Ij

This material is protected by U.S. copyright law. Unauthorized reproduction is prohibited. To purchase quantity reprints,
please e-mail reprints@ons.org or to request permission to reproduce multiple copies, please e-mail pubpermissions@ons.org.
ONLINE EXCLUSIVE
The Effects of Health State, Hemoglobin, Global
Symptom Distress, Mood Disturbance, and Treatment
Site on Fatigue Onset, Duration, and Distress in Patients
Receiving Radiation Therapy
Morris A. Magnan, PhD, RN, and Darlene W. Mood, PhD
Key Points . . .
Purpose/Objectives: To describe the fatigue experience of patients
with cancer receiving radiation therapy and determine to what extent di-
verse correlates of fatigue affect fatigue onset, duration, and distress.
Little is known about the fatigue experience of patients with
Design: Descriptive correlational study completed by secondary data
cancer who are receiving radiation to different areas of the
analysis.
body.
Sample/Setting: Data were obtained from 384 subjects recruited
from two urban, university-affiliated, radiation oncology clinics located
Knowing what factors affect fatigue onset, duration, and se-
in a large, Midwestern city.
verity of distress can help nurses adjust the timing and content
Methods: The effects of health indicators and treatment site on fa-
of patient teaching.
tigue onset, duration, and distress were examined using correlational
Pretreatment screening is needed to identify patients who are
analyses and analyses of variance.
Main Research Variables: Hemoglobin, health status, global symp-
vulnerable to early onset, longer duration, and more distress-
tom distress, mood disturbance, treatment site, and fatigue onset, du-
ing fatigue because of poor health, low hemoglobin, mood
ration, and distress.
disturbance, or symptom distress.
Findings: Fatigue started near the middle of the second week of
treatment, was moderately distressing, and lasted approximately 32
days. Higher levels of health and hemoglobin at the start of therapy were
associated with a delayed onset, shorter duration, and lower levels of fa-
mood disturbance, and area of the body being treated, have
tigue distress. In contrast, higher pretreatment levels of global symp-
been examined in relation to fatigue severity, but the effect of
tom distress and mood disturbance were associated with an earlier
these correlates of fatigue on fatigue onset, duration, and dis-
onset, longer duration, and greater severity of fatigue distress.
tress is virtually unknown. The purposes of this study were to
Conclusion: The fatigue experience in patients undergoing radiation
(a) describe the fatigue experience in terms of its onset, dura-
therapy is highly individualized. Variations in the health states of patients
tion, and distress in adult patients with cancer being treated
as well as the area of the body being treated can influence fatigue on-
with radiation to different areas of the body, and (b) determine
set, duration, and distress.
Implications for Nursing: Pretreatment screening for fatigue and its
to what extent hemoglobin, health state, global symptom dis-
correlates is needed to identify patients at risk for an earlier onset,
tress, mood disturbance, and area of the body being treated
longer duration, and more distressing levels of fatigue.
influence fatigue onset, duration, and severity of distress.
Morris A. Magnan, PhD, RN, is an assistant professor in the College
atigue is a commonly reported and distressing side ef-
F
of Nursing at Wayne State University in Detroit, MI; and Darlene W.
fect of radiation therapy (Munro & Potter, 1996;
Mood, PhD, is a senior research scientist at Karmanos Cancer In-
Oberst, Hughes, Chang, & McCubbin, 1991; Portenoy,
stitute and professor emeritus in the College of Nursing at Wayne
Thaler, Kornblith, Lepore, & Fiedlander-Klar, 1994). Provid-
State University. The parent study for this project was funded by a
ing care to patients experiencing fatigue requires that nurses
grant from the National Cancer Institute (NCI RO5 CA59013-11) to
have, at least, accurate general information about the fatigue
the second author. Direct support for this corollary study came
experience, including its onset, duration, and severity of dis-
through awards made to the first author from the Blue Cross Blue
Shield of Michigan Foundation and the Lambda Chapter of Sigma
tress. This general level of knowledge can be made more spe-
Theta Tau International. (Submitted May 2002. Accepted for publi-
cific by understanding further how factors in people and their
cation October 8, 2002.)
environments affect the symptom experience. Diverse factors,
such as hemoglobin, health state, global symptom distress,
Digital Object Identifier: 10.1188/03.ONF.E33-E39
MAGNAN VOL 30, NO 2, 2003
E33
Literature Review
al.). Also, the literature suggests that patients treated for lung
cancer and those irradiated to the head and neck area have a
Fatigue has been described as a subjective experience of
higher incidence and longer duration of fatigue than individu-
unusual, excessive, or overwhelming tiredness that results in
als receiving therapy to other areas of the body (King et al.;
a decreased desire or capacity for mental or physical activity
Piper et al., 1989). However, the prospective data needed to
(Mock, 1997). In patients with cancer, fatigue is a pervasive,
accurately estimate the duration of fatigue in these seemingly
discouraging, and debilitating symptom that can adversely
more vulnerable groups have yet to be produced.
