Advertisement

Effects of a Full-Body Massage on Pain Intensity, Anxiety, and Physiological Relaxation in Taiwanese Patients with Metastatic Bone Pain: A Pilot Study

Open AccessPublished:December 15, 2008DOI:https://doi.org/10.1016/j.jpainsymman.2008.04.021

      Abstract

      Bone involvement, a hallmark of advanced cancer, results in intolerable pain, substantial morbidity, and impaired quality of life in 34%–45% of cancer patients. Despite the publication of 15 studies on massage therapy (MT) in cancer patients, little is known about the longitudinal effects of MT and safety in cancer patients with bone metastasis. The purpose of this study was to describe the feasibility of MT and to examine the effects of MT on present pain intensity (PPI), anxiety, and physiological relaxation over a 16- to 18-hour period in 30 Taiwanese cancer patients with bone metastases. A quasi-experimental, one-group, pretest-post-test design with repeated measures was used to examine the time effects of MT using single-item scales for pain (PPI-visual analog scale [VAS]) and anxiety (anxiety-VAS), the modified Short-Form McGill Pain Questionnaire (MSF-MPQ), heart rate (HR), and mean arterial pressure (MAP). MT was shown to have effective immediate [t(29)=16.5, P=0.000; t(29)=8.9, P=0.000], short-term (20–30 minutes) [t(29)=9.3, P=0.000; t(29)=10.1, P=0.000], intermediate (1–2.5 hours) [t(29)=7.9, P=0.000; t(29)=8.9, P=0.000], and long-term benefits (16–18 hours) [t(29)=4.0, P=0.000; t(29)=5.7, P=0.000] on PPI and anxiety. The most significant impact occurred 15 [F=11.5(1,29), P<0.002] or 20 [F=20.4(1,29), P<0.000] minutes after the intervention. There were no significant time effects in decreasing or increasing HR and MAP. No patient reported any adverse effects as a result of MT. Clinically, the time effects of MT can assist health care providers in implementing MT along with pharmacological treatment, thereby enhancing cancer pain management. Randomized clinical trials are needed to validate the effectiveness of MT in this cancer population.

      Key Words

      Introduction

      Despite publication of 15 studies on massage therapy (MT) in cancer patients, little is known about the safety and longitudinal effects of MT in cancer patients with bone metastases. Most researchers do not report the inclusion or exclusion of patients with bone metastases, and most include only a single measurement time point.
      • Jane SW
      • Wilkie DJ
      • Gallucci BB
      • Beaton RD
      • Huang HY
      Systematic review of massage intervention for adult patients with cancer: a methodological perspective.
      Bone metastases, the third most common metastatic site after liver and lung,
      • Abrams H.L.
      • Spiro R.
      • Goldstein N.
      Metastases in carcinomas: analysis of 1000 autopsied cases.
      result in intolerable pain, substantial morbidity, and impaired quality of life
      • Mercadante S.
      Prevalence, causes and mechanisms of pain in home-care patients with advanced cancer.
      • Nielsen O.S.
      • Munro A.J.
      • Tannock L.F.
      Bone metastases: pathophysiology and management policy.
      • Coleman R.E.
      Skeletal complications of malignancy.
      in 34%
      • Banning A.
      • Sjogren P.
      • Henriksen H.
      Pain causes in 200 patients referred to a multidisciplinary cancer pain clinic.
      to 45%
      • Mercadante S.
      Prevalence, causes and mechanisms of pain in home-care patients with advanced cancer.
      of cancer patients. The purpose of this study was to describe the feasibility and time effects of MT in Taiwanese cancer patients with bone metastases.
      Despite its long history and popularity in the Chinese culture, a review of the literature yielded only three MT studies conducted in Taiwanese patients, one each in labor pain,
      • Chang M.Y.
      • Wang S.Y.
      • Chen C.H.
      Effects of massage on pain and anxiety during labour: a randomized controlled trial in Taiwan.
      postoperative pain,

      Chin CC. The effects of back massage on surgical stress responses and postoperative pain. Unpublished doctoral dissertation, Case Western Reserve University, Ohio, 1999.

      and chemotherapy-related fatigue.

      Huang TW. Comparison of the effects of back massage and relaxation training in reducing chemotherapy induced fatigue. Unpublished Thesis, Taipei Medical University, Taipei, Taiwan; 2001.

