Identifying Clinically Meaningful Fatigue with the Fatigue Symptom Inventory
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
480 Journal of Pain and Symptom Management Vol. 36 No. 5 November 2008
Original Article
Identifying Clinically Meaningful Fatigue
with the Fatigue Symptom Inventory
Kristine A. Donovan, PhD, Paul B. Jacobsen, PhD, Brent J. Small, PhD,
Pamela N. Munster, MD, and Michael A. Andrykowski, PhD
Health Outcomes and Behavior Program (K.A.D., P.B.J., B.J.S) and Breast Cancer Program (P.N.M.),
Moffitt Cancer Center & Research Institute, Tampa, Florida; Department of Psychology (P.B.J.) and
School of Aging Studies (B.J.S.), University of South Florida, Tampa, Florida; and Department of
Behavioral Science (M.A.A.), University of Kentucky College of Medicine, Lexington, Kentucky, USA
Abstract
The Fatigue Symptom Inventory has been used extensively to assess and measure fatigue in
a number of clinical populations. The purpose of the present study was to further establish its
utility by examining its operating characteristics and determining the optimal cutoff score for
identifying clinically meaningful fatigue. The MOS 36-Item Short Form Vitality scale,
a measure widely used to identify individuals with significant fatigue-related disability, was
used to determine the sensitivity and specificity of the Fatigue Symptom Inventory. Results
indicate that a score of 3 or greater on those items assessing fatigue in the past week is the optimal
cutoff score for identifying clinically meaningful fatigue. Individuals who scored at or above the
cutoff also reported significantly greater fatigue interference, more days of fatigue on average,
and fatigue a greater proportion of each day in the past week. Findings suggest that the Fatigue
Symptom Inventory can be used to discriminate effectively between individuals with and
without clinically meaningful fatigue. J Pain Symptom Manage 2008;36:480e487.
Ó 2008 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved.
Key Words
Fatigue, Fatigue Symptom Inventory
Introduction sclerosis,2 and psychiatric disorders such as de-
pression.3 Among adult cancer patients, fa-
Fatigue is generally defined as a sense of per-
tigue is often the most common symptom
sistent tiredness or exhaustion that is often dis-
reported.4e6 Fatigue also is common in the
tressing to the individual. It is a common
general population.7,8 One epidemiological
symptom of many diseases, including cancer,1
study of working adults found that 98% re-
neurological disorders such as multiple
ported some degree of fatigue and one in
five reported substantial fatigue.9
Fatigue is a subjective phenomenon and is
This work was supported by National Cancer Insti- thus assessed most accurately by individual
tute Grant R01 CA82822.
self-report. To this end, researchers have pub-
Address correspondence to: Kristine A. Donovan, PhD,
lished a plethora of self-report instruments
Health Outcomes and Behavior Program, H. Lee
Moffitt Cancer Center & Research Institute, 12902 designed to assess and measure fatigue. A re-
Magnolia Drive, MRC-PSY, Tampa, FL 33612, USA. cent survey of fatigue measurement scales
E-mail: kristine.donovan@moffitt.org published between 1975 and 2004 identified
Accepted for publication: December 4, 2007. a total of 71 scales focusing specifically on
Ó 2008 U.S. Cancer Pain Relief Committee 0885-3924/08/$esee front matter
Published by Elsevier Inc. All rights reserved. doi:10.1016/j.jpainsymman.2007.11.013Vol. 36 No. 5 November 2008 Fatigue Symptom Inventory 481 fatigue used in 416 studies.10 The information Individual Strength,20 and on disease-specific obtained via these measures depends on the measures such as the Bath Ankylosing Spondyli- developer’s conceptualization of fatigue and tis Disease Activity Index.21 Although there is the respondents’ interpretation of the ques- not an accepted standard for the assessment of tions being asked.11 The utility of any one scale fatigue, the SF-36 Vitality scale is commonly rests ultimately on its reliability and validity. A used to validate instruments designed to assess review by Dittner et al.11 of 30 published fatigue in the general population and in patient fatigue scales noted that many fatigue scales samples (see, e.g., Kleinman et al.22). Thus, re- have been published without basic data about searchers have suggested that using the SF-36 their reliability or evidence of sensitivity to Vitality scores of the general population as refer- change. Further, few scales have