Search

Network meta-analysis compares different interventions for the same condition, by combining direct and indirect evidence derived from all eligible studies. Network metaanalysis has been increasingly used by applied scientists and it is a major research topic for methodologists. This article describes the R package netmeta, which adopts frequentist methods to fit network meta-analysis models. We provide a roadmap to perform network meta-analysis, along with an overview of the main functions of the package. We present three worked examples considering different types of outcomes and different data formats to facilitate researchers aiming to conduct network meta-analysis with netmeta.

Additional file 3.

Additional file 9.

Additional file 5.

Additional file 6.

Additional file 10.

Additional file 2.

Additional file 8.

Background Implications for research (IfR) sections are an important part of systematic reviews (SRs) in healthcare. The PRISMA 2020 statement recommends discussion of the implications of the results for practice, while Cochrane Reviews require an IfR section. However, it is unclear to what extent systematic reviews discuss IfR and if this section is useful. We aimed i) to assess whether systematic reviews include an IfR section and ii) to evaluate which elements informed reporting of IfR. Methods We conducted a meta-research study based on SRs of interventions in advanced cancer patients from a previous project (CRD42019134904). We assessed if the following predefined IfR variables were referred to in the included SRs: patients, intervention, control, outcome, study design (PICOS); Grading of Recommendations, Assessment, Development and Evaluation (GRADE) domains: risk of bias, inconsistency, indirectness, imprecision, publication bias. Data were independently extracted by three reviewers after piloting the data extraction form. Discrepancies were resolved in weekly in-depth discussions. Results We included 261 SRs. The majority evaluated a pharmacological intervention (244, 93.5%); twenty-nine were Cochrane Reviews (11.1%). Four out of five SRs included an IfR section (210, 80.5%). PICOS elements commonly addressed in IfR sections were ‘intervention’ (121, 57.6%), ‘patient‘ (113, 53.8%) and ‘study design’ (107, 51.0%). The most frequent PICOS combinations were ‘patient and intervention‘ (71, 33.8%) and ‘patient, intervention and study design‘ (34, 16.2%). GRADE domains were rarely used for informing IfR recommendations: ‘risk of bias‘ (2, 1.0%), and ‘imprecision‘ (1, 0.5%), ‘inconsistency‘ (1, 0.5%). Additional factors informing IfR recommendations were considerations on cost effectiveness (9, 4.3%), reporting standards (4, 1.9%), and individual patient data meta-analysis (4, 1.9%). Conclusion Although four out of five systematic reviews of our sample included an IfR section, these were mostly considering only selected PICOS elements and GRADE domains were used rarely to guide future research recommendations. In the absence of established reporting standards, we propose more comprehensive and structured IfR sections in SRs informed by PICOS elements and GRADE domains. We suggest further empirical and theoretical work to eventually develop guidance on what IfR elements to consider and how to report IfR in SRs of interventions.

Additional file 7.

Additional file 1.

Additional file 4.

Regularly updated (every two months) living synthesis to identify and summarize the available literature on interventions to foster mental health, psychosocial support, resilience and/or stress management in healthcare workers in the face of the current pandemic. This work is part of the CEOsys project (https://www.covid-evidenz.de; https://www.ceosys.de) which is funded by the German Federal Ministry for Education and Research (BMBF) as part of the national network of academic medical research into COVID-19, Germany (Nationales Netzwerk Universitätsmedizin, NUM).

Background/research question: Implications for research (IfR) in systematic reviews are of high importance to indicate research gaps, guide meaningful future research, and to avoid research waste. Besides, IfR is a mandatory chapter in Cochrane Reviews and part of PRISMA [ref:1]. [for full text, please go to the a.m. URL]

Regularly updated (every two months) living synthesis to identify and summarize the available literature on the impact of the SARS-CoV-2 pandemic on the mental health and work ability in healthcare workers as well as possible (demographic, psychosocial etc.) risk and protective factors for mental health. This work is part of the CEOsys project (https://www.covid-evidenz.de; https://www.ceosys.de) which is funded by the German Federal Ministry for Education and Research (BMBF) within the national network of academic medical research into COVID-19, Germany (Nationales Netzwerk Universitätsmedizin, NUM).

