Control groups in clinical trials of non-pharmacological interventions for cancer pain: a meta-research study

Cancer pain is a common challenge in patients with cancer, arising from tumor progression or adverse effects of treatment. For its management, non-pharmacological interventions have been investigated through randomized controlled trials (RCTs), which require appropriate comparative strategies (placebo or sham) to isolate the effects of the intervention. However, limitations in blinding inherent to this type of intervention may influence the interpretation of results. The objective of this study was to identify the strategies used in the control groups of randomized controlled trials evaluating non-pharmacological interventions for cancer pain. This is a meta-research study (research on how science is produced) conducted in the MEDLINE and LILACS databases, using MeSH/DeCS descriptors and free-text terms related to “Cancer pain” and “Complementary therapies,” with no restrictions on language or year of publication. Non-randomized studies and those involving surgical interventions were excluded. Study selection and data extraction were performed in duplicate. Information on the type of blinding and the strategies used in the control groups was extracted. The search retrieved 311 articles; 52 were included after screening, of which 22 had full-text availability. Of these, 15 were single-blind, 6 were open-label, 1 was described as double-blind, and 12 discussed the impact of performance bias. The most frequently evaluated interventions included acupuncture, therapeutic massage, and music therapy. Control groups varied between usual care, placebo, no intervention, and sham procedures. It is concluded that there is substantial variability in control strategies and that the absence of discussion regarding performance bias may compromise the interpretation of results and clinical decision-making.