affect daily functioning, quality of life, health, and personal
Significant relationships between global symptom distress
well-being (Ferrell, Grant, Dean, Funk, & Ly, 1996; Irvine,
and fatigue, as well as mood disturbance and fatigue, have
Vincent, Graydon, & Bubela, 1998; Irvine, Vincent, Graydon,
been reported across a number of studies involving patients
Bubela, & Thompson, 1994; Kobashi-Schoot, Hanewald, van
with cancer receiving radiation (Blesch et al., 1991; Green-
Dam, & Bruning, 1985; Longman, Braden, & Mishel, 1999;
berg et al., 1992; Holmes, 1991; Irvine et al., 1998, 1994;
Magnan, 2001; Oberst et al., 1991; Vogelzang, Breitbart,
Jamar, 1989; McCorkle & Young, 1987; Mock et al., 1997;
Cella, Curt, & Groopman, 1997). The fatigue experienced by
Pickard-Holley, 1991; Piper et al., 1989; Visser & Smets,
patients with cancer differs from the ordinary fatigue of day-
1998). Also, a relationship between anemia and fatigue has
to-day living in that it tends to be more severe, more distress-
been proposed for patients undergoing different kinds of can-
ing, prolonged, and unrelieved by ordinary measures such as
cer treatment (Maxwell, 1984), but research to date has failed
sleep and rest (Holmes, 1991; Oberst et al.; Piper, Lindsey, &
to demonstrate a significant relationship between these vari-
Dodd, 1987; Richardson & Ream, 1997; Richardson, Ream,
ables when studied in patients with cancer receiving radiation
& Wilson-Barnett, 1998).
(Blesch et al.; Glaus, 1993; Greenberg et al., 1992; Irvine et
The etiologic mechanisms of cancer fatigue have not been
al., 1994).
fully elucidated, but it appears to be caused both by the dis-
Few studies have examined the effect of treatment site on
ease and its treatment (Portenoy et al., 1994). Although re-
the fatigue experience. As noted earlier, several studies have
ports of an association between radiation exposure and fatigue
linked fatigue duration to the area of the body treated. In ad-
appeared in the literature as early as the late 19th century
dition, Piper et al. (1989) reported that fatigue was more se-
(Walsh, 1897), the systematic study of fatigue in patients with
vere for patients with lung cancer than those with breast can-
cancer receiving radiation therapy is still in its infancy. Dur-
cer. However, the effects of treatment site on fatigue onset and
ing the last two decades, estimates of fatigue prevalence in this
distress are virtually unknown.
population have ranged from 65%100% (Graydon, Bubela,
Methods
Irvine, & Vincent, 1995; Greenberg, Sawicka, Eisenthal, &
R o s s , 1992; Haylock & Hart, 1979; Hickok, Morrow,
Sample and Setting
McDonald, & Bellg, 1996; Irvine et al., 1998; King, Nail,
Kreamer, Strohl, & Johnson, 1985; Kubricht, 1984).
The research questions for this corollary study were an-
The onset of fatigue in relation to the start of radiation
swered by secondary data analysis. The data source, referred
therapy is not clearly delineated (Glaus, 1993; Irvine et al.,
to as the parent study, was a longitudinal, controlled clinical
1994; King et al., 1985). Research suggests an early onset,
trial of nursing interventions designed to enhance self-care
with fatigue starting near the end of the first week or the be-
knowledge and performance in patients receiving radiation.
ginning of the second week of treatment (Greenberg, Gray,
Data for the parent study were obtained from two university-
Mannix, Eisenthal, & Carey, 1993; Greenberg et al., 1992;
affiliated, outpatient radiation oncology clinics located in a
Haylock & Hart, 1979; Irvine et al., 1998). The proximity of
large Midwestern city. Participants in the parent study were 18
a recent surgical intervention to the start of radiation therapy
years of age or older, English-speaking, and scheduled to re-
has been associated with an earlier onset of fatigue in patients
ceive a minimum of 20 treatments. Subjects were excluded
treated for lung cancer (Hickok et al., 1996), but this finding
from the parent study if irradiated to an extremity only or ir-
has not been corroborated by studies involving patients under-
radiated to the brain with anticipated cognitive compromise.
going radiotherapy treatment to different areas of the body
The inclusion/exclusion criteria of the parent study applied to
(breast, cervical, or endometrial) (Irvine et al., 1994).
this corollary study, with the added stipulation that subjects
The fatigue experienced by patients undergoing radiation
reported whether fatigue started or got worse during treat-
therapy often extends beyond the treatment period. How far
ment.
it extends beyond treatment is difficult to say. Evidence from
Data Collection
the literature suggests that the area of the body being treated,
as well as the size of the treatment field, influence fatigue
Data collection for the parent study started in April 1995
duration (Devlen, Maguire, Phillips, Crowther, & Chambers,
and ended in December 1999. Data collectors trained in con-
1987; Fobair et al., 1986; King et al., 1985). Patients with non-
ducting semistructured interviews obtained information on
Hodgkin's lymphoma, for example, treated by whole body
major study variables at five points in time: pretreatment (T1),
radiation have reported fatigue lasting as long as one year
at the second (T2) and final weeks (T3) of treatment, and
beyond the end of treatment (Devlen et al.; Fobair et al.).
