, was verified in one study on 72 dying cancer patients from Taiwan (Chiu, et. al.,

2004) and another study on 171 patients with lung cancer from Japan (Tanaka, Akechi,

Okuyama, Nishiwaki, & Uchitomi, 2002).

Directions of the Relationships between Breathing Effort and Dyspnea, and between

Anxiety and Dyspnea

This study demonstrated a positive relationship between breathing effort and

dyspnea, as well as between anxiety and dyspnea sensation. When dyspnea was regressed

on anxiety and on breathing effort, the analysis indicated that breathing effort and anxiety

explained dyspnea sensation. The correlation coefficients among the three factors,

breathing effort, anxiety, and dyspnea, were between 0.62 and 0.64. In an attempt to

explain whether relationships were in the positive direction (i.e. dyspnea explaining

anxiety), analysis was reversed. Anxiety was regressed on dyspnea and breathing effort

regressed on dyspnea. However, due to small sample size, the regression assumption of

independence was violated in these two analyses. A larger sample might allow this

possibility to be examined in the further.

Study Model versus Analytical Model

Chapter one described the physiological conceptual framework used to guide the

exploration of two hypotheses. The two hypotheses were: (1) increasing respiratory

muscle work effort is positively related to individual’s reporting of dyspnea sensation;


115

and (2) anxiety influences the relationship between respiratory muscle effort and

individuals perceived dyspnea. Employing statistical analyses then sought to examine

these two hypotheses by answering three research questions: (1) does breathing effort

significantly predict individuals’ perceived severity of dyspnea; (2) is anxiety positively

related to individuals’ rating their severity of dyspnea sensation; and (3) is anxiety a

moderator influencing the relationship between respiratory work effort and individuals’

perceived the severity of dyspnea, in a positive direction? Figure 5-1 presents this present

study’s model

Figure 5-1.
Study Model

Anxiety c.

b .
Breathing a. The Intensity of
Effort a. Dyspnea Sensation

Note: a. regression analysis was used to examine the hypothesis one


& research question one.
b. a hierarchal multiple moderate regression analysis was used to examine
hypothesis two and research question three.
c. Pearson correlation was used to examin research question two

In spite of significant statistical conclusions, results from this present study, in fact,

supported hypothesis one, but did not support hypothesis two ( which was examined by

research question three). Research question 1 “does breathing effort significantly predict
116

individuals’ perceived severity of dyspnea” was supported in that only consciousness of

breathing effort significantly explained the intensity of dyspnea sensation. Findings from

the statistical analyses provided empirical evidence to support research question 2 “is

anxiety positively related to individuals’ rating of their severity of dyspnea sensation”,

but the findings did not support research question 3 “is anxiety a moderator influencing

the relationship between respiratory work effort and individuals’ perceived severity of

dyspnea, in a positive direction?” However, findings did support that anxiety functioned

as another predictor in explaining the intensity of dyspnea sensation. Figure 5-2 presents

the analytical model of this present study.


117

Figure 5-2.

Analytical Model of this Present Study

β * = 0.39a,b
Anxiety
Gender
Intensity of
r** = 0.62c Dyspnea
Sensation
Subjective
Awareness
of Breathing β* = 0.40a
Effort

Note: *p < 0.05; **p <0.01


a. results supported hypothesis one, yet did not support hypothesis two
(research question three)
b. anxiety function as another predictor, rather than moderator, of predication
the intensity of dyspnea sensation.
c. anxiety was associated with subjective awareness of breathing effort.

Conclusions derived from this analytical model were:

1. Subjective awareness of breathing effort was positively associated with anxiety

in relation to dyspnea.

2. Anxiety and subjective awareness of breathing effort equally explained dyspnea

sensation in patients with cancer.

Strength

This theoretically- based study sought to examine factors influencing

cancer-related dyspnea. Framework variables were conceptualized and operationalized


118

using existing theories and empirical evidence. The propositions of this framework were

properly examined with sophisticated methods of statistical analyses. Importantly, the

study generated some reliable results that reached statistical significance; these were

compared with those from previous empirical works.

The statistical validity of this study was evaluated with effect size and observed

power, which were calculated using the parameter of R2 from two regression analyses.

The first regression analysis, regressing dyspnea on objective and subjective measures of

breathing effort, had an R2 of 0.31. The effect size of this regression analysis was 0.45

with an observed power of 0.69. The second regression analysis, hierarchal multiple

regression, examined an interaction effect to clarify whether anxiety moderated the

relationship between breathing effort and cancer dyspnea sensation. Results revealed no

interaction effect, so that the final model was chosen as the first step regressed dyspnea

on breathing effort and anxiety. Results yielded a significant regression model with an R2

of 0.51. This final regression model then had an effect size of 1.03 and observed power of

0.99. Thus, this study had acceptable statistical validity.

Limitations

The generalizability of this study was limited by age, gender and sample size.

