Women Experience Higher Levels of Fatigue Than Men at the End of Life: A Longitudinal Home Palliative Care Study

Transcription

1 Vol. 33 No. 4 April 2007 Journal of Pain and Symptom Management 389 Original Article Women Experience Higher Levels of Fatigue Than Men at the End of Life: A Longitudinal Home Palliative Care Study Amna F. Husain, MD, MPH, Kay Stewart, MD, Rita Arseneault, RN, Rahim Moineddin, PhD, Victor Cellarius, MD, S. Lawrence Librach, MD, and Deborah Dudgeon, MD Temmy Latner Centre for Palliative Care (A.F.H., V.C., S.L.L.), Mount Sinai Hospital, and Division of Family and Community Medicine (A.F.H., R.M., V.C., S.L.L.), University of Toronto, Toronto, Ontario, Canada; Sheffield Macmillan Unit for Palliative Care (K.S.), Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, UK; Hospital for Sick Children (R.A.), University of Toronto, Toronto, Ontario, Canada; and Palliative Medicine & Supportive Care Program (D.D.), Kingston Regional Cancer Center, Queen s University, Kingston, Ontario, Canada Abstract Few studies have evaluated sex differences in the prevalence, severity, and correlates of fatigue at the end of life. The Brief Fatigue Inventory, McGill Quality of Life (MQOL) Questionnaire, and Karnofsky Performance Scale were administered at two-week intervals to 102 patients in a home palliative program. Outcomes in the sample and a regional palliative database (n ¼ 3,096) were analyzed. Cancer was the diagnosis in 96% of patients enrolled. Prevalence (P ¼ ) and severity of fatigue (P < 0.001) were higher in women at entry and in a repeated measures analysis over time (severity, P ¼ ). Performance status did not explain this difference. MQOL scores were inversely correlated to fatigue (Spearman coefficient ¼ 0.48, P < ), but did not differ by sex. There was no difference in fatigue interference with MQOL in women and men. Although depression was higher in women (P ¼ 0.042) and related to fatigue at entry, it did not explain the sex difference in fatigue scores. Of the sociodemographic variables examined, neither education nor living situation contributed to the fatigue difference. This study shows a sex effect in the fatigue experienced by patients with advanced illnesses, which is not explained by baseline differences in performance, depression, MQOL, education, or living situation. That fatigue interference with MQOL is the same for men and women suggests that higher fatigue scores in women reflect not only a difference in the dimension of fatigue severity, but are also relevant in relation to impact on QOL. Assessment of fatigue should include the dimension of QOL important for both women and men. J Pain Symptom Manage 2007;33:389e397. Ó 2007 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved. This study was supported by a grant from the Physicians Services Incorporated (PSI) Fund and the Frankfort Fund, Mount Sinai Hospital, Toronto, Ontario, Canada. Address reprint requests to: Amna F. Husain, MD, MPH, Temmy Latner Centre for Palliative Care, Mount Ó 2007 U.S. Cancer Pain Relief Committee Published by Elsevier Inc. All rights reserved. Sinai Hospital, 60 Murray Street, Toronto, Ontario, M5T 3L9, Canada. amna.husain@ utoronto.ca Accepted for publication: September 16, /07/$esee front matter doi: /j.jpainsymman

2 390 Husain et al. Vol. 33 No. 4 April 2007 Key Words Fatigue, gender, sex, palliative care, end of life, symptom interference, quality of life Introduction Fatigue is a highly prevalent symptom in men and women at the end of life, regardless of the primary diagnosis. 1e4 The burden of the symptom may be devastating, negatively affecting their quality of life (QOL), 5,6 and encroaching on precious time. In current practice, a majority of adult patients in palliative care programs have a cancer diagnosis, 7 while small proportions have diagnoses of AIDS, neurodegenerative disease, or heart, renal, or hepatic failure. Although some elements of fatigue are disease specific and treatment specific, there is an increasing recognition of the common aspects of the fatigue experienced by patients with different life-threatening diagnoses. 8 Similar to cancer-related fatigue, the fatigue at the end of life associated with HIV disease, heart failure, renal failure, and chronic obstructive lung disease is highly prevalent, highly distressing, multidimensional, and multifactorial. 7,9e11 The study of sex or gender difference in fatigue is an important strategy in improving our understanding of this common symptom. A report at the 2004 National Institutes of Health State-of-the-Science Conference on symptom management in cancer identified the need for sex or gender studies in symptom assessment and management and specifically called for longitudinal studies of fatigue prevalence and severity by sex groups. 