A comprehensive search of literature was performed in September 2021 to identify eligible guidelines published during the last five years (1): six online databases, including PubMed, Cochrane Library, Medline, PsycINFO, Web of Science, and CINAHL (2): nine guideline repositories, including the Guideline International Network, the National Guideline Clearinghouse, the Scottish Intercollegiate Guidelines Network, the National Comprehensive Cancer Network, the National Institute for Health and Care Excellence, the Australian Clinical Practice Guidelines Portal, the New Zealand Guidelines Group, the Canadian Medical Association Infobase, and the Trip Medical Database; and (3) official websites from five professional cancer associations, including the Multinational Association of Supportive Care in Cancer, Cancer Council Australia, the Oncology Nursing Society (ONS), the American Cancer Society (ACS), and the American Society of Clinical Oncology (ASCO). A representative search strategy in PubMed is presented in Table 1 .

The included CPGs met the following criteria (1): CPGs that were published in refereed English academic journals, collected guideline databases, or published by relevant professional cancer associations in the last five years (since January 2016) (2): CPGs that focused on breast cancer survivors, regardless of the stages of diagnosis and types of antineoplastic therapies (3): CPGs that presented any type of non-pharmacological strategies that can be self-managed by breast cancer survivors regardless of the types of delivery method and format, such as yoga, physical exercise, music therapy, meditation, massage, relaxation, acupressure, etc (4): when there are multiple editions of a CPG, only the latest version was included (5): when there are different language/translation versions of a CPG, only the English version was included. CPGs were excluded if they (1): included pharmacological or surgical interventions only (2): were patient-used guidelines, which offer evidence-based recommendations in general without providing detailed evidence analysis, auditing criteria, grade of recommendation, etc.

Duplications were identified and removed via the literature management software EndNote X9. Two independent reviewers (JZ and TW) read the titles and abstracts of the remaining CPGs to select and analyse those that could be potentially included. Then, full-text reviews of the potentially eligible CPGs were conducted by the same two reviewers. Eligible CPGs were eventually included based upon the inclusion and exclusion criteria. Key information in each CPG was extracted using predefined tables, including (1): the characteristics of the included guidelines, such as the name of the CPG, developer, year of publication, whether publication was in a journal, evidence analysis, quality tool referral, etc.; and (2) the contents of the non-pharmacological interventions that were recommended by the included guidelines, such as the form, duration and frequency, level of evidence (LoE), source of evidence (SoE), and grade of recommendation (GoR). In any case of disagreement, a team meeting was organised to resolve the issues during the data retrieval and extraction process.

To critically assess the quality of the included guidelines, the Appraisal of Guidelines for Research and Evaluation, 2nd Edition (AGREE II) was utilised. The AGREE II consists of 23 items, which can be utilised to evaluate the quality of CPGs’ development, transparency, and methodological rigor in six domains: “scope and purpose”, “stakeholder involvement”, “rigor of development”, “clarity and presentation”, “applicability”, and “editorial independence” (33). The AGREE II uses a 7-point Likert scale, from 1 (strongly disagree) to 7 (strongly agree), to assess each item (33). For the global quality and level of the recommendations, a CPG would be considered a grade of “recommended” (high quality) when the mean percentages of the six standardised domains was greater than 70%, a grade of “recommended with modifications” (moderate quality) when the standardised percentages were between 40% and 70% in over three domains, and a grade of “not recommended” (low quality) once the standardised percentages were lower than 40% in three domains or more (34). Four independent assessors critically appraise the quality of each included CPG. All four assessors were experienced academics and health practitioners with over 10 years of professional experience in evidence-based practice, cancer research, and CPG quality appraisal. Each assessor participated the AGREE II Overview Tutorial and the online AGREE II Tutorial and Practice Exercise (33) to effectively apply the instrument to GPG quality appraisal.

Consistency among the assessors in the quality assessment of the CPGs was examined using the intraclass correlation coefficient (ICC): ICC of 0.75 or greater suggests satisfactory consistency as per the recommendations (35). The statistical analyses for the ICC were conducted using the SPSS 25. A value of p < 0.05 indicated statistical significance. Content analysis (36) was adopted to summarise and categorise the self-managed non-pharmacological approaches in the 14 guidelines. Symptoms (e.g., anxiety/depression, fatigue, pain, etc.), quality of life, and risk of recurrence were predetermined themes for the content analysis based on the aims and scopes of the included CPGs that made recommendations on a range of clinical outcomes. the “health promotion” theme was further added after multiple iterative, deductive, and inductive processes (36).