About 60% of the participants in this study were elderly females. However, among those

51 patients with lung cancer who were not eligible for this study due to absence of self

reporting moderate to severe dyspnea sensation, the majority were men. Despite a low
119

refusal rate (4 refusals out of 32 eligible potential subjects); most of those who refused to

participate in this study were men. The most common reason for refusal was, that in spite

of having observable dyspnea, these patients were able to tolerate it and thus did not

report dyspnea. From those observations, men may have a different prospective with

regard to rating the severity of their dyspnea sensation. Thus, results could only be

generalized to female patients with lung cancer who experienced dyspnea.

Although the small sample size did not jeopardize the statistical analyses used to

answer the research questions, it did limit further definition of the variables of interest. In

order to construct a comprehensive theoretical framework for explaining the phenomenon

of cancer-related dyspnea, each variable needed to be fully examined with respect to its

attributes as well as factors associated with it. While previous empirical evidence

revealed possible associations among demographic characteristics, dyspnea sensation,

and anxiety, this study’s small sample size limited verification of those associations.

The major factor resulting in the small sample size was that few patients with lung

cancer reported having moderate to severe dyspnea sensation. The physician and nurse

clinicians were enthusiastic about the study and referred many potential subjects.

However, during the screening process, a large number of potential subjects stated that

they were not experiencing dyspnea, in spite of direct observations to the contrary (e.g.

use of accessory muscles, inability to complete a sentence without stopping for breath,

shallow but rapid respiration, etc.). This observation demonstrated that some individuals
120

could have had various degrees of tolerance of the dyspnea sensation. Another possibility

could be that patients were afraid to acknowledge the presence of dyspnea because of its

association with disease progression.

Implications & Recommendations

Results of the descriptions of the variables of interest and conclusions from the

analytical model facilitated constructing a new conceptual framework for explaining

cancer-related dyspnea sensation. The following sections address this conceptual

framework. Recommendations for further research, clinical practice, and nursing

education are also addressed.

Implication for Theory

According to the analytical model of this study (figure 5-2) and adding the

evidence derived from previous empirical works, figure 5-3 presents a novel conceptual

framework for explaining cancer-related dyspnea sensation. This novel framework

consists of covariate variables, independent variables, and dependent variables. In this

conceptual framework the covariate variables are the demographic characteristics of age,

gender, and ethnicity. In addition, this conceptual framework includes three independent

variables: the language used to describe dyspnea sensation, the affective responses to

dyspnea sensation, and the consciousness of breathing effort. The definitions of the

independent variables are: (1) language used to describe dyspnea sensation: the

individual who experienced dyspnea uses his or her own words to describe this sensation.
121

The descriptions may include words about this sensation, uncomfortable feelings of

breathing movement, affected activity, and other symptoms that the individual links to

dyspnea; (2) the affective component: the psychological reaction or emotional status,

particularly anxiety, in response to the dyspnea experience; and (3) the subjective

awareness of breathing effort: individuals’ awareness of their use of extra effort to

breathe, including either expiratory movement, inspiratory movement, or both. The

dependent variable, the severity of dyspnea sensation was the individual’s self-reporting

of his or her perceived intensity of dyspnea.

The premises of this framework include: (1) dyspnea in cancer patients could

also be influenced by their state of mind when the physiological impairment occurred; (2)

how an individual reports his or her severity of dyspnea sensation is dependent on the

language, the affective component, and the awareness of breathing effort; (3) uses of

language in describing dyspnea have various meanings and are unique to individuals, and

choice of language to describe the dyspnea sensation may be different between genders

and among different cultures.

Further efforts needed to deductively examine this novel framework can begin with

examining variables, verifying the unidentified propositions and examining the premises.

Adding new empirical evidence then is expected to guide the probing of the phenomenon

of cancer-related dyspnea sensation.


122

Figure 5-3.

The Proposed Conceptual Framework of Cancer-Related Dyspnea Sensation

Language of
Describing
Demographic Dyspnea
Characteristics: Affective
Age Component:
Gender Anxiety
Ethnicity

The Severity
Self awareness of Dyspnea
of Breathing Sensation
Effort

Note: Variables in solid-lined boxes were with conceptualization and operationalization in this
present study
Variables in dot-lined boxes required further inquiry.
indicated unidentified proposition which required further inquiry.
& indicated the identified propositions which were verified from this present
study
integrating with previous evidence.

Recommendations for Research

In addition to recruiting more men and increasing the sample size, further research

to explore the phenomenon of cancer-related dyspnea sensation in patients with various


123

cancer diagnoses should also consider ethnicity. Five recommendations for further

investigations address the extensions of major findings, the clarification of the

unidentified information in this study, and examining the premises of the new conceptual

framework for cancer-related dyspnea sensation.