12 The goal of studies in this area is to test whether clinically relevant sex differences exist in fatigue and to elucidate the underlying mechanisms. 13 Ultimately, this work is relevant if there are implications for preventive, diagnostic, therapeutic, and policy strategies. This study aimed to explore sex differences in fatigue at the end of life. We hypothesized that a sex difference in fatigue exists and that sex-specific correlates underlie this difference. In this work, we use the terms sex and gender as proposed by Torgrimson and Minson. 14 Sex is used when referring to the male or female phenotype, while gender is used to refer to behaviorally and psychosocially determined maleness and femaleness. Methods Procedures This three-month longitudinal study drew from new referrals to a home palliative care program serving a major urban center over a 17-month period. The sampling method used was to identify and screen all patients newly admitted by the palliative care team. Patients had to be admitted to the program to be eligible to participate. Based on the admission assessment, the clinician could deem patients ineligible to be approached for the study. Patients aged 18 years or older who were able to complete the self-report questionnaires in English were included. Patients were excluded for cognitive impairment based on clinical assessment and the Bedside Confusion Scale. 15 A research associate visited patients in their homes to obtain informed consent and for subsequent data collection. The research associate was trained in standardized administration of the questionnaires and performance status measurement. Blood was collected at entry, but patients had the option of refusing. The laboratory analyses included hemoglobin; white blood cell count; platelets; sodium; potassium; chloride; serum albumin; and renal, liver, and thyroid function tests. Questionnaires were administered at 0, 1, 3, 5, 7, 9, and 11 weeks. The disposition of patients was followed for 12 months after completion of the study. At each data collection visit, a list of all current medications was recorded. Total daily opioid dose was calculated for the 24-hour period prior to the visit using a morphine equivalency table. 16 Measures The Brief Fatigue Inventory (BFI) is a nineitem, self-report, symptom severity and functional interference scale that was validated in an oncology patient population. 17 It measures

3 Vol. 33 No. 4 April 2007 Sex Differences in Fatigue at End of Life 391 the single domain of fatigue severity. Cutoff levels for mild, moderate, and severe fatigue on the BFI were determined by comparing the worst level of fatigue item with patients self-reported interference with function. A BFI score of >3 was used as a cutoff to identify moderate-to-severe fatigue. 17 The McGill Quality of Life (MQOL) Questionnaire was developed and validated in patients receiving palliative care for advanced illnesses. 18 Scores range from 0 to 10 and higher scores mean better QOL. The MQOL is multidimensional and includes physical, psychological, support, and existential domains. 19 The physical domain consists of an item reporting overall symptom distress and three items asking patients to list three symptoms experienced over the previous two days, each scored on a 0e10 scale ranging from no problem to tremendous problem. The physical domain of the MQOL may overlap the BFI and, therefore, was controlled for in the analysis. The MQOL has been shown to distinguish between good, fair, and bad days for groups of patients receiving palliative care. 20 The Karnofsky Performance Scale is a performance status assessment tool that is widely validated and used across many medical disciplines, including advanced cancer. 21e25 The Karnofsky Performance Scale was chosen for its proven psychometric properties. It was not feasible to use a multi-item depression tool due to patient burden. Therefore, depression was measured by the single-item screen, Are you depressed?, as initially validated by Chochinov et al. in a palliative care population. 26 This depression screen has a dichotomous yes or no response. The Bedside Confusion Scale, a validated delirium screening tool developed in an advanced cancer population, was used for cognitive impairment or delirium. 15 Sample Size A sample size calculation was performed using the mean (4.7) and standard deviation (2.8) of the BFI (scale 0e10) in a previously reported oncology population. 