Fatigue in gynaecological cancer patients: a pilot study.
G. Prue, J. Rankin, F. Cramp, J. Allen, J. Gracey
G. Prue, J. Gracey, Ph.D, Prof J. Allen, Ph.D
Health and Rehabilitation Sciences Research Institute, University of Ulster, Shore
Road, Newtownabbey, Co. Antrim, N. Ireland, BT37 0QB
F. Cramp, Ph.D
Faculty of Health & Social Care, University of the West of England, BS16 1DD
J. Rankin
Belvoir Park Hospital, Hospital Road, Belfast, N. Ireland, BT8 8JR
Address for correspondence:
Dr. J. Gracey
Health and Rehabilitation Sciences Research Institute
University of Ulster
Shore Road
Newtownabbey
Co. Antrim
N. Ireland BT37 0QB
UNITED KINGDOM
Tel:
028 9036 8284
Fax:
028 9036 8914
E-mail:
jh.gracey@ulster.ac.uk
Abstract
Rationale: Fatigue is a frequent complaint of women with cancer. However, the
incidence of fatigue has not been well studied, in particular gynaecological cancer,
which despite its prevalence has received minimal investigation.
Goals of work: The study aims were (1) to explore the symptoms experienced in a
gynaecological cancer population, primarily fatigue, and (2) to determine the
acceptability of a fatigue questionnaire for use in a longitudinal survey.
Patients and Methods: Over the course of one month, women with gynaecological
cancer attending a Regional Cancer Centre completed a demographic and symptom
questionnaire, and the Multidimensional Fatigue Symptom Inventory-Short Form
(MFSI-SF).
Main Results: Of the 32 individuals approached, 30 agreed to participate (mean age
= 61; most common treatment received was surgery followed by chemotherapy n =
11, mean time from commencement of treatment (including surgery) = 3 months). All
participants completed the MFSI-SF. Tiredness was the most commonly reported
symptom, experienced by 90% of subjects and the most frequently stated worst
symptom, reported by 23.3%. Furthermore, 23 of 27 subjects reported that tiredness
interfered completely with their daily living. The MFSI-SF mean total fatigue score
was 14.4 (SD 15.9), ranging from –15 to 50. The possible total fatigue score ranges
from –24 to 96.
Conclusion: Despite the heterogeneous nature of the group, all participants completed
the MFSI-SF. The study suggests that fatigue could be a problem for this population
group. Thus a longitudinal survey utilizing the MFSI-SF to investigate the
phenomenon further would appear feasible and justified.
Keywords: gynaecological, cancer, fatigue, fatigue measurement
1
Introduction
Cancer-related fatigue (CRF) is an abstract concept that is predominately a subjective
experience. It is one of the most recurrent complaints of people with cancer, having a
profound effect on the whole person (1) and directly influencing the desire to continue
with treatment (6). The aetiology remains to be fully elucidated, however it is known
to differ from the fatigue experienced by the general population, which is a protective
response to physical and psychological stress (1).
A lack of consensus exists regarding the optimal means of assessing CRF (4). To date,
there is no universally accepted standard (7); nevertheless, as CRF is a
multidimensional experience, it should be assessed using a multidimensional measure
(5). For the current pilot study, the Multidimensional Fatigue Symptom InventoryShort Form (MFSI-SF) was selected. The MFSI-SF is short, does not assume the
presence of fatigue and makes no reference to a specific illness (11). Therefore it is
ideal for use a future longitudinal survey investigating the prevalence of fatigue in
gynaecological cancer patients, a population group that has received minimal
investigation in this area.
The aims of the pilot study were (1) to explore the incidence and magnitude of fatigue
in a gynaecological cancer population, (2) to test the acceptability and feasibility of
the MFSI-SF in a gynaecological cancer population, as previous use of the MFSI-SF
has largely involved breast cancer patients (2,10,11) and (3) to chart the symptoms
experienced by this population, to inform the choice of symptom assessment tool for
use in the longitudinal survey.