demonstrated ence data is a valid approach for establishing an ability to discriminate clinical cases of fa- cutoff scores on measures of fatigue.21 To in- tigue from noncases, with acceptable levels of dicate significant health-related limitations, sensitivity and specificity.11 That is, few scales previous studies23e25 dichotomized the Vitality have established cutoff scores to determine scale based on the 25th percentile. That is, clinically meaningful fatigue. individuals scoring at or below the 25th percen- The Fatigue Symptom Inventory (FSI), first tile were considered to be experiencing limita- published in 1998,12 has been used extensively tions due to fatigue while those scoring above to assess fatigue, especially among cancer the 25th percentile were not considered to be patients. Its psychometric properties were orig- suffering such limitations. Once the optimal inally established in women undergoing treat- FSI cutoff score was identified, we sought to ex- ment for breast cancer, women who have plore whether interference related to fatigue, completed treatment for breast cancer, and the duration of fatigue, and demographic fac- women with no history of cancer.12 It was fur- tors differentiated individuals who scored above ther validated in a study of males and females or below this cutoff score. with a variety of different cancer diagnoses.13 The scale has been used since to assess fatigue in a number of clinical populations including Methods breast cancer patients,14 patients undergoing Participants hematopoietic stem cell transplantation,15 Participants were recruited as part of a larger hepatocellular cancer patients undergoing study comparing quality of life in women being stereotactic radiotherapy,16 and patients with treated for early stage breast cancer and chronic fatigue syndrome.17 The FSI has women with no history of cancer. Eligibility cri- proven to be a valid and reliable measure of teria for women with no history of cancer were fatigue in medically ill patients and healthy in- that they must (a) be within five years of the dividuals, and reviewers have suggested that it age of the breast cancer patient to whom is a useful tool for the assessment of fatigue.11 they would be matched in the larger study; The purpose of the present study was to fur- (b) reside within the same zip code as the ther establish the usefulness of the FSI by patient to whom they would be matched; (c) examining its operating characteristics and have no discernable psychiatric or neurologi- determining the optimal cutoff score for iden- cal disorders that would interfere with study tifying clinically meaningful fatigue. To accom- participation; (d) be able to speak and read plish this, we recruited a relatively large sample standard English; (e) report no history of of women with no history of cancer who com- cancer (other than basal cell skin carcinoma) pleted both the FSI and the MOS 36-Item or other potentially life-threatening diseases Short Form Vitality scale (SF-36).18 We used (e.g., AIDS); and (f) report no history of a receiver operating characteristic (ROC) curve condition in which fatigue is a prominent analyses of FSI scores to determine the optimal symptom (e.g., multiple sclerosis or chronic FSI cutoff score relative to the established SF- fatigue syndrome). 36 Vitality scale. ROC analysis has been used previously to establish cutoff scores on general Procedure measures of fatigue including the Schedule of Potential participants were identified using Fatigue and Anergia19 and the Checklist a database maintained by Marketing Systems
482 Donovan et al. Vol. 36 No. 5 November 2008
Group, Inc. (Fort Washington, PA) that draws which fatigue interfered with their general ac-
from all listed telephone households in the tivity, ability to bathe and dress, normal work
United States and is estimated to include activity, ability to concentrate, relations with
demographic and contact information for ap- others, enjoyment of life, and mood (FSI inter-
proximately two-thirds of the U.S. population. ference); participants’ ratings of the number
For each patient who completed the six-month of days in the past week (0e7) they felt
assessment in the larger study, up to 25 women fatigued (FSI days); and participants’ ratings
who resided in the same zip code and were of what percent of each day (0e100), on aver-
within five years of the patient0 s age were age, they felt fatigued in the past week (FSI
selected randomly from the database. One of percent).