again at one month (T4) and three months (T5) post-treat-
Among women irradiated for breast cancer, two studies have
ment. Data pertinent to this corollary study were retrieved
reported a return to baseline levels of fatigue by three months
from the computerized databases of the parent study.
post-treatment (Greenberg et al., 1992; Irvine et al., 1994). On
Measurement
the other hand, patients receiving treatment to different areas
Variables in this corollary study included hemoglobin, health
of the body, such as the pelvis and thorax, have continued to
state, global symptom distress, mood disturbance, and fatigue
report fatigue as long as sixth months post-treatment (King et
ONF VOL 30, NO 2, 2003
E34
onset, duration, and distress. Hemoglobin, measured at the start
tigue. Fatigue duration was calculated based on patients' self-
--
of therapy, ranged from 8.1017.00 (X = 13.09, SD = 1.59).
reports of fatigue onset and cessation and quantified as the
--
The mean level for men was about 1 g higher (-- = 13.66,
actual number of calendar days. Fatigue distress was the level
X
of anguish experienced from fatigue. It was measured by self-
SD = 1.62) than the mean level for women ( X = 12.64,
report using a five-point Likert-type scale (5 = severe).
SD = 1.4), but the means for both men and women were at lev-
els recognized as normal for their gender. Five areas classified
Data Analysis
the treatment site: head and neck, lung, breast, prostate, and fe-
Data were analyzed using SPSS PC 9 (SPSS, Inc., Chi-
male pelvic area.
cago, IL). Both descriptive and inferential statistics were used
Global symptom distress was the level of anguish experi-
to analyze the data. Inferential statistics included correlational
enced from a set of commonly reported health complaints. It
analyses, paired sample t tests, and analyses of variance
was assessed using a modified, pretreatment administration of
(ANOVA). In the parent study, subjects were randomly as-
Mood's Symptoms Scale (Mood, 1994). This instrument
signed to one of three experimental conditions: a control
uses a 02 response format to obtain self-assessments of how
group, an educational intervention, or an intervention that
bothersome symptoms are across 13 commonly reported
included education plus contingency contracting. Preliminary
health complaints. The three response choices are worded to
ANOVA showed no significant between-group differences in
fit each specific item and correspond essentially to no trouble,
pretreatment hemoglobin, health state, global symptom dis-
some, or a lot. To ensure independence in the measurement of
tress, or mood disturbance or in the level of fatigue distress
the variables (independent and dependent), this instrument
experienced by subjects randomized to different experimen-
was modified for this study by removing one item that as-
tal condition in the parent study. Therefore, subjects from
sesses energy. The theoretical range for the modified version
these different groups were pooled for analysis. Fatigue dis-
was 024, with higher scores indicating higher levels of glo-
tress was expected to be at its worst during the final week of
bal symptom distress. Scores for the current study ranged
--
treatment; therefore, these final week (T3) measures of fatigue
from 020 (X = 4.53, SD = 3.28). The total score provided an
distress were used for correlational analyses and ANOVA.
internally consistent measure of global symptom distress with
The alpha for statistical significance was set at 0.05.
a Cronbach's alpha of 0.72.
Mood disturbance was measured pretreatment using a
Results
shortened version of the Profile of Mood States (POMS)
(Mood, 1994). This 14-item instrument was adapted from the
Sample Characteristics
64-item version (McNair, Lorr, & Droppleman, 1971) by the
The 384 subjects in this corollary study included 175 (46%)
second author to reduce the burden that patients with cancer
men and 209 (54%) women ranging in age from 2487 years
experienced by answering the longer version. The shortened
--
(X = 58.94, SD = 11.73). Subjects in the sample were ethni-
POMS contains at least one item from each of the original
cally, educationally, socially, and economically diverse,
subscales, except vigor and fatigue. The design of the short-
treated to different areas of the body, and heterogeneous on
ened version follows that of the longer one. A 04 response
cancer stage (see Table 1). The greatest number of subjects
format on 14 paired items is used to generate a total score
was treated for breast cancer (n = 143, 37%), followed by
measuring the affective state. The theoretical range is 056,
prostate cancer (n = 122, 32%). Subjects receiving treatment
with higher scores indicating higher levels of mood distur-
to one of these two sites comprised 69% of the sample. The
bance. In the current study, scores ranged from 056, with a
remaining 31% of the sample included subjects receiving
mean of 7.96 (SD = 9.65). The instrument was internally con-
treatment to the head and neck region (n = 55, 14%), lung
sistent with a Cronbach's alpha of 0.95.
