Screening for endometrial hyperplasia and endometrial cancer in premenopausal women with polycystic ovary syndrome: a systematic review protocol Authors: Rajalakshmi Rajendran, Marva Abdul Khader, Priya Shivanna, Prathiksha M Shet, Niyam Hegde, Deeya Karkera, Girish Thunga Published: 2026 ABSTRACT AbstractIntroduction Endometrial hyperplasia (EH) is a condition that arises due to prolonged exposure of the endometrium to unopposed oestrogen, resulting in abnormal proliferation of endometrial tissue. Atypical hyperplasia, a subtype of EH, is considered a precursor lesion for endometrial cancer (EC). Polycystic ovary syndrome (PCOS) is a recognised risk factor for the development of EH and EC due to chronic anovulation and prolonged oestrogen exposure without progesterone opposition. Despite this increased risk, current clinical guidelines do not recommend routine screening for EH or EC among asymptomatic women with PCOS. The study aims to determine the effectiveness of screening for EH and EC in women with PCOS, particularly among premenopausal women, compared with no screening, reporting the occurrence and clinical outcomes of endometrial pathology among women with PCOS.Methods and analysisThis systematic review will adhere to the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines. A comprehensive literature search will be conducted across multiple electronic databases from inception until February 2026. Two reviewers will independently perform study screening, data extraction and methodological quality assessment of the included studies. Discrepancies between reviewers will be resolved through discussion or consultation with a third reviewer. Extracted data will be summarised narratively, and where appropriate, quantitative synthesis will be performed using Joanna Briggs Institute System for the Unified Management, Assessment and Review of Information (JBI SUMARI). Heterogeneity among studies will be assessed, and subgroup and sensitivity analyses will be conducted when feasible.Ethics and disseminationEthical approval is not required for this systematic review and meta-analysis, as it will use data from previously published studies. The results of the review will be published in academic journals and presented at international conferences.PROSPERO registration numberCRD420251274598. STRENGTHS AND LIMITATIONS OF THIS STUDY The review is designed to inform evidence-based guideline development using predefined and transparent methods. The review is designed to inform evidence-based guideline development using predefined and transparent methods. A comprehensive search strategy will be implemented across multiple bibliographic databases and supplementary sources to maximise study identification. A comprehensive search strategy will be implemented across multiple bibliographic databases and supplementary sources to maximise study identification. Independent duplicate screening, data extraction and risk-of-bias assessment will be undertaken using established systematic review methodology and reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 recommendations. Independent duplicate screening, data extraction and risk-of-bias assessment will be undertaken using established systematic review methodology and reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 recommendations. The review may rely substantially on indirect evidence because studies directly comparing screening versus no screening among women with polycystic ovary syndrome (PCOS) are expected to be limited. The review may rely substantially on indirect evidence because studies directly comparing screening versus no screening among women with polycystic ovary syndrome (PCOS) are expected to be limited. Heterogeneity in PCOS diagnostic criteria, screening modalities and outcome definitions may limit quantitative synthesis and comparability across studies. Heterogeneity in PCOS diagnostic criteria, screening modalities and outcome definitions may limit quantitative synthesis and comparability across studies. Introduction Endometrial hyperplasia (EH) is characterised by excessive proliferation of the endometrial glands relative to the stroma, primarily resulting from prolonged exposure of the endometrium to unopposed oestrogen.1 This hormonal imbalance stimulates continuous endometrial growth and increases the risk of abnormal cellular changes over time. If untreated, certain forms of EH may progress to endometrial cancer (EC), making early identification and appropriate management clinically important.2 Two major classification systems are currently used to categorise EH. The traditional WHO 1994 Classification of EH divides EH into four categories: simple hyperplasia without atypia, complex hyperplasia without atypia, simple atypical hyperplasia and complex atypical hyperplasia.3 4 This classification is based on architectural complexity and cytological atypia.5 However, due to challenges with reproducibility and clinical applicability, the WHO 2014 Classification of EH introduced a simplified two-tier system: hyperplasia without atypia and atypical hyperplasia (also referred to as endometrial intraepithelial neoplasia).6 This revised system better reflects the differing malignant potential of these lesions and supports more standardised clinical decision-making. EH is most commonly diagnosed through histopathological evaluation following endometrial biopsy, particularly among women presenting with abnormal uterine bleeding (AUB).7 Because diagnostic procedures are generally performed only in symptomatic individuals, reliable