      Whether MT is also regarded as safe, beneficial, and acceptable to Taiwanese patients with a cancer metastatic to bone needs further examination. In terms of the time effects of MT, numerous studies have reported immediate
      • Ahles T.A.
      • Tope D.M.
      • Pinkson B.
      • et al.
      Massage therapy for patients undergoing autologous bone marrow transplantation.
      • Ferrell-Torry A.T.
      • Glick O.
      The use of therapeutic massage as a nursing intervention to modify anxiety and the perception of cancer pain.
      • Meeks S.
      Effects of slow stroke back massage on relaxation in hospice clients.
      • Post-White J.
      • Kinney M.E.
      • Savik K.
      • et al.
      Therapeutic massage and healing touch improve symptoms in cancer.
      • Sims S.
      Slow stoke back massage for cancer patients.
      • Wilkinson S.
      Aromatherapy and massage in palliative care.
      • Wilkie D.J.
      • Kampbell J.
      • Cutshall S.
      • et al.
      Effects of massage on pain intensity, analgesics and quality of life in patients with cancer pain: a pilot study of a randomized clinical trial conducted within hospice care delivery.
      or short-term (ranging from five to 20 minutes)
      • Ferrell-Torry A.T.
      • Glick O.
      The use of therapeutic massage as a nursing intervention to modify anxiety and the perception of cancer pain.
      • Meeks S.
      Effects of slow stroke back massage on relaxation in hospice clients.
      • Ironson G.
      • Field T.
      • Scafidi F.
      • et al.
      Massage therapy is associated with enhancement of the immune system's cytotoxic capacity.
      • Grealish L.
      • Lomasney A.
      • Whiteman B.
      A nursing intervention modify the distressing symptoms of pain and nausea inpatients hospitalized with cancer.
      • Field T.
      • Peck M.
      • Krugman S.
      • et al.
      Burn injuries benefit from massage therapy.
      • Field T.
      • Hernandez-Reif M.
      • Hart S.
      • et al.
      Pregnant women benefit from massage therapy.
      benefits on pain, anxiety, or autonomic arousal. In contrast, studies of the intermediate or long-term effects of MT, lasting from one to two hours
      • Kaada B.
      • Torsteinbo O.
      Increase of plasma β-endorphins in connective tissue massage.
      • Weinrich S.
      • Weinrich C.
      The effects of massage on pain in cancer patients.
      or a few hours
      • Totch M.
      • Kahn J.
      • Walton T.
      • et al.
      Therapeutic massage intervention for hospitalized patients with cancer: a pilot study.
      to 24 hours,
      • Corner J.
      • Cawley N.
      • Hildebrand S.
      An evaluation of the use of massage and essential oil on the well-being of cancer patients.
      have yielded inconsistent results. From a methodological perspective, most studies used pretest-post-test research designs, with single time point measures. In studies with repeated measure designs, the measurement intervals varied from one or two hours to 24 hours.
      Given the limitation of the research, little is known about the longitudinal effects of MT. Therefore, the specific aims of this study were to: 1) describe the feasibility of implementing full-body MT in hospitalized Taiwanese cancer patients with bone metastases; and 2) describe the effects of massage, if any, across time, on pain intensity, anxiety, and physiological measures, including heart rate (HR) and mean arterial pressure (MAP).

      Methods

      Sample and Setting

      This pilot study was a quasi-experimental, one-group, pretest-post-test design with repeated measures. It was conducted at five inpatient oncology units in a Taiwanese teaching medical center. To be eligible for inclusion in the study, the patient had to be 1) 18 years or older; 2) diagnosed with radiologically evident bone metastases; 3) experiencing at least moderate metastastic bone pain with an intensity of 4 or more on a 0–10 scale; and 4) admitted to the inpatient unit at least 48 hours before data collection. Patients were excluded if they were 1) undergoing major surgeries or procedures during this admission; 2) having physical or psychological impairments; 3) experiencing any physical condition contraindicating MT, including generalized infection, thrombocytopenia (below 100,000/mm3), spinal cord compression syndrome, pathological fracture, or deep vein thrombosis.
      During a six-month period, 36 patients met the eligibility criteria (16% eligibility rate), and 191 patients were deemed ineligible because of the following: bone pain less than 4 (66%); physical-psychological distress (15%); pathological fracture (14%); and other (5%). Thirty patients consented to participate (17% refusal rate). The reasons for declining included having too much pain or allodynia (n=3), feeling physical discomfort (n=2), and lacking interest (n=1). No subjects withdrew from the study. The consenting sample included 19 women (63%) and 11 men (37%), with a mean age of 52 years (standard deviation [SD]=11.6) and an age range of 33–75 years. Table 1 displays the sample's demographic characteristics.
      Table 1Demographic Characteristics of the Sample (n=30)
      Characteristicsf (%)
      Age
      Mean±SD.
      51.7±11.6 (33
      Minimum.
      –75
      Maximum.
      )
      Gender
       Male11 (37)
       Female19 (63)
      Ethnicity
       Taiwanese21 (70)
       Haka5 (16)
       Mainland Chinese2 (7)
       Aborigines2 (7)
      Religion
       None8 (27)
       Buddhism or Taoism19 (63)
       Christian or Catholic3 (10)
      Education
       None7 (23)
       Primary7 (23)
       Junior high2 (7)
       High school11 (37)
       College or university3 (10)
      Cancer diagnosis
       Lung11 (36)
       Breast11 (36)
       Head and neck2 (7)
       Gastrointestinal cancer4 (14)
       Genitourinary cancer
      Time since initial diagnosis (years)
      Mean±SD.
      Current treatment regimen (>100%)
       Chemotherapy19 (63)
       Radiation therapy14 (47)
       Hormone therapy
      Life expectancy
      Life expectancy is assessed by the patient's primary physician.
       <6 months19 (63)
       >6 months
      Sites of bone metastases
      Sites of bone metastases is determined by results of the Tc-99m (bone scan).
      (>100%)
       Total number of sites of bone metastases
      Mean±SD.
      5.7±4.6 (1
      Minimum.
      –14
      Maximum.
      )
       Head, face, and neck15 (50)
       Shoulders and arms10 (33)
       Chest and rib cage14 (47)
       Upper back trunk21 (70)
      a Mean±SD.
      b Minimum.
      c Maximum.
      d Life expectancy is assessed by the patient's primary physician.
      e Sites of bone metastases is determined by results of the Tc-99m (bone scan).

