Methods for non-proportional hazards in clinical trials: A systematic review.

For the analysis of time-to-event data, frequently used methods such as the log-rank test or the Cox proportional hazards model are based on the proportional hazards assumption, which is often debatable. Although a wide range of parametric and non-parametric methods for non-proportional hazards has been proposed, there is no consensus on the best approaches. To close this gap, we conducted a systematic literature search to identify statistical methods and software appropriate under non-proportional hazard. Our literature search identified 907 abstracts, out of which we included 211 articles, mostly methodological ones. Review articles and applications were less frequently identified. The articles discuss effect measures, effect estimation and regression approaches, hypothesis tests, and sample size calculation approaches, which are often tailored to specific non-proportional hazard situations. Using a unified notation, we provide an overview of methods available. Furthermore, we derive some guidance from the identified articles.

Safety outcomes of direct oral anticoagulants in older adults with atrial fibrillation: a systematic review and meta-analysis of (subgroup analyses from) randomized controlled trials.

Balancing stroke prevention and risk of bleeding in patients with atrial fibrillation (AF) is challenging. Direct oral anticoagulants (DOACs) are by now considered standard of care for treating patients with AF in international guidelines. Our objective was to assess the safety of long-term intake of DOACs in older adults with AF. We included RCTs in elderly (≥ 65 years) patients with AF. A systematic search in MEDLINE and EMBASE was performed on 19 April 2022. For determination of risk of bias, the RoB 2 tool was applied. We pooled outcomes using random-effects meta-analyses. The quality of evidence was assessed using GRADE. Eleven RCTs with a total of 63,374 patients were identified. Two RCTs compared apixaban with either warfarin or aspirin, four edoxaban with either placebo, aspirin, or vitamin K antagonists (VKAs), two dabigatran with warfarin and three rivaroxaban with warfarin. DOACs probably reduce mortality in elderly patients with AF (HR 0.89 95%CI 0.77 to 1.02). Low-dose DOACs likely reduce bleeding compared to VKAs (HR ranged from 0.47 to 1.01). For high-dose DOACS the risk of bleeding varied widely (HR ranged from 0.80 to 1.40). We found that low-dose DOACs probably decrease mortality in AF patients. Moreover, apixaban and probably edoxaban are associated with fewer major or clinically relevant bleeding (MCRB) events compared to VKAs. For dabigatran and rivaroxaban, the risk of MCRB varies depending on dose. Moreover, subgroup analyses indicate that in the very old (≥ 85) the risk for MCRB events might be increased when using DOACs.Registration: PROSPERO: CRD42020187876.

Time since liver transplant and immunosuppression withdrawal outcomes: Systematic review and individual patient data meta-analysis.

BACKGROUND & AIMS: Successful immunosuppression withdrawal (ISW) is possible for a subfraction of liver transplant (LT) recipients but the factors that define the risk of ISW failure are largely unknown. One candidate prognostic factor for ISW success or operational tolerance (OT) is longer time between LT and ISW which we term "pre-withdrawal time". To clarify the impact of pre-withdrawal time span on subsequent ISW success or failure, we conducted a systematic review with meta-analysis. METHODS: We systematically interrogated the literature for LT recipient ISW studies reporting pre-withdrawal time. Eligible articles from Embase, Medline, and the Cochrane Central Register of Controlled Trials were used for backward and forward citation searching. Pre-withdrawal time individual patient data (IPD) was requested from authors. Pooled mean differences and time-response curves were calculated using random-effects meta-analyses. RESULTS: We included 17 studies with 691 patients, 15 of which (620 patients) with IPD. Study-level risk of bias was heterogeneous. Mean pre-withdrawal time was greater by 427 days [95% confidence interval (CI) 67-788] in OT compared to non-OT patients. This increase was potentiated to 799 days (95% CI 369-1229) or 1074 days (95% CI 685-1463) when restricting analysis to adult or European study participants. In time-response meta-analysis for adult or European ISW candidates, likelihood of OT increased by 7% (95% CI 4-10%) per year after LT (GRADE low- and moderate-certainty of evidence, respectively). CONCLUSIONS: Our data support the impact of pre-withdrawal time in ISW decision-making for adult and European LT recipients. PROSPERO REGISTRATION: CRD42021272995.

Ambulatory Long-Term Oxygen Therapy in Patients with Severe COPD: A Randomized Crossover Trial to Compare Constant-Minute-Volume and Constant-Bolus Systems.