(n = 41, 11%), or female pelvic region for cervical or endome-
Health state was assessed using a pretreatment administra-
trial cancers (n = 23, 6%). Treatment was predominantly for
tion of the Functional Assessment of Cancer Therapy (Cella,
localized cancers (stages 0, I, II), but 68 (18%) of the subjects
Tulsky, Gray, Sarflan, & Linn, 1993). This quality-of-life mea-
had regional extension (stage III), and another 48 (12%) of the
sure was used as a proxy to determine health state because it
subjects had distant metastasis (stage IV). Data on cancer
separates aspects of general health unrelated to the cancer ex-
staging were not reported for three subjects.
perience from dimensions of functioning pertinent to cancer
and its treatment. The instrument consists of 27 core items
Fatigue Onset, Duration, and Distress
rated on a five-point scale, with lower scores indicating lower
Fatigue onset varied widely. When determined by the num-
levels of functional quality of life. Four dimensions of quality
ber of actual treatment days, using a five-day treatment week,
of life are assessed: physical (7 items), social (7 items), emo-
patients reported fatigue onset as early as the first and as late
tional (6 items), and functional (7 items). The instrument was
--
as the 38th day of treatment. The average onset (X = 7.69
modified by removing one item from the physical subscale
treatment days, SD = 6.89) was approximately halfway
that measures energy. The theoretical range for the modified
through the second week of treatment.
version was 0104. Subjects' scores ranged from 5104, with
Figure 1 shows the daily distribution of fatigue onset dur-
a mean of 79.29 (SD = 15.17). The internal consistency reli-
ing treatment weeks one and two. Reports of onset were great-
ability of the modified instrument was 0.87.
est on the first treatment day of each week, then peaked on the
Attributes of fatigue--its presence, date of onset, and sever-
third and fifth treatment days of each week. Although the
ity of distress--were measured at the second (T2) and final
average onset of fatigue was near the middle of the second
weeks (T3) of treatment and again at one month (T4) and three
week of treatment, 168 subjects (44%) reported fatigue onset
months (T5) post-treatment. Fatigue onset was determined by
during the first week of treatment, with another 124 (32%)
patients' self-reports and quantified as the number of calendar
reporting onset by the end of the second week. Thus, three-
days from the start of therapy to the first reported day of fa-
MAGNAN VOL 30, NO 2, 2003
E35
Table 1. Demographic Characteristics of the Sample
90
90
Day 1
80
80
%
Characteristic
n
80
80
Day 2
70
70
Gender
Day 3
54
Women
209
60
60
Day 4
46
Men
175
50
50
Ethnicity
41
Day 5
41
51
Anglo American
196
40
40
32
46
African American
178
3
28
28
3
Other
10
30
30
22
2
20
20
20
Marital status
20
18
1
16
20
15
20
1
15
61
Partnered
235
39
Not partnered
149
10
10
Education
0
00
8
Grades 18
29
Week 1
We  1
Week 22
W ek
41
Grades 912
158
n = 168
n ==124
n  124
37
Grades 1316
143
12
Grades 1722
48
2
Unreported
6
Figure 1. Reports of Daily Onset of Fatigue Within
Income
Treatment Weeks One and Two
7
< $5,000
26
15
$5,000$14,999
59
13
$15,000$29,999
51
17
$30,000$49,999
67
gressed, which was confirmed by a paired sample t test (t(383)
13
$50,000$74,999
49
= 10.27, p < 0.001).
13
> $75,000
51
Correlates of Fatigue Onset, Duration, and Distress
Unreported
21
81
Socioeconomic statusa
The first correlate examined was health state. Analyses
9
Class I (highest)
34
showed that pretreatment measures of health state were nega-
21
Class II
81
tively and significantly related to fatigue distress (r = 0.19,
27
Class III
103
p < 0.001) and fatigue duration (r = 0.15, p = 0.004) but posi-
31
Class IV
120
tively and significantly related to fatigue onset (r = 0.15,
10
Class V (lowest)
37
p = 0.004). Similarly, hemoglobin was negatively and signifi-
2
Not classified
9
Treatment site
cantly related to fatigue distress (r = 0.14, p = 0.012) and du-
37
Breast
143
ration (r = 0.14, p = 0.012) but positively and significantly
32
Prostate
122
related to onset (r = 0.19, p = 0.001). The magnitudes of these
4
Head/neck
55
correlations were small. The results suggested that both subjec-
11
Lung
41
tive and objective indicators of positive health at the beginning
6
Cervical/endometrial
23
of treatment were associated with a lower severity of fatigue
Cancer stageb
distress, a delayed onset, and a shorter duration of fatigue.
7
Stage 0
27
On the other hand, two correlates that were negative health
28
Stage I
107
indicators were associated with an earlier onset of fatigue,
34
131
Stage II
more severe fatigue distress, and a longer duration of fatigue.
18
68
Stage III
13
48
Stage IV
Global symptom distress was positively and significantly re-
1
3
Unreported
lated to fatigue distress (r = 0.20, p < 0.001) and duration
(r = 0.11, p = 0.037) but negatively and significantly related
N = 384
to fatigue onset (r = 0.11, p = 0.029). Mood disturbance, like
a
Socioeconomic status was based on Hollingshead's two-factor index
symptom distress, was positively and significantly related to
wherein higher-class rankings reflect lower socioeconomic status (Miller,
fatigue distress (r = 0.16, p = 0.002) and fatigue duration
1991).