these women was selected at random and The Acute (past week) Version of the MOS
sent a letter of introduction describing the 36-Item Short Form18,26 (SF-36) is a widely
study. If this woman did not opt out by calling used self-report measure designed to assess per-
a toll-free telephone number or returned ceived health and functioning. The instrument
a postcard expressing interest in the study, consists of eight scales: Physical Functioning,
telephone contact was initiated to further Role-Physical; Bodily Pain; General Health;
determine eligibility. If she met all eligibility Vitality; Social Functioning; Mental Health;
criteria and verbally agreed to participate, an and Role-Emotional. Each scale is standardized
appointment was set up to obtain written in- on a 0e100 metric, with higher scores indicat-
formed consent and conduct an assessment. ing better functioning. Analyses focused on
If the first woman selected could not be the Vitality scale, which consists of four items
reached, was ineligible, refused to participate, assessing how much of the time in the past
or did not keep the appointment, another week participants felt ‘‘full of pep,’’ had ‘‘a lot
woman on the list was selected randomly until of energy,’’ felt ‘‘worn out,’’ and felt ‘‘tired.’’
a woman matched to the patient was recruited The latter two items are reverse coded prior
and completed the assessment. to scoring. Responses range from ‘‘all of the
time’’ to ‘‘none of the time.’’ In analyses fo-
Measures cused on the operating characteristics of the
Demographic data were obtained via a stan- FSI, participants were classified as fatigued if
dardized self-report questionnaire. Variables their Vitality scale score was less than or equal
assessed were age, race/ethnicity, marital sta- to 45. This score corresponds to the 25th per-
tus, annual household income, educational centile for females in the U.S. general popula-
level, height, weight, and menopausal status. tion,18 and is consistent with previous
The FSI12 is a 14-item measure that assesses research demonstrating that the 25th percen-
the frequency and severity of fatigue and its tile is the most appropriate dichotomous indi-
perceived interference. The measure includes cator of health-related limitations.23 Although
three items specific to fatigue severity in the previous research has demonstrated that a score
past week. Participants rate on 11-point scales of 50 is indicative of biologic and psychologic
(0 ¼ not at all fatigued, 10 ¼ as fatigued as I differences in fatigue,27e32 we chose the more
could be) their level of fatigue: (a) on average stringent score of 45 as the criterion to increase
in the past week (FSI average), (b) on the day the robustness of our results.
they felt most fatigued in the past week (FSI
most), and (c) on the day they felt least
fatigued in the past week (FSI least). A compos- Results
ite fatigue score (FSI composite) was derived by Demographic Characteristics
calculating the average across the three severity The demographic characteristics of the sam-
items. This composite fatigue score showed ple are presented in Table 1. The mean age of
high internal consistency (alpha ¼ 0.84). Anal- the women was 56 years (range, 28e79). The
yses focused on the operating characteristics of vast majority was white, married, and nearly
the FSI average score and FSI composite score. half had a college degree. More than two-thirds
Analyses also were conducted using partici- had annual household incomes $$40,000. The
pants’ average rating of the degree (0 ¼ no average body mass index was 27 and 72% of the
interference, 10 ¼ extreme interference) to women were postmenopausal.Vol. 36 No. 5 November 2008 Fatigue Symptom Inventory 483
Table 1 Table 3
Demographic Characteristics of the Sample Frequency Distribution of FSI Composite Scores
(n ¼ 265) Score Frequency % Cumulative %
Characteristic n (%)
0 27 10.2 10.2
Age in years (mean SD) 56.34 9.42 >0 # 1 47 17.7 27.9
>1 # 2 54 20.4 48.3
Race/ethnicity >2 # 3 48 18.1 66.4
White 252 (95.1) >3 # 4 41 15.5 81.9
Nonwhite 13 (4.9) >4 # 5 25 9.4 91.3
Marital status >5 # 6 11 4.2 95.5
Married or marriage-like 184 (69.4) >6 # 7 8 3.0 98.5
Not married 81(30.6) >7 4 1.5 100.0
Education
College degree 126 (47.5)
Less than college degree 139 (52.5)
overall discriminative accuracy of these items
relative to the established cutoff score for the
Household income
45.18,23 The ROC curves are
graphic representations of the trade-off be- 1
tween the sensitivity (true-positive rate) and
specificity (true-negative rate) for every possi-
ble cutoff score on FSI average and FSI com- 0.8
posite. The area under the curve (AUC) in FSI average = 3
each ROC curve provides an estimate of the
Sensitivity
0.6
Table 2
Frequency Distribution of FSI Average Scores 0.4
Score Frequency % Cumulative %
0 43 16.2 16.2 0.2
1 65 24.5 40.8
2 48 18.1 58.9
3 40 15.1 74.0 0
4 31 11.7 85.7 0 0.2 0.4 0.6 0.8 1
5 18 6.8 92.5 1 - Specificity
6 7 2.6 95.1
7 6 2.3 97.4 Fig. 1. Receiver operating characteristic curve anal-
8 6 2.3 99.6 ysis comparing FSI average scores with established
10 1 0.0 100.0
Vitality cutoff score of >45.484 Donovan et al. Vol. 36 No. 5 November 2008
1 FSI average cutoff score. Similarly, none of
the demographic characteristics were associ-
0.8
ated with the FSI composite cutoff score of 3.