Recommendation 1. Based on major findings of this present study and the novel

framework of cancer-related dyspnea (figure 5-3), the intensity of dyspnea sensation was

hypothesized be related to the subjective awareness of breathing effort, the affective

component, and the language used to describe the dyspnea sensation. Thus, crafting a

multidimensional measure of dyspnea sensation and then validating it with psychometric

reports would be appropriate next steps for further research in this of interest.

Recommendation 2. Further investigation also can focus on examining the

effectiveness of interventions to manage dyspnea. This study established evidence that

anxiety and consciousness of breathing effort together significantly explained the

intensity of dyspnea sensation. This study also verified a significant association between

anxiety and consciousness of breathing effort. Thus, three targeted preliminary mediating

effects can be used in assessing the efficacy of interventions to decrease the intensity of

dyspnea sensation. The first two targeted mediating effects are single effects focused on

decreasing consciousness of breathing effort and reducing anxiety. The third targeted

measure of mediating effect in decreasing the intensity of dyspnea sensation is an

aggregated effect. The aggregated effect usually refers to the weight of the components
124

within a summary scale (Kane, 1997). In this study, the aggregation is the sum of

consciousness of breathing effort and anxiety. How to calculate this aggregated summary

scale requires further inquiry of this association conceptually as well as operationally.

Thus, preliminarily testing of these three targeted mediating effects can be proposed to

evaluate the effectiveness of interventions in managing cancer-related dyspnea sensation.

Suggestions for possible interventions with regard to decreasing anxiety and

consciousness of breathing effort are nonpharmacologic strategies including relaxation

techniques, such as music, guided image, massage, etc. In addition to relaxation

techniques, because anxiety occurs in response to cancer-related dyspnea and sometimes

varies from personal concerns, using psychosocial interventions may facilitate the

management of the individual dealing with personal concerns that affect anxiety. Thus,

the psychosocial intervention would be expected to decrease anxiety, which could then

decrease the intensity of dyspnea sensation.

Recommendation 3. This study verified that anxiety in patients with cancer is a

continuous distress from the subjects’ concerns regarding the fact that dyspnea was

related to the cancer diagnosis. In addition to clarifying whether women and men report

anxiety differently, further investigations can be proposed to elaborate the affective

component responding to dyspnea sensation. Recommendations include: (1) identifying

and examining whether any other affective components responding to dyspnea influence

the severity of dyspnea sensation; (2) determining to what extent affective components
125

may be related to the cancer diagnosis and cancer related symptoms; and (3) whether the

affective components responding to dyspnea are different between men and women with

cancer diagnosis.

Recommendation 4. This study only revealed possible relationships between

demographic characteristics (age and gender) and the severity of dyspnea sensation. The

propositions regarding the relationship between the language used to describe dyspnea

sensation with demographic characteristics was based on little empirical evidence and

field observations. In addition, no evidence exists concerning the relationship between

the language used to describe the dyspnea sensation and the severity of dyspnea sensation.

Thus, examining potential relationships among demographic characteristics, language

used to describe dyspnea sensation, and the severity of dyspnea sensation are proposed to

possibly provide a more comprehensive understanding of cancer-related dyspnea

sensation.

Recommendation 5. In addition to quantitative designs, the hermeneutic

phenomenology focus suggests examining the third premise of the proposed conceptual

framework, how the language used in describing dyspnea has various meanings unique to

the individual, and choice of language to describe dyspnea sensation may be different

between genders and among cultures.


126

Recommendation for Clinical Practice

Dyspnea is a symptom that requires full assessment and effective management

strategies in cancer care. The assessment of dyspnea sensation should include all the

aspects: subjective rating of its intensity, personal descriptions, breathing effort and

anxiety, and other possible affective components. These assessments should be

streamlined to not take too much nursing practice time nor exhaust the patient, yet be able

to provide important baseline data for evaluating the effectiveness of various dyspnea

management strategies. Although a focus of nursing care, comprehensive management of

dyspnea requires the coordinated effort of an interdisciplinary team. Nurses, as the health

care professionals most in contact with individuals experiencing dyspnea, may serve as

team coordinators in cancer-related dyspnea management.

Recommendations for Nursing Education

Integration of the results of this present study with previous empirical studies may

help to clarify the phenomenon of dyspnea sensation in patients with cancer. Taking this

knowledge to nursing education will allow students to build fundamental knowledge

regarding dyspnea in cancer care. More importantly, future nursing students are expected

to be able to distinguish dyspnea sensation in patients with cancer from the dyspnea

experience in patients with other pulmonary disease diagnoses. Integrating this evidence

of cancer-related dyspnea also can help to improve the understanding of the phenomenon

for nurses in clinical practice.