population-level prevalence estimates remain limited. Hospital-based studies in India that evaluated women presenting with AUB have reported that approximately 22–38% of patients undergoing endometrial biopsy were diagnosed with EH.8 These findings suggest that EH may represent a substantial but under-recognised clinical burden among women presenting with gynaecological symptoms. Globally, EC represents a significant public health concern. It is currently the sixth most common cancer affecting women worldwide, with more than 417 000 new cases reported since 2020.9 In India, EC is the third most common gynaecological malignancy, with an estimated incidence rate of approximately 2.1 per 10 000 women.10 The relationship between EH and EC is well established, as EH, particularly atypical hyperplasia, is considered a precursor lesion in the pathogenesis of many ECs.11 Women with atypical hyperplasia are at substantially greater risk of EC than those with non-atypical hyperplasia. A meta-analysis reported a 32.6% prevalence of concurrent EC and an annual progression rate of 8.2% among women with atypical hyperplasia, whereas evidence for non-atypical hyperplasia was limited, with only one study reporting an annual progression rate of 2.6%.12 Premenopause itself is not considered an independent risk factor for the development of EH or EC. However, several underlying clinical conditions that are common in reproductive-age women may increase the risk of these endometrial abnormalities. These include obesity, chronic anovulation, prolonged or irregular menstrual cycles, nulliparity and hereditary cancer syndromes such as Lynch syndrome. A particularly important endocrine condition associated with increased endometrial risk is polycystic ovary syndrome (PCOS).13 14 PCOS is one of the most common endocrine disorders affecting reproductive-aged women and is frequently characterised by irregular ovulation or chronic anovulation.15 In the absence of ovulation, the cyclical production of progesterone does not occur, leaving proliferative effects of the oestrogen on the endometrium unopposed. This persistent oestrogenic stimulation promotes endometrial thickening and cellular proliferation, thereby increasing the likelihood of developing EH and potentially EC over time.16 Several meta-analyses and cohort studies have reported that women with PCOS have approximately threefold to fivefold higher risk of developing EC compared with women without PCOS.1719 Importantly, this elevated risk appears to be particularly pronounced among premenopausal women. Despite this association, current clinical guidelines do not recommend routine EC screening for asymptomatic women with PCOS. Major guidelines, including those from the Endocrine Society, as well as international PCOS consensus collaborations, advise that screening procedures such as transvaginal ultrasound or endometrial biopsy should not be performed routinely in asymptomatic individuals. Instead, endometrial evaluation is recommended for women with PCOS who present with AUB or additional risk factors, such as obesity or hereditary cancer syndromes like Lynch syndrome.20 In recent years, research efforts have focused on developing improved tools for identifying individuals at higher risk of endometrial pathology. These include predictive risk models, circulating biomarkers such as Anti-Müllerian Hormone and emerging approaches based on artificial intelligence and machine learning.2123 Although these strategies show promise for enhancing early detection, they have not yet been fully validated or integrated into routine clinical screening for asymptomatic PCOS populations. Existing systematic reviews and meta-analyses consistently highlight the need for more robust prospective research to better define the magnitude of endometrial risk among women with PCOS.24 Much of the currently available evidence is derived from retrospective or case–control studies, which are subject to important methodological limitations including confounding, selection bias and variability in diagnostic criteria for both PCOS and EH.25 These limitations complicate the interpretation of existing findings and hinder the development of evidence-based screening recommendations. Current international PCOS guidelines do not recommend routine screening for EH or EC in women with PCOS; however, these recommendations are based on low-certainty evidence.26 Moreover, direct studies evaluating the effectiveness of screening vs no screening for EH or EC in women with PCOS are lacking. Nevertheless, a substantial body of observational evidence, including cohort, case–control and cross-sectional studies, has examined the occurrence, prevalence, incidence and risk of EH and EC in this population. In the absence of direct comparative evidence, these data may provide valuable insights into disease burden and help identify subgroups at increased risk who may potentially benefit from screening strategies. Accordingly, this review aims to systematically synthesise the available evidence on the association between PCOS and endometrial pathology, including both direct and indirect evidence relevant to screening. By evaluating the current evidence base, this review seeks to clarify the magnitude of risk, identify key evidence gaps and inform future research, clinical decision-making and guideline development aimed at improving long-term gynaecological health outcomes. Objectives The primary objective of this review is to determine the effectiveness of screening for EH and EC in women with