Patients and Methods
Ethical approval was obtained from the Belfast City Hospital Trust and the University
of Ulster Research Ethics Committee.
Patients
Over the course of one month, a convenience sample of women diagnosed with
gynaecological cancer, attending the Belfast City Hospital Trust for surgery,
2
chemotherapy and/or radiotherapy on an inpatient or outpatient basis were approached
via their medical Consultant.
Eligible participants were female, aged 18 years or over, fully aware of their diagnosis
and able to read and understand English. All participants gave written, informed
consent prior to study entry. Participants with cognitive impairment or incompetence,
with co-morbidities that would interfere with study participation, those who had
received anti-neoplastic treatment for a previous cancer diagnosis, and those
participating in other research studies were excluded. Individuals with a longstanding
history of depression, that is, subjects who had been medically diagnosed with
depression before their cancer diagnosis were also excluded. This decision was made
as a diagnosis of depression may affect the individual’s interpretation of their cancer
experience, as they come upon it with an already depressed attitude. Those
participants that were diagnosed with depression following their cancer diagnosis
were not excluded, as this depression could be associated with having cancer, and can
be considered as part of the cancer symptom experience.
Procedure
For this cross-sectional study, subjects who met the eligibility criteria and provided
informed consent completed two questionnaires; the investigator developed
demographic and symptom experience questionnaire and a fatigue questionnaire; the
MFSI-SF.
Subject’s medical records were accessed to gain information on their tumour site and
stage, relevant past medical history, drug history, haemoglobin levels and treatment
received. For chemotherapy, cycle dates and drug regime were recorded, for
radiotherapy, type, dates, dose and fractionation were documented.
Questionnaires
The investigator developed questionnaire obtained details from three areas. Firstly,
demographic information, that is, age, marital status and employment status.
Secondly, participants were asked to give details of any previous cancer diagnosis,
any other relevant past medical history and their current medication. Finally, each
subject was asked to record the symptoms they had experienced and associated with
3
the ordeal of cancer, and the effect these symptoms had on their daily living (Table 1).
They were asked to rate on a 5 point Likert scale ranging from strongly agree to
strongly disagree if a symptom interfered completely with their daily living. The
decision to use an investigator developed questionnaire was made to ensure all
possible symptoms were recorded. This was facilitated by the inclusion of an open
question inviting participants to list any other symptoms that they associated with
having cancer. This will be used to inform the choice of an appropriate validated
symptom assessment tool in a future longitudinal survey.
Fatigue assessment
A number of the available multidimensional fatigue scales are limited by their length
and complicated response formats (11). To address this limitation the authors of the
MFSI developed a shorter fatigue scale, the MFSI-SF (10,11). The MFSI consists of
both rationally and empirically derived subscales, with each item measured on a five
point Likert scale. The MFSI-SF consists of the 30 items of the empirically derived
subscales of general fatigue (GF), physical fatigue (PF), emotional fatigue (EF),
mental fatigue (MF) and vigor (V). The respondent is required to indicate the extent
to which they have experienced each symptom during the preceding week (0 = not at
all; 4 = extremely) (11). Ratings are summed to obtain scores for each of the
subscales detailed previously. In addition, a total fatigue (TF) score can be generated
by totalling the four fatigue subscales and subtracting from this the vigor scale (11).
The analysis of the MFSI has been undertaken on a sample of breast cancer patients
(9). This showed the empirically derived subscales have good internal consistency
(GF 0.96, EF 0.93, PF 0.85, MF 0.90, V 0.88) and test-retest correlations ranging
from 0.51 – 0.64 (0 to 3-4 weeks) and from 0.60 – 0.70 (6-8 weeks) (10). This
satisfied the criteria for test-retest reliability considering that fatigue fluctuates on a
regular basis. Construct validity has been verified, as has convergent and divergent (8)
and concurrent validity (10). This psychometric analysis was further verified in a
larger mixed sample, which focused specifically on the MFSI-SF (11).
The MFSI-SF is a brief, valid, reliable measure that provides information on the
overall level of fatigue and the extent to which an individual is experiencing fatigue in
each of the five domains listed (11).