All included CPGs (26–30, 37–45) were published or updated between 2016 and 2021 ( Table 2 ), of which four were from Europe, four from the US, two from the UK, two from Spain, and one each from German and Canada. Eight guidelines (57.14%) were updated versions, while the rest were newly developed. Most the guidelines (12/14, 85.71%) were specifically designed for breast cancer survivors, while the other two also focused on prostate cancer and colorectal cancer. Twelve of the 14 (85.71%) guidelines were published in a journal, while two guidelines were published on the NICE website (28, 42).

Half of the included CPGs engaged with cancer survivors as the key stakeholder during the guideline development. A systematic review approach was adopted in only four CPGs as part of the development methodology with comprehensive database search, specific inclusion criteria, data selection and synthesis. Only two guidelines (NICE guidelines) adopted the AGREE II tool to inform the guideline development process (28, 42).

The ICC between the four reviewers for each guideline ranged from 0.956 to 0.995, with an average of 0.980, indicating a high consistency of rating scores among the four reviewers. According to the AGREE II instrument, the mean overall standardised percentage of the guidelines was 67.49%, ranging from 36.12% to 93.82% ( Table 2 ). The following five guidelines were rated as high quality, with an indication of recommended for use: the NICE (CG81) guideline, the NICE (NG101) guideline, the Society for Integrative Oncology (SIO) guideline, the ONS guideline, and the ACS-ASCO guideline (26, 28, 30, 42, 45). The two SEOM guidelines were rated unsatisfactory quality overall (27, 37), while the remaining guidelines (50%) were rated as moderate quality (recommended with modifications).

The AGREE II domain scores for each guideline varied ( Table 3 ). The results indicated that the “scope and purpose” domain had the highest mean scores (mean: 84.61%; range: 69.4% -100%), whereas the “applicability” domain had the lowest mean scores (mean: 51.04%; range: 12.5% - 91.7%). More heterogeneous scores were demonstrated in the “stakeholder involvement” domain (mean: 71.43%; range: 36.1% - 100%), “clarity and presentation” domain (mean: 66.15%; range: 31.9% - 97.2%), “rigour of development” domain (mean: 62.32%; range: 18.2% - 92.7%), and “editorial independence” domain (mean: 72.19%; range 33.3% - 91.7%).

The 14 included CPGs recommended a number of self-managed non-pharmacological interventions for breast cancer survivors. Details of the non-pharmacological interventions recommended by the included guidelines are summarised in Table 4 .

This review highlighted that the quality of the included CPGs had much room for improvement, which was particularly obvious in the “applicability” and “rigour of development” domains in the AGREE II tool. It was reported that the median scores of the “scope and purpose” domain for most of the CPGs were > 70%, indicating that most of the included CPGs had clear purposes for guideline development. In contrast, the “applicability” domain had the lowest median scores, suggesting that the facilitation of and the barriers to the CPGs’ implementation were not appropriately addressed. To facilitate the implementation of the CPGs, some barrier analysis and/or pilot studies should be conducted to identify the barriers to their implementation (48). In addition, engaging end-users and other stakeholders (e.g., patients, patient advocacy, policymakers, etc.) in the CPGs’ development of non-pharmacological interventions for cancer survivors could help to enhance the incorporation of CPGs and to ensure that the interventions and therapies are sustainable and clinically feasible (49).

Regarding the “rigour of development” domain, this review found that the majority of the CPGs did not use a systematic approach in the formulation of the guidelines, indicating a dearth of systematic review methodologies for the synthesis of evidence in the analysed CPGs. A systematic review approach can be used to identify relevant evidence and to sufficiently describe the methodologies for developing the recommendations. In addition, the CPGs update frequency varied, with only approximately half of the guidelines indicating schedules for updating their guidelines. Scientific evidence will have advanced much quicker than the scheduled updates of the guidelines; hence, more timely updates underpinned by the latest evidence would enhance the implementation and acceptance of these CPGs (50).