PCOS, particularly among premenopausal women, compared with no screening, using both direct evidence from studies evaluating screening interventions and indirect evidence from studies reporting the occurrence and clinical outcomes of endometrial pathology among women with PCOS. Effectiveness will be assessed using outcomes including stage of EC at diagnosis, EC-related mortality, 5-year survival and potential harms associated with screening, including psychological distress, false-positive results, unnecessary biopsies, overdiagnosis and procedure-related complications. In addition, the incidence and prevalence of EH and EC among women with PCOS will be synthesised as indirect evidence relevant to screening decisions. The secondary objectives are to examine whether the effectiveness of screening varies across specific subgroups, including obese versus lean women and other clinically relevant subgroups where data are available, and to compare outcomes between screened and unscreened women with PCOS where such data are reported. Guided by the Population, Intervention, Comparator, Outcome (PICO) framework, the review addresses the following question: ‘Should all women diagnosed with PCOS be offered screening for EH & EC?’ Methods This systematic review will be conducted following the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) 2015 for preparing the protocol27 and refer to PRISMA 2020 guidelines to report the review, to ensure transparency and completeness in reporting.28 Additionally, the study protocol is registered in PROSPERO with the registration number CRD420251274598. Eligibility criteria Population The target population for this review will include women clinically diagnosed with PCOS using recognised diagnostic criteria (eg, Rotterdam, National Institute of Health (NIH) or Androgen Excess-PCOS (AE-PCOS) Society criteria) or with a confirmed diagnosis documented in scientific literature. Where data are available, subgroup analyses will be conducted to examine potential differences between premenopausal and postmenopausal women. In addition, stratification based on body mass index will be considered to explore variations in outcomes between obese and lean women. Studies will be excluded if they include women without a confirmed clinical or research diagnosis of PCOS (for example, women with ovarian cysts due to other causes) unless results for participants with confirmed PCOS are reported separately. Pregnant women at the time of screening will also be excluded to avoid confounding effects related to pregnancy-associated hormonal changes. Furthermore, studies involving women with other untreated endocrine disorders that may mimic PCOS or independently influence relevant outcomes (such as untreated Cushing’s syndrome, hyperprolactinaemia or thyroid disorders) will be excluded unless PCOS is reported as a co-diagnosis and data for women with PCOS can be clearly distinguished. Intervention Studies will be included if they evaluate screening programmes (either universal or targeted) for EH or EC among women diagnosed with PCOS. Eligible screening modalities include transvaginal ultrasound (TVUS) for the assessment of endometrial thickness, endometrial biopsy or sampling, hysteroscopy, endometrial cytology, molecular or biomarker-based tests or combinations of these approaches. Targeted screening programmes based on clinical characteristics, risk factors or PCOS phenotypes/subphenotypes will also be considered eligible. Diagnostic accuracy studies evaluating screening tests (eg, studies assessing endometrial thickness thresholds using TVUS) conducted in premenopausal women with PCOS will also be considered. Additionally, studies reporting clinical outcomes following screening such as detection rates of EH or EC, stage at diagnosis, treatment outcomes or morbidity will be included. Studies that evaluate implementation-related outcomes, including acceptability, feasibility, screening uptake, harms associated with screening, referral rates and costs, will also be eligible. On the other hand, studies will be excluded if they include women with PCOS but do not report outcomes related to EH or EC. In addition, studies in which data for women with PCOS cannot be separated from those of participants without PCOS will be excluded. Comparator The comparator for this review will include women with PCOS who did not undergo screening for EH or EC. Studies comparing screened vs unscreened populations will be included to assess the effectiveness of screening. In addition, studies comparing different screening strategies, thresholds or modalities (eg, varying TVUS endometrial thickness thresholds or different diagnostic approaches) will also be considered. Outcomes The outcomes of interest for this review are classified as critical and important outcomes. Critical outcomes Incidence of EC among women with PCOS. Incidence of EC among women with PCOS. Stage of EC at the time of diagnosis (eg, early-stage vs advanced-stage disease). Stage of EC at the time of diagnosis (eg, early-stage vs advanced-stage disease). Incidence of EH including atypical and non-atypical EH where reported. Incidence of EH including atypical and non-atypical EH where reported. EC-related mortality, including both all-cause mortality and cancer-specific mortality. EC-related mortality, including both all-cause mortality and cancer-specific mortality. 