Volume 31, Issue 2 , Pages 148-161, February 2006
Supportive Intervention for Fatigue in Patients Undergoing Chemotherapy: A Randomized Controlled Trial
Article Outline
Abstract
This study evaluated a supportive intervention for fatigue in patients undergoing chemotherapy. One hundred three chemotherapy-naïve patients were recruited, stratified by treatment regimen, and randomly allocated to intervention or usual care. The intervention was conducted over three months. Recipients were provided with an investigator-designed information pack and Fatigue Diary that they completed during the week following each treatment. Additionally, support nurses visited them monthly at home. They assessed fatigue, provided psychological support, and coached participants in self-care. The intervention group reported significantly less fatigue (P
<0.05), lower associated distress (P<0.05), and less impact of fatigue on valued pastimes (P<0.05) than the control group. Further, they reported significantly less anxiety (P<0.05) and depression (P<0.05) and displayed more adaptive coping (P<0.05). The intervention enabled patients to adapt to living with fatigue and contributed to their psychological/emotional well-being and ability to cope with their illness and treatment.
Key Words: Cancer, chemotherapy, fatigue, management, intervention, supportive care
Introduction
Fatigue is a symptom commonly experienced by patients during the course of cancer and its treatment. Particularly prevalent in patients undergoing chemotherapy, it is reported to affect 82–100% of such patients.1, 2 The patients describe it as one of their most troublesome symptoms, not least as it compounds symptom distress and impacts greatly on quality of life.3, 4
Historically, fatigue has been poorly managed in this patient group. While it is increasingly acknowledged as a common problem, many health care professionals appear reluctant to treat it. It is perceived as an inevitable consequence of disease and its treatment that patients have to endure.2, 5 Thus, although cancer care teams are becoming more cognizant of the problem that fatigue poses for many patients, it often remains undertreated and poorly managed.6
Patients report that it can be difficult to access information on how to manage fatigue7 and typically resort to resting and sleeping more or curtailing activities in attempting to manage it.4, 8, 9 Unfortunately, such approaches are ineffective and frequently compound their fatigued state.10
Determining the best way to treat cancer-related fatigue and facilitate patients' self-management of it has been the objective of a growing number of published intervention studies. Pharmacological trials have sought to determine the efficacy of a range of treatments including corticosteroids, methylphenidate, donepezil, and erythropoietin.11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 However, all of these, except erythropoietin, have been evaluated predominantly in patients with advanced cancer.
Studies evaluating the effect of prednisolone in advanced cancer are largely inconclusive. One randomized controlled trial reported increased activity levels,11 another documented enhanced quality of life,12 and a further suggested short-term (2 weeks) improvements in self-reported ‘weakness’.13 However, these studies did not specifically investigate subjective fatigue, and thus the impact of prednisolone on this symptom is not clear.
Evidence supporting the use of either methylphenidate or donepezil is more limited because trials into these agents are relatively novel. Pilot studies evaluating the effects of methylphenidate in advanced cancer are encouraging; it appears that it may provide a rapid (within 2 weeks) improvement in fatigue.14, 15 A review of the palliative uses of methylphenidate suggests that it ameliorates opioid-induced somnolence and improves cognitive function.16
Pilot trials have also been conducted with donepezil to determine its effect on sedation and other symptoms in patients receiving opioids.17 Bruera et al.17 suggest that it is a well-tolerated medication with potential to improve fatigue and sedation associated with opioid medication. However, randomized trials are required before the effects of donepezil can be confirmed.
Erythropoietin has been the more extensively tested pharmacological intervention for fatigue. Outcomes of a series of randomized controlled trials18, 19 and large open label studies20, 21 suggest that it is effective in reducing fatigue associated with anemia and enhancing overall quality of life in patients receiving chemotherapy. To date, however, few studies have investigated its use outside this context, and thus its wider role in the management of cancer-related fatigue has not been ascertained.
A number of nonpharmacological interventions for cancer-related fatigue have also been developed and evaluated. Most evidence has been published in support of exercise. The effects of exercise has been tested in women with breast cancer,22, 23 including those receiving either chemotherapy or radiotherapy.24, 25 It has been evaluated in individuals undergoing high-dose chemotherapy prior to autologous peripheral blood stem cell transplantation26 and more recently in people with advanced disease.27, 28 It appears that a regular program of exercise may have a role across the disease trajectory in reducing fatigue. However, compliance with exercise can prove problematic, and it appears that exercise is neither suitable for nor acceptable to all patients.23
More recently, psychoeducational programs have been developed and evaluated. Energy conservation and activity management were the focus of one,29 while others have taken a broader supportive-educative approach.30, 31, 32, 33 A further study has taken a cognitive behavioral approach to managing fatigue and associated symptoms.33, 34 Many were pilot studies with small samples.30, 31, 32 Interventions were provided in groups,30, 32 one-to-one,31 over the telephone,29 or to patients together with their caregivers.33, 34 These suggest that psychoeducational programs may facilitate management of fatigue and reduce its impact on functioning. Further, they describe patients' appreciation of tailored interventions for fatigue management. However, the evidence base is as yet insufficient to determine conclusively the impact of such supportive care on the management of this symptom,35 or to determine the best means of providing this.
