To identify and synthesize the evidence for mind-body interventions for which the evidence suggests benefit for at least two of the three cluster symptoms of pain, fatigue, and sleep disturbance.

Databases searched were CINAHL, MEDLINE, and PsycINFO through March 2009.

Search keywords were guided imagery, hypnosis, relaxation, biofeedback, cognitive behavioral therapy, coping skills training, meditation, virtual reality, music AND cancer AND fatigue, sleep disturbance, sleep difficulty, insomnia, and pain.

Studies were included in the review if they

• Were limited to research
• Included adults aged 18 years and older
• Included mind-body activities that involved primarily mental activity that could be performed by almost all patients
• Included pain, fatigue, or sleep among study dependent variables.

Studies were excluded if they

• Involved the use of yoga
• Involved patients in whom a diagnosis of cancer was not yet established
• Had a sample that included people without cancer.

A total of 47 studies were identified. In four of those, all testing virtual reality, only the symptom of fatigue was measured, so these were eliminated.

The final sample included 43 studies. Study sample sizes and total patients involved across studies were not reported.

Relaxation

Six studies examined relaxation interventions in hospitalized patients, outpatients with chronic pain, and women with early-stage breast cancer.

• Significantly greater pain relief was obtained with progressive muscle relaxation compared to massage, usual treatment, mood manipulation, distraction, and controls.
• One study found no difference in pain between a daily relaxation exercise and distraction.
• Training in muscle relaxation did not improve fatigue in one study compared to provision of information.
• In one study, muscle relaxation improved sleep compared to usual treatment controls.

Imagery and Hypnosis

Six studies examined imagery and hypnosis.

• In four studies, imagery was used in hospitalized patients with cancer pain, and beneficial effects were reported.
• One study found no differences in pain or fatigue between patients with an imagery intervention and those receiving standard care.
• Four studies used imagery in comparison to cognitive-behavioral therapy (CBT) and combined imagery with relaxation. Of those, one study reported no significant effect, two reported significant pain reduction, and one reported significant reduction in fatigue and sleep disturbance.

Cognitive Behavioral Therapy (CBT)/Coping Skills Training (CST)

Twenty-one studies tested CBT/CST.

• In three studies, fatigue was the primary focus. Significantly more improvement in fatigue was reported with a six- to 12-week CBT/CST intervention compared to usual treatment and controls.
• Three studies evaluated CST effects on the combination of pain and fatigue. In all of these, a one-session CST intervention resulted in no difference in symptoms compared to controls.
• Seven studies evaluated the effects of CBT/CST on fatigue and sleep disturbance. One study reported a decreased incidence of fatigue and sleep disturbance using an audio recording for coping skills training prior to chemotherapy. Two studies reported improvement in sleep with a four- to eight-week CBT intervention, but only one of these also reported improvement in fatigue. One study reported improvement in sleep and fatigue with a five-session CBT intervention, two other studies showed improvement in sleep but no change in fatigue, and one study reported no improvement in either of these two symptoms.
• Four studies reported effects of CBT/CST on all three symptoms concurrently. One showed improvement in fatigue and sleep but no impact on pain. One study reported less sleep disturbance but no difference in pain or fatigue. One reported lower ratings of worst pain immediately after the CBT program and greater reduction in pain and fatigue six months after the intervention compared to controls. One study found no differences in any of the three symptoms with a CST intervention.

Meditation

Four studies were included.

• Three of these studies used mindfulness-based interventions. One study reported significant improvements in both fatigue and sleep among outpatients who participated in an eight-week intervention.

Music

• Four studies looked at the effect of music on pain. Two studies found significant improvements in a pre-/posttest design using 30 minutes of preferred music among hospitalized patients. Two other studies found no difference in pain with listening to music compared to control groups.
• Two studies tested a music intervention on fatigue. One found a significant effect, and one found no difference in fatigue between intervention and control groups.

Findings of this review were equivocal.