Beschreibung: Das dynamische Publikationsgeschehen in vielen medizinischen Fachgebieten erfordert zunehmend strukturierte und aussagekräftige Zusammenfassungen der Evidenz durch Systematische Reviews und Metaanalysen. Anwendungsorientierte Softwares erleichtern die schnelle Durchführung von [zum vollständigen Text gelangen Sie über die oben angegebene URL]

Background/research question: The methodological quality of systematic reviews (SRs) can be evaluated using a measurement tool to assess systematic reviews 2 (AMSTAR 2). Although the tool is well-designed and recommended by experts, its rating scheme for the overall confidence has been criticized. [for full text, please go to the a.m. URL]

Network meta-analysis (NMA) is widely used in evidence synthesis to estimate the effects of several competing interventions for a given clinical condition. One of the challenges is that it is not possible in disconnected networks. Component network meta-analysis (CNMA) allows technically 'reconnecting' a disconnected network with multicomponent interventions. The additive CNMA model assumes that the effect of any multicomponent intervention is the additive sum of its components. This assumption can be relaxed by adding interaction component terms, which improves the goodness of fit but decreases the network connectivity. Model selection aims at finding the model with a reasonable balance between the goodness of fit and connectivity (selected CNMA model). We aim to introduce a forward model selection strategy for CNMA models and to investigate the performance of CNMA models for connected and disconnected networks. We applied the methods to a real Cochrane review dataset and simulated data with additive, mildly, or strongly violated intervention effects. We started with connected networks, and we artificially constructed disconnected networks. We compared the results of the additive and the selected CNMAs from each connected and disconnected network with the NMA using the mean squared error and coverage probability. CNMA models provide good performance for connected networks and can be an alternative to standard NMA if additivity holds. On the contrary, model selection does not perform well for disconnected networks, and we recommend conducting separate analyses of subnetworks., Comment: 17 pages, 5 figures, 5 tables

Background Network meta-analysis (NMA) has attracted growing interest in evidence-based medicine. Consistency between different sources of evidence is fundamental to the reliability of the NMA results. The purpose of the present study was to estimate the prevalence of evidence of inconsistency and describe its association with different NMA characteristics. Methods We updated our collection of NMAs with articles published up to July 2018. We included networks with randomised clinical trials, at least four treatment nodes, at least one closed loop, a dichotomous primary outcome, and available arm-level data. We assessed consistency using the design-by-treatment interaction (DBT) model and testing all the inconsistency parameters globally through the Wald-type chi-squared test statistic. We estimated the prevalence of evidence of inconsistency and its association with different network characteristics (e.g., number of studies, interventions, intervention comparisons, loops). We evaluated the influence of the network characteristics on the DBT p-value via a multivariable regression analysis and the estimated Pearson correlation coefficients. We also evaluated heterogeneity in NMA (consistency) and DBT (inconsistency) random-effects models. Results We included 201 published NMAs. The p-value of the design-by-treatment interaction (DBT) model was lower than 0.05 in 14% of the networks and lower than 0.10 in 20% of the networks. Networks including many studies and comparing few interventions were more likely to have small DBT p-values (less than 0.10), which is probably because they yielded more precise estimates and power to detect differences between designs was higher. In the presence of inconsistency (DBT p-value lower than 0.10), the consistency model displayed higher heterogeneity than the DBT model. Conclusions Our findings show that inconsistency was more frequent than what would be expected by chance, suggesting that researchers should devote more resources to exploring how to mitigate inconsistency. The results of this study highlight the need to develop strategies to detect inconsistency (because of the relatively high prevalence of evidence of inconsistency in published networks), and particularly in cases where the existing tests have low power. Supplementary Information The online version contains supplementary material available at 10.1186/s12874-021-01401-y.