Although many CPGs have been developed for breast cancer diagnosis and treatment, few CPGs have addressed non-pharmacological interventions for breast cancer survivors (45). Due to the significantly heterogeneous LoE presented in the included CPGs, the majority of the evidence bases was not adequate to warrant a strong recommendation for the effectiveness of the self-managed non-pharmacological interventions in various clinical outcomes. Physical activity, particularly in the form of regular physical exercise, was the only core self-managed non-pharmacological intervention for psychological and physical symptom management recommended in all the included CPGs. However, the form, duration, and frequency of the different physical activities were not adequately described. A recent pilot study showed that a physical exercise rehabilitation program was effective in reducing breast cancer fatigue at both post-intervention and follow-up time points (11). This physical exercise rehabilitative protocol incorporated 10 min of warm-up, followed by aerobic exercise and strength training (40 min), and then cool-down (10 min), twice a week in a 4-week period of time. Future CPGs might consider providing more details of the recommended interventions, particularly the recommended duration, frequency, delivery method, and sessions, etc.

Similarly, some evidence bases supported the application of other self-managed non-pharmacological interventions, such as meditation, relaxation, stress management, music therapy, yoga, massage, and acupressure, for breast cancer survivors. Nevertheless, a clear understanding of which specific modality of each intervention was effective and acceptable to breast cancer survivors was lacking. Rather, most of the recommendations, such as yoga, meditation, relaxation, etc., were largely deemed as possibly effective self-management strategies for breast cancer survivors given the limitations in relation to low LoE and inconsistency in the GoR of the evidence bases. Some interventions, such as qigong and stress management for reducing clinical symptoms were demonstrated as effective in trial settings only; however, few evidence bases of these interventions were successfully translated to a wide range of populations. The review highlighted that the effectiveness of self-managed non-pharmacological interventions in bone health management among breast cancer patients receiving aromatase inhibitors was under-represented in the included CPGs. One pilot randomized study reported that physical exercises together with whole-body vibration demonstrated a significant effect on pain, muscle functioning, strength, and QoL in breast cancer survivors with aromatase inhibitor-induced musculoskeletal symptoms (51). More evidence base supporting the effect of non-pharmacological strategies on bone health in breast cancer survivors receiving aromatase inhibitors are warranted.

Current evidence bases of the included CPGs supported the efficacy of physical activity, weight management, nutrition, limited alcohol consumption, and smoking cessation in improving breast cancer survivorship outcomes in the “health promotion” domain. Nevertheless, the quality of the evidence was inconsistent or poor for particular topics within the “health promotion” domain, such as losing weight and a reduction/cessation of alcohol and tobacco consumption. For example, although weight loss was recommended by some guidelines, there were some ambiguities in terms of the specific amount of weight that breast cancer survivors should lose. Teras, Patel (52) showed that in women aged 50 years and over, women with sustained weight loss (at least 2 kg) had a lower risk of breast cancer, and those who lost at least 9 kg had the lowest risk, compared with women with a stable weight.

Clinicians might consider using the results of this review as a potential guide in choosing high-quality guidelines to inform the self-managed non-pharmacological interventions that could be recommended to breast cancer survivors. For example, when designing an intervention programme to alleviate fatigue in breast cancer survivors, physical activity/exercise with a duration and frequency equal to three to five hours of moderate walking per week could be considered and adjusted based on the breast cancer survivors’ actual condition. In addition, the results of the quality appraisal using the AGREE II assessment tool can help guideline developers to determine which domain needs to be further strengthened. In particular, the domain of “applicability” should be addressed adequately. To enhance overall quality, the development of CPGs may consider taking the AGREE II’s “rigor of development” domain into account.

The review findings also provided some directions for further research. This study emphasised the inconsistency of evidence bases for some of the recommendations in the analysed CPGs. It is necessary to conduct more large-scale and rigorously designed RCTs to consolidate the evidence bases and further define the effectiveness of various non-pharmacological interventions whose evidence bases are not adequately strengthened or are contradictory, such as the effectiveness of qigong and stress management on QoL in breast cancer survivors. In addition, given the current person-centred care approach, considering the views of breast cancer survivors as the end-users on the flexibility and usefulness of the CPGs would be crucial to ensure that the self-managed non-pharmacological interventions are properly designed to meet their needs. The review findings also encourage future research to explore strategies to support the best translation of research evidence to clinical practice, such as strategies to support clinicians in adhering to guidelines to provide evidence-based treatment and care (53).

This review has some limitations. Only guidelines published in English were included, therefore non-English-language guidelines might have been missed. One presumed limitation of this review could be the subjective process of assessment, which might have had an impact on the rating of the items, the global guideline appraisal, and the degree of the recommendations. However, in this review, the ICC was > 98%, indicating excellent agreement among the four experienced assessors.