127

Subject Burden

This present study did not seem to be burdensome to participants. The average

length of time for the second face-to-face interview, which included the use of two

physiological measures and administration of the survey questionnaires, was about 28

minutes. One interview took 60 minutes to complete because the participant shared her

concerns regarding her illness and dyspnea experience with the investigator. Even with

the extended interview period, this participant neither appeared exhausted nor did she

express any negative consequences from the interview.

Most participants completed their interview without requesting rest. Rest was

requested during six interviews; five participants were emotionally overwhelmed and one

participant was physically exhausted. Nevertheless, these participants were able to

continue and complete the interviews after a brief (3 to 5 minute) break.

Eight participants, however, were unable to take the physiological measures,

mostly due to their muscle wasting, emotional exhaustion from the dyspnea experience,

and various personal reasons. The physiological evaluation used two hand-held devices

that measured ventilation outputs and respiratory pressure. These measurements all

required effort to blow air into the devices. Three participants with body mass indices of

less then 20m2 were unable to successfully blow air into the devices. One participant was

frightened that the devices would block the airway and he would be unable to breathe. In

addition, one participant stated that chemotherapy already took too much of his personal
128

“chi” so that he did not have the extra “chi” needed to blow into the device.

(Chinese/Taiwanese’s perspective regarded breathing activity is to bring “chi” inside

body which then generates energy to live.) These patients acknowledged the importance

of preserving enough energy in order to survive the lung cancer diagnosis. The energy

preservation for these patients is to store their “chi” by not using any energy for activities,

such as blowing into a tube.

Difficulty

The conceptualization and operationalization of this study were mainly based on

western culture while subjects were from Taiwan. The greatest difficulty encountered

while conducting this study was culture difference. For example, the gold standard of

measuring subjective perception, such as dyspnea sensation and breathing effort is the use

of the visual analogous scale. The concept of attempting to quantify a type of discomfort

or perception, however, is not a current part of this culture’s experience. Individuals agree

or disagree that a symptom exists and do not attempt to identify to what degree it is

present when it does exist. Thus, questions and measures used to attempt to generate

quantifying evidence may not have been understood or appreciated. Thus, in Taiwan, the

VAS measurement is not widely used clinically. Most of the subjects were not familiar

with it. In addition, most subjects would expect the investigator to assess their dyspnea

sensation by how the symptom affects their daily life, as opposed to generating a number

of unknown relevance.
129

Conclusion

This theory-derived cross-sectional study, with minimum subject burden, explored

factors associated with cancer-related dyspnea. Results supported the hypothesis that

increasing respiratory muscle work effort is positively related to the individual’s reporting

of dyspnea sensation. The hypothesis that anxiety influences the relationship between

respiratory muscle effort and individuals’ perceived dyspnea was not support by the

statistical conclusion. Nevertheless, results revealed that anxiety is another factor

explaining the dyspnea sensation. Although this present study explored cancer-related

dyspnea in Taiwan, most major findings were consistent with previous empirical

evidence that was conducted worldwide. Integrating major findings with previous

empirical evidence facilitated constructing a new conceptual framework for explaining

cancer-related dyspnea sensation. Further recommendations to explore the phenomenon

of cancer-related dyspnea sensation include: developing and validating multidimensional

instruments measuring dyspnea sensation; developing effective interventions to manage

dyspnea in cancer care; examining the affective components responding to dyspnea; and

testing the premises of the novel conceptual framework. Evaluating the efficacy of

intervention can be based on preliminary testing of three targeted mediating effects

(decreasing consciousness of breathing effort, reducing anxiety, and the aggregation

effects of both). In addition, the strategies of managing dyspnea with regard to clinical

practice should emphasize assessing personal aspects associated with dyspnea, such as
130

the language, the affective components, and individual’s awareness of breathing effort.

Nursing education should include introducing this phenomenon to provide evidence upon

which to continue current clinical practices in dyspnea management strategies, to develop

new care strategies, and to evaluate the effectiveness of these strategies.


131

Appendix

ID ___________

Enrollment

1. Age at the first interview ________

2. Gender † Male † Female

3. Height __________ (in centimeter)

4. Current diagnosis † primary † Metastastic

If Metastastic, the original: _______________

5. Stage † Stage I † Stage II † Stage III

† Stage IV † other (specify) __________

6. Hemoglobin __________

7. History of COPD † Yes † No

8. Employment † Employed † Retired † Not employed

† Other (please specify) ____________

9. Education † no education † elementary school

†Junior High School † High school

† Associate Degree † Bachelor † Post graduate


132

Rapid Shallow Breathing ______________

FEV1 ______________

FVC ______________
133

ID ___________

Katzman’s Short-memory concentration Scale

Error

What year is it now?

