Comparative effect of different mindfulness-based intervention types and deliveries on depression in patients with breast cancer: a protocol for a systematic review and network meta-analysis

STRENGTHS AND LIMITATIONS OF THIS STUDY

This systematic review and network meta-analysis will include various types and deliveries of mindfulness-based interventions to evaluate their effectiveness in mitigating depression in patients with breast cancer.
The study will be carried out according to the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions and the Preferred Reporting Items for Systematic Reviews and Network Meta-Analysis Protocol guidelines.
The quality of this network meta-analysis may be limited by the heterogeneity among the included trials in terms of design (e.g., randomised controlled trial (RCT), non-RCT), quality and risk of bias.
Due to the potential lack of direct comparisons between different types and delivery methods of mindfulness-based interventions, the credibility and interpretability of the results may be reduced.
Including only trials published in Chinese and English may lead to the omission of important studies published in other languages, thereby affecting the comprehensiveness and representativeness of the results.

Introduction

Breast cancer (BC) is the most common cancer worldwide, leading to the main cause of cancer-related mortality in women.1 The WHO reported that approximately 2.3 million women were diagnosed with BC globally in 2022, and male BC accounts for approximately 0.5%–1% of all BC cases.2 Early detection of cancer and advances in primary cancer treatment (surgery plus chemotherapy or radiotherapy) and endocrine therapy have resulted in a growing number of BC survivors.3 However, because of the numerous adverse effects related to traditional chemotherapy and radiation therapy, patients with BC often develop co-occurring symptoms with cancer treatments, such as fatigue, depression, anxiety, stress and pain, which can last for months or even years after the treatment is concluded.4–7 Among these, depression, with a global prevalence of 32.2% among patients with BC, has been suggested as a strong risk factor for cancer progression and death,8 9 seriously impairing the quality of life in patients with BC. Therefore, more appropriate management strategies are urgently needed to alleviate depression in patients with BC.

Mindfulness-based interventions (MBIs), as a complementary therapy for symptom management among cancer survivors, have been increasingly popular.10 Mindfulness, with a core construct of self-regulation of attention, was to help individuals pay attention to the present moment consciously and to be curious, open and accepting of the present-moment experience in a nonjudgmental way.11 12 Currently, MBIs include mindfulness-based stress reduction (MBSR), mindfulness-based cognitive therapy (MBCT), mindfulness-based cancer recovery (MBCR), mindfulness meditation training (MMT) and other adapted versions of mindfulness.13 Among them, MBSR and MBCR are the most common structured MBIs and are suggested to be effective in alleviating depression in patients with BC by current meta-analyses.7 14 15

Conventional MBIs were mainly conducted in face-to-face groups. Patients with BC are more likely to feel safe in face-to-face groups, which makes it easier for them to trust each other, and learn how to recognise their emotions.16 However, the face-to-face group format may lead to scheduling conflicts for patients with BC who have completed primary treatment, as this period is associated with the return to work and human interaction for them.17 Especially, the COVID-19 pandemic is also a hindrance to the face-to-face group format MBIs. As a result, internet-delivered MBIs (iMBIs) such as internet-delivered MBCT (iMBCT) and internet-delivered MBSR (iMBSR)18 19 are being increasingly used, as their benefits include decreasing cost, saving time, helping researchers recruit more participants.20–22 Nevertheless, the current studies are insufficient to guide clinical decision-making well, as recent evidence7 10 14 15 18 19 mainly focuses on the comparisons between face-to-face MBIs (or iMBIs) and usual control, the direct comparisons between different types or deliveries of MBIs are insufficient. In other words, we still do not know which type of MBI is the most effective or which delivery (face to face or internet delivered) of MBI is the most effective.

The network meta-analysis (NMA) allows three or more interventions to be compared simultaneously in a single analysis by combining direct and indirect evidence across a network of studies.23 To the best of our knowledge, there was no NMA that synthesised various types of MBI to compare their relative efficacy in alleviating depression in patients with BC. Therefore, in this systematic review and NMA, we will treat some kind of MBI for different delivery methods as different interventions. For example, we refer to conventional face-to-face group format MBSR as face-to-face MBSR and internet-delivered iMBSR as iMBSR, we also refer to MBSR delivered through a combination of face-to-face and internet formats as hybrid MBSR. Then, these MBIs will be included in the NMA to compare their relative efficacy in alleviating depression in patients with BC.

The purpose of this systematic review and NMA is to establish a hierarchy of the efficacy of various types (and deliveries) of MBIs in mitigating depression in patients with BC.

Methods

Design and registration

This protocol was registered in PROSPERO (CRD42024495996) and was reported according to the guidelines of the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P).24 Additionally, this systematic review and NMA will be conducted following the PRISMA-NMA.24 Any amendments to the protocol will be made through PROSPERO. The systematic review will be conducted in accordance with the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions.