17 Since a twounit difference in BFI corresponds to a oneunit difference in performance status (Eastern Cooperative Oncology Group), 17 this was used as a clinically relevant effect. Assuming 0.05 type I error, the required sample size for detecting at least a two-unit difference in the mean of BFI with 90% power is n ¼ 86 in the overall sample, with n ¼ 43 in each sex group. Statistical Analyses The relationship of relevant variables to sex was explored through univariate analyses first, and then through multivariate regression analyses to adjust for covariates. Interference of fatigue with QOL by sex group was examined using regression analysis and a best fit, nonparametric curve to plot mean QOL for each value of BFI score for males and females. A repeated measures analysis was performed to study the relationship of fatigue to sex over time. Survival analysis, including a Cox proportional hazards model, was performed. An analysis of the palliative program s administrative database for the period of time that the study was conducted yielded information to determine if the study sample was representative of the patient population we serve. The study was approved by the Research Ethics Board of the University of Toronto. Results Patients were recruited from new referrals to a home palliative care program serving the Greater Toronto Area, Ontario, Canada. Of 3,096 new referrals on whom administrative data are available, 2,137 were admitted to the program during the recruitment period of the study. Commonly reported reasons for the clinician identifying a patient as ineligible were patient too ill, cognitive impairment, language barrier, and psychosocial and family-related problems that preclude participation. Of 600 subjects identified as potentially eligible during the recruitment period, 102 agreed to be enrolled. The reasons cited for refusal to participate were too much to undertake study with everything else that is going on, feel too weak to participate, and have too many medical appointments already. Table 1 reports patient profile information for the study sample and the population it was drawn from. Over 96% of the study population had a cancer diagnosis, while the remainder had diagnoses of chronic obstructive pulmonary disease, neurodegenerative diseases, HIV,

4 392 Husain et al. Vol. 33 No. 4 April 2007 or heart, renal, or liver failure. Men and women in the sample had similar demographic profiles with regard to language fluency and education, but differed in their living situation. Table 1 also reports on disposition and attrition rates. Attrition rates were equally distributed by sex. At entry (Week 0), results showed that the prevalence of moderate-to-severe fatigue, as defined by a BFI score of >3, 17 was significantly higher in women (P ¼ ). In univariate analysis, fatigue severity at this time was also significantly higher in women than Table 1 Patient Profiles By Sex in Study Sample and Palliative Database Study Patients Palliative Database (During Study Period) N 102 3,096 Mean age (SD) 67.7 (12.3) 69.3 Male Female P-value Male Female Group sample size (%) 47 (46) 55 (54) 1,553 (50.2) 1,543 (49.8) Mean age (SD) 69.8 (11.9) 65.9 (12.3) ns Social support: home support Lives alone (%) 14 (29.8) 14 (25.5) 178 (11) 235 (15) Lives with partner (%) 28 (59.6) 24 (43.6) N/A N/A Lives with other (%) 5 (10.6) 17 (30.9) N/A N/A Fluent in English (%) 45 (95.7) 54 (98.2) ns 484/629 (77) 487/611 (80) Education <High school 6 (12.8) 8 (14.6) 95/409 (23.2) 86/401 (21.4) High school 13 (27.7) 25 (45.5) 122/409 (29.8) 151/401 (37.7) Post-secondary 28 (59.6) 22 (40.0) 190/409 (46.5) 164/401 (40.9) Completed study: Week (44.7) 31 (58.5) N/A N/A Dropout: illness progression or death (%) N/A N/A Week (3.6) Week 1 3 (6.4) 5 (9.4) Week 3 7 (14.9) 5 (9.4) Week 5 3 (6.4) 4 (7.6) Week 7 4 (8.5) 4 (7.6) Week 9 3 (6.4) 2 (3.8) Week 11 6 (12.8) 2 (3.8) Disposition/# data available (%) Death at home 16/46 (34.7) 13/51 (25.5) 425/1,075 (39.5) 424/977 (43.4) Death in acute care setting 13/46 (28.3) 10/51 (19.6) 403/1,075 (37.5) 338/977 (34.6) Death in PCU or long-term care unit 12/46 (26.1) 15/51 (29.4) 193/1,075 (18.0) 172/977 (17.6) Lost to follow-up 1/47 (2.1) 4/55 (7.6) 130/1,205 (10.8) 170/1,147 (14.8) Diagnoses (%) Non-cancer 3 (6.4) 1 (1.8) 224 (14.4) 272 (17.6) All Cancer 44 (93.6) 54 (98.2) 1,329 (85.6) 1,271 (82.4) Breast 0 18 (32.7) (16.3) Prostate 7 (14.9) (10.4) 0 Median survival in weeks N/A N/A Median Karnofsky Score 70 (60e70) 60 (60e70) ns N/A N/A (interquartile range) a Depressed (%) b 18 (39.1) 32 (61.5) N/A N/A Anemia: Hb <11 mmol/l (n of sample who consented to blood work) 15 (29) 11 (31) ns N/A N/A Prevalence moderate-severe fatigue c 29 (63.0) 46 (86.8) N/A N/A Mean fatigue scoredbfi score (SD) 3.91 (2.21) 5.51 (2.02) N/A N/A Mean Quality of Life scoredmqol (SD) 7.03 (1.09) 6.82 (1.22) ns N/A N/A Radiotherapy or chemotherapy in the 2 14 (29.8) 14 (25.9) ns N/A N/A weeks prior to data collection Mean daily opioid dose by morphine equivalency in mg d ns N/A N/A a Karnofsky Performance Scale. b Single-item screen. c BFI score >3. d Ref. 16, p. 41.