Background and Methods: Constant-minute-volume and constant-bolus devices serve as two different means of portable oxygen conservation. A prospective randomised crossover study was conducted in COPD GOLD IV patients to investigate the effect of these two devices on dyspnea, oxygenation and 6-minute walking test (6MWT) distance. The primary endpoint was the final operating level required (operating level range 1-5 for both devices) by either device to meet the success criteria for mobile oxygen therapy, as outlined in the British Thoracic Society guidelines (SpO2 ≥90% throughout 6MWT; ≥10% increase in walking distance from baseline; improvement in BORG of at least 1 point from baseline). Results: Twenty-five patients were enrolled in the study and randomly assigned to one of two sequences involving the use of each type of portable oxygen conservation device. 14 female, 67.9 years (±7.8); FEV1: 27.3%pred. (±8.4); PaO2 at rest without oxygen: 50.3mmHg (±5.9). For both systems, 24/25 patients (96%) were successfully recruited. The mean operating-level difference when success criteria were met was -0.58 in favor of the constant bolus device (95% CI: -0.88 to -0.28, P <0.001). Secondary endpoints (walking distance, respiratory rate and BORG dyspnea) showed no statistically significant or clinically relevant differences. An algorithm created especially for this study showed a high success rate in terms of titration for the required operating level. Conclusion: Both portable oxygen-conserving devices met the success criteria in 96% of patients in the 6MWT when they were titrated to the correct level. The constant-bolus device required a significantly lower operating level to achieve the success criteria, hereby reducing energy consumption. Individual titration of the respective device is recommended, which can be facilitated by the novel titration algorithm described here.

Adherence enhancing interventions for pharmacological and oxygen therapy in patients with COPD: protocol for a systematic review and component network meta-analyses.

BACKGROUND: Chronic obstructive pulmonary disease (COPD) is characterised by hyperinflation and expiratory airflow limitation due to long-term exposure to irritants. The variety and complexity of COPD treatment and the possible added comorbidities may make the patients find it difficult to cope with the required medications. That is why supporting patients' adherence is critical because not taking medications correctly increases the risk of complications and creates an additional financial burden. A range of interventions aiming to improve patient adherence were used, and most of them are complex since they involve a mix of elements. Furthermore, despite the variety of available tools, assessing adherence is challenging because clinicians usually do not get a concrete judgement if their patients followed their treatment plan reliably. We aim to evaluate the effectiveness of adherence-enhancing interventions for COPD patients, explore which intervention (component) works for which patients and check the factors influencing the implementation and participant responses. METHODS: We will perform a comprehensive literature search (Medline, Embase, Cochrane Library, trial registries) without restrictions on language and publication status, and we will include all controlled studies investigating the effect of adherence-enhancing intervention on patients with COPD. We plan to involve COPD patients in the systematic review development through two patient interviews (one before and one after the systematic review). Two reviewers will perform the screening, data extraction and risk of bias (ROB) assessment. For ROB, we will use ROB 2.0 to assess randomised controlled trials, and ROBINS-I to assess non-randomised studies. We will perform pair-wise random-effects meta-analyses and component network meta-analyses to identify the most effective components and combinations of components. We will use the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach to assess the quality of evidence. To determine the degree of complexity, we will use the iCAT_SR checklist, and then, following a logical model, we will group the interventions according to prespecified criteria. DISCUSSION: This systematic review aims to point out the most effective and implementable adherence-enhancing interventions by using methods for synthesising evidence on complex interventions and involving COPD patients all along with the review process. SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42022353977.

AMSTAR 2 is only partially applicable to systematic reviews of non-intervention studies: a meta-research study.

OBJECTIVES: A measurement tool to assess systematic reviews 2 (AMSTAR 2) was originally developed for systematic reviews (SRs) of health-care interventions. The aim of this study was to assess the applicability of AMSTAR 2 to SRs of non-intervention studies. STUDY DESIGN AND SETTING: This was a meta-research study. We used 20 SRs for each of the following four types of SRs: Diagnostic Test Accuracy reviews, Etiology and/or Risk reviews, Prevalence and/or Incidence reviews, and Prognostic reviews (80 in total). Three authors applied AMSTAR 2 independently to each included SRs. Then, the authors assessed the applicability of each item to that SR type and any SR type. RESULTS: Researchers unanimously indicated that 7 of 16 AMSTAR 2 items were applicable for all four specific SR types and any SR type (items 2, 5, 6, 7, 10, 14 and 16), but 8 of 16 items for any SR type. These items could cover generic SR methods that do not depend on a specific SR type. CONCLUSION: AMSTAR 2 is only partially applicable for non-intervention SRs. There is a need to adapt/extend AMSTAR 2 for SRs of non-intervention studies. Our study can help to further define generic methodological aspects shared across SR types and methodological expectations for non-intervention SRs.