• Although the authors stated a criterion for inclusion of examination of at least two of the three symptoms of interest, the review appeared to include studies in which only one of these symptoms was reported.
• Few investigators used multisymptom interventions and evaluations.
• Measures of symptom clusters were not been well identified.
• Some instruments were stated to potentially be more sensitive; however, the scales and individual items that were most useful to measure this symptom cluster were not determined.
• Timing, dosage, and frequency of interventions varied among studies, making it difficult to draw systematic conclusions. Most music interventions were very brief.
• This review did not provide study details, such as clear sample descriptions, sample sizes, or actual statistical results, and no effect sizes were calculated, although some studies used the same outcome measures.

Although the findings did not clearly demonstrate the effects of these interventions across studies, the authors concluded that these interventions hold promise. Although such interventions carry minimal risk to patients, some interventions would require substantial time and resource commitment to provide.

To assess, among hospitalized patients with cancer-related pain, responses to progressive muscle relaxation (PMR) and analgesic imagery, including the impact of these interventions on pain-related distress and perceived control over pain

To examine the influence of ability, outcome expectancy, previous experience, and concurrent symptoms on the effectiveness of the interventions

Scripts relating to \"control conditions,\" PMR, and guided imagery interventions were recorded on a CD. The recorded voice was male, and the recording did not include music. The day 1 control recording consisted of identification of health team members, explanation of patient rights, and a description of various hospital services. The second control recording described exercise and other activities to maintain strength while hospitalized and presented issues for consideration in discharge planning. The PMR recording guided listeners in tensing and relaxing muscles in a series of 12 major muscle groups from head to feet. The analgesic, or guided imagery, recording asked listeners to scan their bodies to identify areas of pain and to imagine replacing pain with comforting sensations. Then the recording asked listeners to imagine putting their hand in an anesthetic fluid, feeling it go numb, and then transfer the numbness to painful areas of the body. All recordings were of similar length. Trials were done when current pain was rated 2–8 on a 10-point scale. Trials were delayed if patients had received an oral analgesic within one hour or an IV analgesic within 30 minutes. Patients with the option of patient-controlled analgesia were asked to refrain from increasing doses during the 15-minute trial and were maintained on basal continuous infusion. A control trial was the first trial done each day. If pain persisted at the end of the control trial, the research nurse continued with whichever experimental trial was assigned for that day. Two intervention trials were done each day, at least one hour apart. Patients completed pre- and post-trial ratings of pain intensity and related distress. For each outcome measured, scores were averaged across two control, two PMR, and two analgesic imagery trials. In cases where only one trial was completed, this score was used in analysis. Patients were randomly assigned to two different groups. In one group, PMR preceded imagery interventions; in the other, the sequence was reversed. All patients received the same control trials.

• The study reported on a sample of 32 patients.
• Mean patient age was 48.9 years (SD = 16.29 years).
• The sample was 55% female and 45% male.
• Of all patients, 55% had hematologic malignancy.
• All patients were Caucasian.
• The average number of months since diagnosis was 18.36 (SD = 33.32 months). Of all patients, 33% were receiving chemotherapy, 13% were undergoing radiotherapy, and 20% were undergoing combined therapy; 78% were being treated with a combination that included a strong opioid, a weak opioid, and nonopioid medications for pain management.

• Single site
• Inpatient setting
• Academic medical center in the Midwest, United States

The study used a crossover design with control.

• Scale of pain intensity (0 = no pain, 10 = worst pain imaginable)
• Scale of pain distress (0–10)
• Control Subscale of the Survey of Pain Attitudes
• Imaging Ability Questionnaire
• Outcome Expectancy Scale
• Edmonton Symptom Assessment Scale

