Wilkinson, S., Barnes, K., & Storey, L. (2008). Massage for symptom relief in patients with cancer: Systematic review. Journal of Advanced Nursing, 63, 430–439.

To assess the evidence regarding massage as a means of improving the physical and psychological well-being of patients with cancer

• Databases searched were Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, CINAHL, EMBASE, Allied and Complementary Medicine (AMED), System for Information on Grey Literature (SIGLE), PsycINFO, and CancerLIT. Investigators also searched dissertation abstracts.
• Year of publication, as defined by the search strategy, varied by database. The most recent date was September 2006.
• Search keywords were massage, aromatherapy, therapeutic touch, essential oil, volatile oil and cancer or neoplasm or oncolog* or palliate* or terminal or hospice.
• Studies were included in the review if they
  • Were randomized controlled trials (RCTs)
  • Involved adult patients with cancer who were receiving care in any healthcare setting
  • Included any type of massage, with a systemic goal, provided by a therapist with a recognized qualification
  • Produced patient-reported outcomes, reported by means of reliable and valid assessment tools, of physical or psychological symptoms and quality of life
  • Included the means by which the reliability and validity of assessment tools were evaluated.
• Studies were excluded if the purpose of the massage was a specific localized physical effect without a specific systemic aim. For example, a study involving prostatic massage to obtain a semen sample was excluded.

After elimination of duplicates from the initial search, investigators considered 1,321 references. Of these, investigators chose 10 studies for analysis, using the Jadad scoring approach and CONSORT Statement to appraise study quality. Investigators also used sample size and duration of follow-up to evaluate studies. Given the range of massage techniques and patients, meta-analysis was impossible. Three investigators reviewed studies independently.

• The review reported on 386 patients from nine studies.
• One of the initial references was a duplicate report of a single study.
• One study was based on personal communication, rather than a final published report.
• Study samples included more females than males.
• Various cancer diagnoses were represented.
• Sample sizes ranged from 6 to 87 participants.
• Samples included patients receiving surgery, radiotherapy, chemotherapy, and combined treatments.

• Patients were assessed 4–16 times, with all patients assessed before and after the intervention or control period.
• Follow-up was limited. One study assessed patients three weeks after the last massage.
• Five trials assessed psychological outcomes, including outcomes related to anxiety and depression.
• Seven trials assessed physical symptoms and quality of life.
• One study found a significant post-massage decrease in pain for males only. One study found a significant reduction in pain after massage but not after the control period. One study showed pain reduction after the first and third massage, but not after the second and fourth, and a nonsignificant trend of greater pain improvement in the massage group than in the group that did not receive massage.
• One study showed significant reduction of depression with massage, and one study showed no change.
• Three studies demonstrated significant post-massage improvement in anxiety.
• One study associated massage with short-term reduction in nausea.
• One study associated massage with reduction in fatigue.
• One study associated massage with improvement in quality of life.
• Most studies reported decreases in anxiety and other psychological benefits; however, results relating to depression were equivocal across studies.
• Adverse effects included a single case of skin rash in one trial and, in another trial that involved the use of essential oils, a higher incidence of digestive problems.

The authors stated that results were inconclusive; however, reported results showed that massage had a preponderantly positive effect in regard to anxiety reduction.

• This review did not discuss the patients' treatment phase, so effects relating to specific groups of patients were unspecified.
• Authors did not know how and if variations in massage style affected outcomes.
• Trials that included the use of oils used different oils, so the authors could draw no conclusions regarding the effect of aromatherapy.
• No evidence was available to assess long-term effects of massage. Authors found that research of long-term effects, as reported in the studies, was of low quality.
• Although inclusion criteria demanded that all studies be RCTs, results suggest that one study did not include a separate control group.

Two studies included in the review used a crossover design, and one of these showed significant results after massage but not after the control period. This suggests that crossover design may be particularly applicable to this area of research and that the effects of massage are immediate and time sensitive.