5 Vol. 33 No. 4 April 2007 Sex Differences in Fatigue at End of Life 393 in men (P # 0.001). This difference was maintained over time, as illustrated in Fig. 1. We analyzed the correlates of fatigue at Week 0 that may explain this difference. Performance status, as measured by the Karnofsky Performance Scale, was not distributed differently by sex (Wilcoxon Rank Sum, P ¼ 0.13). Correlational analysis at Week 0 showed that fatigue and MQOL scores (Spearman coefficient ¼ , P < ) were correlated in the overall sample and in both sex groups. However, MQOL did not differ by sex at any time point or longitudinally. Depression was related to fatigue scores in the overall sample at Week 0 (P ¼ ) and was more prevalent in women (P ¼ 0.042), but it did not explain the difference in fatigue scores by sex. As the blood sampling was optional, there were limited laboratory data available. For those who had data available (n ¼ 60), the prevalence of anemia was distributed equally between men and women and was not related to fatigue scores. Likewise, no significant relationship was found with other laboratory findings. Fatigue interference with MQOL is illustrated in Fig. 2, which is a plot of mean QOL scores by values of BFI score by sex. There was no significant difference in the interference curves by sex. Regression analysis revealed that, for the overall sample, each unit difference in BFI score resulted in a 1.6 unit difference in MQOL score (parameter estimate ¼ 1.6, P ¼ ). Multivariate analyses at Week 0 showed the contributions made by covariates to fatigue scores overall and in each sex group. Based on the univariate analyses and previously published data on relevant covariates, 2e4,27e35 the Mean BFI Score male female Week Fig. 1. Fatigue scores by sex. Mean fatigue scores (BFI) were significantly higher in women than men at entry (Week 0, P # 0.001) and over time. Mean MQOL Score BFI Score Fig. 2. Fatigue interference with QOL by sex. Women and men had similar levels of fatigue interference with QOL (P ¼ ns). covariates included in the regression model were total MQOL score, sex, age, depression, living situation, psychological domain of the MQOL (MQOLsub3), physical symptom domain of the MQOL (MQOLsub1), performance status, recent chemotherapy or radiotherapy, and daily opioid dose. Depression, found to be related to fatigue by sex in the univariate analysis, was no longer significant in the multivariate model. Sex (P ¼ ) and total MQOL (P ¼ 0.012) were the only significant covariates. A regression analysis in each sex group was performed on the relationship of fatigue scores to a limited number of covariates: total MQOL, MQOLsub1, performance status, depression, and total opioid dose. Total opioid dose (P ¼ 0.017) and physical symptoms (MQOLsub1, P ¼ ) were related to fatigue scores in women but not in men (Table 2). The covariates in this model account for 54% (R-squared ¼ 0.544) of the variation in fatigue scores in women but only 22% (R-squared ¼ 0.218) in men. To analyze the relationship of fatigue to sex and other covariates longitudinally, we used repeated measures ANOVA. The sex difference seen in adjusted fatigue scores at baseline was maintained over the duration of the study (P ¼ ). In this model, total MQOL score (P # ) and depression (P ¼ ) were significant when sex was controlled. There were no additional increases in fatigue change scores by sex over the study period. Survival curve analysis showed a survival advantage to being female (P ¼ )

6 394 Husain et al. Vol. 33 No. 4 April 2007 despite no significant difference in performance status by sex at Week 0 (Fig. 3). However, in a Cox proportional hazards model adjusting for covariates, performance status (P ¼ 0.013, hazard ratio 0.97, confidence interval ¼ 0.94e0.99) was the only significant predictor of survival. Discussion This is the first study focusing on a population with advanced illness receiving palliative care at home that shows that women experienced a higher prevalence of fatigue and greater fatigue severity. The difference was maintained over the three-month study period, and was not explained by baseline differences in performance status, depression, or QOL. To our knowledge, it is also the first study to compare fatigue interference with Survival Distribution Function Table 2 Gender-Specific Regression Analyses at Week 0 Variable Females Parameter Estimate P- Value Males Parameter Estimate P- Value MQOL (total score) Depression (not depressed) MQOLsub (physical domain subscale) Karnofsky Performance Status Total