• Compared to the control conditions, PMR resulted in greater change in pain intensity (Z = –2.15, p < 0.05) and pain-related distress (Z = –1.82, p < 0.05) and greater perceived control over pain (Z = –3.02, p < 0.01).
• Compared to the control conditions, guided imagery resulted in greater change in pain intensity (Z = –2.7, p < 0.01), change in pain-related distress (Z = –2.11, p < 0.05), and greater perceived control over pain (Z = –2.4, p < 0.01).
• PMR was associated with meaningful reduction in pain for 41% of participants. Imagery was associated with meaningful improvement in pain for 52% of patients. Authors defined meaningful improvement as an improvement greater than or equal to 30% reduction in pain intensity along with improvement in pain-related distress or perceived control.
• Compared to other patients in the study, those who experienced meaningful improvement with imagery had greater imaging ability (p < 0.05), more positive outcome expectancy (p < 0.05), and fewer concurrent symptoms (p < 0.05).
• The control condition was associated with some reduction in pain intensity and pain-related distress.

In only 50% of patients did PMR and guided imagery appear to have a more positive short-term effect on pain intensity, pain-related distress, and perceived control of pain than did simple distraction. Variables examined did not appear to influence the effects of PMR, but expectations and number of concurrent symptoms appeared to influence the effects of guided imagery.

• The study had a small sample size, with fewer than 100 participants.
• Interventions had very short-term effects on the outcomes examined. Authors did not state how long patients had been experiencing pain; whether patients were suffering acute, chronic, or episodic pain was unclear.
• Of all patients approached for participation, 41% declined to enter the study.
• Despite the crossover design, the study had risk of bias due to several factors.
  • Pain intensity measures were self-reported by patients.
  • Patients certainly knew the difference between the control and intervention recordings.
  • Analysis was limited to comparisons between responders and nonresponders; therefore, the capacity of the investigators to identify influencing variables was limited.
  • While changes in pain intensity showed differences that were statistically significant, whether the changes were clinically significant is unclear.
  • Many factors can affect the results evaluated (e.g., time of day, nature of hospital care, activities of the day). Investigators could not have accounted for or controlled for these factors.
  • The trial was extremely brief, only two days.

PMR and analgesic imagery, facilitated by a recording on CD, might be helpful to some patients to reduce short-term pain intensity, decrease pain-related distress, and increase sense of control over pain. However, much more evidence is needed before investigators can draw firm conclusions. These interventions could be implemented easily in a hospital setting and do not appear to harm patients. They appear to be more effective for individuals with greater ability to imagine, greater expectations of effectiveness, and fewer additional symptoms. This fact might lead clinicians to identify that group of patients most likely to benefit from such interventions.

To evaluate the feasibility of a patient-controlled, cognitive-behavioral intervention for pain, fatigue, and sleep disturbance during treatment for advanced cancer and to assess the initial efficacy of the intervention.

Patients provided baseline measures, such as measures relating to demographics and a symptom inventory, received education, and underwent training to use an mp3 player loaded with 12 cognitive-behavioral strategies (relaxation exercises, guided imagery, nature sounds). Patients used the strategies as needed for symptom management for two weeks and kept a log of symptom ratings with each use. Following the two-week intervention, patients completed a second symptom inventory and an evaluation of the intervention. Clinic staff identified patients who met the eligibility criteria based on diagnosis and treatment and then were briefly introduced to the study and asked if a research nurse could visit to provide additional information. The research nurse met with interested patients, assessed symptoms, and completed eligibility screening. Study purpose and procedures were explained, and written informed consent was obtained.

• The sample was comprised of 30 patients (20% male, 80% female).
• Mean age was 56.27 years (range 36–79).
• All patients had advanced (recurrent or metastatic) cancer and were receiving chemotherapy or radiotherapy. Half of the patients had a diagnosis of gynecologic cancer. Patients had been diagnosed with lung cancer (n = 8), colorectal cancer (n = 6), and prostate cancer (n = 1).
• Most patients were Caucasian, women, and had earned a bachelor’s degree or higher.
• Almost all patients (n = 27) were taking two or more prescribed supportive medications to treat symptoms.
• Studies were included if they included at least two of the following three symptoms:  pain rated 3 or higher on a 0-to-10 scale in the previous 48 hours, fatigue in the previous week, and sleep disturbance in the previous week. 
• Studies were excluded if they involved postoperative or neuropathic pain.