4
Statistical Analysis
Data analysis was performed using the Statistical Package for the Social Sciences
(SPSS) Version 11 for Windows. The small size and heterogeneous nature of the
sample resulted in demographic, clinical and symptom data summarized using
descriptive statistics. The MFSI-SF was scored as described above.
Results
Of the 32 participants who were eligible to participate, two declined due to emotional
reasons, one participant had suffered a family bereavement; the second had just
undergone surgery and become aware of her diagnosis. Complete data was collected
from 30 participants.
Information about demographic characteristics of the subjects are presented in Table
2. The mean age of the sample was 60.9 years (SD 13.5) ranging from 31 years to 84
years. The majority of participants were married (56.7%). Regarding employment
status, 28 of the 30 subjects were not working, with 8 of this 28 citing their cancer as
the reason for not working.
The most common gynaecological malignancy was endometrial carcinoma (40%).
The sample was heterogeneous as regards anti-neoplastic treatment received, thus at
the time of questionnaire completion, subjects were at a variety of stages in their
cancer management. The majority of participants had received surgery and were in
the process of receiving radiotherapy (n = 8), chemotherapy (n = 11) or chemotherapy
and radiotherapy (n = 3). Three had just received surgery, and five were undergoing
anti-neoplastic treatment with no surgery: radiotherapy (n = 1), chemotherapy (n = 2)
and chemotherapy and radiotherapy (n = 2). Surgical procedures consisted of radical
hysterectomy and pelvic node dissection (n = 2) with oophrectomy (n = 1); total
abdominal hysterectomy and bilateral salpingo-oophrectomy (n = 9) with
omentectomy (n = 5); radical or partial vulvectomy and bilateral groin node excision
(n = 1 respectively) and omentectomy and oophrectomy (n = 2). Three had undergone
an examination under anaesthetic and one subject a hemicolectomy. The
chemotherapy
regimes
administered
were
Doxorubicin/Cisplatin
(n
=
3),
5
Carboplatin/Taxol (n = 9), Doxorubicin/Ifosfamide (n = 1) and weekly Cisplatin (n =
5). For radiotherapy the total dose was 45 Gy (n = 9), 42 Gy (n = 1) and 50.4 Gy
followed by internal radiation of 24 Gy (n = 1), for 3 participants inadequate detail on
radiotherapy dose was provided.
The average time from commencement of
intervention (including surgery) was 3.39 months (SD 2.54), ranging from one week
to 10 months.
Symptom experience
Two participants stated that they did not associate any symptoms with the
gynaecological cancer experience.
Tiredness was the most common symptom reported and the most frequently reported
worst symptom. Other common symptoms and regularly reported worst symptoms are
listed in Table 3. Participants were also given the opportunity to list any other
symptoms they associated with the cancer experience. These were pins and needles in
the extremities (n = 3); depression (n = 1); increased appetite (n = 1); palpitations (n =
1); staining (n = 1); breathlessness (n = 1); bladder problems (n = 1); lack of taste (n =
1) and shock (n = 1).
Impact on daily living
This section was irrelevant for two participants who had not experienced any
symptoms up to this stage in their cancer experience. One participant completed the
psychological questions only (anxiety and feelings of sadness). Therefore 27
participants completed the physical symptom questions, with 28 completing the
psychological symptom questions.
23 of the 27 participants (85%) agreed that tiredness interfered completely with their
daily living; this was the highest level of agreement across any of the symptoms. The
incidence of other listed symptoms reported to interfere completely with daily living
are detailed in Table 3.
Emotional aspect
Throughout their cancer experience, a higher number of participants felt they had not
experienced anxiety (n = 16) or feelings of sadness (n = 20) than those that had.
6
Furthermore, 22 of the 28 (79%) that completed this section disagreed that anxiety
interfered with their daily living, regarding feelings of sadness 23 of the 28 (82%)
disagreed.
MFSI-SF
The MFSI-SF was self-assessed by participants who rated their fatigue experience
over the past week. There was a 100 percent completion for all items of the MFSI-SF.