This study entailed evaluation of a one-to-one, in person, intervention that aimed to educate and support patients in initiating self-care measures for managing fatigue during chemotherapy. Pilot work had indicated that this approach could offer patients many benefits.31 This paper details outcomes of the evaluation.
Methods
Study Objectives
The research tested four hypotheses. It tested whether the intervention
Study Design
The intervention was implemented in two regional cancer centers in the United Kingdom following ethical approval from the relevant local research ethics committees. It was evaluated through conduct of a randomized, two-arm (standard versus experimental treatment), unblinded, controlled trial.
Blinding is important for reducing bias, but full blinding is not always easy or possible in trials of different styles of patient management or with alternative therapies and other forms of nonpharmacological trial.36, 37 In this study, it was not possible to achieve full blinding. The principal researcher, support nurse providing the intervention, and the patients themselves were aware of the groups subjects were allocated to. However, blindness was maximized as far as possible; the researcher coding and entering the data was unaware of the group subjects had been randomized to, and thus bias in inputting data was minimized.
It may have been possible to blind patients to the treatment they had through delivery of a placebo to the control arm. However, as noted by Boutron et al.37 this would have been unlikely to succeed; patients themselves took part in the intervention and would most likely have been able to determine the treatment they had. Blinding must succeed to reap benefits38 and, given the likelihood of failure in concealing group allocation, was not incorporated into the study design. Further, it is argued that many nonpharmaceutical interventions embody specific effects, such as empathy, that if delivered to the control would lead to underestimation of the treatment effect.39
However, allocation concealment was possible. Many authors believe this to be more important than blinding in reducing bias as it prevents selection bias through shielding those involved in a trial from knowing upcoming assignments in advance.40 Subjects admitted to the trial were stratified according to the center where they received treatment and the chemotherapy regimen they were given. Individuals within these strata were then allocated at random between the intervention and control groups by a computer-generated randomization table that was referred to after individuals had consented to the study.
Participants
A consecutive convenience sample of 90 individuals was sought. This sample was based on power calculations that determined that 45 patients were needed in each study arm to yield 80% power to detect a significant (one-sided α=0.05) reduction in fatigue of 10%. Eligible individuals were aged between 18 and 70 years, chemotherapy-naïve, and had been diagnosed with non-Hodgkin's lymphoma or gastrointestinal, nonsmall cell lung, colorectal, breast, or unknown primary cancer. They were due to commence their first cycle of cytotoxic chemotherapy. These patient populations were selected because the research team had previously studied and was aware of the pattern and impact of fatigue in these patient groups.41 Further, to be eligible they had to understand, speak, read, and write English. Patients were excluded if they were being treated for psychiatric illness. They were recruited from the inpatient or outpatient services prior to commencing their first cycle of treatment, and they provided written informed consent stating their willingness to take part in the trial.
Outcome Measures
Instruments for this study had published evidence of acceptable reliability, validity, factor structure, and sensitivity with patients with cancer. Fatigue was the primary outcome variable, and three characteristics of this symptom were assessed by four visual analogue scales (VASs): subjective quantification of fatigue, subjective distress because of fatigue, and subjective assessment of effects of fatigue on chores/work and on pastimes/hobbies (see Appendix). It has been suggested that these dimensions are salient characteristics to assess when measuring fatigue,42 and previous research by the authors found fatigue VASs to provide valid measures of fatigue that were sensitive to change.31, 41 Emotional well-being was measured by the Hospital Anxiety and Depression (HAD) Scale, a psychometrically sound 14-item scale used frequently in patients with cancer.43 General health status was measured using the SF-36 General Health Status questionnaire44 and coping was measured by a single VAS of perceived general coping and the COPE (shortened version).45 Both the SF-36 and COPE questionnaires have favorable psychometric properties and have been used successfully in populations with cancer. The instruments outlined above comprised the questionnaire battery administered at baseline, and then prior to the second, third, and fourth treatment cycles. On each occasion, subjects were instructed to record their experiences over the previous month.
Intervention
The intervention evaluated in this study comprised four principle components: assessment/monitoring of fatigue; education on fatigue; coaching in self-care; and provision of emotional support (Fig. 1). It aimed to enable patients to manage fatigue through energy conservation and management and to optimize activity and functioning. It was principally underpinned by Winningham's46 Psychobiological Entropy model. This defines fatigue as an energy deficit that, if persistent, leads to a cycle of decreased activity, fatigue, and reduced function, which according to Winningham, results in disability. This model proposes that interventions should aim to either reduce the onset of fatigue (i.e. address the factors giving rise to it, including in this case the disease and chemotherapy administered) or prevent secondary fatigue through achieving an optimal balance between restorative rest and restorative energy.47 The intervention addresses this second aspect.
Fig. 1.
Components of the Beating Fatigue intervention.