• Multisite
• Outpatient
• Midwestern United States
   

• Patients were undergoing multiple phases of care.
• The study has clinical applicability for late effects and survivorship, end of life care, and palliative care.

The study used a one-group, pre- and postintervention design.

• Demographic questionnaire and chart review form to record age, gender, education, race, ethnicity, cancer diagnosis, current treatments, and supportive medications prescribed, such as analgesics, steroids, psychostimulants, hypnotics, or sedatives
• Symptom inventory, to measure
  • Pain severity, by means of four, 0-to-10 numeric rating scales for pain now, worst pain, least pain, and average pain in the last 24 hours
  • Fatigue severity, by means of four, 0-to-10 scales for fatigue now, worst fatigue, least fatigue, and average fatigue in the last 24 hours
  • Sleep disturbance, by means of a 0-to-10 scale rating sleep disturbance in the last 24 hours and, from the Pittsburgh Sleep Quality Index (PSQI), a verbal rating of sleep quality
• Additional symptoms and their effects on daily functioning, as assessed by the MD Anderson Symptom Inventory (MDASI), which includes items measuring 13 common cancer-related symptoms and their effect on general activity, mood, work, relations with others, walking, and enjoyment of life
• Treatment log book, in which patients tracked each time they used a cognitive-behavioral strategy. Data tracked were time of day; strategy used; and immediate pre- and posttreatment scores, on a 0-to-10 scale, regarding severity of pain, fatigue, and sleep disturbance
• Poststudy evaluation, a 12-item survey created, for the current study, to assess patients’ perceptions of the acceptability of the study procedures. Among the items evaluated were time commitment; equipment; and questionnaires and factors relating to the patient-controlled, cognitive-behavioral intervention, such as length, variety, and usefulness.
   

• Symptom scores at two weeks did not differ significantly from baseline scores; however, significant reductions in pain, fatigue, and sleep disturbance severity were found in ratings made immediately before and after the use of a cognitive-behavioral strategy.
• Mean pain scores decreased from 4.54 pretreatment to 2.77 posttreatment.
• Mean fatigue scores decreased from 4.9 pretreatment to 3.44 posttreatment.
• Average sleep disturbance scores decreased from 5.05 pretreatment to 2.81 posttreatment.

The patient-controlled, cognitive-behavioral intervention is feasible and may reduce the day-to-day severity of co-occurring pain, fatigue, and sleep disturbance.

• The study had a small sample size, with less than 100 patients.
• The study had a risk of bias due to no control group.
• The study used a convenience sample comprised of well-educated, primarily female patients—a sample that may not be representative of all populations.
• Uncontrolled extraneous variables, such as supportive medication use, could have influenced symptom outcomes.

The findings support nurse education and the recommendation of the specified patient-controlled, cognitive-behavioral interventions for the management of pain, fatigue, and sleep disturbance. In regard to patient care and symptom management at all stages of cancer, nurses are the front-line educators of patients. This intervention supports the principle of autonomy for patients able to participate actively in care. Further study—a randomized, controlled trial to test the efficacy of the intervention for co-occurring pain, fatigue, and sleep disturbance—was under way at the time this study was published.

To test the effects of a psychoeducational intervention on pain, fatigue, and sleep disturbance.

Patients were randomly assigned to the intervention group or the wait-list control group. The intervention consisted of a single one-on-one training session with a research nurse, which was followed by recorded guidance that provided imagery, relaxation exercises, and nature sounds. Most exercises lasted 20 minutes and were delivered via an mp3 player. The study lasted two weeks.

• The sample was comprised of 78 patients (41% male, 59% female).
• Mean age was 60.29 years (standard deviation = 11.09 years).
• Patients had lung, prostate, colorectal, or gynecologic cancer.
• Patients were receiving chemotherapy or radiation therapy and had multiple concurrent symptoms at baseline.
• At study entry, all patients had to have fatigue, sleep disturbance, and pain scores of at least 3 on an 11-point numeric scale.
• Of the patients, 71% were taking steroids, 59% were taking opioids, and 86% were taking antiemetics.
   