What month is it now?

Repeat this phrase:

John Brown, 42 market street, Chicago

About what time is it? (within one hour)

Count backwards 20 to 1

Say the months in reverse order

Repeat the phrase just given


134

ID___________

Second Interview

Breathing Status

Rapid Shallow Breathing ______________

FEV1 ______________

FVC ______________

Respiratory Pressure

First Second Third The

time time time best count

(IP)

Rest (T1~T2)
(T2~T3)
Time
135

First time Second time Third time The best

count

Expiratory

Pressure

(EP)

Rest Time (T1~T2)


(T2~T3)
136

ID ___________
VVAS
Please mark how you feel that how hard you have to breathe
137

VVAS
Please mark the intensity of your breathing difficulty
138

ID: _________________
CDS

Not at A Some- Consi- Very


all Little what derably Much
1. Do you feel breathing
1 2 3 4 5
difficulty
by palpitation and
sweating?
2. Do you feel such
1 2 3 4 5
breathing difficulty
that you don’t know
what to do?

3. Do you feel your breath 1 2 3 4 5

may stop?

4. Do you feel as if your 1 2 3 4 5

are drowning?
139

ID: _______________
The State Scale of STAI

Not at all Somewhat Moderately Very Much

1. I feel
1 2 3 4
calm
2. I am
1 2 3 4
tense

3. I feel 1 2 3 4

upset

4. I am 1 2 3 4

relaxed

5. I feel 1 2 3 4

content

6. I am 1 2 3 4

worried
140

ID __________

Subject Burden

How long does the second interview take place? (In min) ___________

During the second interview, did subject asked to stop?


† Yes † no

(If yes) How many times did the second interview stop: ________

The stop period (In min) : _____ _______ ______


141

References

Adams, L., Chronos, N., Lane, R., & Guz, A. (1986). The measurement of

breathlessness induced in normal subjects: individual differences. Clinical science,

70(2), 131-140.

Aiken, L. S., & West, S. G. (1991). Multiple Regression: testing and interpreting

interactions. California: Sage Publication, Inc.

Altman, D., Edmonds, P., Higginson, I., Mcdonnell, M., & Sen-Gupta, G. (2000). Is the

presence of dyspnea a risk factor for morbidity in cancer patients? Journal of Pain

and Symptom Management, 19(1), 15-22.

American Thoracic Society (1998). Dyspnea: Mechanism, assessment, and management.

American Journal of Respiratory Critical Care Medicine, 159, 351-359.

Aubier, M., Peiffer, C., Poline, J-B., Samson, Y. & Thivard, L. (2001). Neural substrates

for the perception of acutely induced dyspnea. American Journal of Respiratory

Critical Care Medicine, 163(4), 951-957.

Bailey P. H. (2004). The dyspnea-anxiety-dyspnea cycle--COPD patients' stories of

breathlessness: "It's scary/when you can't breathe".” Qualitative Health Research,

14 (6), 2004, 760-778.

Banzett, R.B., Im, B.S., Lansing, r. W., Legedza, A.T.R., & Thwing, J.I. (2000). The

perception of respiratory work and effort can be independent of the perception of

air hunger. American Journal of Respiratory critical Care Medicine, 162,


142

1690-1696.

Bideian A. G., & Mossholder, K. W. (1994). Simple question, not so simple answer:

interpreting interaction terms in moderated multiple regression. Journal of

Measurement, 20(1), 159-165.

Borg, G. A. (1982). Psychophysical bases of perceived exertion. Medicine and Science in

Sports and Exercise,14(5),377-381.

Brand P. L, Rijcken B. , Schouten J. P. , Koeter, G.. H., Weiss, S. T, & Postma, D. S.


HU UH HU UH HU UH HU UH HU UH HU UH

(1992). Perception of airway obstruction in a random population sample.

Relationship to airway hyperresponsiveness in the absence of respiratory symptoms.

American Review of Respiratory Disease, 146(2), 396-401.

Broadley, K.E., Davies, A. N., Higgins, J. P. T., & Jennings, A. (2002). A systematic

review of the use of opioids in the management of dyspnea. Thorax, 87(11),

939-944.

Bruera, E., Schmitz, B., Pither, J., Neumann, C.M., & Hanson, J. (2000). The frequency

and correlates of dyspnea in patients with advanced cancer. Journal of Pain and

Symptom Management, 19(5), 357-362.

Burns, N., & Grove, S. K. (2001). The Practice of Nursing Research: Conduct, Critique,

& Utilization. Philadelphia: W.B. Saunders Company.

