Article Text
PDF
Abstract
Introduction Pain management in patients with cancer is a critical aspect of oncological care, yet remains challenging with current pharmacological therapies. Non-pharmacological interventions, offering potential benefits without the adverse effects of drugs, have gained attention. However, the effectiveness of these diverse non-pharmacological interventions is not well understood, creating a gap in clinical practice. This study aims to conduct a systematic review and network meta-analysis (NMA) to evaluate the efficacy of various non-pharmacological interventions for pain management in patients with cancer, providing evidence-based guidance for clinicians and patients.
Methods and analysis A systematic review and Bayesian NMA will be performed. To assess the efficacy of interventions for cancer pain, we will search six electronic databases: Cochrane Library, Web of Science, PubMed, EMBASE, PsycINFO and the Cumulative Index to Nursing and Allied Health Literature, focusing on identifying randomised controlled trials. Literature screening should be independently performed by two reviewers. A NMA will evaluate the efficacy of various non-pharmacological interventions for cancer pain. A second NMA will compare the efficacy of different non-pharmacological interventions in relieving pain interference in patients with cancer pain. Bayesian 95% credible intervals will be used to estimate the pooled mean effect size for each treatment, and the surface under the cumulative ranking area will be employed to rank the effectiveness of the treatments.
Ethics and dissemination Ethical approval is not required for this systematic review of the published data. Findings will be disseminated via peer-reviewed publication.
PROSPERO registration number CRD42024483025.
- psychosocial intervention
- cancer pain
- network meta-analysis
- physical therapy modalities
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
The study employs a Bayesian network meta-analysis approach, allowing for a comprehensive comparison of multiple non-pharmacological interventions simultaneously, which enhances the depth and applicability of the findings.
Using six major electronic databases ensures a wide coverage of the literature, increasing the likelihood of capturing relevant randomised controlled trials and enhancing the robustness of the systematic review.
The employment of the surface under the cumulative ranking area to rank the interventions offers a clear, hierarchical understanding of treatment effectiveness, facilitating straightforward clinical decision-making.
Due to the nature of network meta-analysis, this study focuses solely on randomised controlled trials, which may exclude relevant studies with different designs that could offer additional insights into the efficacy of the interventions.
The reliance on published studies in English might overlook significant research conducted in other languages, potentially introducing language bias into the findings.
Introduction
Pain in patients with cancer is a common symptom, characterised not only by its high prevalence but also by its significant impact on the patients.1–3 A systematic review has indicated varying prevalence rates of pain among patients with cancer.1 The prevalence of pain is reported to be 66.4% in patients with advanced, metastatic or terminal stages of cancer, and 50.7% across all stages of the disease.1 Furthermore, approximately 38% of patients experience moderate to severe pain.1 Beyond its high prevalence, cancer pain causes considerable physical discomfort and affects psychological and social functions, leading to symptoms such as fatigue, anxiety and depression.2–4 This, in turn, reduces the quality of life of patients and impacts their work efficiency.2 The causes of pain in patients with cancer are diverse, including tumour growth, side effects of treatment, among others.5 6 The pain experience also varies depending on the type of cancer, its stage and individual factors.6 Therefore, it is essential to adopt personalised pain management approaches for patients.
Clinical interventions for alleviating pain in patients with cancer include both pharmacological and non-pharmacological methods. In terms of pharmacological interventions, the efficacy of opioid medications and non-steroidal anti-inflammatory drugs (NSAIDs) has been widely validated, making them standard treatments in clinical practice.7 8 Building on this, the WHO’s analgesic ladder provides comprehensive guidelines for a rational use of these medications, recommending a stepwise approach to cancer pain management.9 This strategy starts with non-opioid analgesics, such as NSAIDs and acetaminophen, for mild pain, then moves to weak opioids for persistent discomfort, and finally employs strong opioids, like morphine, for severe pain.9 However, pharmacological interventions have many side effects and limitations, such as causing gastrointestinal reactions and increasing cardiovascular risks.7 8 As a result, non-pharmacological treatments are gaining increasing attention. These non-pharmacological interventions include psychological therapies (such as cognitive behavioural therapy and mindfulness), physical interventions (like exercise and acupuncture) and assistive techniques (such as guided imagery and yoga).10–12 Research indicates that certain non-pharmacological methods may contribute to pain relief in patients with cancer and potentially enhance their overall quality of life.13 Therefore, it is essential to identify and select effective non-pharmacological interventions.