The intervention program was provided over the first three treatment cycles. This time period rather than an entire course of treatment was selected because some patients discontinue treatment after three cycles if their response to treatment is poor. Further, the intervention was provided at the beginning of treatment because it aimed to prepare patients in the management of fatigue and facilitate their adoption of self-care activities that would assist its amelioration. Through providing the intervention over the first three treatment cycles, it was envisaged that the intervention would be delivered over the period when fatigue was likely to be developing.
The educational element of the intervention was underpinned by an investigator-designed information pack that was provided to patients prior to chemotherapy. This pack presented information on exercising, balancing activity with rest, prioritizing and delegating activities, dietary supplements, relaxation, diversion, and sleep-enhancement techniques. While for some of these activities there are some data regarding their efficacy,29, 48, 49 most have face value as potentially useful interventions.
Assessment of fatigue was aided by patients' completion of a Fatigue Diary developed for use in this study. Participants completed this diary for the first seven days of each of the three treatment cycles over which the intervention ran. This week-long period was selected because fatigue generally peaks during this phase of the treatment cycle. It was perceived that the Fatigue Diary would enable patients' recognition of periods during their treatment when they were susceptible to fatigue and facilitate their insight into aggravating and relieving factors.
Diary entries were reviewed by support nurses who visited patients at home once during each treatment cycle. They assessed and discussed the extent to which fatigue impeded patients' lives and reviewed the efficacy of self-care patients adopted in an attempt to relieve it. Strategies promoted in the information pack were reviewed at these meetings, and the support nurse coached patients in the use of these strategies. Furthermore, these meetings allowed patients to explore the meaning of fatigue in the lives, their hopes, and future goals. The aim was to tailor the intervention to deal with particular concerns around fatigue that subjects expressed, and help them find the most suitable approach to managing it. However, the intervenor followed a basic protocol to ensure the integrity of the intervention. The support nurse in this study was an experienced cancer nurse with a counseling qualification and knowledge of cancer-related fatigue.
The hypothesized relationship between intervention and outcome tested in the study is presented in Fig. 2. Individuals not in receipt of the intervention received standard care delivered in each center. This did not entail regular or systematic assessment of patients' fatigue. Rather, assessment of this symptom and education regarding its management were provided in an ad hoc manner. Further, when the study was conducted there were no written resources available to patients at either center.
Fig. 2.
Analytical model of the relationship between intervention and outcome.
Statistical Methods
Various statistical tests were performed using the Statistical Package for Social Sciences Computer Program50 to analyze data derived from the outcome measures. Data from the Fatigue and Coping VASs were considered to be interval level and sufficiently normally distributed to allow the use of t-tests to explore between-group differences. The other outcome measures produced ordinal level data that were analyzed utilizing Mann–Whitney tests. All significance levels were set at P<0.05. The data presented in this paper compare the control and intervention at two points: baseline and prior to the fourth treatment cycle. The latter is the primary outcome. The differences over time were investigated and will be reported in a later paper.
In addition to analyzing the four fatigue VAS items independently, a global measure of fatigue (Mean Fatigue Score) was calculated to determine the impact of the intervention on the overall feelings of fatigue. This was calculated through finding the mean of the four VAS items.
There were few missing data (<1%), and where possible with the HAD, SF-36, and COPE scales, missing values were substituted with the average for the subscale that they belonged to. It was not possible to substitute a mean value for the fatigue VASs. However, if three of the four VASs were correctly completed, the Mean Fatigue Score was obtained by calculating the average of the other three.
Results
Recruitment ran for a period of 18 months. During this time, 103 individuals were recruited to the study. The flow of participants through each stage of the study is shown in Fig. 3. Unfortunately, some individuals withdrew from the study as data collection progressed, and thus statistical power was marginally lower than intended. However, despite this, statistically significant results were identified with regard to the primary outcome (fatigue), as detailed below. However, it is possible that other effects of the intervention failed to be identified due to insufficient statistical power.
Fig. 3.
Flow diagram of participants' progress through study.
Baseline Data
Baseline data were collected and analyzed to determine differences between the study groups prior to introduction of the intervention. This revealed no statistically significant differences (Table 1). Patients were typically cohabiting (66%); they were in occupational groups C1/C2 (61%) but during the trial were not working (70%).