      Procedures

      The study protocol involved three days (Table 2). On Day 1, the primary investigator (PI) explained the massage procedure and preintervention preparations after participants provided informed consent and completed initial pain and anxiety measures. On Day 2, baseline readings were obtained, a 45-minute full-body massage was administered, and postintervention measures were carried out at the designated intervals. On Day 3, the final postintervention measures and safety evaluation were carried out, along with the patients' perceptions of receiving the MT. Table 3 outlines the timeline for all measurements used in this study.
      Table 2Study Protocol for Cancer Patients with Bone Metastases
      DayMajor Procedures
      Preparation day (Day 1)
      • Explained the study procedures to eligible participants.
      • Obtained participants' informed consent.
      • Instructed the participants how to use the VAS for pain and anxiety measures and completed the initial assessment of the pain and anxiety measures.
      • Explained the massage procedure and preintervention preparations.
      Intervention day (Day 2)
      • Obtained three prebaseline readings of HR and MAP to allow participants to become accustomed to the digital probe of the pulse oximeter and the feeling of automatic inflation and deflation of the blood pressure cuff.
      • Obtained all baseline readings (T0) and premassage assessments.
      • Instructed participants to take an additional 10-minute rest while the PI drew the curtain, posted an “In Session” sign, dimmed the lights, and kept the door closed to ensure the stability of physiological measures and privacy, and minimize distractions.
      • Removed the physiologic devices from the participants to assure the thoroughness of massage administered.
      • A 45-minute full-body massage provided by the PI after the standardized massage protocol.
      • Immediately after the massage, encouraged the participants to quietly rest for an additional 30 minutes before administering postintervention measures across time at the designated intervals and safety evaluation.
      Follow-up day (Day 3, 16–18 hours after massage therapy)
      • Obtained the final postintervention readings of pain and anxiety, and safety evaluation around 10–11 am (T10).
      • Interviewed participants regarding perceptions of receiving the massage intervention.
      Table 3Timelines for Collecting Outcome Measures
      MeasuresT0MassageT1T2T3T4T5T6T7T8T9T10
      5”10”15”20”30”60”90”150”16–18 hours
      Time effects
      PPI-VAS×××××××××××
      Anx-VAS×××××××××××
      HR×××××××
      MAP×××××
      MSF-MPQ××
      Feasibility
      Safety×××
      Acceptability×
      Perception×
      X=collected timepoint for measures; T0=10 minutes of rest; T1=immediately after the massage; ”=minutes after intervention.

      Massage Intervention

      Because there are no licensed massage therapists available in Taiwan, and the safety and consistency of the intervention was concerned, the first author provided all massages. The first author is a registered nurse specializing in oncology nursing and was trained in MT for a four-month period. Each massage lasted from 38 to 50 minutes and followed a standardized protocol that was intended to target both highly innervated (i.e., head, hand, and feet) and less innervated (i.e., back and limbs) regions (Table 4). Five patients received a modified massage protocol, with omission of some massage motions, such as traction of neck and compression of shoulder. The massage was completed without any interruptions in 25 patients (84%); however, for 16% of the patients, the massage was interrupted because of phone calls, family members, or relative or physician visits.
      Table 4Massage Protocol
      The primary goal of the full-body massage was to decrease pain intensity and anxiety, and improve physiological relaxation for patients with metastatic bone pain during an inpatient hospitalization. To maximize therapeutic effects, this intervention included the following protocol:

        Premassage preparation

      • 1.
        The subject was instructed to have the evening meal at least one hour before the intervention or to have the meal after the intervention, remain in bed for at least 30 minutes immediately before the intervention, wear loose clothes, and empty his/her bladder.
      • 2.
        Each massage was performed between 4 and 5 pm (before the administration of regular analgesics in the evening) in his/her unit, with the door closed and curtain drawn.
      • 3.
        Before the intervention, the subject was instructed to lie first on his/her side or in a prone position, then face up, close his/her eyes, take three deep breaths, and refrain from any conversation during the entire intervention.
      • 4.
        Each massage session included unscented massage lotion as a skin lubricant.
      • 5.
        The subject was asked about their comfort with the room temperature. A fan or a heater was used based on the individual's comfort.
      • 6.
        The massage therapist assessed sites of metastases and any local skin conditions that would contraindicate a massage intervention.