Network meta-analysis (NMA) can synthesise both direct and indirect evidence from various studies regarding treatment methods, including those not directly compared in randomised controlled trials (RCTs).14 This enables researchers to rank the relative efficacy of a range of non-pharmacological interventions.14 This approach addresses a major limitation of traditional meta-analysis, which can only compare two treatments at a time, thus enabling a more comprehensive assessment of the effectiveness of various non-pharmacological methods for alleviating cancer pain. Consequently, compared with the studies by Ruano et al and Cuthbert et al,11 12 this research is expected to reveal specific differences in effectiveness among these methods and provide a ranked assessment of their efficacy, offering more comprehensive evidence to support clinical decision-making. Furthermore, the literature search for these two traditional meta-analyses focusing on non-pharmacological interventions for cancer pain concluded in 2020. Following the recommendations of Cochrane, it is prudent to update systematic reviews when they surpass a certain age, considering their academic value and practical utility.15 16
Methods and analysis
This systematic review has been officially registered with PROSPERO (registration number CRD42024483025) and its protocol reporting has been following the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols items.17 We commit to conducting the systematic review in alignment with the directives provided in the Cochrane Handbook for Systematic Reviews of Interventions.18 Should there be any necessary modifications to the protocol, these will be duly updated through PROSPERO.
Inclusion and exclusion criteria
Studies will be eligible for inclusion if they meet the following PICOS elements:
P (Population): the study will include adults aged 18 and over who have been diagnosed with cancer and suffer from pain caused by the cancer, with no restrictions on gender.
I (Intervention): to systematically assess the impact of various non-pharmacological interventions on pain management among patients with cancer, we will include psychological interventions such as cognitive behavioural therapy (CBT), relaxation training and mindfulness meditation, designed to help manage and alleviate pain through psychological regulation. Additionally, we will incorporate physical therapies like thermotherapy, cryotherapy and massage therapy that alleviate pain through physical contact and temperature variations. Complementary therapies to be included in our study are music therapy, aromatherapy, and acupressure, which stimulate the senses and promote relaxation to reduce pain. All interventions selected for this study will be non-invasive and function independently of pharmacological effects.
C (Control): given the comprehensive nature of NMA, this study will impose no restrictions on the choice of comparison groups. This approach will allow the inclusion of a wide range of control interventions, from standard treatments and placebos to active therapies, thereby facilitating a broad evaluation of the relative therapeutic effects of various interventions.
O (Outcome): the reduction of pain related to cancer is to be measured using the following validated methods—Visual Analogue Scale, Brief Pain Inventory, The Numeric Rating Scale, Multidimensional Pain Inventory and The Numeric Rating Scale. The relief of pain interference in patients with cancer is to be measured using the following validated methods—Multidimensional Pain Inventory, Brief Pain Inventory, The Psychological Inflexibility in Pain Scale, The Pain Interference Index, Multidimensional Pain Inventory, 10-cm Visual Analogue Scales.
S (Study design): the study will focus on RCTs published in English. Our literature search will be confined to publications in English, primarily due to the limitations of our team’s resources and language proficiency. Given that English is the predominant language in our field of research and covers the vast majority of innovative and comprehensive studies, this approach will help ensure manageability and high quality of data, thereby maintaining rigorous analytical standards. This practice is consistent with similar language restrictions adopted within our field.
Studies will be excluded from our systematic review if they meet the following criteria: (1) published in languages other than English; (2) not an RCT design; (3) use or apply medical or pharmacological interventions; (4) include patients awaiting biopsy/diagnosis or who have already overcome the disease; (5) solely include children or adolescents; (6) involve pain studies related to pathologies other than cancer or diseases other than cancer; (7) have excessively broad inclusion criteria, where participants do not meet the standards for this systematic review; (8) include participants without a basic level of education or cognitive abilities impaired by disease or any form of mental disability; and (9) do not include any outcomes relevant to the purpose of this review.