daily opioid dose For the objective of detecting trends and hypothesis generation, statistical significance defined as P # Bold values indicate statistically significant values by stated criteria. STRATA: Gender=Female Gender=Male Censored Gender=Female Censored Gender=Male Weeks to Death Fig. 3. Survival curves by sex. Women had a significant survival advantage over men (Wilcoxon rank sum, P ¼ ). QOL in men and women in a palliative population. Exploratory univariate analyses suggested that depression and living situation may be variables relevant to the relationship between sex and fatigue. However, these variables were not found to be significant in multivariate analyses. The domain of the MQOL that captures physical symptom burden was significantly related to fatigue in women in both univariate and multivariate analyses. This is likely explained by the overlap in reporting of fatigue between this MQOL domain and the BFI. Although there was no absolute difference in mean opioid dose between men and women, total opioid dose was related to fatigue in women in the multivariate analysis at Week 0. Although no causal relationship can be inferred, this association raises the question that differences in opioid dosing relative to weight or sex-related differences in susceptibility to side effects may exist. A limitation of this study is its inability to adjust for the effect of diagnosis, and hence specific cancer treatments, due to inherent differences in certain cancers between the sexes. Further disease-specific studies are needed to elucidate the role of the underlying diagnosis and the treatment associated with it on the fatigue reported by men and women. Additionally, since high levels of fatigue are reported in the general population and at least one study from the primary care setting found that women report higher fatigue severity scores than men, 36 comparison with an appropriate nonpalliative population should be included in future studies of fatigue prevalence and severity. Few studies in oncology have explored the issue of sex difference in fatigue. One of these studies with 50 lung cancer patients receiving radiotherapy showed no sex difference in prevalence. 29 Other oncology studies examining fatigue severity have shown either increased severity in women, 28,37e39 or no difference in fatigue severity across sex; it is noteworthy that there are none that report greater fatigue in men. 12,40,41 Beyond the oncology setting, a study in an HIV population showed that women experience higher levels of fatigue. 35 Additionally, a study in outpatients with chronic obstructive pulmonary disease showed that despite women in the group having better

7 Vol. 33 No. 4 April 2007 Sex Differences in Fatigue at End of Life 395 pulmonary function indicators than the men, women had similar levels of fatigue. 42 Of the oncology studies, only two have focused on a palliative cancer population. Stone et al. showed no sex difference in fatigue severity in a sample of 95 palliative inpatients with advanced cancer; however, their study sample had a notably shorter median survival than our home palliative sample. 40 In another, Walsh et al., reporting on a sample of 1,000 oncology inpatients and outpatients who were referred to palliative care, showed no sex difference in fatigue prevalence. 43 It is important to note that the nature of this sample is also different from ours, and their methodology did not include a validated assessment tool. Increasingly, studies are exploring sex differences in muscle fatigability, exercise capacity, disease susceptibility, pharmacokinetics, and patterns of symptom reporting. The underlying mechanisms are conceivably psychological, societal, hormonal, or due to differential genetic susceptibility to disease and treatment. A counterargument is that the sex differences in symptom severity in cancer and other diseases are a reflection of the baseline differences in women s reporting of symptoms in the general population. Although this difference in reporting may be a result of underlying psychosocial mechanisms, this point may be raised to discount the existence of biologic mechanisms. Since fatigue, like pain, is subjective, it is important to examine whether a difference in reporting can or ought to be distinguished from a difference in experience. The analysis performed in this study, which shows similar levels of fatigue interference with QOL in men and women, sheds light on this discussion. Unlike findings in oncology, fatigue did not interfere with performance status