• Multisite
• Outpatient
• Midwestern United States

Patients were undergoing the active antitumor treatment phase of care.

The study used a randomized, controlled trial, pre- and postintervention design.

• Selected items from various instruments, including the Brief Fatigue Inventory (BFI), Brief Pain Inventory (BPI), and Pittsburgh Sleep Quality Index (PSQI)
• Eleven-point numeric scales, to measure symptoms
• Patient diaries, which recorded use of cognitive-behavioral therapy (CBT) exercises
• Symptom cluster score, calculated by averaging the pain, fatigue, and sleep disturbance scores
• Symptom interference subscale of the MD Anderson Symptom Inventory (MDASI)

• Comparison of pre- and postintervention symptom severity scores showed a significant reduction in the severity of pain, fatigue, and sleep disturbance and symptom interference (p = 0.000).
• Symptom cluster scores and individual symptom scores declined in both the intervention and control groups.
• The reductions in symptom cluster severity (p < 0.05), pain (p < 0.01), and fatigue (p < 0.05) were significantly greater in the intervention group than in the control group; however, the effect sizes (calculated as partial eta) were extremely small (range 0.041–0.093).

The intervention demonstrated a small statistically significant effect on the symptoms of pain and fatigue and the overall symptom cluster of pain, fatigue, and sleep disturbance.

• The study had a small sample size, with less than 100 patients.   
• The study had risks of bias due to no blinding and no appropriate attentional control condition.
• The sample included baseline and group differences of import.
  • The control group had higher depression scores at baseline than did the intervention group.
  • A greater number of those in the intervention group dropped out of the study, suggesting that the intervention was not well accepted. Intention-to-treat analysis used the last value carried forward. If symptoms worsened, this would produce biased results.
• Although the intervention was called a CBT intervention, whether cognitive reframing or problem solving was a part of the intervention was unclear. The intervention appeared to have been a relaxation or imagery therapy.
• Measurement validity and reliability were questionable.
  • Authors used Z-scores to compare sleep disturbance severity, rather than actual scores; the reason for this was unclear. The actual change in Z-scores for this symptom was larger than the score changes associated with other symptoms, but the Z-score change was not statistically significant.
  • Whether average individual symptom scores were more meaningful than the total score for the cluster was unclear.  
• The intervention may be too expensive or impractical, in terms of training needs, to be feasible.

The intervention was a relaxation or imagery therapy rather than a true CBT. The intervention was associated with short-term statistically significant benefits, but the actual size of the effect was small. Findings suggested that approaches using relaxation and imagery may result in some small benefit for patients, but the effect was weak.

To assess the efficacy of a cognitive-behavioral (CB) intervention for the pain, fatigue, and sleep disturbance symptom cluster

Patients were randomized to the CB group or a waitlist control group. The intervention was a single one-on-one training session with a research nurse providing information about causes of pain, fatigue, and sleep disturbance; rational about how CB strategies could affect symptoms; overview of 12 strategies; and recommendations for patients to practice. CB strategies were relaxation and imagery approaches. Patients were provided scripted recordings to use on an MP3 player and written instructions. Patients were followed for two weeks and completed study assessments at baseline and at two weeks. The control group had usual care. Patients in the intervention group were to keep a log recording each use of CB strategies, and all patients in both groups were taught to complete a daily symptom diary. Follow-up phone calls were made to both groups on the second study day, the seventh study day, and at the end of the two-week study period. Post-intervention measures of symptom severity were mailed to participants to complete and provide to clinic staff.