Data collection and search methodology
This systematic review is scheduled to be conducted over a period of approximately 9 months, aiming for completion by September 2024. It will entail a comprehensive search of electronic databases, conducted by two independent reviewers. The literature retrieval will cover the period from the inception of the databases until July 2024. Relevant studies will be identified from major databases including the Cochrane Library, Web of Science, PubMed, EMBASE, PsycINFO and the Cumulative Index to Nursing and Allied Health Literature (CINAHL). In addition, reference lists of selected articles and related review articles will be manually examined to uncover any potential relevant publications. The comprehensive search strategy is detailed in online supplemental appendix A.
Supplemental material
Screening and data extraction
We will use EndNote X9 software to manage the database search results. The first step is to remove duplicate studies using the software. Then, two independent reviewers will screen the titles and abstracts of the retrieved studies according to the predefined criteria and exclude those that do not meet the criteria. Next, the same reviewers will read the full texts of the remaining studies and further exclude those that do not meet the criteria, resulting in the final selection of studies. The reviewers will work independently and will discuss any disagreements. If no consensus is reached, a third reviewer will arbitrate the disputed outcome.
Assessment of risk of bias
The researchers will use the Cochrane risk of bias tool to assess the methodological quality of the included studies.19 Two reviewers will independently evaluate the included studies. Any discrepancies in the evaluation will be resolved through discussions or by consulting a third reviewer. Each item will be assessed as having low, unclear or high risk of bias. The Cochrane risk of bias tool will evaluate the included studies in the following seven areas: random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias) and other bias.
Missing data
In instances where data are presented as medians or using alternative measures of variability, these will be transformed into mean and SD values using established mathematical formulae.20 For data depicted solely in graphical format, without accompanying numerical details in the text, extraction will be performed using ImageJ software (available at https://imagej.nih.gov/ij/).21 This process involves measuring the pixel length of graph axes for calibration, followed by determining the pixel length corresponding to the data points of interest.21 In cases where direct data extraction proves unfeasible, attempts will be made to obtain the necessary information from the study authors. Any arising discrepancies during the data extraction process will be resolved through consultation with an appointed adjudicator.
Statistical analysis
Effect measurements
For continuous outcomes, if the outcome is measured using the same scale, the mean difference and the corresponding 95% CI will be used. If different scales are used to measure related outcomes, the standardised mean difference (SMD) will be calculated.18 Subsequently, we will conduct pairwise meta-analyses and use a random-effects model for each comparison, followed by the generation of network graphs and evaluation to determine the feasibility of the NMA. Afterwards, we will conduct NMA analysis using Bayesian methods.
Network meta-analysis
The execution of the NMA will use a Bayesian framework employing Markov chain Monte Carlo (MCMC) methods. This analysis will be implemented using the R Statistical Software, specifically through the Meta, and Gemtc package. Our approach involves running four parallel MCMC chains in the model, and conducting two distinct MCMC simulations to facilitate the comparison of convergence. The Bayesian model will be operationalised through an initial phase of 5000 burn-in iterations, followed by a subsequent phase of 100 000 simulation iterations. The assessment of convergence will be conducted using the potential scale reduction factor, with an anticipated reduction target of below 1.05.22 23
Heterogeneity, specifically pertaining to direct evidence, and model consistency (direct vs indirect comparisons) will be evaluated using the node split function within the Gemtc package. This process will facilitate the exploration of sources of heterogeneity across studies. The comparative analysis of each intervention will be conducted using SMDs along with 95% credible intervals, which are akin to the Bayesian equivalent of CIs. Rank probabilities will be used to quantify the likelihood of each intervention’s effectiveness. Additionally, the surface under the cumulative ranking score will be employed to estimate the probability of an intervention being the most effective.22 23
Quality of evidence
The grading of evidence derived from the NMA will be executed in four distinct stages, using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) framework, facilitated by GRADE Pro software.24 The initial step involves presenting the effect sizes and CIs for both direct and indirect comparisons across interventions. Subsequently, the second step entails an independent assessment of the quality of evidence for each comparison. The third stage involves the presentation of the NMA results. In the final stage, the quality of evidence underpinning the NMA outcomes is evaluated. For direct comparisons, we apply the same grading approach as the traditional meta-analysis using the GRADE system for evidence grading. When it comes to indirect comparisons, the quality of evidence is downgraded if the direct comparisons are of lower quality. The overall quality of evidence for NMA results, which includes both direct and indirect comparisons, is determined based on the highest quality of evidence found in either comparison type. This method ensures that the final grading reflects the most reliable evidence available.