        During massage intervention

      • 1.
        Each massage must be a direct hands-on and skin-to-skin manipulation of the soft tissue of the backside of the body and must last for 45 minutes, if possible.
      • 2.
        Each massage should include selected strokes with certain amount of time to eight different body areas, including head, neck, back, gluteus muscles, and the four extremities.
      • 3.
        Selected strokes included:
        • Effleurage: rhythmic, long, firm, and gliding stroke conforming to the contours of the body using the therapist's whole palm of the hand or the tips of the middle fingers, five to 10 times.
        • Light petrissage: rolling, squeezing, and kneading movement of the fingers and thumb done slowly and lightly using the therapist's tips of thumbs or three middle fingers, five to 10 times.
        • Nerve stroke: very light brushing of the skin with the therapist's fingers or full hand, three to five times.
        • Light compression: fingertips are used to lightly compress selected areas of tension using only mild to moderate pressure to the plantar surfaces, shoulders, palms, and possibly to the sacral area: 0–15 seconds duration.
      • 4.
        Modification of massage: administration of very light and gentle pressure with effleurage stroke in the bone metastases sites and avoidance of sites of superficial tumors, infection, and hyperesthesias.
      Considering that patients with bone metastases were theoretically at risk for pathological fractures, the safety of delivering massage was ensured by: 1) confirming sites of bone metastases with the patients' radiological tests and consultation with their primary physicians for the appropriateness of participation; 2) initiating a preintervention evaluation and modifying the intervention protocol with administration of very light and gentle pressure with effleurage stroke in the bony metastatic sites and avoidance of sites of superficial tumors; and 3) monitoring consistently for the presence of escalating pain during and after the massage.

      Instruments

      Feasibility

      An investigator-designed eligibility-screening tool and massage logs were used to document the rate of referral and reason for ineligibility, and the fidelity of giving and the acceptability of receiving massage, respectively. The perception of receiving massage was elicited with an audiotaped interview using three open-ended questions: “How do you feel about receiving massage?;” What's your general comment about receiving the massage?;” and “What physical environmental factors do you consider to be a good or bad for a massage experience?” An interrater agreement of qualitative responses was 96% for independent coding of 25 patients by the primary author and a co-author.

      Pain

      A single item of present pain intensity (PPI) with a vertical form of the visual analog scale (PPI-VAS) was used to examine the massage effects across time. The PPI-VAS wasa 100 mm line upward anchored by “pain as bad as it could be” and downward anchored by “no pain.” The choice of the vertical form of VAS was a cultural consideration, because Chinese people traditionally read and write vertically downward and from right to left. The vertical form results in smaller response errors than the horizontal one,
      • Aun C.
      • Lam Y.M.
      • Collett B.
      Evaluation of the use of visual analogue scale in Chinese patients.
      and is more sensitive and easier for patients to use.
      • Gift A.G.
      • Plaut D.M.
      • Jacox A.K.
      Psychologic and physiologic factors related to dyspnea in subjects with chronic obstructive pulmonary disease.
      • Gift A.G.
      Visual analogue scales: measurement of subjective phenomena.
      In this study, the test-retest reliability of the PPI-VAS was 0.97.
      The long-term massage effects (16–18 hours) on pain were examined with a modification both of the Short-Form McGill Pain Questionnaire
      • Melzack R.
      The short-form McGill Pain Questionnaire.
      and the Brief Pain Inventory (BPI)
      • Cleeland C.S.
      • Ryan K.M.
      Pain assessment: global use of the brief pain inventory.
      (MSF-MPQ). The MSF-MPQ consists of a body schematic outline to describe pain locations, 15 descriptive words with a 0–3 scale describing the quality of pain, and four items with a 0–10 scale describing pain intensity. The Cronbach's alpha for the quality of pain was 0.78 and the test-retest reliability of pain intensity was 0.82 in the current study.

      Anxiety

      A single item of anxiety-VAS from a six-item linear analog self-assessment profile of mood states
      • Sutherlan H.J.
      • Lockwood G.
      • Cunningham A.J.
      A simple, rapid method for assessing psychological distress in cancer patients: evidence of validity for linear analog scales.
      was adapted to examine the anxiety variable across time. It contained a 100 mm line anchored by the two extremes of anxiety, “not at all anxious” (downward) and “extremely anxious” (upward). The test-retest reliability of anxiety-VAS in this study was 0.96.

      Physiological Relaxation

      The Ohmeda Biox 3700 Pulse Oximeter was used to indirectly measure HR. Critikon DINAMAP 8100 Adult Vital Signs Monitor was used for the measurement of MAP. The accuracy of the oximeter was validated by comparing the manual count of the subject's apical pulse with the count from the oximeter, with 0.77 technical error of measurement (TEM=Sdi22n).
      • Engstrom J.L.
      Assessment of the reliability of physical measures.
      The precision and stability of this measure was 0.48 TEM by comparing readings from the oximeter and the DINAMAP, and 0.63 TEM for comparison of two readings one minute apart, respectively.

      Data Analysis

      We examined massage effects on pain intensity, anxiety, HR, and MAP across time with paired t-tests, separately, comparing each value of the postintervention measures (T1–T10) with the value of the preintervention measures (T0). A Bonferroni correction was used to control cumulative Type I error for the familywise comparisons, and the alpha level at 0.005 was considered significant (0.05/10×4=0.005; 10 comparison times and four classes of measures). A repeated-measures analysis of variance (RANOVA) with an alpha level at 0.013 (0.05/4=0.013, four classes of measures) was used to capture the patterns of massage effects on outcome measures over time.