• N = 78
• MEAN AGE = 60.29 years (SD = 11.09 years)
• MALES: 41%, FEMALES: 59%
• KEY DISEASE CHARACTERISTICS: Lung, prostate, colorectal, and gynecologic cancers; 72% on chemotherapy
• OTHER KEY SAMPLE CHARACTERISTICS: 86% on antiemetics, 71% on steroids, 59% on opioids; 93% white; 74% had at least undergraduate college education

• SITE: Multi-site 
• SETTING TYPE: Outpatient 
• LOCATION: Wisconsin

• PHASE OF CARE: Active antitumor treatment

• RCT

• 0–10 numeric scale for worst, least, and average pain over 24 hours, and pain now, fatigue, and worst sleep disturbance
• Brief Fatigue Inventory
• One item from the Pittsburgh Sleep Quality Index
• Summary scores were calculated as the average for each symptom, and cluster severity score was calculated as the z-transformation of the average of symptom summary scores.
• MD Anderson Symptom Inventory scores of overall average calculated across six items
• Profiles of Mood States short form

CB strategies were used an average of 13.65 times (SD = 6.98) during the two weeks. Patients in the control group were significantly more depressed at baseline (p = .003). Total cluster scores post-intervention were lower in the intervention group (eta² 0.052, p = .032). Examination of individual symptoms showed differences between groups in pain (P = .006) and fatigue, but not in sleep.

Use of various patient self-controlled relaxation and imagery recordings appeared to have a beneficial effect in reducing pain and fatigue compared to controls. No clear effect was seen on sleep disruption.

• Small sample (less than 100)
• Baseline sample/group differences of import
• Risk of bias (no blinding)
• Risk of bias (no appropriate attentional control condition)
• Measurement validity/reliability questionable
• Other limitations/explanation: More study dropouts in the intervention group; short follow-up period; validity and reliability of cluster measurement methods is unclear.

Findings suggest that patient use of relaxation and imagery instructions and recordings may improve management of pain and fatigue during cancer treatment.

To assess the efficacy of a cognitive-behavioral (CB) therapy intervention for the symptom cluster of pain, fatigue, and sleep disturbance

Patients were randomized to the CB group or a waitlist control group. The intervention was a single, one-on-one training session with a research nurse providing information about causes of pain, fatigue, and sleep disturbance. Patients learned how CB strategies could affect symptoms and received an overview of 12 strategies and recommendations to practice. CB strategies included relaxation and imagery approaches. Patients were provided scripted recordings to use on an MP3 player and written instructions. Participants were followed for two weeks, and study assessments were completed at baseline and at two weeks. The control group received usual care. Intervention patients kept a log recording each use of CB strategies, and all patients in both groups were taught to complete a daily symptom diary. Follow-up phone calls were made to both groups on study days 2 and 7 and at the end of the two-week study period. Postintervention measures of symptom severity were mailed to participants to complete and provide to clinic staff members.

• N = 78
• MEAN AGE = 60.29 years (SD = 11.09 years)
• MALES: 41%, FEMALES: 59%
• KEY DISEASE CHARACTERISTICS: Lung, prostate, colorectal, and gynecologic cancers; 72% receiving chemotherapy
• OTHER KEY SAMPLE CHARACTERISTICS: 86% took antiemetics, 71% took steroids, and 59% took opioids; 93% were white; 74% had at least an undergraduate college education

• SITE: Multi-site
• SETTING TYPE: Outpatient
• LOCATION: Wisconsin, United States

• PHASE OF CARE: Active antitumor treatment

Randomized, controlled trial

• 0–10 numeric scales for worst, least, and average pain over 24 hours and for current pain and fatigue, and worst sleep disturbance
• Brief Fatigue Inventory (BFI)
• Pittsburgh Sleep Quality Index (PSQI) (one item used)
• Summary scores were calculated as an average for each symptom, and cluster severity scores were calculated as the z-transformation of average of symptom summary scores.
• The MD Anderson Symptom Interference (MDASI) scores overall average were calculated across six items.
• Profiles of Mood States Short Form (POMS-SF)

CB strategies were used an average of 13.65 times (SD = 6.98) during the two weeks. Patients in the control group were significantly more depressed at baseline (p = 0.003). Total cluster scores postintervention were lower in the intervention group (eta² = 0.052, p = 0.032). An examination of individual symptoms showed differences between the groups in pain and fatigue but not in sleep.