(r = 0.12, p = 0.014) but negatively and significantly related
b
Cancer staging was based on the American Joint Committee on Cancer's
to fatigue onset (r = 0.12, p = 0.016).
staging guidelines (1998).
The last correlate to be examined was treatment site. The
effect of treatment site on fatigue distress, onset, and duration
quarters (n = 292, 76%) of the subjects reported fatigue onset
was evaluated using one-way ANOVAs to compare means on
by the end of the second week of treatment. Notably, 80
fatigue onset, duration, and severity of distress. The results of
(20%) of the subjects reported that fatigue started or got worse
this analysis showed that the mean level of fatigue onset did
on the very first day of treatment.
not differ by treatment site (see Table 2). However, the main
Subjects reported fatigue lasting from 178 calendar days
effect of treatment site was significant for both fatigue dura-
--
(X = 32.22, SD = 12.98). On average, level of fatigue distress
tion, F(4, 379) = 3.09, p = 0.016, and fatigue distress, F(4,
--
reported during the second week of treatment (X = 2.39,
379) = 3.01, p = 0.018. Post hoc analyses using a Bonferonni
SD = 1.31) was less severe than the level of fatigue distress re-
correction (alpha = 0.05) showed that the only significant dif-
--
ported at the final week of treatment (X = 3.16, SD = 1.17).
ference in mean duration was between patients with prostate
This suggested a worsening of fatigue distress as therapy pro-
cancer and those treated for gynecologic cancer (mean differ-
ONF VOL 30, NO 2, 2003
E36
Table 2. Analyses of Variance Comparisons of Fatigue Onset, Duration, and Distress by Treatment Site
Head/Neck
Lung
Breast
Prostate
Gynecologic
--
--
--
--
--
Fatigue
SD
SD
df
p
SD
SD
SD
F
X
X
X
X
X
6.22a
9.00b
4.74a
0.029c
Onset
6.45
6.13
4,379
7.24
6.60
7.64
7.99
5.33
2.74
34.87a
30.25b
39.48a
Duration
12.73
11.77
4,379
0.016
32.93
14.76
31.69
13.36
13.10
3.09
2.78a
2.09b
2.61a
Distress
2.56
1.45
1.26
2.43
1.28
1.23
1.44
4,379
0.018
3.01
Note. In post hoc analyses using a Bonferroni correction (alpha = 0.05), "a" was significantly greater than "b."
c
Equal variances could not be assumed, Levene's F (4,379) = 3.029, p = 0.018, therefore, a more conservative value (alpha = 0.025) was used to test the sig-
nificance of the omnibus F.
ence = 9.22, p = 0.017), indicating that fatigue duration in
though it seems clear that a search for the etiologic mecha-
patients with gynecologic cancer was longer than fatigue du-
nisms underlying this more immediate experience of fatigue
is warranted, it seems equally clear that clinicians and re-
ration in patients with prostate cancer. With respect to fatigue
distress, the only significant mean difference (mean difference
searchers need be alerted to the fact that a substantial percent-
= 0.63, p = 0.034) was between subjects treated for prostate
age (20%) of patients in this sample experienced fatigue on
the very first day of treatment.
cancer or lung cancer, indicating that fatigue distress was
An exploration of the influential effects of symptom dis-
greater in subjects treated for lung cancer than it was in sub-
tress and mood disturbance on fatigue demonstrated that
jects treated for prostate cancer.
higher levels of the former correlated with higher levels of the
In summary, an examination of the relationship of diverse
latter. The magnitudes of these correlations were small (0.10
correlates of fatigue to fatigue onset, duration, and distress dem-
onstrated that correlates indicative of positive health, such as a
0.20), which is consistent with reports from the literature re-
garding the relationship between global symptom distress and
higher health state and hemoglobin, were associated with a later
fatigue severity (Irvine et al., 1998, 1994) and mood distur-
onset of fatigue, lower levels of fatigue distress, and a shorter
duration of fatigue. In contrast, correlates indicative of negative
b a n c e and fatigue severity (Irvine et al., 1998, 1994;
McCorkle & Young, 1987; Mock et al., 1997).
health, such as global symptom distress and mood disturbance,
Results of the current study demonstrated that both symp-
were associated with an earlier onset of fatigue, a longer dura-
tion of fatigue, and higher levels of fatigue distress. Also, fa-
tom distress and mood disturbance had a small but significant
negative relationship with fatigue onset. This suggested that
tigue distress was more severe in patients treated for lung can-
higher pretreatment levels of symptom distress and mood dis-
cer compared to patients treated for prostate cancer, whereas
fatigue duration was greater for women treated for gynecologic
turbance were risk factors for an earlier onset of fatigue.
cancers compared to men treated for prostate cancer.