5-year survival among women diagnosed with EC. 5-year survival among women diagnosed with EC. Harms associated with screening, including false-positive results, unnecessary diagnostic procedures (eg, endometrial biopsy), overdiagnosis, psychological distress and procedure-related complications. Harms associated with screening, including false-positive results, unnecessary diagnostic procedures (eg, endometrial biopsy), overdiagnosis, psychological distress and procedure-related complications. Important outcomes Prevalence of EH among women with PCOS. Prevalence of EH among women with PCOS. Prevalence of EC among women with PCOS. Prevalence of EC among women with PCOS. Where available, outcomes will be reported separately for women with PCOS who underwent screening and those who did not. Given the anticipated paucity of direct screening-versus-no-screening studies, indirect evidence describing the burden, occurrence, progression and outcomes of EH and EC among women with PCOS will also be synthesised to inform the assessment of screening effectiveness. Studies that do not assess or report any of the outcomes listed above will be excluded. Study design This review will include randomised controlled trials (RCTs) comparing screening vs no screening for EH or EC among women with PCOS. Cluster randomised controlled trials and controlled before–after studies evaluating the impact of screening interventions will also be considered. In addition, observational studies will be eligible if they report screening outcomes or provide evidence on the occurrence, prevalence, incidence, progression or clinical outcomes of EH or EC among women with PCOS. Studies using qualitative methods without standardised measurement instruments, reviews, meta-analyses, case reports, opinion pieces, conference abstracts and other forms of secondary literature will be excluded. Information resources In designing the search strategies for this systematic review, we balanced pragmatic considerations with scientific rigour. We selected databases such as PubMed/MEDLINE, Scopus, Embase, Web of Science, Cumulative Index to Nursing and Allied Health Literature (CINAHL) and the Cochrane Library databases due to their relevance to biomedical literature, ensuring comprehensive coverage of the topic. Additionally, we will consult official websites such as the International Clinical Trials Research Platform (ICTRP) and the Clinical Trials Registry-India (CTRI) to gather information on ongoing research. Search strategy Query strategies will be developed for each database in consultation with an experienced information specialist and adapted to the indexing terms and search functionalities of individual databases. The search strategy will combine controlled vocabulary terms (eg, MeSH and Emtree terms, where applicable) and free-text keywords such as ‘Polycystic ovary syndrome’, ‘PCOD’, ‘endometrial hyperplasia’ and ‘EC’, with appropriate Boolean operators, truncation and field-specific searches to optimise the sensitivity and specificity of the search. A detailed search strategy in PubMed/MEDLINE is given in online supplemental file 1. For this review, we will also employ a search validation procedure to ensure the search strategy accurately retrieves relevant studies. We will first identify a test set of key articles known to be directly relevant to the research question. Using this set as a benchmark selection, we will conduct an initial trial search in each database and review whether the search captures all key articles. If any articles are missing, we will adjust search terms, Boolean operators or field codes as needed—incorporating synonyms, alternative spellings or modifying terms for broader or narrower scopes as required. After making modifications, we will re-run the adjusted search and confirm that it captures the entire test set. All adjustments and their rationales will be documented to ensure transparency and reproducibility of the search strategy.2932 This iterative validation process will help maximise both the sensitivity and specificity of the search, ensuring comprehensive and relevant results. For this review, additional search strategies will be employed to ensure comprehensive coverage of relevant studies. These include the ascendancy approach, where we will examine the reference lists of included studies to identify additional sources that may not have been captured by the primary search, thus uncovering foundational or related literature. Additionally, we will use the descendency approach through citation tracking tools like Google Scholar and CrossRef to find studies that cite our included sources, allowing us to capture more recent studies that reference key findings in the field. The search will be conducted from inception till February 2026. No restrictions will be applied for language or time. Any non-English literature will be translated using Microsoft Translator and will be verified using a language expert. Additionally, we will apply the Peer Review of Electronic Search Strategies (PRESS) checklist to systematically review and refine the search strategy.31 The PRESS checklist will guide us in identifying potential issues and areas for improvement in the search terms, Boolean operators, and field selections. In addition to the checklist, we will consult a librarian with expertise in systematic reviews to refine the use of controlled vocabulary (eg, MeSH terms, Emtree) and database-specific indexing strategies, ensuring a more precise and comprehensive search. This added layer of validation will further enhance the precision and reliability of the search. Data management References retrieved through the search process will be imported into Zotero 7.0.30, where