      Results

      The results of the statistical analyses are summarized in Table 5.
      Table 5Means, Standard Deviations, and Results of Paired t-Tests for PPI-VAS, Anxiety-VAS, HR, and MAP byTimePoints
      Time PointPPI-VAS/100 mmAnxiety-VAS/100 mmHR/beats per minuteMAP/mm Hg
      Mean±SDtP
      Bonferroni corrected P-value (P=0.05/10×4=0.001).
      Mean±SDtP
      Bonferroni corrected P-value (P=0.05/10×4=0.001).
      Mean±SDtPMean±SDtP
      T05.4±1.25.4±2.083.7±17.288.8±14.2
      T1 (immediate)3.1±1.416.50.0002.5±2.18.90.00082.9±15.10.90.3590.1±14.5−1.20.26
      T2 (5 minutes after T1)2.9±1.512.60.0002.3±1.710.90.00083.4±15.60.30.81
      T3 (10 minutes after T1)2.9±1.612.10.0002.3±1.810.60.00083.0±15.50.50.6190.4±14.7−1.30.22
      T4 (15 minutes after T1)2.7±1.711.10.0002.1±1.711.30.00083.8±15.1−0.10.93
      T5 (20 minutes after T1)2.6±1.810.80.0002.2±1.910.40.00083.5±15.40.20.8890.6±13.3−1.10.29
      T6 (30 minutes after T1)2.8±1.89.30.0002.2±1.710.10.00082.3±15.81.40.1691.3±14.0−1.70.10
      T7 (60 minutes after T1)3.1±1.98.60.0002.3±1.99.30.000
      T8 (150 minutes after T1)3.0±1.89.50.0002.3±2.09.80.000
      T9 (2.5 hours after T1)3.2±2.07.90.0002.5±2.08.90.000
      T10 (16–18 hours after IT0)3.8±2.04.00.0003.0±2.55.70.000
      df (29).
      a Bonferroni corrected P-value (P=0.05/10×4=0.001).

      Present Pain Intensity-Visual Analog Scale and Anxiety-Visual Analog Scale

      All nine postintervention measures of PPI-VAS and anxiety-VAS over a 2.5-hour period were statistically reduced as compared with the baseline (T0). RANOVA revealed statistically significant time effects on PPI-VAS [F(10, 20)=24.6, P<0.001] and anxiety-VAS [F(10, 20)=10.3, P<0.001] across time. As shown in Fig. 1, PPI-VAS declined immediately and consistently for 20 minutes, with the lowest mean at T5 (20″), then gradually began to return at T6 (30″), but stayed below the baseline throughout the 16- to 18-hour period. Anxiety-VAS declined immediately after the intervention and consistently declined for 15 minutes with the nadir at T4 (15″), then gradually began to return at T5 (20″), but not regressing to the baseline throughout the 16- to 18-hour period (Fig. 2).
      Figure thumbnail gr2
      Fig. 2Mean scores for anxiety-VAS over time.

      Modified Short-Form McGill Pain Questionnaire

      Overall, the mean number of pain locations at 16–18 hours was statistically decreased from baseline [t(29)=2.1, P<0.04]. The mean total score of the quality of pain at 16–18 hours approached statistical difference from baseline (T0) [t(29)=1.8, P<0.08]. The pain descriptor of gnawing pain was statistically reduced and hot-burning pain was trending toward a reduction.
      Both the present pain and worst pain were statistically different between the T0 and 16–18 hour measures. Least pain and acceptable level of pain intensity, however, were statistically nonsignificant.

      Physiological Relaxation

      None of the six postintervention measures of HR was statistically different from baseline (T0). Similarly, no time effect or linear trend on HR was noted. HR fluctuated across time, with a slight decrease immediately after massage, then an increase at T2, followed by decreases at T3 and T5, and end with a great decrease at T6 (30″). Furthermore, we found no statistically significant difference between baseline (T0) and any of the four postintervention measures of MAP. Results from RANOVA also showed no time or linear trend effect on MAP [F(1, 29)=1.8, P=0.19]. MAP consistently increased across time, but this trend effect is not at a statistically significant level.