Clinical utility of overviews on adverse events of pharmacological interventions.

BACKGROUND: Overviews (i.e., systematic reviews of systematic reviews, meta-reviews, umbrella reviews) are a relatively new type of evidence synthesis. Among others, one reason to conduct an overview is to investigate adverse events (AEs) associated with a healthcare intervention. Overviews aim to provide easily accessible information for healthcare decision-makers including clinicians. We aimed to evaluate the clinical utility of overviews investigating AEs. METHODS: We used a sample of 27 overviews exclusively investigating drug-related adverse events published until 2021 identified in a prior project. We defined clinical utility as the extent to which overviews are perceived to be useful in clinical practice. Each included overview was assigned to one of seven pharmacological experts with expertise on the topic of the overview. The clinical utility and value of these overviews were determined using a self-developed assessment tool. This included four open-ended questions and a ranking of three clinical utility statements completed by clinicians. We calculated frequencies for the ranked clinical utility statements and coded the answers to the open-ended questions using an inductive approach. RESULTS: The overall agreement with the provided statements was high. According to the assessments, 67% of the included overviews generated new knowledge. In 93% of the assessments, the overviews were found to add value to the existing literature. The overviews were rated as more useful than the individual included systematic reviews (SRs) in 85% of the assessments. The answers to the open-ended questions revealed two key aspects of clinical utility in the included overviews. Firstly, it was considered useful that they provide a summary of available evidence (e.g., along with additional assessments, or across different populations, or in different settings that have not been evaluated together in the included SRs). Secondly, it was found useful if overviews conducted a new meta-analysis to answer specific research questions that had not been answered previously. CONCLUSIONS: Overviews on drug-related AEs are considered valuable for clinical practice by clinicians. They can make available evidence on AEs more accessible and provide a comprehensive view of available evidence. As the role of overviews evolves, investigations such as this can identify areas of value.

Cost-effectiveness of prehabilitation prior to elective surgery: a systematic review of economic evaluations.

BACKGROUND: Prehabilitation aims at enhancing patients' functional capacity and overall health status to enable them to withstand a forthcoming stressor like surgery. Our aim was to synthesise the evidence on the cost-effectiveness of prehabilitation for patients awaiting elective surgery compared with usual preoperative care. METHODS: We searched PubMed, Embase, the CRD database, ClinicalTrials.gov, the WHO ICTRP and the dissertation databases OADT and DART. Studies comparing prehabilitation for patients with elective surgery to usual preoperative care were included if they reported cost outcomes. All types of economic evaluations (EEs) were included. The primary outcome of the review was cost-effectiveness based on cost-utility analyses (CUAs). The risk of bias of trial-based EEs was assessed with the Cochrane risk of bias 2 tool and the ROBINS-I tool and the credibility of model-based EEs with the ISPOR checklist. Methodological quality of full EEs was assessed using the CHEC checklist. The EEs' results were synthesised narratively using vote counting based on direction of effect. RESULTS: We included 45 unique studies: 25 completed EEs and 20 ongoing studies. Of the completed EEs, 22 were trial-based and three model-based, corresponding to four CUAs, three cost-effectiveness analyses, two cost-benefit analyses, 12 cost-consequence analyses and four cost-minimization analyses. Three of the four trial-based CUAs (75%) found prehabilitation cost-effective, i.e. more effective and/or less costly than usual care. Overall, 16/25 (64.0%) EEs found prehabilitation cost-effective. When excluding studies of insufficient credibility/critical risk of bias, this number reduced to 14/23 (60.9%). In 8/25 (32.0%), cost-effectiveness was unclear, e.g. because prehabilitation was more effective and more costly, and in one EE prehabilitation was not cost-effective. CONCLUSIONS: We found some evidence that prehabilitation for patients awaiting elective surgery is cost-effective compared to usual preoperative care. However, we suspect a relevant risk of publication bias, and most EEs were of high risk of bias and/or low methodological quality. Furthermore, there was relevant heterogeneity depending on the population, intervention and methods. Future EEs should be performed over a longer time horizon and apply a more comprehensive perspective. TRIAL REGISTRATION: PROSPERO CRD42020182813.