Inclusion criteria

Types of studies

The study will include randomised controlled trials (RCTs) as well as other types of trials published in English or Chinese.

Types of participants

Patients with BC will be included if they were identified with depression at baseline using standardised diagnostic criteria or a clinical cut-off score on a validated rating scale. Patients with BC with depressive symptoms that did not meet diagnostic criteria will also be included, regardless of gender, ethnicity, types of BC, cancer stages and current treatments. Trials will be included if at least 30%–50% of the general cancer population are patients with BC.25

Types of interventions

Trials’ interventions must meet at least one of the following criteria:

MBCT, MBCR, MBSR, MMT and other adapted versions of mindfulness in face-to-face format.
MBCT, MBCR, MBSR, MMT and other adapted versions of mindfulness in internet-delivered format.
MBCT, MBCR, MBSR, MMT and other adapted versions of mindfulness are delivered via a combination of internet and face-to-face formats in one study arm.

Types of control groups

Treatments in the comparison groups can be standard treatment including usual care (e.g., basic medical services, psychological support, self-management support),26 no intervention and waitlist; or other types or deliveries of MBIs. Additionally, based on a previous meta-analysis27 on this topic, we will include supportive-expressive therapy as an active control.

Types of outcome

The outcome will be depression, assessed by applying standardised and validated depression scales (e.g., Hamilton Depression Rating Scale, Beck Depression Inventory, Patient Health Questionnaire-9, Hospital Anxiety and Depression Scale).

Exclusion criteria

We will also apply the exclusion criteria as follows:

(1) trials without full text; (2) trials without available or convertible data (non-convertible data including graphical data, categorical data); (3) duplicate or retracted trials. Trials that meet any of the criteria above will be excluded.

Search strategy

The literature will be comprehensively retrieved from the following databases: PubMed, Web of Science, the Cochrane Library, Embase, Google Scholar and the China National Knowledge Infrastructure. The search period will range from the date of establishment of the database to December 2023. The studies will be limited to English or Chinese. A combination of Medical Subject Headings (MeSH) and free words related to “breast cancer”, “depression”, “mindfulness-based interventions” and “RCT” will be used. Unpublished studies will be searched via OpenGrey and conference proceedings; and we will also search the international trial register platforms such as Clinical Trial Registration and WHO–International Clinical Trials Registry Platform to identify the ongoing trials. The detailed search strategies are provided in online supplemental file 1.

Supplemental material

Study selection

Two researchers (DZ and WZ) will collaboratively search the literature in the above-mentioned databases. The search strategies will be peer-reviewed by a suitably qualified and experienced medical/healthcare librarian or information specialist prior to execution, following the Cochrane Handbook recommendations.28 EndNote V.X9 software will be used to remove duplicates. After removing the duplicates, two researchers (DZ and WZ) will independently examine the titles and abstracts of identified studies for eligibility following the inclusion criteria, and the irrelevant studies will be excluded. Then, the full texts of retained studies will be reviewed and evaluated for final inclusion independently by two researchers (DZ and WZ). All disagreements will be resolved by a third independent researcher (XJ). The selection of studies in the systematic review will be illustrated in figure 1, based on the PRISMA 2020 statement for reporting new systematic reviews.24

Figure 1

The flow chart of studies selection.

Based on our preliminary search, we have identified a sufficient number of studies to conduct a robust NMA.29 This ensures that our analysis will be comprehensive and statistically reliable.

Data extraction

Two researchers (DZ and WZ) will independently extract the relevant data from the included articles based on the Cochrane handbook.28 The extracted data will be as follows: study general data (e.g., authors, year of publication, country of study, study design, methods of randomisation and blinding); participant characteristics (e.g., sample size, age, cancer type, treatment phase); intervention and control (e.g., intervention type, duration, frequency, intensity, follow-up, control group); outcome (e.g., measures of effect) and access time. If possible, the corresponding author will be contacted for further information or data as necessary. From each study, the mean and standard deviation (SD) of the outcomes, and sample sizes for both the intervention and control groups will be collected to calculate the Hedges’ effect size using the formula provided by the Cochrane Handbook for Systematic Reviews of Interventions.30

Risk of bias assessment

Two researchers (DZ and WZ) will independently examine the methodological quality of the included studies using the Cochrane Risk of Bias Tool for Randomised Trials 2.0 (RoB 2.0).31 RoB 2.0 evaluates the following domains: randomisation process, deviations from the intended interventions, missing outcome data, outcome measurement and reporting bias. Each domain will be assigned a category characterising the potential risk of bias as either ‘low risk’, ‘some concerns’ or ‘high risk’. The discrepancy will be resolved by a third researcher (XJ).