Findings of the current study demonstrated that hemoglo-
bin correlated negatively with fatigue distress and positively
Discussion
with fatigue onset. These findings suggested that lower lev-
els of hemoglobin were consistent with higher levels of fa-
Results of this study placed the average onset of fatigue near
tigue distress and an earlier onset of fatigue. Differences in the
the middle of the second week of treatment, but a breakdown
findings of this study compared to the findings of others
of onset by treatment week showed that 43% of the subjects
(Blesch et al., 1991; Glaus, 1993; Greenberg et al., 1992;
actually reported fatigue onset during the first week of treat-
Irvine et al., 1994) and may be accounted for, in part, by the
ment. Thus, for a percentage of patients, fatigue started earlier
fact that different dimensions of fatigue (distress versus sever-
in the treatment trajectory than what might be expected based
ity) were being examined. Also, in the studies previously
on the aggregate findings and reports from the literature. This
mentioned, power to detect an effect may have been limited
early onset for fatigue may have been a function of the timing
because of small sample sizes (N < 110). In the current study,
of measurement. In the parent study, the first symptom inter-
intercorrelations between hemoglobin and attributes of fatigue
view occurred between the 8th and 12th day of treatment. Con-
were examined among 303 subjects, a sample size large
sequently, errors of recall regarding the onset of fatigue were
enough to detect a small to medium effect with power of 0.97
likely to be less in the parent study than in studies using assess-
(p = 0.05).
ment points further removed from the start of treatment. On the
An exploration of the influence of initial health status on
other hand, the fact that 80 subjects (20% of the sample) re-
fatigue demonstrated that it correlated positively with fatigue
ported that fatigue started or got worse on the very first day of
onset but negatively with fatigue distress. This suggested that
treatment raises questions about the etiologic mechanisms of
starting radiation therapy in a better state of health was asso-
fatigue in patients with cancer receiving radiation.
ciated with a delayed onset of fatigue and lower levels of fa-
The physiologic mechanisms of a so-called "radiation-in-
tigue distress.
duced fatigue" have not been fully elucidated. Therefore, it is
Few studies have examined the relationship between the
difficult to know whether patients reporting fatigue on the
area of the body being treated and fatigue onset, duration, or
very first day of treatment started therapy with a baseline fa-
distress. Results of this study demonstrated that both fatigue
tigue as a result of other causes or were especially vulnerable
distress and fatigue duration varied according to the area of
to the "fatigue-inducing" effects of radiation, or both. Al-
the body being treated. In particular, fatigue distress was
MAGNAN VOL 30, NO 2, 2003
E37
of health, as well as the area of the body being treated, af-
worse for patients with lung cancer than it was for patients
fected the fatigue experience. Therefore, it seems prudent to
with prostate cancer, but the fatigue distress experienced by
advise patients to avoid the trap of comparing their fatigue
patients with lung cancer was not significantly different than
experience to the experiences of others.
the distress experienced by patients treated to other areas of
the body (e.g., breast, head/neck, female pelvis). Piper et al.
Limitations
(1989) reported that fatigue was more severe for patients with
Fatigue, a key construct in this study, was not a key con-
lung cancer compared to patients with breast cancer. How-
struct in the parent study and was not measured using a psy-
ever, in the current study, no differences were noted in the
chometrically sound instrument. Consequently, the psycho-
levels of fatigue distress experienced between these two
metric properties of the fatigue measure, its reliability and
groups. Still, findings from the current study in conjunction
validity, could not be subjected to rigorous evaluation. Using
with findings reported by other researchers (Piper et al., 1989)
less than optimal operations of constructs is a recognized limi-
suggest that the fatigue experience is more severe and more
tation of secondary data analysis (McCall & Appelbaum,
distressing for patients with lung cancer than it is for patients
1991). A replication study using a psychometrically sound
treated for breast or prostate cancer.
measure of fatigue is needed to determine whether the ob-
Implications for Nursing Practice
served relationships are repeatable across studies.
Clinically, knowing when fatigue is likely to start can help
Future Research
nurses adjust both the timing and content of preparatory infor-
mation. Because the average onset of fatigue was near the
The fact that 20% of the subjects in this study reported fa-
middle of the second week of treatment, one might conclude
tigue on the very first day of treatment raises important ques-
that patient counseling on symptom occurrence and manage-
tions for future research. Investigating in what ways these
ment should be accomplished no later than the end of the first
subjects differ from other subjects on characteristics such as
week of treatment. However, findings of this study suggest
age, extent of disease, and prior or concurrent treatment might
that assistance is needed sooner.
help expand the understanding of the mechanisms contribut-
Given the substantial percentage of patients reporting fa-
ing to fatigue in this population. In addition, how this more
tigue on the first day of treatment, pretreatment screening
immediate experience of fatigue affects the fatigue trajectory,
seems warranted. This screening assessment could be per-
the requirement for fatigue-related self-care, the ability to
formed either during the initial consultation with the radiolo-
engage in self-care, the achievement of fatigue control, and
gist or at the time of simulation. Because patients with lower
the quality of health-related outcomes are questions that need
levels of hemoglobin, poorer health, and higher levels of glo-
to be researched.
bal symptom distress and mood disturbance seem especially
vulnerable to an early onset of fatigue, pretreatment screening
Conclusion
should include an assessment of these parameters.