Systematic reviewers used various approaches to data extraction and expressed several research needs: a survey.

OBJECTIVE: Data extraction is a prerequisite for analyzing, summarizing, and interpreting evidence in systematic reviews. Yet guidance is limited, and little is known about current approaches. We surveyed systematic reviewers on their current approaches to data extraction, opinions on methods, and research needs. STUDY DESIGN AND SETTING: We developed a 29-question online survey and distributed it through relevant organizations, social media, and personal networks in 2022. Closed questions were evaluated using descriptive statistics, and open questions were analyzed using content analysis. RESULTS: 162 reviewers participated. Use of adapted (65%) or newly developed extraction forms (62%) was common. Generic forms were rarely used (14%). Spreadsheet software was the most popular extraction tool (83%). Piloting was reported by 74% of respondents and included a variety of approaches. Independent and duplicate extraction was considered the most appropriate approach to data collection (64%). About half of respondents agreed that blank forms and/or raw data should be published. Suggested research gaps were the effects of different methods on error rates (60%) and the use of data extraction support tools (46%). CONCLUSION: Systematic reviewers used varying approaches to pilot data extraction. Methods to reduce errors and use of support tools such as (semi-)automation tools are top research gaps.

The use of digital web-based video training for correct inhalation technique during the COVID-19 pandemic.

Scientifically validated web-based training videos for proper inhalation technique were increasingly used by an international audience during the pandemic. Translations into additional languages would support a larger patient population. https://bit.ly/3lYQwsD.

Evidence-based health policy in Germany: lack of communication and coordination between academia and health authorities?

Health-care decision making should consider the best available evidence, often in the form of systematic reviews (SRs). The number of existing SRs and their overlap make their identification and use difficult. Decision makers often rely on de novo SRs instead of using existing SRs. We describe two cases of duplicate reviews (minimum volume threshold of total knee arthroplasties and lung cancer screening) and one case of duplicate primary data analysis (transcatheter aortic valve implantation). All cases have in common that unintended duplication of research occurred between health authorities and academia, demonstrating a lack of communication and coordination between them.It is important to note that academia and health authorities have different incentives. Academics are often measured by the number of peer-reviewed publications and grants awarded. In contrast, health authorities must comply with laws and are commissioned to deliver a specific report within a defined period of time. Most replication is currently unintended. A solution may be the collaboration of stakeholders commonly referred to as integrated knowledge translation (IKT). The IKT approach means that research is conducted in collaboration with the end users of the research. It requires active collaborations between researchers and decision-makers or knowledge users (clinicians, managers, policy makers) throughout the research process. Wherever cooperation is possible in spite of requirements for independence or confidentiality, legal regulations should facilitate and support collaborative approaches between academia and health authorities.

Protocol for the development of a core outcome set for studies on centralisation of healthcare services.

INTRODUCTION: Centralisation defined as the reorganisation of healthcare services into fewer specialised units serving a higher volume of patients is a potential measure for healthcare reforms aiming at reducing costs while improving quality. Research on centralisation of healthcare services is thus essential to inform decision-makers. However, so far studies on centralisation report a variability of outcomes, often neglecting outcomes at the health system level. Therefore, this study aims at developing a core outcome set (COS) for studies on centralisation of hospital procedures, which is intended for use in observational as well as in experimental studies. METHODS AND ANALYSIS: We propose a five-stage study design: (1) systematic review, (2) focus group, (3) interview studies, (4) online survey, (5) Delphi survey. The study will be conducted from March 2022 to November 2023. First, an initial list of outcomes will be identified through a systematic review on reported outcomes in studies on minimum volume regulations. We will search MEDLINE, EMBASE, CENTRAL, CINHAL, EconLIT, PDQ-Evidence for Informed Health Policymaking, Health Systems Evidence, Open Grey and also trial registries. This will be supplemented with relevant outcomes from published studies on centralisation of hospital procedures. Second, we will conduct a focus group with representatives of patient advocacy groups for which minimum volume regulations are currently in effect in Germany or are likely to come into effect to identify outcomes important to patients. Furthermore, two interview studies, one with representatives of the German medical societies and one with representatives of statutory health insurance funds, as well as an online survey with health services researchers will be conducted. In our analyses of the suggested outcomes, we will largely follow the categorisation scheme developed by the Cochrane EPOC group. Finally, a two-round online Delphi survey with all stakeholder groups using predefined score criteria for consensus will be employed to first prioritise outcomes and then agree on the final COS. ETHICS AND DISSEMINATION: This study has been approved by the Research Ethics Committee at the Brandenburg Medical School Theodor Fontane (MHB). The final COS will be disseminated to all stakeholders involved and through peer-reviewed publications and conferences.