FSI composite = 3
Relation of Fatigue $3 Cutoff Score
Sensitivity
0.6 to Fatigue Interference
Analyses of variance indicated that women
0.4 who scored above the FSI average cutoff re-
ported significantly greater FSI interference
compared to women who scored below the cut-
0.2
off (2.29 1.80 vs. 0.41 0.58, P < 0.0001).
Similarly, women who scored above the FSI
0 composite cutoff reported significantly greater
0 0.2 0.4 0.6 0.8 1
fatigue interference compared to women who
1- Specificity
scored below the cutoff (2.31 1.82 vs. 0.42
Fig. 2. Receiver operating characteristic curve anal- 0.61, P < 0.0001).
ysis comparing FSI composite scores with estab-
lished Vitality cutoff score of >45. Relation of Fatigue $3 Cutoff Score
to Fatigue Duration
25th percentile of the Vitality scale. On FSI Analyses of variance also indicated that the
composite, it yielded a sensitivity of 0.81 and FSI average cutoff score of 3 was significantly
specificity of 0.70 relative to the 25th percen- associated with differences in both FSI days
tile of the Vitality cutoff score. Other cutoff and FSI percent. Women who scored above
scores yielded less optimal results. For exam- the FSI average cutoff reported that they felt
ple, a cutoff score of 4 on FSI average yielded fatigued an average of 4.11 1.97 days in
a sensitivity of 0.62 and a specificity of 0.83 the past week vs. 1.56 1.65 days for women
relative to the 25th percentile of the Vitality below the cutoff (P < 0.0001). Compared to
scale. On FSI composite, it yielded a sensitivity women below the cutoff, women above the cut-
of 0.56 and specificity of 0.83 relative to the off also reported significantly greater FSI per-
25th percentile of the Vitality cutoff score. cent; they felt fatigued a significantly greater
proportion of the day in the past week: an
Relation of Fatigue $3 Cutoff Score average of 36.9% vs. 14.0%, (P < 0.0001).
to Demographic Characteristics Similar results were obtained for the FSI
Chi-squared analyses and analysis of variance composite cutoff. Compared to women below
were conducted to explore the relation of the the cutoff, women above the cutoff reported
FSI average and FSI composite cutoff score of significantly more days of fatigue on average:
3 to demographic characteristics. As shown in 4.06 2.03 vs. 1.6 1.70 (P < 0.0001). Women
Table 5, none of the demographic characteris- above the cutoff also reported that they felt fa-
tics assessed were related significantly to the tigued a significantly greater proportion of the
day in the past week: an average of 37.5% vs.
14.2% (P < 0.0001).