in the overall sample and in either sex group. These findings suggest that the domain of QOL is more relevant than performance status when measuring fatigue at the end of life. At any given level of fatigue score, women and men appeared to have similar interference with QOL scores. A larger study is required to validate this finding and to determine cutoff levels for fatigue severity based on interference with QOL in each sex group. Whether the sex difference in fatigue is due to reporting or based on underlying mechanisms, using a scale of symptom severity alone may result in misclassification of symptom cutoff levels. This has relevance to inclusion criteria for symptom research and monitoring of adverse effects. It also has implications for the emerging area of symptom-clusters research, as cutoff levels for inclusion of a symptom in a cluster have yet to be determined. In clinical practice, responding to symptom severity without measuring interference with a global measure of well being may risk overtreatment and adverse effects. Conversely, not responding appropriately to fatigue interference with QOL may result in undertreatment. Therefore, if the sex differences observed in this study are substantiated, practitioners will need to consider the implication that this has on prevention, assessment, and management of fatigue in men and women. Our study concludes that in a sample of patients receiving palliative care at home, women have higher levels of fatigue. Although the analysis suggested some associations between living situation, total opioid dosing, physical symptoms, and women s fatigue, the sex difference was largely unexplained by the physical, psychological, and sociodemographic variables examined. These and an expanded set of covariates need to be examined in a prospective population-based study, with an appropriate healthy control group. References 1. Stone P, Ream E, Richardson A, et al. Cancerrelated fatigueea difference of opinion? Results of a multicentre survey of healthcare professionals, patients and caregivers. Eur J Cancer Care (Engl) 2003;12:20e Stone P, Richards M, A Hern R, et al. A study to investigate the prevalence, severity and correlates of fatigue among patients with cancer in comparison with a control group of volunteers without cancer. Ann Oncol 2000;11:561e Respini D, Jacobsen PB, Thors C, et al. The prevalence and correlates of fatigue in older cancer patients. Crit Rev Oncol Hematol 2003;47:273e Vogelzang NJ, Breitbart W, Cella D, et al. Patient, caregiver, and oncologist perceptions of cancer-related fatigue: results of a tripart assessment survey. The Fatigue Coalition. Semin Hematol 1997; 34:4e12.

8 396 Husain et al. Vol. 33 No. 4 April Krishnasamy M. Fatigue in advanced cancerd meaning before measurement? Int J Nurs Stud 2000;37:401e Lindqvist O, Widmark A, Rasmussen BH. Meanings of the phenomenon of fatigue as narrated by 4 patients with cancer in palliative care. Cancer Nurs 2004;27:237e Anderson H, Ward C, Eardley A, et al. The concerns of patients under palliative care and a heart failure clinic are not being met. Palliat Med 2001;15:279e Solano JP, Gomes B, Higginson IJ. A comparison of symptom prevalence in far advanced cancer, AIDS, heart disease, chronic obstructive pulmonary disease and renal disease. J Pain Symptom Manage 2006;31:58e Bakshi R. Fatigue associated with multiple sclerosis: diagnosis, impact and management. Mult Scler 2003;9:219e Olsson M, Lexell J, Soderberg S. The meaning of fatigue for women with multiple sclerosis. J Adv Nurs 2005;49:7e McCann K, Boore JR. Fatigue in persons with renal failure who require maintenance haemodialysis. J Adv Nurs 2000;32:1132e Miaskowski C. Gender differences in pain, fatigue, and depression in patients with cancer. J Natl Cancer Inst Monogr 2004;32:139e Becker JB, Arnold AP, Berkley KJ, et al. Strategies and methods for research on sex differences in brain and behavior. Endocrinology 2005;146: 1650e Torgrimson BN, Minson CT. Sex and gender: what is the difference? J Appl Physiol 2005;99: 785e Sarhill N, Walsh D, Nelson KA, et al. Assessment of delirium in advanced cancer: the use of the bedside confusion scale. Am J Hosp Palliat Care 2001; 18:335e Librach SL, Squires BP. The pain manual: Principles and issues in cancer pain management. Montreal, Quebec: Pegasus Healthcare International, Mendoza TR, Wang XS, Cleeland CS, et al. The rapid assessment of fatigue severity in cancer patients: use of the Brief Fatigue Inventory. Cancer 1999;85:1186e Cohen SR, Mount BM, Strobel MG, et al. The McGill