The observed mean scores and range for TF and each fatigue subscale are presented in
Table 4. The mean score for TF was 14.4 (SD 15.9) ranging from –15 to 50.
Comparisons of the TF and fatigue subscales for surgery versus no surgery and for
undergoing chemotherapy and/or radiotherapy versus awaiting chemotherapy and/or
radiotherapy are presented in Table 5 and Table 6 respectively.
The mean TF score for those who had undergone surgery was 11.3 (SD 14.7), for
those who had not undergone surgery the mean was 29.8 (SD 13.4). In all of the
subscales the non-surgical group had higher fatigue scores.
The mean TF score for those currently receiving chemotherapy and/or radiotherapy
was 13.0 (SD 16.1), in comparison to 26.7 (SD 7.0) for those awaiting chemotherapy
and/or radiotherapy. The subscale scores for GF and PF were similar for these two
groups (Table 6), for EF and MF, the group awaiting chemotherapy and/or
radiotherapy had higher mean scores (EF 10.7 compared with 4.3 and MF 5.3
compared with 2.7).
Discussion
In this cross sectional study, the symptom experience of 30 women diagnosed with
gynaecological cancer was examined. Special attention was given to CRF, including
the application of the MFSI-SF, to determine if a longitudinal survey to investigate
the phenomenon further using this assessment tool is feasible and justified. Indeed,
tiredness was the most common symptom reported by the subjects, the most
frequently reported worst symptom and the largest majority of subjects agreed that it
interfered completely with their daily living. This data adds to past research that
7
indicates that people with cancer perceive fatigue as a real and prevalent problem (12)
and underscores the high prevalence of the symptom, warranting further research in
this population.
No attempt was made to recruit a homogenous sample of participants with respect to
tumour type, stage, or treatment regime. With regards to stage of intervention,
including surgery, participants ranged from one week to ten months of treatment.
Recruitment of a heterogeneous group aimed to ensure the MFSI-SF was a suitable
measure for use with participants at various treatment stages, thus providing some
support for its use in a longitudinal survey. Despite the heterogeneous nature of the
group, reflected by the wide range of MFSI-SF scores reported, all participants were
able to complete the MFSI-SF, thus this measure appears to be a user-friendly tool for
use with this population.
The possible subscales scores for the MFSI-SF range from 0 – 24 with the possible
TF range from –24 – 98. Therefore although there was a wide range of scores shown
for each of the subscales, no subject scored the maximum fatigue score. The wide
range in scores reflects the heterogeneous nature of the sample. Subjects varied with
respect to tumour site, stage and treatment, thus their total fatigue score had high
variability, however there were no extreme scores. When the scores are compared
with the mean scores of healthy volunteers (10), the gynaecological individuals scored
higher in all subscales, except in the mental fatigue subscale. This is a very crude
comparison as it is unknown if the two groups are comparable as they are from two
separate studies.
Those participants who had yet to commence chemotherapy and/or radiotherapy had a
higher fatigue score than those undergoing treatment, as a result of high EF and MF
subscale scores. This suggests that individuals might be anxious while awaiting
treatment (4). Moreover, high EF and MF scores may reflect the fact that women with
gynaecological cancer do suffer more on an emotional than on a physical level
possibly indicating a coping issue. These hypotheses would require further
investigation in a larger, methodologically sound longitudinal survey, with adequate
documentation of pertinent variables that would have an effect on fatigue such as time
since surgery, weight loss and depression. Further investigation into the higher fatigue
8
scores reported by the non-surgical group, possibly due to advanced tumour stage (14)
and treatment progression (18) is also warranted.
Inconsistencies were found in the results with regards to the prevalence and impact of
symptoms. Loss/lack of appetite and gastrointestinal problems were the second and
third most commonly experienced symptoms, yet the majority of participants
disagreed that these symptoms interfered with their daily living. This may indicate
that although these symptoms are common, participants can deal with them on a dayto-day basis. Interestingly the accepted common symptoms of pain, nausea and
vomiting were not prevalent and did not impact on the subjects’ daily living. This
may indicate that these symptoms are anticipated and thus well managed (3). In
contrast, the high prevalence of fatigue may suggest that individuals may not know
how to manage their fatigue effectively.