If pretreatment screening demonstrates that the patient is
General knowledge of fatigue onset, duration, and severity
positive for fatigue, an evaluation of the patient's fatigue-re-
of distress is needed to plan nursing care for patients with
lated self-care system should be undertaken to detect and rec-
cancer receiving radiation. The nursing plan of care can be
tify any existing limitations in its content or use. Also, it
tailored to the patient's situation when this general knowledge
should be recognized that the complexity of the nursing plan
of the fatigue experience is made more specific by under-
of care increases when patients receiving radiotherapy start
standing the effects of diverse correlates of fatigue on fatigue
treatment with lower levels of hemoglobin, in poorer states of
onset, duration, and distress.
health, or with higher levels of global symptom distress or
mood disturbance. Consequently, to optimize patient out-
comes, the nursing plan of care needs to address these other
areas of concern as well.
Author Contact: Morris A. Magnan, PhD, RN, can be reached at
Findings of this study suggest that fatigue is a highly indi-
mamagnan@wayne.edu, with copy to editor at rose_mary@
vidualized experience. Both positive and negative indicators
earthlink.net.
References
American Joint Committee on Cancer. (1998). AJCC cancer staging handbook
ing Forum, 23, 15391547.
(5th ed.). Philadelphia: Lippincott-Raven.
Fobair, P., Hoppe, R.T., Bloom, J., Cox, R., Varghese, A., & Speigel, D.
Blesch, K.S., Paice, J.A., Wickham, R., Harte, N.H., Schnoor, D.K., & Purl,
(1986). Psychosocial problems among survivors of Hodgkin's disease.
S. (1991). Correlates of fatigue in people with breast or lung cancer. On-
Journal of Clinical Oncology, 4, 805814.
cology Nursing Forum, 18, 8187.
Glaus, A. (1993). Assessment of fatigue in cancer and non-cancer patients and
in healthy individuals. Support Care Cancer, 1, 305315.
Cella, D.F., Tulsky, D.S., Gray, G., Sarflan, B., & Linn, E. (1993). The func-
tional assessment of cancer therapy scale: Development and validation of
Graydon, J.E., Bubela, D.I., Irvine, D.M., & Vincent, L. (1995). Fatigue-re-
the general measure. Journal of Clinical Oncology, 11, 570579.
ducing strategies used by patients receiving treatment for cancer. Cancer
Nursing, 18, 2328.
Devlen, J., Maguire, P., Phillips, P., Crowther, D., & Chambers, T. (1987).
Psychological problems associated with diagnosis and treatment of lympho-
Greenberg, D.B., Gray, J.L., Mannix, C.M., Eisenthal, S., & Carey, M. (1993).
mas. BMJ, 295, 953957.
Treatment-related fatigue and serum interleukin-1 levels in patients during
external beam irradiation for prostate cancer. Journal of Pain and Symptom
Ferrell, B.R., Grant, M., Dean, G.E., Funk, B., & Ly, J. (1996). "Bone tired":
The experience of fatigue and its impact on quality of life. Oncology Nurs-
Management, 3, 196200.
ONF VOL 30, NO 2, 2003
E38
Mood, D.W. (1994). Psychosocial interventions in radiation therapy (grant
Greenberg, D.B., Sawicka, J., Eisenthal, S., & Ross, D. (1992). Fatigue syn-
drome due to localized radiation. Journal of Pain and Symptom Manage-
proposal #RO-1 CA59013 to the National Cancer Institute). Unpublished
ment, 7, 3845.
manuscript, Wayne State University, Detroit, MI.
Munro, A.J., & Potter, S. (1996). A quantitative approach to the distress
Haylock, P.J., & Hart, L.K. (1979). Fatigue in patients receiving localized
caused by symptoms in patients treated with radical radiotherapy. British
radiation. Cancer Nursing, 2, 461467.
Journal of Cancer, 74, 640647.
Hickok, J.T., Morrow, G.R., McDonald, S., & Bellg, A.J. (1996). Frequency
Oberst, M.T., Hughes, S.H., Chang, A.S., & McCubbin, M.A. (1991). Self-
and correlates of fatigue in lung cancer patients receiving radiation therapy:
Implications for management. Journal of Pain and Symptom Management,
care burden, stress appraisal, and mood among persons receiving radio-
therapy. Cancer Nursing, 14, 7178.
11, 370377.
Pickard-Holley, S. (1991). Fatigue in cancer patients: A descriptive study.
Holmes, S. (1991). Preliminary investigations of symptom distress in two
Cancer Nursing, 14, 1319
cancer patient populations: Evaluation of a measurement instrument. Jour-
nal of Advanced Nursing, 16, 439446.