Quality of Life Questionnaire: a measure of quality of life appropriate for people with advanced disease. A preliminary study of validity and acceptability. Palliat Med 1995;9:207e Cohen SR, Mount BM, Bruera E, et al. Validity of the McGill Quality of Life Questionnaire in the palliative care setting: a multi-centre Canadian study demonstrating the importance of the existential domain. Palliat Med 1997;11:3e Cohen SR, Mount BM. Living with cancer: good days and bad daysdwhat produces them? Can the McGill quality of life questionnaire distinguish between them? Cancer 2000;89: 1854e O Dell MW, Riggs RV, Turner JL, et al. Interrater reliability of the Karnofsky Performance Status in an HIV-infected sample. AIDS 1991;5:1396e Brezinski D, Stone PH, Muller JE, et al. Prognostic significance of the Karnofsky Performance Status score in patients with acute myocardial infarction: comparison with the left ventricular ejection fraction and the exercise treadmill test performance. The MILIS Study Group. Am Heart J 1991;121: 1374e Schaafsma J, Osoba D. The Karnofsky Performance Status Scale re-examined: a cross-validation with the EORTC-C30. Qual Life Res 1994;3: 413e Yates JW, Chalmer B, McKegney FP. Evaluation of patients with advanced cancer using the Karnofsky performance status. Cancer 1980;45:2220e Mor V, Laliberte L, Morris JN, et al. The Karnofsky Performance Status Scale. An examination of its reliability and validity in a research setting. Cancer 1984;53:2002e Chochinov HM, Wilson KG, Enns M, et al. Are you depressed? Screening for depression in the terminally ill. Am J Psychiatry 1997;154:674e Egner A, Phillips VL, Vora R, et al. Depression, fatigue, and health-related quality of life among people with advanced multiple sclerosis: results from an exploratory telerehabilitation study. Neuro- Rehabilitation 2003;18:125e Akechi T, Kugaya A, Okamura H, et al. Fatigue and its associated factors in ambulatory cancer patients: a preliminary study. J Pain Symptom Manage 1999;17:42e Hickok JT, Morrow GR, McDonald S, et al. Frequency and correlates of fatigue in lung cancer patients receiving radiation therapy: implications for management. J Pain Symptom Manage 1996; 11:370e Irvine D, Vincent L, Graydon JE, et al. The prevalence and correlates of fatigue in patients receiving treatment with chemotherapy and radiotherapy. A comparison with the fatigue experienced by healthy individuals. Cancer Nurs 1994;17:367e Lerdal A, Celius EG, Moum T. Fatigue and its association with sociodemographic variables among multiple sclerosis patients. Mult Scler 2003;9: 509e Okuyama T, Tanaka K, Akechi T, et al. Fatigue in ambulatory patients with advanced lung cancer: prevalence, correlated factors, and screening. J Pain Symptom Manage 2001;22:554e Servaes P, Verhagen C, Bleijenberg G. Fatigue in cancer patients during and after treatment:

9 Vol. 33 No. 4 April 2007 Sex Differences in Fatigue at End of Life 397 prevalence, correlates and interventions. Eur J Cancer 2002;38:27e Sullivan PS, Dworkin MS. Prevalence and correlates of fatigue among persons with HIV infection. J Pain Symptom Manage 2003;25:329e Voss JG. Predictors and correlates of fatigue in HIV/AIDS. J Pain Symptom Manage 2005;29: 173e Kroenke K, Wood DR, Mangelsdorff AD, et al. Chronic fatigue in primary care. Prevalence, patient characteristics, and outcome. JAMA 1988;260: 929e Heinonen H, Volin L, Uutela A, et al. Gender-- associated differences in the quality of life after allogeneic BMT. Bone Marrow Transplant 2001;28: 503e Pater JL, Zee B, Palmer M, et al. Fatigue in patients with cancer: results with National Cancer Institute of Canada Clinical Trials Group studies employing the EORTC QLQ-C30. Support Care Cancer 1997;5:410e Redeker NS, Lev EL, Ruggiero J. Insomnia, fatigue, anxiety, depression, and quality of life of cancer patients undergoing chemotherapy. Sch Inq Nurs Pract 2000;14:275e290. discussion 291e Stone P, Hardy J, Broadley K, et al. Fatigue in advanced cancer: a prospective controlled cross-sectional study. Br J Cancer 1999;79:1479e Servaes P, van der Werf S, Prins J, et al. Fatigue in disease-free cancer patients compared with fatigue in patients with chronic fatigue syndrome. Support Care Cancer 2001;9:11e Carter R, Holiday DB, Stocks J, et al. Peak physiologic responses to arm and leg ergometry in male and female patients with airflow obstruction. Chest 2003;124:511e Walsh D, Donnelly S, Rybicki L. The symptoms of advanced cancer: relationship to age, gender, and performance status in 1,000 patients. Support Care Cancer 2000;8:175e179.