Background: The Peto method is a classic fixed effect approach for pairwise meta-analysis of binary outcomes [ref:4] and is available, for example, in Review Manager 5 by Cochrane. The Peto odds ratio – calculated using the observed and expected number of events in one treatment group[for full text, please go to the a.m. URL], 65th Annual Meeting of the German Association for Medical Informatics, Biometry and Epidemiology (GMDS), Meeting of the Central European Network (CEN: German Region, Austro-Swiss Region and Polish Region) of the International Biometric Society (IBS)

Background: Component network meta-analysis (CNMA) is an extension of standard network meta-analysis (NMA). It can be used when treatments are composed of common components, such as combinations of drugs or combinations of a drug with psychotherapy. In addition to knowledge of the structure of the network,[for full text, please go to the a.m. URL], 65th Annual Meeting of the German Association for Medical Informatics, Biometry and Epidemiology (GMDS), Meeting of the Central European Network (CEN: German Region, Austro-Swiss Region and Polish Region) of the International Biometric Society (IBS)

Additional file 1: Appendix 1. Eligibility criteria, screening, study selection, and data abstraction. Appendix 2. Model description. Appendix Figure 1. Flowchart for network meta-analysis study inclusion. Appendix Table 1. Number of consistent and inconsistent networks at 0.05 and 0.10 significance levels using DL and REML heterogeneity estimators. Appendix Table 2. Multivariable regression analysis results. Appendix Figure 2. Stacked bar plot of consistent (green bars) and inconsistent (blue bars) networks at α = 0.05 per heterogeneity estimator and year of study publication‡. Appendix Figure 3. Plot of p-values (fourth-root scale) of the DBT model against network structural characteristics (logarithmic scale). Appendix Figure 4. Plot of p-values (fourth-root scale) of the DBT model against ratios of network structural characteristics (logarithmic scale). Appendix Figure 5. Plot of the between-study standard deviation estimated in the consistency model against the ratio of the number of studies to the number of interventions in a network (logarithmic scale). Appendix Figure 6. Box plot of the p-values (fourth-root scale) of the DBT model per (a) type of outcome, (b) type of intervention comparison, c) presence of complex interventions, and d) presence of at least one direct intervention comparison with a single study. Appendix Figure 7. Plot of the p-values (fourth-root scale) of the DBT model against the I-squared. Appendix Figure 8. Plot of the between-study standard deviation in consistency against the inconsistency model. Appendix Figure 9. Plot of the between-study standard deviation in inconsistency against the degrees of freedom of the Wald chi-square test (logarithmic scale).

In this project we provide the R code and data for reproducing results of the article "Reevaluation of statistically significant meta-analyses in advanced cancer patients using the Hartung-Knapp method and prediction intervals – a methodological study" published in Research Synthesis Methods

Network meta-analysis is a more recent development that provides methods to combine direct and indirect evidence and thus constitutes an extension of pairwise meta-analysis for combining the results from studies that used different comparison groups (Salanti, 2012). R package netmeta (Rücker et al., 2020) implements a frequentist approach for network meta-analysis based on graph-theoretical methods (Rücker, 2012) and is one of the most popular R packages for network meta-analysis. Main objective of netmeta is to provide a comprehensive set of R functions for network meta-analysis in a user-friendly implementation.In this presentation, we will give a brief overview of implemented methods in netmeta as well as planned future extensions. References: Salanti G (2012): Indirect and mixed-treatment comparison, network, or multiple-treatments meta-analysis: many names, many benefits, many concerns for the next generation evidence synthesis tool. Research Synthesis Methods. 3(2):80-97. Rücker G (2012): Network meta-analysis, electrical networks and graph theory. Research Synthesis Methods. 3(4):312-24. Gerta Rücker et al. (2020): netmeta: Network Meta-Analysis using Frequentist Methods. R package version 1.2-1. https://CRAN.R-project.org/package=netmeta

Additional file 2. Search strategy for the electronic database PubMed/MEDLINE.