duplicates will be removed. Study selection and screening Before initiating the study selection process, all reviewers will undergo a structured training protocol to ensure consistency in applying the inclusion and exclusion criteria. This training will include familiarisation with the eligibility criteria and a pilot screening exercise on a subset of studies for blinded and independent selection. The screening process for this review will follow a two-stage approach to efficiently manage and evaluate the search results: Title and abstract screening: In the first stage, the authors will manually screen the titles and abstracts of the retrieved articles. Human reviewers will assess each entry for relevance to our review question based on our inclusion criteria. Articles that do not meet the criteria will be excluded at this stage. Title and abstract screening: In the first stage, the authors will manually screen the titles and abstracts of the retrieved articles. Human reviewers will assess each entry for relevance to our review question based on our inclusion criteria. Articles that do not meet the criteria will be excluded at this stage. Full text screening: In the second stage, we will perform a full-text review of the articles that passed the title and abstract screening. Human reviewers will comprehensively evaluate the content of each study to confirm its eligibility for inclusion in the review. Any articles that do not meet the established criteria will be excluded. Full text screening: In the second stage, we will perform a full-text review of the articles that passed the title and abstract screening. Human reviewers will comprehensively evaluate the content of each study to confirm its eligibility for inclusion in the review. Any articles that do not meet the established criteria will be excluded. During the screening process, two reviewers (DK and NH) will independently conduct both the title and abstract screening and the full-text screening of the retrieved studies. The reviewers will perform these assessments independently and without prior discussion to minimise potential bias. To evaluate the reliability of the screening decisions, Cohen’s Kappa Statistic will be used to measure the level of agreement between the two reviewers.33 Established interpretation thresholds for Kappa values will guide further actions, with values below 0.40 indicating poor agreement. If poor agreement is observed (Kappa<0.40), the reviewers will discuss discrepancies and resolve differences through consensus. If disagreements persist, a third reviewer (RR) will be consulted to facilitate resolution and ensure consistency in study selection. The results of the literature search, study selection, and inclusion process will be reported in detail in the final systematic review. The screening and selection process will be illustrated using the PRISMA Flow Diagram, in accordance with the recommendations of the PRISMA 2020 guidelines. Data extraction Data from the included primary studies will be extracted manually using a validated data extraction form developed specifically for this review in Microsoft Excel 2016. The extraction form will include key elements such as study characteristics, demographic information, outcome measures and indicators related to risk of bias. These variables will be selected to directly address the research questions and facilitate a comprehensive and structured analysis of the included evidence. The structure and variables included in the extraction form are provided in online supplemental file 2. The data extraction process will follow a multi-stage approach designed to enhance reliability and minimise potential errors. Initially, two reviewers (PS and PMS) will independently extract data from a predetermined subset of included studies to ensure consistency and reduce the risk of biased extraction. The extracted data will then be compared with identify any discrepancies. Differences in extracted information will be discussed between the reviewers and resolved through consensus. In cases where disagreements persist or substantial discrepancies arise, a senior researcher (GT) will be consulted to provide guidance and facilitate resolution. To evaluate the consistency of the extraction process, inter-rater reliability will be assessed using Cohen’s Kappa Statistic.33 A Kappa value greater than 0.70 will be considered indicative of acceptable agreement between the reviewers, reflecting a high level of consistency in the extracted data. Masking will not be applied during the data extraction process. The reviewers will have full knowledge of the research questions, hypotheses and the contextual significance of the variables being extracted. This approach allows reviewers to better understand the relevance and importance of the extracted information, thereby supporting more accurate and contextually meaningful data capture. The extraction process will be guided by detailed protocols developed in accordance with the predefined inclusion and exclusion criteria of the review. These protocols will be clearly communicated to the reviewers prior to the extraction process. Regular discussions among the review team will also be conducted to address questions, clarify uncertainties and ensure consistent application of the extraction criteria. Through this structured and transparent approach, supported by continuous communication and adherence to standardised procedures, the review aims to ensure high-quality and reliable data extraction that aligns with the overall objectives of the