Limitations
The small sample size and cross sectional nature of the survey are limitations of the
study. The small sample and the heterogeneity of the treatment received prevented
meaningful statistics from being performed; furthermore, the small sample reduced
the ability to generalize the results to the population as a whole. Moreover, the ‘oneoff’ administration of the MFSI-SF does not provide the full spectrum of the
participants’ fatigue. The results reflect only the fatigue experienced during the
preceding week. However the aim of this study was to determine if fatigue was a
problem in this population group, and confirm the MFSI-SF as an acceptable measure
for use with this population. This in turn provides some justification for a longitudinal
survey to chart fully the fatigue experienced by a gynaecological cancer population.
This survey will incorporate, for comparison purposes, a group of healthy female
volunteers. This is a more effective method of investigating the fatigue experience, as
although costly in terms of time and resources, longitudinal studies provide more
information on the true manifestation of fatigue (5). Moreover, as fatigue is a
common complaint among the general population this must be taken into account
when reporting the prevalence of this symptom among cancer patients (13).
Conclusions
9
The 100% completion rate would indicate that the MFSI-SF is an acceptable measure
for the study of patients with gynaecological cancer. Furthermore it would appear that
fatigue is a prevalent problem in this population group as in spite of the variance in
time since initiation of treatment and the type of treatment received, 90% of the
participants experienced fatigue that they associated with having cancer. The exact
cause of this fatigue remains to be elucidated. Therefore further research into fatigue
in this population group using the MFSI-SF is warranted. Finally, the use of an
investigator-developed symptom questionnaire enabled symptoms experienced by this
population to be documented, to inform the choice of an appropriate validated
symptom assessment tool for future research in this population.
Acknowledgement(s)
The authors would like to thank the staff of the Belfast City Hospital Trust for their
help during data collection, and those who took part in the study for their time and
effort.
10
References
1. Alhberg K Ekman T Gaston-Johansson F Mock V (2003) Assessment and
management of cancer-related fatigue in adults. Lancet 362: 640-650
2. Broeckel JA Jacobsen PB Horton J Balducci L Lyman GH (1998)
Characteristics and correlates of fatigue after adjuvant chemotherapy for
breast cancer. J Clin Oncol 16: 1689-1696
3. Curt GA (2001) Fatigue in cancer. BMJ 322: 1560
4. Jacobsen PB Hann DM Azzarello LM Horton J Balducci L Lyman GH (1999)
Fatigue in women receiving adjuvant chemotherapy for breast cancer:
characteristics, course and correlates. J Pain Symptom Manage 19: 14 - 27
5. Jacobsen PB Thors CL (2003) Fatigue in the radiation therapy patient: current
management and investigations. Semin Radiat Oncol 13: 372-380
6. Morrow GR Andrews PLR Hickok JT Roscoe JA Matheson S (2001) Fatigue
associated with cancer and its treatment. Support Care Cancer 10: 389-398
7. National Cancer Institute (2003). Fatigue [online], National Cancer Institute.
Available from: http://www.cancer.gov/ [Accessed 8 July 2003]