      Perceptions of Receiving Massage

      The major themes elicited from the taped interviews indicated that patients described MT as: muscle relaxing or generalized relaxing, comfortable or good, secure, with improved pain, sleep, circulation, and decreased anxiety or stress (Table 6). Examples of comments included: “It made me comfortable and let me feel that I was back to Mom's womb, a sense of security;” “I almost forgot the existence of pain and it was like a sort of distraction;” “My arms and shoulder were not as tense as before;” “Although massage couldn't decrease my pain as much as analgesics, it did give a lot of relief of my emotional tension;” and “It helped me easily enter deep sleep because I felt generalized comfort after the massage.”
      Table 6Themes and Selected Quotations of the Benefits of Receiving MT
      Themesf (%)Selected Quotations
      Muscle relaxation26 (87)To me, the best benefit of having massage was muscle relaxation.
      Less pain26 (87)At the beginning, I really doubted the effects of massage, but you know, my pain was not as painful as before and my pain has been reduced from 6 to 3 or4.
      Feeling comfort or good24 (80)It was hard to describe the feeling, but overall, I just felt so good about something that couldn't be gained with money.
      Generalized relaxation19 (73)Due to the skin-to-skin interpersonal touch, massage helped me feel more calm, easy, and relaxed in general.
      Reducing anxiety orstress13 (43)My anxiety level had been decreased a little bit, as my focus was on enjoying the massage rather [than thinking] too much. I think it was a function of distraction.
      Improving sleep11 (37)I used to have an insomnia problem; somehow, I slept quite early last night and easily fell in[to] deep sleep. Overall, I got a good night's sleep after the massage.
      Feeling secure or safe8 (27)I just felt secure because of the tactile touch and having a professional around. It was just like [a] return to Mom's womb.
      Improving circulation6 (20)The massage might stimulate my circulation, so I felt generalized warm[th]. Also, I felt more energetic, because it boosted my circulation.
      Reducing other symptoms distress4 (14)To me, the most precious benefit of massage was to ease the tightness [in my] chest, so I felt more [relaxed] while I breathed.
      Feeling warmth [cozy] while staying in the hospital2 (7)The hospital was not as cold as before and it was more homelike. The bed was not a hospital bed, rather it was my home bed.

      Discussion

      This is the first study to examine the longitudinal effect of massage with short, multiple-interval measures in hospitalized patients with metastatic bone pain. To date, most studies of massage measured outcomes at only one or two time points postintervention, whereas this study demonstrated that a single session of full-body massage decreased pain intensity and anxiety across time. Massage was shown to have immediate, short-term (10–20 or 30 minutes), intermediate (1–2.5 hours) and even long-term (16–18 hours) effects on decreasing PPI and anxiety, but there were no significant time effects for any of the physiological measures used (HR and MAP). More importantly, this study demonstrated the feasibility of implementing a full-body massage, and the acceptability and perceptions of benefit in patients with metastatic bone pain.
      Similar to other studies,
      • Grealish L.
      • Lomasney A.
      • Whiteman B.
      A nursing intervention modify the distressing symptoms of pain and nausea inpatients hospitalized with cancer.
      • Field T.
      • Peck M.
      • Krugman S.
      • et al.
      Burn injuries benefit from massage therapy.
      • Cassileth B.R.
      • Vickers A.J.
      Massage therapy for symptom control: outcome study at a major cancer center.
      • Hernandez-Reif M.
      • Martinez A.
      • Field T.
      • et al.
      Premenstrual symptoms are relieved by massage therapy.
      we found immediate and short-term effects on pain and anxiety. Yet, the degree of reduction in this study was greater than 30%–39% reported in other studies.
      • Grealish L.
      • Lomasney A.
      • Whiteman B.
      A nursing intervention modify the distressing symptoms of pain and nausea inpatients hospitalized with cancer.
      • Field T.
      • Peck M.
      • Krugman S.
      • et al.
      Burn injuries benefit from massage therapy.
      This discrepancy perhaps is a function of the baseline values. In Grealish et al.'s study
      • Grealish L.
      • Lomasney A.
      • Whiteman B.
      A nursing intervention modify the distressing symptoms of pain and nausea inpatients hospitalized with cancer.
      for instance, the average of pain intensity at baseline was mild pain (25.1±21.7 mm), whereas in our study, it was moderate pain (54±12 mm). In contrast, the results of intermediate effect are inconsistent with previous studies,
      • Kaada B.
      • Torsteinbo O.
      Increase of plasma β-endorphins in connective tissue massage.
      • Weinrich S.
      • Weinrich C.
      The effects of massage on pain in cancer patients.
      in which other authors claimed that massage effect did not last more than an hour. We speculate that because the baseline was somewhat lower (31±28 mm),
      • Weinrich S.
      • Weinrich C.
      The effects of massage on pain in cancer patients.
      the effect was less than that in our study.
      We found a long-term effect (16–18 hours after massage) on reduction of pain intensity and anxiety from baseline. However, the interpretation of our findings should be cautious as other factors might account for this result, such as variation of analgesic consumption and circadian effect on pain and anxiety. In our study, the baseline was measured at some point between 4 and 5 pm, suggesting a time three to four hours postanalgesic, whereas the 16- to 18-hour measurement was taken at some point between 10 and 11 am, which would be one to two hours after medication. Thus, analgesic concentration might have been lower at the baseline compared with the 16- to 18-hour measure, and points to the need to control this variable in future studies.
      In addition, the rhythmicity of pain and anxiety for some oncology patients could vary throughout the day. Researchers have found that patients reported improvement in their fatigue, associated with decreased pain and depression, after a good night of sleep, and reported increasing symptom distress later in the day.
      • Blesch K.S.
      • Paice J.A.
      • Wickham R.
      • et al.
      Correlates of fatigue in people with breast or lung cancer.
      • Lee K.
      • Hicks G.
      • Nino-Murcia G.
      Validity and reliability of a scale to assess fatigue.
      • Miaskowski C.
      • Lee K.A.
      Pain, fatigue, and sleep disturbances in oncology outpatients receiving radiation therapy for bone metastasis: a pilot study.
      The baseline and 16- to 18-hour measurements in this study, however, were measured at different times of the day, the late afternoon, and midmorning, respectively. Massage effects potentially were influenced by the time of the day and the intervention itself.
      In contrast to PPI-VAS, the quality of pain measured with MSF-MPQ in this study was not statistically different compared with the baseline, except the gnawing and hot-burning pain. The lack of significance might be related to a floor effect, sensitivity of the tool, timing of measurements, or a combination of these issues. With the same mean pain intensity at baseline (5.3 on a 0–10 scale), our mean quality of pain after intervention (13.1±5.7 on a 0–45 scale) was much lower than that in Field et al.'s study
      • Field T.
      • Peck M.
      • Krugman S.
      • et al.
      Burn injuries benefit from massage therapy.
      (23.6, SD not reported). We also hypothesized that this low magnitude of quality of pain was the result of the cultural sensitivity of the tool, indicating that MSF-MPQ does not seem to adequately reflect the nature of bone pain in this Taiwanese population.
      The most striking and prevalent perceived benefit of receiving massage was subjective muscle relaxation or generalized relaxation. However, this subjective relaxation report was not reflected in decreased HR and MAP. Potential reasons accounting for this discrepancy may include the nature of disease progression of this population, environmental distractions, and measurement methods (manual vs. automatic measures). We postulate that patients in our study tended to have a high level of physiological arousal because of the progression of disease, their moderate to severe bone pain, and the effects of anticancer treatments during the current hospitalization. Environmental distractions could also account for the physiological arousal, as 97% of patients received their massage in a hospital room with a two- or three-bed capacity. Thirty percent of patients suggested improving the privacy and decreasing noise level while delivering massage. In addition, most studies, but one,
      • Meeks S.
      Effects of slow stroke back massage on relaxation in hospice clients.
      manually measured HR and MAP and mentioned strategies to obtain a stable baseline of those measures; hence, the possibility of measurement error from previous studies cannot be completely ignored.
      In terms of feasibility, no patient reported any adverse effects as a result of the study procedures. Forty-seven percent of patients were, however, hesitant to participate in the study when initially approached, because they were concerned about the safety of the massage. The predominant reason for refusal given in our study was the patients' experiences with allodynia, and their biggest concern was whether bone pain would be provoked by light tactile touch, such as massage. Therefore, we would suggest assessment of allodynia at the initial screening procedure and consider this as another exclusion criteria.
      Despite the impressive time effects of massage on pain intensity and anxiety found in this study, cautious interpretation of study findings is warranted because of some limitations, such as the convenience sampling, small sample size, and a quasi-experimental design without a control condition, no blinding of the data collector, and response bias. With a repeated measure design using short-term intervals, a response bias might have occurred because of the memory-bias effect, remembering the location of the mark on the prior VAS. Given the limitations noted in this study, future studies with single blinding and randomized allocation of study patients and a control group are important to validate the effectiveness of massage interventions in patients with metastatic bone pain. If successful, these data will give clinicians great confidence in appropriately implementing the massage intervention in conjunction with pharmacological treatments for improving metastatic bone pain.