Table 1. Sample Demographics
| Experimental Group (n=48) | Control Group (n=55) | All Patients (n=103) | P-Value | |
|---|---|---|---|---|
| Age: mean (SD) | 55.1 (11.4) | 57.8 (8.5) | 56.5 (10.1) | 0.17 |
| Sex | n % | n % | n % | |
| Male | 25 (52) | 32 (58) | 57 (55) | 0.53 |
| Female | 23 (48) | 23 (42) | 46 (45) | |
| Disease group | ||||
| Colorectal | 5 (10) | 5 (9) | 10 (10) | 0.23 |
| Stomach | 5 (10) | 8 (14) | 13 (13) | |
| Esophageal | 7 (15) | 16 (29) | 23 (22) | |
| Pancreatic/cholangio | 4 (8) | 7 (13) | 11 (11) | |
| Lung | 8 (17) | 9 (16) | 17 (16) | |
| NH lymphoma | 2 (4) | 3 (6) | 5 (5) | |
| Breast | 10 (21) | 5 (9) | 15 (14) | |
| Unknown primary | 7 (15) | 2 (4) | 9 (9) | |
| Chemotherapy type | ||||
| CMF | 9 (19) | 5 (9) | 14 (14) | 0.63 |
| CHOP/CEOP | 2 (4) | 3 (5) | 5 (5) | |
| ECF | 24 (50) | 34 (62) | 58 (56) | |
| 5Fu+folinic acid | 5 (10) | 5 (9) | 10 (10) | |
| MVP | 8 (17) | 8 (15) | 16 (15) | |
| Disease status | ||||
| No apparent disease | 18 (37) | 13 (24) | 31 (30) | 0.29 |
| Loco/regional disease | 20 (42) | 26 (47) | 46 (45) | |
| Metastatic disease | 10 (21) | 16 (29) | 26 (25) | |
| Intention of treatment | ||||
| Curative/first line | 20 (42) | 23 (42) | 43 (42) | 0.99 |
| Palliative | 28 (58) | 32 (58) | 60 (58) | |
Primary Outcome Data
Primary outcome data for this study were the measures of fatigue taken prior to the subjects' fourth treatment cycle. It was envisaged that the intervention would require time to attain maximum effect, and these were the final measures taken. Secondary data were those relating to emotional, general health, and coping status at this same time point.
Primary outcome data were favorable (Table 2). They provided evidence that the intervention exerted a statistically significant effect on patients' experiences of fatigue. On average, the experimental group enjoyed a 20% pretest–post-test fall in the different dimensions of fatigue, while the control group reported a negligible fall. The intervention impacted on particular dimensions more than others. It was most potent in lowering distress evoked by fatigue (P<0.01) and reducing its impact on favored pastimes (P<0.02). It proved less able to reduce the extent of fatigue experienced by patients or lower its impact on work and chores. With the latter two dimensions, there were visible between-group differences but these were not of a magnitude to reach significance.
Table 2. Between-Group Differences in Fatigue Dimensions at Baseline and Prior to Treatment Cycle 4
| Study Group | Mean | Standard Deviation | T-Value | P-Value | 95% CI | |
|---|---|---|---|---|---|---|
| Extent of fatigue | ||||||
| Baseline | Experimental | 44.3 | 28.5 | −0.28 | 0.78 | −13.1, 9.8 |
| Control | 45.9 | 29.0 | ||||
| Prior to Cycle 4 | Experimental | 36.4 | 29.5 | −0.88 | 0.19 | −19.2, 7.5 |
| Control | 42.3 | 31.2 | ||||
| Distress caused by fatigue | ||||||
| Baseline | Experimental | 28.4 | 30.1 | −0.84 | 0.40 | −16.7, 6.8 |
| Control | 33.3 | 33.3 | ||||
| Prior to Cycle 4 | Experimental | 24.0 | 27.1 | −2.25 | 0.01b | −27.2, −1.6 |
| Control | 38.4 | 31.0 | ||||
| Disruption to work/chores by fatigue | ||||||
| Baseline | Experimental | 42.8 | 34.5 | −0.38 | 0.71 | −16.1, 11.0 |
| Control | 45.4 | 33.5 | ||||
| Prior to Cycle 4 | Experimental | 33.3 | 31.4 | −1.29 | 0.10 | −22.5, 4.8 |
| Control | 42.1 | 30.8 | ||||
| Disruption to pastimes/hobbies by fatigue | ||||||
| Baseline | Experimental | 39.7 | 37.4 | 0.46 | 0.46 | −19.7, 8.9 |
| Control | 45.1 | 34.3 | ||||
| Prior to Cycle 4 | Experimental | 28.7 | 28.8 | −2.20 | 0.02a | −28.5, −1.5 |
| Control | 43.6 | 32.6 | ||||
| Mean Fatigue Score | ||||||
| Baseline | Experimental | 38.8 | 28.9 | −0.66 | 0.51 | −15.3, 7.7 |
| Control | 42.6 | 28.8 | ||||
| Prior to Cycle 4 | Experimental | 30.6 | 27.7 | −1.74 | 0.04a | −23.6, 1.5 |
| Control | 41.6 | 29.4 | ||||
ap < 0.05. |
bp < 0.01. |
The level of distress evoked by fatigue was typically lower in both study arms than the other three dimensions of fatigue. Although the extent of fatigue and its impact on chores/work and on pastimes/hobbies were relatively high, distress caused by fatigue was consistently lower.
Analysis of the global measure of fatigue (Mean Fatigue Score) confirmed the statistically significant between-group difference (P<0.04). Confidence intervals (95% CI) were calculated and again provided evidence of a beneficial experimental effect. The mean difference in Mean Fatigue Score reported by the two groups on completion of the study was 9.0mm (95%CI, −23.6, 1.5).
Secondary Outcome Data
Secondary outcome data were similarly favorable (Table 3). Levels of anxiety and depression were not dissimilar in the two study arms prior to intervention. However, by the end of the trial, patients in the experimental group reported significantly less anxiety (P<0.05) and depression (P<0.05) than those in the control group. Additionally, it should be noted that although the median levels of anxiety and depression reported by the sample fell within normal limits (subscale scores 0–7), some individuals in both arms of the study reported severe anxiety and/or depression (subscale scores of 15–21).