Table 4
Correspondence of FSI Average and FSI
Composite with the Vitality Scale of the SF-36 Relation of Fatigue $3 Cutoff Score to Vitality
SF-36 Vitality Scale Frequency (%) Finally, analysis of variance was conducted to
examine whether there were differences in the
> 45 # 45 Vitality continuous score between women below
FSI average a
and above the FSI average and FSI compositeVol. 36 No. 5 November 2008 Fatigue Symptom Inventory 485
Table 5
Relation of the FSI Average and Composite Cutoff Score of 3 to Demographic Characteristics
FSI Average FSI Composite486 Donovan et al. Vol. 36 No. 5 November 2008
We compared the FSI to the SF-36 Vitality transplantation. Bone Marrow Transplant 1999;24:
scale, a measure that has been widely used 1121e1129.
with both healthy and medically ill populations 5. Baker F, Denniston M, Smith T, West MM. Adult
and has established norms for identifying indi- cancer survivors: how are they faring. Cancer 2005;
viduals with significant fatigue. In addition, we 104:2565e2576.
used statistical methods appropriate for the 6. Fox SW, Lyon DE. Symptom clusters and quality
identification of an optimal cutoff score. The of life in survivors of lung cancer. Oncol Nurs
current study also has several noteworthy limi- Forum 2006;33:931e936.
tations. Only women were included in the 7. Loge JH, Ekeberg O, Kaasa S. Fatigue in the
study sample and the majority was peri- or general Norwegian population: normative data
and associations. J Psychosom Res 1998;45:53e65.
postmenopausal. There also was limited diver-
sity within the sample with respect to ethnicity, 8. Watt T, Groenvold M, Bjorner JB, et al. Fatigue
education, and socioeconomic status. Thus, in the Danish general population. Influence of soci-
odemographic factors and disease. J Epidemiol
the operating characteristics of the FSI cutoff Community Health 2000;54:827e833.
score are unknown in men and in minority
9. Bültmann U, Kant I, Kasl SV, Beurskens AJ, van
populations and low-literacy populations of den Brandt PA. Fatigue and psychological distress in
women. Finally, the finding that a cutoff score the working population: psychometrics, prevalence,
of 3 on FSI average or the FSI composite mea- and correlates. J Psychosom Res 2002;52:445e452.
sure yielded the optimal combination of sensi- 10. Hjollund NH, Andersen JH, Bech P. Assessment
tivity and specificity was not cross-validated in of fatigue in chronic disease: a bibliographic study
a second sample of individuals. Findings that of fatigue measurement scales. Health Qual Life
a similar cutoff score was obtained in another Outcomes 2007;5:12e17.
sample of healthy individuals would increase 11. Dittner AJ, Wessely SC, Brown RG. The assess-
confidence in our findings. ment of fatigue: a practical guide for clinicians
In conclusion, the present study further es- and researchers. J Psychosom Res 2004;56:157e170.
tablishes the usefulness of the FSI by determin- 12. Hann DM, Jacobsen PB, Azzarello LM, et al.
ing that an FSI average or FSI composite score Measurement of fatigue in cancer patients: develop-
of 3 or greater is indicative of clinically mean- ment and validation of the Fatigue Symptom Inven-
tory. Qual Life Res 1998;7:301e310.
ingful fatigue. This cutoff score yielded the op-
timal sensitivity and specificity relative to the 13. Hann DM, Denniston MM, Baker F. Measure-
ment of fatigue in cancer patients: further valida-
widely used SF-36 Vitality scale. The cutoff tion of the Fatigue Symptom Inventory. Qual Life
score also classified effectively those individ- Res 2000;9:847e854.
uals with greater fatigue-related interference
14. Donovan KA, Jacobsen PB, Andrykowski MA,
and fatigue duration. These findings support et al. Course of fatigue in women receiving chemo-
the continued use of the FSI not only as therapy and/or radiotherapy for early stage breast
a means of assessing fatigue but also as a means cancer. J Pain Symptom Manage 2004;28:373e380.
of distinguishing those individuals with clini- 15. Hacker ED, Ferrans CE. Ecological momentary
cally meaningful fatigue. assessment of fatigue in patients receiving intensive
cancer therapy. J Pain Symptom Manage 2007;33:
267e275.