      Acknowledgments

      The authors wish to acknowledge all participants for their dedication in improving the quality of care for patients with metastatic bone pain and their time and effort contributed to this study. A special thanks goes to Chang Gung Memorial Hospital (CGMH) for its administrative support and assistance, and the health care providers from oncology units for their generous patient consultation and referral.

      References

        • Jane SW
        • Wilkie DJ
        • Gallucci BB
        • Beaton RD
        • Huang HY
        Systematic review of massage intervention for adult patients with cancer: a methodological perspective.
        Cancer Nurs. 2008; 36: E24-E35
        • Abrams H.L.
        • Spiro R.
        • Goldstein N.
        Metastases in carcinomas: analysis of 1000 autopsied cases.
        Cancer. 1950; 3: 74-85
        • Mercadante S.
        Prevalence, causes and mechanisms of pain in home-care patients with advanced cancer.
        Pain Clin. 1994; 7: 131-136
        • Nielsen O.S.
        • Munro A.J.
        • Tannock L.F.
        Bone metastases: pathophysiology and management policy.
        Clin J Oncol Nurs. 1991; 9: 509-524
        • Coleman R.E.
        Skeletal complications of malignancy.
        Cancer. 1997; 80: 1588-1594
        • Banning A.
        • Sjogren P.
        • Henriksen H.
        Pain causes in 200 patients referred to a multidisciplinary cancer pain clinic.
        Pain. 1991; 45: 45-48
        • Chang M.Y.
        • Wang S.Y.
        • Chen C.H.
        Effects of massage on pain and anxiety during labour: a randomized controlled trial in Taiwan.
        J Adv Nurs. 2002; 38: 68-73
      1. Chin CC. The effects of back massage on surgical stress responses and postoperative pain. Unpublished doctoral dissertation, Case Western Reserve University, Ohio, 1999.

      2. Huang TW. Comparison of the effects of back massage and relaxation training in reducing chemotherapy induced fatigue. Unpublished Thesis, Taipei Medical University, Taipei, Taiwan; 2001.