Table 3. Between-Group Differences in Emotional Well-Being, General Health Status, and Coping
| Study Group | Mean Rank | U-Value | P-Value | |
|---|---|---|---|---|
| Anxiety—HAD Scale | ||||
| Baseline | Experimental | 44.4 | 966.0 | 0.99 |
| Control | 44.5 | |||
| Prior to Cycle 4 | Experimental | 38.0 | 693.0 | 0.03a |
| Control | 47.9 | |||
| Depression—HAD Scale | ||||
| Baseline | Experimental | 40.2 | 781.5 | 0.12 |
| Control | 48.6 | |||
| Prior to Cycle 4 | Experimental | 37.6 | 677.0 | 0.02a |
| Control | 48.3 | |||
| Vitality subscale—SF-36 | ||||
| Baseline | Experimental | 46.3 | 892.0 | 0.53 |
| Control | 42.8 | |||
| Prior to Cycle 4 | Experimental | 47.1 | 695.0 | 0.05a |
| Control | 38.2 | |||
| Mental health subscale—SF-36 | ||||
| Baseline | Experimental | 45.3 | 934.5 | 0.78 |
| Control | 43.8 | |||
| Prior to Cycle 4 | Experimental | 48.4 | 639.0 | 0.02a |
| Control | 36.9 | |||
| Behavioral disengagement—COPE | ||||
| Baseline | Experimental | 44.7 | 833.0 | 0.36 |
| Control | 41.4 | |||
| Prior to Cycle 4 | Experimental | 39.7 | 763.0 | 0.02a |
| Control | 46.3 | |||
| Humor—COPE | ||||
| Baseline | Experimental | 48.7 | 789.0 | 0.13 |
| Control | 40.5 | |||
| Prior to Cycle 4 | Experimental | 51.0 | 568.5 | 0.00b |
| Control | 35.2 | |||
| Coping—Coping VAS | ||||
| Baseline | Experimental | 18.2 | T-value | 0.24 |
| Control | 24.0 | −1.19 | ||
| Prior to Cycle 4 | Experimental | 11.9 | T-value | 0.05a |
| Control | 18.3 | −1.66 | ||
ap < 0.05. |
bp < 0.01. |
The general health status data determined that the experimental group scored more highly than the control group across all subscales and all time points. However, the only dimensions reported here are those where the differences were statistically significant. Notably, by the end of the trial the intervention group reported greater vigor and psychological well-being than the control.
Both groups reported to be coping near to the best of their ability. However, by the end of the study, perceived coping was greater in the experimental group of patients who displayed more adaptive coping than in the control. The specific dimensions of coping, measured by the COPE (short form), where there were statistically significant between-group differences are reported. These data determine that the intervention group used significantly more humor (a coping response associated with less distress and greater well-being51) and significantly less behavioral disengagement strategies relative to the control. Behavioral disengagement is synonymous with giving up or withdrawing effort from attempting to deal with the stressor, in this case with the diagnosis and treatment of cancer.
Discussion
Fatigue is a symptom that is rarely assessed systematically or managed effectively.2, 5, 52 This study introduced a supportive intervention for this symptom that comprised both systematic and detailed assessment, and provision of information and support in its management. Unlike previous studies that delivered these elements within a group intervention30, 32 or by telephone,29 this intervention was provided one-to-one, in person.
Across all measures of fatigue, both VASs and the vitality subscale within the SF-36, the experimental group reported less fatigue by the end of the study than the control. These differences attained significance in the following dimensions: distress associated with fatigue; ability to perform favored pastimes; mean (global) fatigue score, and vitality. Differences were particularly evident with regard to the distress caused by fatigue. Although it appeared that the intervention did not significantly reduce the intensity of the experience, the emotional consequences of the symptom were reduced in those receiving it. Many factors can be advocated in elucidating this finding.
First, given the physiological side effects of treatment, it may have been unrealistic to suppose that a supportive education, like the one evaluated, could lower the intensity of the experience. The limited impact of the intervention on this dimension may therefore be unsurprising.
Second, patients in the experimental arm were encouraged not to view fatigue with distrust or fear, but to consider it a normal and expected side effect of treatment. Distress evoked by a symptom relates not only to the quantity, quality, and duration of the experience,53 but also to the manner in which patients interpret its onset and impact.54 Further, those in the experimental arm appeared to adapt to living with it and it would seem logical that they would find it less disruptive and distressing as a result.
Finally, the support nurse talked with patients about the guilt they may have when surrendering daily tasks to others. This guilt is alluded to in the literature.3 Individuals were encouraged to view delegation of tasks as a method of gaining control over their lives, rather than as a failing.
It is interesting that Holley and Borger30 reported similar reductions in distress associated with fatigue in their pilot study investigating the impact of group education and support in managing cancer-related fatigue. Like the current study, their group sessions provided a sharing and support component. Together, these preliminary findings suggest that supportive intervention, incorporating education in managing cancer-related fatigue alongside sharing and support, serves particularly to diminish the distressing impact of this symptom.