8. Schwartz AH (2002) Validity of cancer-related fatigue instruments.
Pharmacotherapy 22: 1433-1441
9. Steginga SK Dunn J (1997) Women’s experiences following treatment for
gynecologic cancer. Oncol Nurs Forum 24: 1403-1408
10. Stein KD Martin SC Hann DM Jacobsen PB (1998) A multidimensional
measure of fatigue for use with cancer patients. Cancer Pract 6: 143-152
11. Stein KD Jacobsen PB Blanchard CM Thors C (2004) Further validation of
the multidimensional fatigue symptom inventory – short form. J Pain
Symptom Manage 27: 14-23
12. Stone P Richardson A Ream E Smith AG Kerr DJ Kearney N (2000) Cancerrelated fatigue: inevitable, unimportant and untreatable? Results of a
multicentre patient survey. Ann Oncol 11: 971-975
13. Stone P Richards M A’Hern R Hardy J (2001) Fatigue in patients with cancers
of the breast and prostate undergoing radical radiotherapy. J Pain Symptom
Manage 22: 1007-1015
14. Stone P (2002) The measurement, causes and effective management of cancerrelated fatigue. International Journal of Palliative Nursing 8: 120-128
15. Visovsky C Schneider SM (2003) Cancer-related fatigue. Online Journal of
Issues in Nursing 8:8
11
Table 1 Symptoms listed in investigator developed questionnaire
Symptoms
Physical
Tiredness
Pain
Nausea and vomiting
Fever/sweats/hot flushes
Gastrointestinal problems
Lymphodema
Itching
Loss/lack of appetite
Sleep problems
Psychological
Anxiety
Feelings of sadness
12
Table 2 Patient Demographics
N
Tumour type
Cervix
Vagina
Vulva
Uterus
Ovary
6
1
3
12
8
1
2
3
4
Unknown
8
6
9
5
2
Radiotherapy
Chemotherapy
Surgery
Surgery, radiotherapy, chemotherapy
Surgery, radiotherapy
Surgery, chemotherapy
Radiotherapy, chemotherapy
Time from commencement of
treatment
≤ 1 month
≤ 2 months
≤ 3 months
≤ 4 months
≤ 5 months
≤ 6 months
≤ 7 months
≤ 8 months
≤ 9 months
≤ 10 months
1
2
3
3
8
11
2
Tumour stage
Treatment received
6
6
5
5
3
1
1
0
1
2
13
Table 3 Symptom experience and impact on daily living
Symptom
N
Tiredness
Loss/lack of appetite
GI problems
Nausea and vomiting
27
20
18
16
Tiredness
Unable to differentiate
Pain
7
6
3
Tiredness
Nausea and vomiting
Loss/lack of appetite
Sleep problems
GI problems
Pain
Fever/sweats/hot flushes
Itching
23
11
9
8
7
5
5
2
Most common symptom experienced?
Worst symptom experienced?
Did symptom interfere completely with daily
living?
14
Table 4 Observed MFSI-SF scores
Total fatigue score
General fatigue score
Physical fatigue score
Mental fatigue score
Emotional fatigue score
Vigor score
n
30
30
30
30
30
30
Range
Minimum
Maximum
-15
50
2
22
0
14
0
10
0
18
0
18
Mean
14.37
10.80
3.50
2.97
4.93
7.83
SD
15.91
5.96
4.07
3.00
5.00
5.15
15
Table 5 Mean MFSI-SF scores for surgery versus no surgery
Group 1 (Surgery)
n = 25
Group 2 (No Surgery)
n= 5
11.28
14.71
29.80
13.44
10.08
5.66
14.40
6.80
2.88
3.90
6.60
3.78
2.64
2.77
4.60
3.91
3.96
4.44
9.80
5.22
8.28
5.42
5.60
2.97
Total fatigue score
Mean
SD
General fatigue score
Mean
SD
Physical fatigue score
Mean
SD
Mental fatigue score
Mean
SD
Emotional fatigue score
Mean
SD
Vigor score
Mean
SD
16
Table 6 Mean MFSI-SF scores for undergoing treatment versus still to start
treatment
Group 1 (receiving
treatment)
n = 27
Group 2 (awaiting
treatment)
n=3
13.00
16.10
26.67
7.02
10.78
6.03
11.00
6.56
3.44
3.95
4.00
6.08
2.07
2.83
5.33
4.16
4.30
4.76
10.67
3.51
8.22
5.26
4.33
2.08
Total fatigue score
Mean
SD
General fatigue score
Mean
SD
Physical fatigue score
Mean
SD
Mental fatigue score
Mean
SD
Emotional fatigue score
Mean
SD
Vigor score
Mean
SD
17