        • Ahles T.A.
        • Tope D.M.
        • Pinkson B.
        • et al.
        Massage therapy for patients undergoing autologous bone marrow transplantation.
        J Pain Symptom Manage. 1999; 18: 157-163
        • Ferrell-Torry A.T.
        • Glick O.
        The use of therapeutic massage as a nursing intervention to modify anxiety and the perception of cancer pain.
        Cancer Nurs. 1992; 16: 93-101
        • Meeks S.
        Effects of slow stroke back massage on relaxation in hospice clients.
        Image J Nurs Scholarsh. 1993; 25: 17-21
        • Post-White J.
        • Kinney M.E.
        • Savik K.
        • et al.
        Therapeutic massage and healing touch improve symptoms in cancer.
        Integr Cancer Ther. 2003; 2: 332-344
        • Sims S.
        Slow stoke back massage for cancer patients.
        Nurs Times. 1986; 82: 47-50
        • Wilkinson S.
        Aromatherapy and massage in palliative care.
        Int J Palliat Nurs. 1995; 1: 21-30
        • Wilkie D.J.
        • Kampbell J.
        • Cutshall S.
        • et al.
        Effects of massage on pain intensity, analgesics and quality of life in patients with cancer pain: a pilot study of a randomized clinical trial conducted within hospice care delivery.
        Hosp J. 2000; 15: 31-53
        • Ironson G.
        • Field T.
        • Scafidi F.
        • et al.
        Massage therapy is associated with enhancement of the immune system's cytotoxic capacity.
        Int J Neurosci. 1996; 84: 205-217
        • Grealish L.
        • Lomasney A.
        • Whiteman B.
        A nursing intervention modify the distressing symptoms of pain and nausea inpatients hospitalized with cancer.
        Cancer Nurs. 2000; 23: 237-243
        • Field T.
        • Peck M.
        • Krugman S.
        • et al.
        Burn injuries benefit from massage therapy.
        J Burn Care Rehabil. 1998; 19: 241-244
        • Field T.
        • Hernandez-Reif M.
        • Hart S.
        • et al.
        Pregnant women benefit from massage therapy.
        J Psychosom Obstet Gynaecol. 1999; 20: 31-38
        • Kaada B.
        • Torsteinbo O.
        Increase of plasma β-endorphins in connective tissue massage.
        Gen Pharmacol. 1989; 20: 487-489
        • Weinrich S.
        • Weinrich C.
        The effects of massage on pain in cancer patients.
        Appl Nurs Res. 1990; 3: 140-145
        • Totch M.
        • Kahn J.
        • Walton T.
        • et al.
        Therapeutic massage intervention for hospitalized patients with cancer: a pilot study.
        Altern Complement Ther. 2003; 9: 117-124
        • Corner J.
        • Cawley N.
        • Hildebrand S.
        An evaluation of the use of massage and essential oil on the well-being of cancer patients.
        Int J Palliat Nurs. 1995; 1: 67-73
        • Aun C.
        • Lam Y.M.
        • Collett B.
        Evaluation of the use of visual analogue scale in Chinese patients.
        Pain. 1986; 25: 215-221
        • Gift A.G.
        • Plaut D.M.
        • Jacox A.K.
        Psychologic and physiologic factors related to dyspnea in subjects with chronic obstructive pulmonary disease.
        Heart Lung. 1986; 15: 595-601
        • Gift A.G.
        Visual analogue scales: measurement of subjective phenomena.
        Nurs Res. 1989; 38: 286-288
        • Melzack R.
        The short-form McGill Pain Questionnaire.
        Pain. 1987; 30: 191-197
        • Cleeland C.S.
        • Ryan K.M.
        Pain assessment: global use of the brief pain inventory.
        Ann Acad Med. 1994; 23: 129-138
        • Sutherlan H.J.
        • Lockwood G.
        • Cunningham A.J.
        A simple, rapid method for assessing psychological distress in cancer patients: evidence of validity for linear analog scales.
        J Psychosoc Oncol. 1989; 7: 31-43
        • Engstrom J.L.
        Assessment of the reliability of physical measures.
        Res Nurs Health. 1988; 11: 383-389
        • Cassileth B.R.
        • Vickers A.J.
        Massage therapy for symptom control: outcome study at a major cancer center.
        J Pain Symptom Manage. 2004; 28: 244-249
        • Hernandez-Reif M.
        • Martinez A.
        • Field T.
        • et al.
        Premenstrual symptoms are relieved by massage therapy.
        J Psychosom Obstet Gynecol. 2001; 21: 9-15
        • Blesch K.S.
        • Paice J.A.
        • Wickham R.
        • et al.
        Correlates of fatigue in people with breast or lung cancer.
        Oncol Nurs Form. 1991; 18: 81-87
        • Lee K.
        • Hicks G.
        • Nino-Murcia G.
        Validity and reliability of a scale to assess fatigue.
        Psychiatry Res. 1991; 36: 291-298
        • Miaskowski C.
        • Lee K.A.
        Pain, fatigue, and sleep disturbances in oncology outpatients receiving radiation therapy for bone metastasis: a pilot study.
        J Pain Symptom Manage. 1999; 17: 320-332