Although results from the current study identified that the intervention program led to statistically significant improvements in individuals' ability to continue with valued pastimes, it determined that it impacted less on their ability to undertake work or chores. It is likely this arose as it is possible to delay or reschedule hobbies but frequently less easy to alter work or household obligations. For some people, these are inflexible and less open to negotiation.
This study also suggests that supportive interventions for fatigue may have benefits other than reducing fatigue. It appeared that over time the experimental group experienced significantly less anxiety and depression and reported significantly different coping strategies; those in the experimental arm used humor more frequently than the control group and reported fewer avoidance strategies. It is not easy, and it may be of limited value, to attempt to determine which elements of the program induced these effects. Yet, the domiciliary meetings provided time when patients' concerns and frustrations could be discussed; it would appear logical that this could directly impact on psychological well-being. Further, through being able to engage in pastimes they valued, it is likely that those in the experimental arm felt more in control, greater self-esteem, and less emotional distress. It would appear that the analytical model of the proposed relationship between intervention and outcome (Fig. 2) was supported by the study data.
Thus, it appeared that the intervention brought statistically significant improvements in certain outcomes, notably dimensions of fatigue. However, it is equally important to consider the degree to which these changes might be clinically important. This can be gauged through reference to estimates of minimal clinically important differences in fatigue that have been calculated for various fatigue scales.55, 56 It has been suggested that improvements of 2.4 points or greater represent a clinically meaningful change when recorded on a 10-point single-item scale where ‘0’ represents ‘no fatigue’ and a score of ‘10’ the ‘greatest possible fatigue’.56 The current study used 100mm VASs with various anchors; translation of the previous estimate would suggest that changes of 24mm on the VASs could be clinically important. However, in this study, changes were not of this magnitude, and it is possible they were not clinically important. However, equally it may be that minimally important clinical differences for VASs are unlike those of numeric rating scales. Further, the different anchors used in this study may hinder translation from previous estimates. It is also possible that when considering aspects of the fatigue experience other than intensity, including the distress it evokes and its impact on functioning, a change of 24mm is no longer apposite. Further research is needed to identify clinically meaningful minimal changes, measured by VASs, in the extent of fatigue, and the distress and disruption it causes.
In all but two cases, the intervention was delivered as planned. In these two cases, only two domiciliary visits were made rather than the stipulated three following misunderstanding regarding appointments. Thus, compliance with the intervention was very promising and appeared similar to that reported previously for a telephone intervention.57 It also compares favorably with the community educational program evaluated by Grant et al.32 These authors report that with their two-session group program attendance at the second only reached 42%. The authors suggest various reasons for this. However, it could be argued that this may have arisen because the approach failed to meet the needs of the attendees. Further, it may reflect the difficulties that fatigued individuals have with traveling to sessions. Interventions that are delivered to individuals at home, either in person or by telephone, are likely to achieve higher completion.
The research team did record complications and comorbidity that arose during the three treatment cycles when data were collected. This enabled them to consider whether these influenced outcomes. In the event, few complications were reported by patients or recorded in their notes and did not explain the consistent improvement shown by the experimental group when compared with the control. The wide-ranging benefits reported by the intervention group support the patient-centered approach initiated in this study. Further, it could be argued that assessment, giving information, and support should provide the basis for any management program for fatigue in this patient group, irrespective of whether additional approaches including exercise or pharmacological interventions are introduced. Supportive care should be the bottom line. It has the potential to facilitate patients' coping with fatigue, reduce its burden on their lives, and allow them to lead lives of far better quality.
Study Limitations
The study was unable to entirely control the information and support available to participants. Individuals were at liberty to attain information and support from other parties in managing fatigue. However, the consistent improvements across the intervention relative to the control group suggested that the changes could be attributed to the intervention program. There was no equally plausible explanation for the observed changes.
The study was largely unblinded. Only individuals who coded and entered study data were blind to the treatment patients were allocated. Lack of blinding can compromise validity through bias introduced through knowledge of treatment allocation. Although the magnitude of bias introduced by unblinding is uncertain,58 readers need to consider the potential patient and investigator bias that may have impacted on the validity of this study's findings.
All patients were aware of the study group they were randomized to. Thus, it is possible that a placebo effect influenced those in the intervention group and resulted in their overestimation of the treatment effect. Psychological effects can arise in patients allocated to receive an intervention through their knowing they are provided a promising new treatment approach.38 This can alter their perceptions and assessments of its efficacy. However, while this may have arisen in this study, patients may equally have been skeptical of the potential benefits of the fatigue program, a program that required investment of participants' time and effort over a period when they were feeling fatigued. They could also have been more critical of their functioning through expecting fatigue to be resolved. It is difficult to be certain how they perceived the treatment options, and thus, it is difficult to interpret the impact of lack of blinding on patients' responses. The literature also suggests that knowledge of the group patients are assigned to can impact on their adherence to follow-up procedures leading to biased loss to follow-up.38 However, this did not appear to affect this study greatly.