Additional analysis
Given the potential differences in study populations, interventions and outcome measurements among the included studies, we will conduct a series of subgroup analyses if data allow. We will investigate differences across populations residing in different continents (such as Europe, Asia, etc), treatment durations (less than 12 weeks vs more than 12 weeks), age groups (under 20 years, 20–40 years, 40–60 years, over 60 years), and follow-up periods (less than 4 weeks, 5–12 weeks, more than 12 weeks).
Additionally, we intend to conduct a sensitivity analysis to evaluate the potential impact of the quality of the included trials on the overall findings. This analysis will involve reassessing each outcome by specifically excluding those studies deemed to have a high risk of bias, thus gauging the robustness of our results in relation to the quality of the evidence.
Reporting bias evaluation
Assessment for reporting bias in the NMA will be conducted when the included trials number is at least 10.18 This threshold is based on statistical considerations, recognising that with fewer than 10 trials, the power to discern between random variations and genuine asymmetry is significantly compromised. To detect potential reporting bias, we will employ Begg’s test for analysing funnel plots.25 This method involves correlating the magnitude of effect sizes with their variance, where a notable correlation suggests the presence of reporting bias.25 Should asymmetry be observed in the funnel plots, we will consider alternative explanations beyond reporting bias, such as selective reporting of outcomes, inferior methodological quality in smaller studies, or the presence of heterogeneity.
Discussion
The objective of this systematic review is to explore a significant area in cancer treatment—the use of non-pharmacological interventions for pain management in patients with cancer. Pain management remains a key challenge in cancer treatment, significantly affecting patients' quality of life and treatment compliance.7 8 While pharmacological treatments are mainstream, their adverse reactions can greatly diminish patients' well-being and adherence to treatment.7 8 Therefore, exploring effective non-pharmacological treatment options is not only necessary but also urgent in enhancing comprehensive cancer therapy.
Previous studies have shown that non-pharmacological interventions such as CBT, relaxation training and physical therapy vary in their efficacy in managing cancer-related pain.10–12 However, these studies often lack comparative analysis and are unable to rank the effectiveness of these interventions. By employing NMA, this study will systematically compares and ranks these interventions, providing a broader perspective on their relative benefits and making a unique contribution to the existing literature. The findings of this study are expected to have a significant impact on clinical practice. By providing evidence-based rankings of non-pharmacological interventions, clinicians will be able to make the most appropriate decisions based on individual patient needs, thereby enhancing pain management in cancer care. Furthermore, the results of this study may encourage the adoption of a more holistic approach in cancer treatment, integrating non-pharmacological methods into standard treatment protocols to enhance overall patient care and satisfaction. By focusing on non-pharmacological interventions that do not involve pharmacology, this research addresses significant patient-centred concerns such as the desire to reduce medication side effects and the need for interventions that can be used concurrently with drug treatments without interactions. Additionally, the study’s methodology—using Bayesian NMA—adds a robust statistical layer to the evaluation of treatment effects, enhancing the reliability of the conclusions. This not only aids clinical decision-making but also sets a precedent for future research methodologies in the field.
Patient and public involvement
This study will not include patient or public involvement in any stage, encompassing the planning or design phases. No invitation will be extended to patients to provide feedback on the study design, nor will they be consulted for the development of patient-relevant outcomes or in interpreting the results. Furthermore, patient contributions will not be sought for the writing or editing of this document, whether for enhancing readability or ensuring accuracy. This approach will be maintained throughout the research process.
Ethics and dissemination
Ethical committee approval is not necessitated for this protocol, given its nature as a secondary analysis that aggregates data from primary studies. Dissemination of the findings will be conducted through publications in peer-reviewed journals, ensuring academic rigour and credibility in the presentation of the results.
Ethics statements
Patient consent for publication
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
LY and Y-HL contributed equally.
Contributors LY and Y-HL are co-first authors, jointly responsible for the study design and manuscript writing. Y-HP, Y-HH, QCD and Y-XH completed the Prospero registration information and contributed to the manuscript writing. DL reviewed and revised the manuscript, and acted as guarantor. All authors made substantial contributions to the writing and editing of the manuscript and approved the final version prior to submission.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.