Carla, R., & Eduardo, B. (1997). Dyspnea: Pathophysiology and assessment. Journal of

Pain and Symptom Management, 13, 220-232.


143

Carrieri, V.K., Janson-Bjerklie, S., & Jacobs, S. (1984). The sensation of dyspnea: a

review. Heart & Lung, 13 (4), 436-447.

Carrieri-Kohlman, V. , Gormley, J. M. , Douglas, M. K. , Paul, S.M. , Stulbarg, M. S.


HU UH HU UH HU UH HU UH HU UH

(1996) Differentiation between dyspnea and its affective components. Western

Journal of Nursing Research, 18(6), 626-642.

Chan, C. W. H., Richardson, A. & Richardson, J. (2005). A study to access the

extenstence of the symptom cluster of breathlessness, fatigue, and anxiety in

patients with advanced lung cancer. European Journal if Oncology Nursing, 9,

325-333.

Chen, R. C., & Yan, S. (1999). Perceived inspiratory difficulty during inspiratory

threshold and hyperinflationary loadings. American Journal of Respiratory and

Critical Care Medicine, 159(3), 720-727.

Chiu, T.Y., Hu, W. Y., Lue, B. H., Yao, C. A., Chen, C. Y., & Wakai, S. (2004). Dyspnea

and its correlates in taiwanese patients with terminal cancer. Journal of Pain and

Symptom Management, 28(2), 123-132.

Chonan, T., Mulholland, M. B., Cherniack, N. S., Altose, M. D.(1987). Effects of

voluntary constraining of thoracic displacement during hypercapnia. Journal of

applied physiology, 63(5), 1822-1828.

Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences (2nd Ed.). New

Jersey: Lawrence Erlbaum Associates, Publishers.


144

Cohen, J., Cohen, P, West, S. G., &. Aiken, L. S. (2003). Applied Multiple Regression /

Correlation Analysis for the Behavior Sciences (3rd Ed.). New Jersey: Lawrence

Erlbaum Association, Inc., Publishers.

Comore, J. (1966). Some theories in the mechanism of dyspnea. In J.B. Howell & E.J.

Campbell(Eds.) Breathlessness. Oxford: Blackwell Scientific Publications. Pp. 1-7.

Corcoran, R., Fielding, K., Grosby, V., Hughes, A., Tattersfied, A. E., & Wilcock, A.

(2002). Descriptors of breathless in patients with cancer and other cardiorespiratory

disease. Journal of Pain and Symptom Management, 23(3), 182-189.

Cournand, A., Baldwin, E. D., darling, R. C., & Richard, D. W. (1941). Studies on

intrapulmonary of gases, IV. The significance of the pulmonary empting rate and a

simplified open circuit measurement of residual air. The Journal of Clinical

Investigation, 20(6), 681-689.

Doyle, D., Hanks, G. C., & McDonald, N. (1998). Oxford Textbook of Palliative

Medicine (2nd ed.). New York: Oxford University Press.

Dudel, J. (1986). General sensory physiology, psychophysics. In R. F. Schmidt (3rd ed.).

Fundamentals of neurophysiology. New York: Springer-Verlag. PP 1-29.

Dudgeon, D. J., & Lertzman, M. (1998) Dyspnea in the advanced cancer patient. Journal

of Pain and Symptom Management,16 (4), 212-219.

Dudgeon, D.J., Kristjanson, L., Solan, J. A., Lertzman, M., & Clement, K. (2001).

Dyspnea in cancer patients:prevalence and associatd factors. Journal of Pain and


145

Symptom Management, 21(2), 95-102.

Dudley, D. L., Martin, C. J., Holmes, T. H. (1968). Dyspnea: psychologic and

physiology observations. Journal of Psychosomatic Research, 11, 325-339.

Dugeon, D. J., Lertzman, M., & Askew, G. R. (2001). Physiological changes and clinical

correlations of dyspnea in cacner outpatients. Journal of Pain and Symptom

Management, 21(5), 373-379.

Fishbein, D., Kearon, C., & Killian, K. (1989). An approach to dyspnea in cancer subjects.

Journal 0f Pain and Symptom Management, 4(2), 76-81.

Fishman, A. P., & Ledlie, J. F. (1979). Dyspnea. Bulletin European de Physiopathologie

Respiratorie, 15(5), 789-804.

Gift, A. G. (1989). Validation of a vertical visual analogue scale as a measure of clinical

dyspnea. Rehabilitation Nursing, 14(6), 323-325.

Gottfried, S. B., Altose, M.D., Kelsen,S.G., Cherniack, N.S. (1981). Perception of

changes in airflow resistance in obstructive pulmonary disorders. American Review

of Respiratory Disease, 124 (5), 566-570.