16. Lai YH, Shun SC, Hsiao YL, et al. Fatigue expe-
References riences in hepatocellular carcinoma patients during
six weeks of stereotactic radiotherapy. Oncologist
1. Prue G, Rankin J, Allen J, Gracey J, Cramp F. 2007;12:221e230.
Cancer-related fatigue: a critical appraisal. Eur J Can-
cer 2006;42(7):846e863. 17. Siegel SD, Antoni MH, Fletcher MA, et al.
Impaired natural immunity, cognitive dysfunction,
2. MacAllister WS, Krupp LB. Multiple sclerosis-- and physical symptoms in patients with chronic
related fatigue. Phys Med Rehabil Clin N Am fatigue syndrome: preliminary evidence for a sub-
2005;16:483e502. group. J Psychosom Res 2006;60:559e566.
3. Baldwin DS, Papakostas GI. Symptoms of fatigue 18. Ware JE. SF-36 Health Survey: Manual and inter-
and sleepiness in major depressive disorder. J Clin pretation guide. Boston, MA: The Health Institute,
Psychiatry 2006;67:9e15. New England Medical Center, 1993.
4. Andrykowski MA, Cordova MJ, Hann DH, et al. 19. Hadzi-Pavlovic D, Hickie IB, Wilson AJ, et al.
Patients’ psychosocial concerns following stem cell Screening for prolonged fatigue syndromes:Vol. 36 No. 5 November 2008 Fatigue Symptom Inventory 487
validation of the SOFA scale. Soc Psychiatry HIV in 2003: results from the national ANRS-EN12--
Psychiatr Epidemiol 2000;35:471e479. VESPA Study. AIDS 2007;21(Suppl 1):S19eS27.
20. Bültmann U, de Vries M, Beurskens AJHM, 26. Ware JE. The SF-36 Physical and Mental Health
et al. Measurement of prolonged fatigue in the Summary Scales: A user’s manual. Boston, MA: The
working population: determination of a cutoff point Health Institute: New England Medical Center,
for the Checklist Individual Strength. J Occup 1994.
Health Psychol 2000;5:411e416.
27. Bower JE, Ganz PA, Desmond KA, et al. Fatigue
21. Dagfinrud H, Vollestad NK, Loge JH, Kvien TK, in breast cancer survivors: occurrence, correlates,
Mengshoel AM. Fatigue in patients with ankylosing and impact on quality of life. J Clin Oncol 2000;
spondylitis: a comparison with the general popula- 18:743e753.
tion and associations with clinical and self-reported
measures. Arthritis Rheum 2005;53:5e11. 28. Bower JE, Ganz PA, Aziz N, Fahey JL. Fatigue
and proinflmammatory cytokine activity in breast
22. Kleinman L, Zodet MW, Hakim Z, et al. Psycho- cancer survivors. Psychosom Med 2002;64:604e611.
metric evaluation of the Fatigue Severity Scale for
use in chronic hepatitis C. Qual Life Res 2000;9: 29. Bower JE, Ganz PA, Aziz N, Fahey JL, Cole SW.
499e508. T-cell homeostatis in breast cancer survivors with
persistent fatigue. J Natl Cancer Inst 2003;95:
23. Rose SR, Koshman ML, Spreng S, Sheldon R. 1165e1168.
Statistical issues encountered in the comparison of
health-related quality of life in diseased patients 30. Bower JE, Ganz PA, Aziz N. Altered cortisol re-
to published general population norms: problems sponse to psychologic stress in breast cancer survi-
and solutions. J Clin Epidemiol 1999;52(5): vors with persistent fatigue. Psychosom Med 2005;
405e412. 67:277e280.
24. Torres MS, Calderón SM, Dı́az IR, et al. Health-- 31. Bower JE, Ganz PA, Dickers SS, et al. Diurnal
related quality of life in coronary heart disease com- cortisol rhythm and fatigue in breast cancer survi-
pared to norms in Spanish population. Qual Life vors. Psychoneuroendocrinology 2005;30:92e100.
Res 2004;13:1401e1407. 32. Bower JE, Ganz PA, Desmond KA, et al. Fatigue
25. Préau M, Marcellin F, Carrieri MP, et al. Health-- in long-term breast carcinoma survivors. Cancer
related quality of life in French people living with 2006;106:751e758.You can also read