Second, investigator bias may have impacted the results. This arises through both investigators and members of the health care team knowing group allocation.38 Schulz38 singles out clinicians as being particularly susceptible to altering behavior if aware of group allocation. It is suggested that they may differentially administer treatments other than those under study, influence patients willingness to continue in a study, or sway the outcomes they report. However, in this study the clinical team was typically unaware of the outcomes of randomization; the intervention was delivered in patients' homes and was largely masked from them. Thus, these effects were avoided. However, investigators enrolling patients to the study and administering the questionnaire battery over time were also aware of group allocation. Thus, they may have inadvertently spent more time with those in receipt of the intervention and offered differential support to patients accordingly. This may have influenced perceived and reported well-being.
A further limitation to this study relates to its inability to estimate the costs associated with the fatigue program. Cost data were not collected, and thus it is not possible to comment on the cost-effectiveness of the intervention when considered alongside alternative methods of promoting the management of cancer-related fatigue. Many diverse approaches to managing cancer-related fatigue are being developed and tested. However, for clinicians to determine which if any to implement, they need to be aware of their relative costs and benefits.
Conclusion
This was the first U.K. randomized controlled trial to evaluate a supportive/educative approach in managing cancer-related fatigue. Despite its limitations, the study has provided preliminary evidence regarding the benefits that can be afforded to patients when systematic attempts are made at managing this nebulous, yet distressing, symptom of cancer and its treatment. What remains unclear is the benefit and cost of supportive intervention relative to, or in combination with, other pharmacological or nonpharmacological interventions. This will only be realized with conduct of additional sufficiently powered randomized trials. Future research should seek to determine this.
Acknowledgments
This study was funded by Cancer Research UK (formally the Cancer Research Campaign).
Appendix.
Visual Analogue Scales Used to Measure Fatigue
References
- . Survey of the incidence and severity of side effects reported by patients receiving six cycles of FEC chemotherapy. J Cancer Nurs. 1997;1(2):61–73
- Cancer-related fatigue: inevitable, unimportant and untreatable?. Ann Oncol. 2000;11:1–5
- . Fatigue in patients with cancer and chronic obstructive airways disease: a phenomenological enquiry. Int J Nurs Stud. 1997;34(1):44–53
- . Bone Tired: the experience of fatigue and its impact on quality of life. Oncol Nurs Forum. 1996;23(10):1539–1547
- Patient, caregiver, and oncologist perceptions of cancer-related fatigue: results of a tripart assessment survey. Semin Hematol. 1997;34(3 Suppl. 2):4–12
- Impact of cancer-related fatigue on the lives of patients: new findings from the fatigue coalition. Oncologist. 2000;5(5):353–360
- . Quality and efficacy of educational materials on cancer-related fatigue: views of patients from two European countries. Eur J Oncol Nurs. 2003;7(2):99–109
- . In their own words: using the common sense model to analyze patient descriptions of cancer-related fatigue. Oncol Nurs Forum. 2001;28(9):1363–1369
- . Patients' perspectives of fatigue whilst undergoing chemotherapy. Oncol Nurs Forum. 1997;24(1):43–48
- . Self-care behaviours initiated by chemotherapy patients in response to fatigue. Int J Nurs Stud. 1997;34(1):35–43
- . Action of oral methylprednisolone in terminal cancer patients: a randomised double-blind study. Cancer Treat Rep. 1985;69(7–8):751–754
- . Effect of methylprednisolone sodium succinate on quality of life in pre-terminal cancer patients: a placebo-controlled, multicenter study. Eur J Cancer Clin Oncol. 1989;25(12):1817–1921
- . Methylprednisolone as palliative therapy for female terminal cancer patients. The Methylprednisolone Female Preterminal Cancer Study Group. Eur J Cancer Clin Oncol. 1989;25(12):1823–1829
- Patient-controlled methylphenidate for the management of fatigue in patients with advanced cancer: a preliminary report. J Clin Oncol. 2003;21(23):4439–4443
- . Interferon-induced fatigue in patients with melanoma: a pilot study of exercise and methylphenidate. Oncol Nurs Forum. 2002;29(7):E85–E90
- . Palliative uses of methylphenidate in patients with cancer: a review. J Clin Oncol. 2002;20(1):335–339
- The effect of donepezil on sedation and other symptoms in patients receiving opioids for cancer pain: a pilot study. J Pain Symptom Manage. 2003;26(5):1049–1054
- . Recombinant human erythropoietin in the treatment of cancer and chemotherapy induced anaemia: results of double-blind open-label follow-up studies. Semin Oncol. 1994;21(2 Suppl. 3):21–28
- . Effects of epoetin alfa on hematologic parameters and quality of life in cancer patients receiving nonplatinum chemotherapy: results of a randomized, double-blind, placebo-controlled trial. J Clin Oncol. 2001;19(11):2865–2874