Gottfried, S. B., Redline, S., Altose, M.D. (1985). Respiratory sensation in chronic

obstructive pulmonary disease. American Review of Respiratory Disease, 132

(5),954-959.

Guyton, A. C. & Hall, J. E. (2006). Textbook of Medical Physiology(11th ed.). U U

Philadelphia: W. B. Saunders Company. Pp.514-523 & pp. 524-533.


146

Hamnegard. C. H., Wragg, S.,Kyroussis, D., Aquilina, R., Moxham, J., & Green, M.

(1994). Portable measurement of maximum mouth pressures. The European

Respiratory Journal,7 (2) 398-401.

Han, J., Zhu, Y., Li, S., Chen, X., Put, C., Van de Woestijne, K. P. & Can de Bergh, O.

(2005), Respiratory complaints in Chinese: cultural and diagnostic specificities.

Chest, 127, 1942-1951.

Hudgel, D.W., Cooperson, D.M., & Kinsman, R.A. (1982) Recognition of added resistive

loads in asthma. American Review of Respiratory Disease, 126(1) 121-125.

Jaccard, J., & Turrisi, R. (2003). Interaction Effects in Multiple Regression. California:

Sage Publication, Inc.

Janson-Bjerklie, S., Carrieri, V.K., & Hudes, M (1986). The sensations of pulmonary

dyspnea. Nursing Research, 35(3), 154-159.

Jones, G. L., Killian, K. J., Summers, E., & Jones, N. L. (1985). Inspiratory muscle forces

and endurance in maximum resistive loading. Journal of applied physiology,58, (5),

1608-1615.

Kane, R. L. (1997), Approaching the outcome question. in R. L. Kane (ed.)

Understanding Health Care Outcomes Research. London: Jones and Bartlet

Publishers, Inc. PP 1-16.

Katzman,R., Brown, T., Fuld, P., Peck, A . Schechter, R., & Schimmel, H (1983).
147

Validation of a short orientation-memory-concentration test of cognitive

impairment. The American Journal of Psychiatry, 140, 734-739.

Killian K. J. & Jones, N. L. (1988). Respiratory muscles and dyspnea. Clinics in

Chest Medicine, 9(2), 237-248.

Killian, K. J. (1998). Sense of effort and dyspnea. Monaldi Arch Chest Disease 53(6),

654-660.

Killian, K. L., Gandevia, S. C., Summers, E. & Campell, E.M.J. (1984). Effect of

increased lung volume on perception of breathless, effort, and tension. American

Physiological Society, 686-691.

Lansing, R. W., IM, B. S., Thwing, J. I., Legedza, A. T. R., & Banzett, R. B. (2000).

The perception if respiratory work and effort can be independent of the perception

of air hunger. American Journal of Critical Care Medicine, 162, 1690-1696.

Levitzky, M. G. (1999). Pulmonary Physiology (5th Ed.). New York: McGraw-Hill Health

Professions Division.

Lum, L. C. (1981). Hyperventilation and anxiety state. Journal of the Royal Society of

Medicine, 74(1), 1-4.

Lush, M. T., Janson-Bjerklie, S., Carrieri, V. K., & Lovejoy, N. (1988). Dyspnea in the

ventilator-assisted patient. Heart and Lung, 17 (5), 528-535.

Marteau, T. M. & Bekker, H. (1992). The development of a six-item short-form of the

state scale of the Spielberger state-trait anxiety inventory (STAI). British Journal of
148

Clinical Psychology, 31, 301-306.

Meek, P. M. Laureau, S. C., & Hu, J. (2003), Are self-reports of breathing effort and

breathing distress stable and valid measures among persons with asthma, persons with

COPD, and health person? Heat & Lung, 32 (5), 335-346.

Mor,V. & Rebuen, D. B. (1986). Dyspnea in terminally ill cancer patients. Chest, 89(2),

234-246.

Muers, M. F., & Round, C. E. (1993). Palliation of symptom in non-small cell lung

cancer: a study by the Yorkshore regional cancer organization thoracic group.

Thorax, 48(4), 339-343.

Mystakidou, K., Tsilika, E., Parpa, E., Katsouda, E., Galanos, A. & Vlahos, L. (2005).

Assessment of anxiety and depression in advanced cancer patients and their

relationship with quality of life. Quality of Life Research, 14(8), 1825-1833

Newton-Bishop J.A., Nolan C., Turner F., McCabe M., Boxer C., Thomas J.M., Coombes

G., A'Hern R. P., & Barrett J.H., (2004), A quality-of-life study in high-risk

(thickness > = or 2 mm) cutaneous melanoma patients in a randomized trial of

1-cm versus 3-cm surgical excision margins. The Journal of Investigative

Dermatology Symposium proceeding. 9(2), 152-159.