The effect of minimum volume standards in hospitals (MIVOS) - protocol of a systematic review.

BACKGROUND: The volume-outcome relationship, i.e., higher hospital volume results in better health outcomes, has been established for different surgical procedures as well as for certain nonsurgical medical interventions. Accordingly, many countries such as Germany, the USA, Canada, the UK, and Switzerland have established minimum volume standards. To date, there is a lack of systematically summarized evidence regarding the effects of such regulations. METHODS: To be included in the review, studies must measure any effects connected to minimum volume standards. Outcomes of interest include the following: (1) patient-related outcomes, (2) process-related outcomes, and (3) health system-related outcomes. We will include (cluster) randomized controlled trials ([C]RCTs), non-randomized controlled trials (nRCTs), controlled before-after studies (CBAs), and interrupted time-series studies (ITSs). We will apply no restrictions regarding language, publication date, and publication status. We will search MEDLINE (via PubMed), Embase (via Embase), CENTRAL (via Cochrane Library), CINHAL (via EBSCO), EconLit (via EBSCO), PDQ evidence for informed health policymaking, health systems evidence, OpenGrey, and also trial registries for relevant studies. We will further search manually for additional studies by cross-checking the reference lists of all included primary studies as well as cross-checking the reference lists of relevant systematic reviews. To evaluate the risk of bias, we will use the ROBINS-I and RoB 2 risk-of-bias tools for the corresponding study designs. For data synthesis and statistical analyses, we will follow the guidance published by the EPOC Cochrane group (Cochrane Effective Practice and Organisation of Care (EPOC), EPOC Resources for review authors, 2019). DISCUSSION: This systematic review focuses on minimum volume standards and the outcomes used to measure their effects. It is designed to provide thorough and encompassing evidence-based information on this topic. Thus, it will inform decision-makers and policymakers with respect to the effects of minimum volume standards and inform further studies in regard to research gaps. SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42022318883.

Potentially Inadequate Medications in the Elderly: PRISCUS 2.0.

BACKGROUND: The term potentially inadequate medication (PIM) is used to describe substances that may be unsuitable for use inthe elderly and should be avoided. The PRISCUS list, published in 2010, was the first catalog of PIM designed for the Germandrug market to become adopted in practice. While 24% of German patients aged ≥ 65 years were prescribed at least one PIMper year in 2009, the proportion in 2019 was only 14.5%. METHODS: In a three-round Delphi process, experts from clinical practice and research evaluated whether selected substancesare PIM for the elderly. The participants were provided with dedicated literature including systematic reviews carried out for theparticular purposes of this project. RESULTS: Fifty-nine persons took part in the Delphi process and, in addition, contributed comments and therapeutic alternatives.Altogether, 187 substances were classed as PIM. One hundred thirty-three of the substances now listed were not in the originalPRISCUS list: these include some oral antidiabetics, all of the selective COX-2 inhibitors, and moderately long acting benzodiazepinessuch as oxazepam. For some other substances, e.g., proton pump inhibitors (PPI), the advisability of treatment formore than 8 weeks was considered as potentially inappropriate, as was the use of ibuprofen in doses >1200 mg/day and formore than 1 week without PPI. Risperidone for more than 6 weeks is also PIM. CONCLUSION: The new, greatly extended PRISCUS list must now be validated in epidemiological and prospective studies and itspracticability in routine daily use must be verified.

The effects of modifying elements of written informed consent forms for elective surgical or invasive procedures: A systematic review.