The use of various patient-controlled relaxation and imagery recordings appeared to have a beneficial effect in reducing pain and fatigue compared to a control group. There was no clear effect on sleep disruption.

• Small sample (< 100)
• Baseline sample/group differences of import
• Risk of bias (no blinding)
• Measurement validity/reliability questionable
• Other limitations/explanation: More study drop-outs in intervention group; short follow-up period; validity and reliability of cluster measurement methods unclear

The findings of this study suggest that patient use of relaxation and imagery instructions and recordings may improve the management of pain and fatigue during cancer treatment.

This intervention was music versus simple distraction and control (treatment as usual).

The study reported on a sample of 58 patients with cancer having painful, cancer-related medical procedures.

A randomized controlled trial design was used.

• Spielberger State-Trait Anxiety Inventory (STAI-S) pre- and post-procedure
• Pain intensity (0–10)
• Perceived control (single item, 0–10)

There were no significant differences in post-procedure anxiety (no p values reported), pain, or perceived control across conditions.

• The study had a small sample size.
• One-third of the sample used analgesics and anxiolytics.

To examine the evidence and produce recommendations for exercise and lymphedema management in female breast cancer survivors

• Databases searched were MEDLINE, CINAHL, Cochrane Library, PapersFirst, ProceedingsFirst, WorldCat, PEDro, National Guidelines Clearing House, DARE, and ACP Journal Club.
• Search keywords were lymphedema, swelling, edema and keywords for exercise, such as physical activity, physical therapy, exercise, strength training.
• Studies were included if they were randomized controlled, cohort, or case-control studies or meta-analyses or systematic reviews.
• Exclusion criteria were not specified. 

A total of 659 references were retrieved. Two authors used the Oncology Nursing Society Putting Evidence Into Practice (PEP) categories of evidence to evaluate the references.

• The final number of studies included was 19.
• The sample size across studies was 912 patients with breast cancer, with a range of 14–242.

Patients were undergoing multiple phases of care.

Resistance exercises were determined to be \"Likely to be effective.\" Aerobic and resistance exercise were rated as \"Benefits balanced with harms,\" though no clear harms were obvious in the review of the studies. Other exercise approaches studied were deemed \"Effectiveness not established.\"

Benefits of exercise have been reported, and this review suggests that current evidence supports the use of resistance exercising.

• The method of evaluating aerobic exercise was not clear.
• Not all studies included use of compression garments during exercise.

Findings support the use of resistance exercise in women with breast cancer for the management of arm lymphedema.

To examine the perceived effects and experiences of mindfulness-based stress-reduction (MBSR) training as described by patients with cancer

Eight group sessions of a modified MBSR program using relaxation, meditation, and yoga exercises. The instructors were two clinical psychologists. At the end of the intervention, quantitative psychological measures were assessed and participants were invited to participate in a semistructured telephone interview regarding their experiences. This study reports on the data from the qualitative analysis of interviews.

• Mean patient age was 54 years, with a range of 31–65 years.
• The sample was 100% female.
• Seventeen patients had breast cancer and 1 had lymphatic cancer.
• All patients had been treated with chemotherapy, and the patients with breast cancer had also been treated with surgery and radiation therapy.
• Twenty-eight percent of the participants were on antidepressive medication.

• Single site
• Sweden

• Patients were undergoing long-term follow-up.
• The study has clinical applicability for late effects and survivorship.

A qualitative study design was used.

Thematic analysis of audiotaped interview transcripts

Meditation and yoga exercises were experienced as most positive. The group itself offered a positive effect through shared experience, nonjudgmental approach, and acceptance.

The thematic analysis was insufficient to offer support for the researchers’ conclusions.

• The study had a small, self-selected sample.
• Thematic analysis and participant characteristics were not explored sufficiently to recommend the intervention.