Piper, B.F., Lindsey, A.M., & Dodd, M.J. (1987). Fatigue mechanisms in
cancer patients: Developing nursing theory. Oncology Nursing Forum,
Irvine, D., Vincent, L., Graydon, J.E., & Bubela, N. (1998). Fatigue in women
14(6), 1723.
with breast cancer receiving radiation therapy. Cancer Nursing, 21, 127135.
Piper, B.F., Lindsey, A.M., Dodd, M.J., Ferketich, S., Paul, S.M., & Weller,
Irvine, D., Vincent, L., Graydon, J.E., Bubela, N., & Thompson, L. (1994).
S. (1989). The development of an instrument to measure the subjective
The prevalence and correlates of fatigue in patients receiving treatment with
chemotherapy and radiotherapy. Cancer Nursing, 17, 367378.
dimension of fatigue. In S. Funk, E. Tournquist, M. Champagne, L. Copp,
& R. Wiese (Eds.), Key aspects of comfort: Management of pain, fatigue,
Jamar, S.C. (1989). Fatigue in women receiving chemotherapy for ovarian
and nausea (pp. 199208). New York: Springer.
cancer. In S.G. Funk, E.M. Tornquist, M.T. Champagne, L.A. Copp, &
R.A. Wiese (Eds.), Key aspects of comfort: Management of pain, fatigue,
Portenoy, R.K., Thaler, H.T., Kornblith, A.B., Lepore, J.M., & Fiedlander-
and nausea (pp. 224228). New York: Springer.
Klar, H. (1994). Symptom prevalence, characteristics and distress in a can-
cer population. Quality of Life Research, 3, 183189.
King, K.B., Nail, L.M., Kreamer, K., Strohl, R.A., & Johnson, J.E. (1985).
Richardson, A., & Ream, E.K. (1997). Self-care behaviours initiated by che-
Patient's descriptions of the experience of receiving radiation therapy.
motherapy patients in response to fatigue. International Journal of Nursing
Oncology Nursing Forum, 12(4), 5561.
Studies, 34, 3543.
Kobashi-Schoot, J.A., Hanewald, G.J., van Dam, F.S., & Bruning, P.F.
Richardson, A., Ream, E., & Wilson-Barnett, J. (1998). Fatigue in patients
(1985). Assessment of malaise in cancer patients treated with radiotherapy.
receiving chemotherapy: Patterns of change. Cancer Nursing, 21, 1730.
Cancer Nursing, 8, 306313.
Visser, M.R., & Smets, E.M. (1998). Fatigue, depression and quality of life
Kubricht, D.W. (1984). Therapeutic self-care demands expressed by outpa-
in cancer patients: How are they related? Support Care Cancer, 6, 101108.
tients receiving external radiation therapy. Cancer Nursing, 7, 4351.
Vogelzang, N.J., Breitbart, W., Cella, D., Curt, G.A., & Groopman, J.E.
Longman, A.J., Braden, C.J., & Mishel, M.H. (1999). Side-effects burden,
(1997). Patient, caregiver, and oncologist perceptions of cancer-related
psychological adjustment, and life quality in women with breast cancer:
fatigue: Results of a tripart assessment survey. Seminars in Hematology,
Pattern of association over time. Oncology Nursing Forum, 26, 909915.
34(Suppl. 2), 412.
Magnan, M.A. (2001). Self-care and health in persons with cancer-related
Walsh, D. (1897). Deep tissue traumatism from roentgen ray exposure. BMJ,
fatigue: Refinement and evaluation of Orem's self-care framework. Unpub-
2, 272273.
lished doctoral dissertation, Wayne State University, Detroit, MI.
Maxwell, M.B. (1984). When the cancer patient becomes anemic. Cancer
Nursing, 7, 321326.
McCall, R.B., & Appelbaum, M.J. (1991). Some issues of conducting second-
ary analyses. Developmental Psychology, 27, 911917.
For more information . . .
McCorkle, R., & Young, K. (1987). Development of a symptom distress scale.
Cancer Nursing, 1, 373378.
McNair, D.M., Lorr, M., & Droppleman, L.F. (1971). EDITS manual for the
ASTRO Nursing Corner
Profile of Mood States. San Diego: Education and Industrial Testing Service.
www.astro.org/about/nursing_frame.html
Miller, D.C. (1991). Handbook of research design and social measurement
(5th ed.). Newbury Park, CA: Sage.
CancerFatigue.org
Mock, V. (Speaker). (1997). An exercise intervention for management of fa-
www.cancerfatigue.org
tigue and emotional distress during radiotherapy treatment for breast can-
Radiotherapy 4.35
cer [cassette recording]. Pittsburgh, PA: Oncology Nursing Society.
www.radiotherapy.com
Mock, V., Dow, K.H., Meares, C.J., Grimm, P.M., Dienemann, J.A.,
Haisfield-Wolfe, M.E., et al. (1997). Effects of exercise on fatigue, physical
Links can be found using ONS Online at www.ons.org.
functioning, and emotional distress during radiation therapy for breast can-
cer. Oncology Nursing Forum, 24, 9911000.
MAGNAN VOL 30, NO 2, 2003
E39