O'Donnell, D. E., Bertley, J. C., Chau, L. K., Webb, K. A. (1997). Qualitative aspects of

exertional breathlessness in chronic airflow limitation: pathophysiologic

mechanisms. American Journal of Respiratory and Critical Care Medicine, 155(1),


149

109-115

O'Driscoll, M., Corner, J., Bailey, C. (1999). The experience of breathlessness in lung

cancer. European Journal of Cancer Care, 8(1), 37-43

Person, C., Robert, D. K., & Thorne, S. E. (1993). The experience of dyspnea in late

stage cancer. Cancer Nursing, 16(4), 310-320.

Rice, R. L. (1950). Symptom patterns of the hyperventilation syndrome. American

Journal of Medicine, 8(6), 691-700.

Rispoli A., Pavone I., Bongini A., Di Loro F., Ponchietti R., Rizzo M.. (2005),

Genitourinary cancer: psychological assessment and gender differences. Urologia

Internationalis. 74(3), 246-249.

Rochester, D. F. (1988). Tests of respiratory muscle function. Clinics in Chest Medicine,

9(2), 249-261.

Roy, C., & Robinson, C. R. (1992). Physiological nursing research in dyspnea: a

paradigm shift and a metaparadigm exemplar. Scholarly Inquiry for Nursing

Practice: an International Journal, 6(2), 81-104.

Rubinfield, A. R., & Pain, M.C.F. (1976). Perception of asthma. Lacent 24(1) 882-884.

Saunders, D. R. (19560. Moderator variables in prediction. Educational and

Psychological Measurement, 16, 209-222.

Schwartzstein,R.M., Manning, H. L., Weiss, J. W., & Weinberger, S.E. (1990). Dyspnea:

a sensory experience. Lung, 168, 185-199.


150

Skevington, S. M., Pilaar, M., Routh, D., & Macleod, R. D. (1997). On the language of

breathlessness. Psychology and Health, 12, 677-689.

Smith, E. L., Hann, D. M., Ahles, T. A., Furstenberg, C. T., Mitchell, T. A., Meyer, L.,

Maurer, L. H., Rigas, J. & Hammond, S. (2001). Dyspnea, anxiety, body

consciousness, and quality of life in patients with lung cancer. Journal of

Pain and Symptom Management,21(4), 323-329.

Smith, P. E & Royall, R. A. (1992). A portable mouth pressure meter. Anaesthesia,

47(2), 144-145.

Steele B., & Shaver, J. (1992). The dyspnea experience: nociceptive properties and a

model for research and practice. Advanced Nursing Science, 15 (1), 64-76.

Stevens, S. S. (1957). On the psychophysical law. Psychological Review, 64, 153-181.

Tabachnick, B. G., & Fidell, L. S. (2001). Using Multivariate Statistics (4th ed.).

Massachusetts: Allyn & Bacon.

Tanaka, K. Akechi, T., Okuyama, T., Niskiwaki, & Uchitomi, Y. (2000). Development

and validation of the cancer dyspnoea scale: a multidimensional, brief, self-rating

scale. British Journal of Cancer, 82(4), 800-805.

Tanaka, K., Akechi, T., Okuyama,T., Nishiwaki, Y., & Uchitomi, T. (2002). Factors

correlated with dyspnea in advanced lung cancer patients: organic causes and what

else? Journal of Pain and Symptom Management, 23(6), 490-500.

Tarzian, A. J. (2000). Caring for dying patients who have air hunger. Journal of nursing
151

scholarship, 32(2), 137-143.

West, J. B. (1985). Respiratory Physiology: the essential (3rd. ed.). Baltimore: Williams &

Willkins.

Widimisky, J. (1979). Dyspnea. Cor Vasa, 2(2), 128-141.

Wilson, R.C., & Jones, P.W. (1989). A comparison of the visual analogue scale and

modified Borg scale for the measurement of dyspnoea during exercise. Clinical

science,76(3), 277-282

Wilson, S. H., Cooke, N. T., Edwards, R. H. T., & Spiro, S. G. (1984). Predicted

normal values for maximal respiratory pressures in Caussian adults and children.

Thorax, 39(7). 535-538.

Yan, S. (1999). Sensation of inspiratory difficulty during inspiratory threshold and

hyperinflationary loadings, effect of inspiratory muscle strength. American Journal

of Respiratory & Critical Care Medicine, 160(5), 1544-1549.

http://crs.cph.ntu.edu.tw/crs_c/annual.html
H

http://crs.cph.ntu.edu.tw/crs_c/annual.html, 2005
H H

http://www.cdc.gov/nccdphp/dnpa/physical/measuring/perceived_exertion.htm, 2005
H H

http://www.micro-direct.com, 2005

Potrebbero piacerti anche