This intervention requires significant training, time, and effort for implementation.

To test the hypothesis that massage would decrease pain and analgesic medicine use

To explore effects on quality of life and physical and emotional symptom distress

Patients were randomly assigned to a massage treatment group or to a control group in which patients received simple touch controlled for time and attention. Individual baseline data for disease characteristics, pain, symptom distress, quality of life, functional status, expectations from massage, and concurrent interventions were collected within 72 hours of study inclusion and at three weekly visits over the three to four weeks of study participation, for measurement of sustained effects. Data collectors were blinded to study group assignment. Participants received up to three 30-minute treatments over two weeks with at least 24 hours between treatment sessions, according to a schedule jointly determined by the patient and the treatment provider. Treatment providers obtained immediate outcome data prior to and following each treatment. All participants received routine care in addition to study interventions. Massage intervention included gentle effleurage, petrissage, and myofascial trigger point release. The most frequently massaged areas were neck and upper back and arms, hands, lower legs, and feet. Massages were performed by licensed massage therapists who had at least six months’ experience working with patients with advanced cancer. Control touch included placement of both hands on the participant bilaterally on the neck, shoulder blades, lower back, calves, heels, clavicles, lower arms, hands, patellae, and feet with light and consistent pressure. All treatment providers had standardized hands-on training and were evaluated for competency.

• The study reported on 348 patients.
• Mean patient age was 65.2 years (SD = 14.4) in the experimental group, and 64.2 years (SD = 14.4) in the control group.
• The sample was 61% female and 39% male.
• The most common diagnoses were breast and lung cancers.
• All patients had metastatic disease, 27% had bone metastasis, 54% had constant pain, and 26% had neuropathic pain.
• Of the sample, 44% were married or in a committed relationship, 39%–42% had a college level or higher education, and 86% were non-Hispanic white.
• In the experimental group, 77% were receiving care at home, and 81% were receiving care at home in the control group.
• Mean worst pain in 24 hours in both groups was 6.4 or greater at baseline.

• Multisite
• 15 U.S. hospices and the University of Colorado Cancer Center

The study was a randomized, single-blind, controlled trial.

• Memorial Pain Assessment Card (MPAC) using a 0–10 point scale for immediate effect
• Brief Pain Inventory (BPI) for sustained measure
• MPAC Mood Scale
• McGill Quality of Life Questionnaire
• Memorial Symptom Assessment Scale (MSAS)
• Recording of name, dose, and frequency of medication for symptom management

Both massage and touch were associated with significant improvements in immediate and sustained pain outcomes. Massage was superior to touch, but the difference was not statistically significant. Both groups demonstrated statistical, but not clinically significant, improvement in BPI scores. Both massage and simple touch were reported to be associated with statistically significant immediate improvement in mood, with massage showing statistically superior effect compared to touch. Confidence intervals were provided, but significance levels were not reported. Both groups demonstrated improvement in physical and emotional symptom distress and quality of life across weekly evaluations, but there were no differences between groups. There were no adverse effects associated with the interventions, and no differences in general adverse events or mortality between groups. Differences in pain medication use were not reported.

Both massage and simple touch appeared to have immediate beneficial effects on pain and mood in these patients. Both groups experienced slight improvement in pain, quality of life, and symptom distress over time. These changes were minimal, showing statistical significance but not clinical relevance.

• Findings are limited to patients with very advanced cancer, the majority of whom were in hospice programs, and may not be applicable to other patient groups.
• There was no usual care control group.
• Having an appropriate attentional control group was useful, but given the findings that both study groups experienced benefits, the attention itself may be the most relevant factor in changes seen.

Simple touch appeared to have a short-term positive effect on patient mood and pain experience. This is an intervention that should be easy to provide for patients and could be something that caregivers could also be educated to provide. This intervention could be useful for intermittent use as an adjunct to other interventions for pain management. Formal massage did not provide significantly greater effects. Given findings of simple touch in the population studied here, evaluation of this approach in other patient groups can be useful.