- Impact of therapy with epoetin alfa on clinical outcomes in patients with nonmyeloid malignancies during cancer chemotherapy in community oncology practice. J Clin Oncol. 1997;15(3):1218–1234
- . Quality of life benefit in chemotherapy patients treated with Epoetin Alfa is independent of disease response or tumor type: results from a prospective study. J Clin Oncol. 1998;16(10):3412–3425
- . Effects of aerobic interval training on cancer patients' functional capacity. Nurs Res. 1989;38(6):348–351
- Structured exercise improves physical functioning in women with stages I and II breast cancer: results of a randomised controlled trial. J Clin Oncol. 2001;19(3):657–665
- Effects of exercise on fatigue, physical functioning, and emotional distress during radiation therapy for breast cancer. Oncol Nurs Forum. 1997;24(6):991–1000
- Fatigue and quality of life outcomes of exercise during cancer treatment. Cancer Pract. 2001;9(3):119–127
- . Effects of physical activity on the fatigue and psychologic status of cancer patients during chemotherapy. Cancer. 1999;85(10):2273–2277
- . An exercise intervention for advanced cancer patients experiencing fatigue: a pilot study. J Palliat Care. 2000;16(3):30–36
- . The effect of seated exercise on fatigue and quality of life in women with advanced breast cancer. Oncol Nurs Forum. 2004;31(5):977–983
- A randomized clinical trial of energy conservation for patients with cancer-related fatigue. Cancer. 2004;100(6):1302–1310
- . Energy for living with cancer: preliminary findings of a cancer rehabilitation group intervention study. Oncol Nurs Forum. 2001;28(9):1393–1396
- . Facilitating patients' coping with fatigue during chemotherapy—pilot outcomes. Cancer Nurs. 2002;25(4):300–308
- . Developing a community program on cancer pain and fatigue. Cancer Pract. 2000;8(4):187–194
- Pain and fatigue management: results of a nursing randomised clinical trial. Oncol Nurs Forum. 2002;29(6):949–956
- Effect of a cognitive behavioral Intervention on reducing symptom severity during chemotherapy. J Clin Oncol. 2004;22(3):507–516
- . Systematic review of psychological therapies for cancer patients and recommendations for future research. J Natl Cancer Inst. 2002;94(8):558–584
- . Blinding in clinical trials and other studies. BMJ. 2000;321(19–26):504
- . Blinding was judged more difficult to achieve and maintain in nonpharmacologic than pharmacologic trials. J Clin Epidemiol. 2004;57:543–550
- . The landscape and lexicon of blinding in randomized trials. Ann Intern Med. 2002;136(3):254–259
- . Characteristic and incidental (placebo) effects in complex interventions such as acupuncture. BMJ. 2005;330(21 May):1201–1204
- Physicians interpretations and textbook definitions of blinding terminology in randomized controlled trials. JAMA. 2001;285(15):2000–2003
- . Fatigue in patients receiving chemotherapy: patterns of change. Cancer Nurs. 1998;21(1):17–30[Erratum Cancer Nurs 1998;21(3):195]
- Defining and measuring fatigue. Image J Nurs Scholarsh. 1999;31(1):45–50
- . The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand. 1983;67:361–370
- . SF-36 Health Survey. Manual and interpretation guide. Boston, MA: The Health Institute, New England Medical Center; 1993;
- . Assessing coping strategies: a theoretically based approach. J Pers Soc Psychol. 1989;56(2):267–283
- . Fatigue. In: Groenwald S, Hansen Frogge M, Goodman M, Yarbro C editor. Cancer symptom management. Boston: Jones & Bartlett Publishers; 1996;p. 42–58
- Fatigue and the cancer experience: the state of the knowledge. Oncol Nurs Forum. 1994;21(1):23–36
- . Coping with cancer. Can exercise help?. Physician Sports Med. 2000;28(5):1–19
- . An environmental intervention to restore attention in women with newly diagnosed breast cancer. Cancer Nurs. 2003;26(4):284–292
- . SPSS Data Analysis System. Release 11.0. Chicago, Illinois: SPSS Inc.; 2001;
- How coping mediates the effect of optimism on distress: a study of women with early stage breast cancer. J Pers Soc Psychol. 1993;65(2):375–390
- Cancer-related fatigue—a difference of opinion? Results of a multi centre survey of healthcare professionals, patients and caregivers. Eur J Cancer Care. 2002;12:20–27
- . The middle-range theory of unpleasant symptoms: an update. Adv Nurs Sci. 1997;19(3):14–27
- . Cancer-related fatigue. Suffering a different fatigue. Cancer Pract. 2000;8(2):87–95
- . Combining anchor and distribution-based methods to derive minimally clinically important differences on the functional assessment of cancer therapy (FACT) anemia and fatigue scales. J Pain Symptom Manage. 2002;24(6):547–561
- Measurement of fatigue determining minimally important clinical differences. J Clin Epidemiol. 2002;55:239–244
- . A pilot study examining energy conservation for cancer-treatment related fatigue. Cancer Nurs. 2002;25(5):333–341
- Double blind, you are the weakest link—goodbye!. Evid Based Nurs. 2002;5:36–37
Caroline Alexander-Dann, our friend, colleague, and contributor to this paper, died on August 22, 2003.This project was supported by a Cancer Research Campaign Project Grant, number ce1162/0101.
PII: S0885-3924(06)00003-0
doi:10.1016/j.jpainsymman.2005.07.003
© 2006 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved.
Volume 31, Issue 2 , Pages 148-161, February 2006