study. Methodological quality assessment To assess the quality of the included sources in our review, we will use the Cochrane Risk of Bias 2.0 tool to evaluate RCTs for potential biases across several domains, including selection, performance, detection, attrition and reporting biases.34 For non-randomised interventional designs, Risk Of Bias In Non-randomised Studies (ROBINS) will be used to evaluate the quality of the included sources.35 For observational studies, New Castle Ottawa scale will be used to assess the methodological quality.36 Two reviewers (DK and MAK) will independently assess the methodological quality of the included studies. In cases of disagreement, a third reviewer (RR) will be consulted to reach a consensus. Data extraction will be conducted for all included studies. Quantitative analysis will be performed only on studies with high methodological quality, while the remaining studies will be narratively summarised. Data synthesis and integration The analysis procedure to address the research questions will involve both narrative and quantitative synthesis, as appropriate. Data will initially be synthesised narratively to summarise study characteristics, screening approaches and outcomes. Where sufficient homogeneity exists in study design, population, interventions and outcomes, meta-analysis will be conducted using the Joanna Briggs Institute System for the Unified Management, Assessment and Review of Information.37 Depending on the degree of heterogeneity, either fixed-effect or random-effects models will be applied to estimate pooled effect measures. A subgroup analysis will be conducted to examine the impact of moderators such as population characteristics and intervention types. Secondary analyses may include sensitivity analyses to test the robustness of the findings. If any part of the analysis plan encounters data limitations, alternative approaches will be implemented, such as utilising additional data sources or adjusting the analysis strategy as needed. Statistical heterogeneity will be assessed using the I² statistic, where thresholds of 0–25%, 25–50% and 50–75% indicate low, moderate and high heterogeneity, respectively. In the synthesis process, there will be one primary researcher (RR) who will be responsible for synthesising the extracted data. Additionally, a subject expert, methodologist and statistician will be included to provide input and ensure the synthesis is comprehensive and robust. The synthesist will first conduct the analyses independently, applying their own interpretations and conclusions to the data. After completing the individual synthesis, the findings will be discussed among the research team and reconcile any differences in the interpretations. If significant divergences arise during the reconciliation, the team may consult with the third/subject expert to provide additional insight and guidance. Publication bias analyses Initially, a visual inspection of funnel plots will be conducted to identify any asymmetry that may suggest the presence of publication bias. Additionally, Egger’s and Begg’s test will be performed to statistically evaluate funnel plot asymmetry and assess the likelihood of publication bias in the literature.38 If significant publication bias is detected, methods such as trim-and-fill analysis will be applied to estimate the number of missing studies and adjust the overall effect size accordingly. Furthermore, sensitivity analyses will be conducted to determine how robust the findings are to potential publication bias. Sensitivity analyses/robustness checks To assess the robustness of the findings, first, we will evaluate the impact of individual studies on the overall effect size by systematically removing one study at a time from the analysis. This will help determine if any single study disproportionately influences the results. Second, subgroup analyses will be performed based on key characteristics of the studies, such as study design, sample size and quality of the studies, to explore how these factors might affect the overall outcomes. Finally, we will perform alternative modelling approaches, such as fixed-effect vs random-effects models, to evaluate how the choice of model affects the results. The methodological limitations of the included studies will be systematically reported through a descriptive assessment, which will be presented in the Results and Discussion sections of the full review. Key methodological concerns identified in the included studies will also be extracted and narratively summarised. Identifying these gaps will help highlight areas requiring further evidence-based research in the management of PCOS. Certainty of evidence The Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach will be used to evaluate the quality of evidence for each critical outcome across the included studies.39 This process involves assessing five domains: overall risk of bias, inconsistency of results (heterogeneity), indirectness of evidence, imprecision of estimates and publication bias. Evidence derived from RCTs will initially be considered high quality. The quality of evidence may then be downgraded depending on concerns identified within these domains during the assessment of the body of evidence for each outcome. Ethics and dissemination Ethical approval is not required for this systematic review and meta-analysis as it will use data from previously published studies. The results of the review will be published in academic journals and presented at international conferences. Supplementary material