OBJECTIVE: To study the effect of modifying content and design elements within written informed-consent-forms (ICF) for patients undergoing elective surgical or invasive procedures. METHODS: We included (quasi-)randomized trials in which a modified written ICF (e.g. visual aids) was compared to a standard written ICF. We searched PubMed, Web-of-Science and PsycINFO until 08/2021. Risk of Bias was assessed. The complexity of intervention was assessed using the Intervention Complexity Assessment Tool for Systematic Reviews. RESULTS: Eleven trials with 1091 participants were eligible. Effect sizes and levels of evidence varied from trivial to moderate andthere were contradictory findings for some outcomes. Providing patients with more informationin general or specific information on risks and complications mostly increased anxiety. The use of verbal risk presentation decreased anxiety and increased satisfaction.A lower readability level decreased anxiety and improved comprehension and knowledge. CONCLUSION: Our results suggest that providing more information and addressing certain types of risks have differential effects. While more information improved knowledge, it also increased anxiety. We did not find any or only insufficient evidence for many other possible ICF modifications. PRACTICE IMPLICATIONS: When developing ICFs the differential impact of different elements on patient important outcomes should be carefully considered.

Results of the Diaphragmatic Plication Database: 10 Years' Experience.

BACKGROUND: Unilateral diaphragmatic paralysis or paresis (UDP) in adults is an often overlooked disease which relevantly impairs the patient's lung function and quality of life. Particularly in idiopathic UDP, there is no evidence for conservative therapy and only little evidence for surgical therapy. METHODS: The method involves retrospective single-center analysis of patients with UDP persistent for at least 1 year who were operated by diaphragmatic resection, plication, and augmentation with a polypropylene mesh. The patients were tested for lung and diaphragmatic function, six-minute walk test (6MWT), and blood gas analysis before, 3 and 12 months after surgery. RESULTS: In total, 85 patients received surgery for UDP. The most frequent reasons for UDP were idiopathic (67%), iatrogenic (mainly cardiac and cervical spine surgery; 24%), and trauma (9%). The mean operation time was 84 ± 24 minutes, the length of hospital stay 8.4 ± 3.9 days, chest tubes were removed after 11.7 ± 4.1 days. Overall morbidity was 42%, mortality 0%. Forced expiratory volume in one second (FEV1) in supine position improved by 12.4% absolute, vital capacity by 11.8% absolute, and sniff nasal inspiratory pressure by 1.4 kPa 12 months after surgery (p <0.001 each). Total lung capacity increased by 6.8% absolute at 12 months (p = 0.001) The 6MWT distance improved by 45.9 m at 3 months and 50.9 m at 12 months (p = 0.001, each). CONCLUSION: Surgical therapy for UDP is highly effective in the long term. The superiority over conservative treatments needs to be evaluated prospectively with standardized physiotherapeutic protocols. FEV1 in supine position and 6MWT are easy to perform tests and represent statistically and patient-relevant outcomes.

Performance of several types of beta-binomial models in comparison to standard approaches for meta-analyses with very few studies.

BACKGROUND: Meta-analyses are used to summarise the results of several studies on a specific research question. Standard methods for meta-analyses, namely inverse variance random effects models, have unfavourable properties if only very few (2 - 4) studies are available. Therefore, alternative meta-analytic methods are needed. In the case of binary data, the "common-rho" beta-binomial model has shown good results in situations with sparse data or few studies. The major concern of this model is that it ignores the fact that each treatment arm is paired with a respective control arm from the same study. Thus, the randomisation to a study arm of a specific study is disrespected, which may lead to compromised estimates of the treatment effect. Therefore, we extended this model to a version that respects randomisation. The aim of this simulation study was to compare the "common-rho" beta-binomial model and several other beta-binomial models with standard meta-analyses models, including generalised linear mixed models and several inverse variance random effects models. METHODS: We conducted a simulation study comparing beta-binomial models and various standard meta-analysis methods. The design of the simulation aimed to consider meta-analytic situations occurring in practice. RESULTS: No method performed well in scenarios with only 2 studies in the random effects scenario. In this situation, a fixed effect model or a qualitative summary of the study results may be preferable. In scenarios with 3 or 4 studies, most methods satisfied the nominal coverage probability. The "common-rho" beta-binomial model showed the highest power under the alternative hypothesis. The beta-binomial model respecting randomisation did not improve performance. CONCLUSION: The "common-rho" beta-binomial appears to be a good option for meta-analyses of very few studies. As residual concerns about the consequences of disrespecting randomisation may still exist, we recommend a sensitivity analysis with a standard meta-analysis method that respects randomisation.