Statistical analysis

Network and pairwise meta-analysis

This NMA will be conducted using STATA V.16.0 under a frequentist framework with a random-effects model because of the unavoidable heterogeneity of the multiple MBIs and outcome depression scales included. Where there are more than two studies existing on the same MBI, the conventional pairwise meta-analysis will be conducted for direct comparisons using STATA V.16.0. Hedges’ effect size with a 95% confidence interval (CI) will be used to compare the relative effectiveness of the different MBIs included. Hedges’ effect size is preferred in this context because it provides a more accurate estimate of effect size, especially in studies with smaller sample sizes, which can be common in meta-analyses with a limited number of primary studies.32 A league table will be used to conduct the pairwise analysis and analyse the results of the comparisons among the multiple MBIs based on NMA. To explore the best evidence for mitigating depression, a probability-based ranking will be established based on the surface under the cumulative ranking method. A network plot of all comparisons will be generated using STATA V.16.0, where interventions are represented by nodes, the line between nodes represents a direct comparison between the two MBIs, the thickness of this line represents the number of studies included in this direct comparison, and the size of the nodes indicates the sample size. A narrative synthesis of findings will be conducted according to the Synthesis Without Meta-analysis guidelines when meta-analysis is unavailable.33

Heterogeneity assessment

Heterogeneity will be assessed using the χ² test (χ² or Cochrane Q) and inconsistency index (I²) to examine whether the results of individual studies could be combined.34 Statistical heterogeneity (p≥0.1, I²≤50%) indicates no significant heterogeneity among the studies, and a fixed-effects model will therefore be suitable for meta-analysis. Otherwise, heterogeneity (p<0.1, I²>50%) indicates the evidence of significant heterogeneity, and the random effects model will therefore be appropriate for combining the effect size.35

Assumption of transitivity

When conducting an NMA, it is crucial to assess the characteristics of patients and studies, and these characteristics are considered effect modifiers.36 In this study, patient age, symptom duration and severity, intervention types, duration, and intensity will be considered as effect modifiers. Within this context, it will be assumed that the main effect modifiers are as similar as possible among the included studies. The presumed transitivity will be evaluated by comparing the above-mentioned main effect modifiers of the included MBIs.

Inconsistency analysis

The design-by-treatment model will be used to evaluate inconsistency for explaining the various sources of inconsistency (multiarm trial, design-by-treatment interaction, design inconsistency, loop inconsistency). The consistency between direct and indirect evidence will be evaluated using the node-splitting method.37

Additional analysis

Where possible, we will perform subgroup analyses based on various types and deliveries of MBIs, different characteristics of patients with BC and different types of study designs (e.g., RCTs, non-RCTs). These subgroup analyses will help identify whether and how these factors influence the effectiveness of MBIs. Sensitivity analyses will be conducted by excluding each subgroup of different treatment phases in turn to explore the potential impact of treatment phases on the overall findings. Additionally, we will conduct sensitivity analyses by excluding studies with a high risk of bias and those in which 30%–50% of the general cancer population are patients with BC. These sensitivity analyses will help ensure the robustness and validity of our findings. Furthermore, publication bias will be evaluated by Egger’s test and the symmetry of the funnel plot.

Evidence quality assessment

Two researchers (LY and QC) will use The Grading of Recommendation Assessment, Development and Evaluation system38 to assess the quality of evidence independently. The quality of evidence will be judged as very low, low, moderate and high according to the assessments of the risk of bias, indirectness of evidence, inconsistency, imprecision and publication bias.

Patient and public involvement

Patients and the public will not be directly involved in this study.

Discussion

Recently, BC has become the most common cancer in the world.1 Current studies have found that patients with BC suffer from physical and psychological problems.4–7 Among these problems, depression has been suggested as a strong risk factor for cancer progression and death.8 MBIs have been suggested to be promising therapies for this population by previous meta-analyses.7 10 14 15 However, MBIs include various types, including MBCT, MBSR, and MBCR. These interventions can be delivered face-to-face or through innovative methods such as the internet. Therefore, the clinical practice of MBIs presents significant variability. Current research has only demonstrated the effectiveness of some MBIs, but it does not provide guidance for clinicians or therapists on selecting the most effective MBIs.

This study will conduct a systematic review and NMA to integrate both direct and indirect evidence from published studies, compare the relative efficacy of different types and delivery methods of MBIs, identify the most effective MBIs or delivery methods and provide specific recommendations for clinical decision-making and practice. However, the study has several limitations. First, the quality of this NMA may be limited by the heterogeneity among the included trials in terms of design (e.g., RCT, non-RCT), quality and risk of bias. Second, due to the potential lack of direct comparisons between different types and delivery methods of MBIs, the credibility and interpretability of the results may be reduced.39 Third, including only trials published in Chinese and English may lead to the omission of important studies published in other languages, thereby affecting the comprehensiveness and representativeness of the results.

Ethics statements

Patient consent for publication

This post was originally published on https://bmjopen.bmj.com