Additional file 2: Results of the systematic review with meta-analyses. eTable 9. Details on number of included (pandemic and comparative) studies. eTable 10. Study characteristics of the prepandemic comparative studies. eTable 11. Cut-off values reported in included pandemic studies. eResults 1. Forest plots of main analyses. eTable 12. Quality assessment of included pandemic studies. eTable 13. Assessment of level of comparability between pandemic and prepandemic comparative studies. eResults 2. Forest plots of sensitivity analyses. eResults 3. Detailed results of subgroup analyses. eTable 14. Risk factors in the general population, healthcare workers, and patients. eTable 15. Protective factors in the general population, healthcare workers, and patients.

Additional file 1. PRISMA-P 2015 Checklist

Additional file 1: Methods of the systematic review with meta-analyses. eTable 1. MOOSE Checklist. eTable 2. Differences between protocol and review. eMethods 1. Search strategies for SARS-CoV-2 (‘pandemic’) studies. eMethods 2. Search strategy for prepandemic comparative studies. eTable 3. Eligibility criteria for SARS-CoV-2 pandemic studies. eTable 4. Eligibility criteria for prepandemic comparative studies. eTable 5. Eligibility criteria for pairwise meta-analyses. eTable 6. Customized data extraction sheet. eTable 7. Modified quality assessment tool. eTable 8. Rating of comparability between pandemic and prepandemic comparative studies. eMethods 3. Further methodological details of this systematic review and meta-analyses.

Background/research question: Taking heterogeneity in meta-analyses into account is crucial for drawing conclusions, e.g. for treatment guidelines. An analysis of prediction intervals (PIs) and the way of considering heterogeneity in meta-analyses of advanced cancer patients has not yet been performed[for full text, please go to the a.m. URL], Nützliche patientenrelevante Forschung; 21. Jahrestagung des Deutschen Netzwerks Evidenzbasierte Medizin

Additional file 2 Plots resulting from running all commands in Additional File 1.

Standard generic inverse variance methods for the combination of single proportions are based on transformed proportions using the logit, arcsine, and Freeman-Tukey double arcsine transformations. Generalized linear mixed models are another more elaborate approach. Irrespective of the approach, meta-analysis results are typically back-transformed to the original scale in order to ease interpretation. Whereas the back-transformation of meta-analysis results is straightforward for most transformations, this is not the case for the Freeman-Tukey double arcsine transformation, albeit possible. In this case study with five studies, we demonstrate how seriously misleading the back-transformation of the Freeman-Tukey double arcsine transformation can be. We conclude that this transformation should only be used with special caution for the meta-analysis of single proportions due to potential problems with the back-transformation. Generalized linear mixed models seem to be a promising alternative.

Background A meta-analysis as part of a systematic review aims to provide a thorough, comprehensive and unbiased statistical summary of data from the literature. However, relevant study results could be missing from a meta-analysis because of selective publication and inadequate dissemination. If missing outcome data differ systematically from published ones, a meta-analysis will be biased with an inaccurate assessment of the intervention effect. As part of the EU-funded OPEN project (www.open-project.eu) we conducted a systematic review that assessed whether the inclusion of data that were not published at all and/or published only in the grey literature influences pooled effect estimates in meta-analyses and leads to different interpretation. Methods and findings Systematic review of published literature (methodological research projects). Four bibliographic databases were searched up to February 2016 without restriction of publication year or language. Methodological research projects were considered eligible for inclusion if they reviewed a cohort of meta-analyses which (i) compared pooled effect estimates of meta-analyses of health care interventions according to publication status of data or (ii) examined whether the inclusion of unpublished or grey literature data impacts the result of a meta-analysis. Seven methodological research projects including 187 meta-analyses comparing pooled treatment effect estimates according to different publication status were identified. Two research projects showed that published data showed larger pooled treatment effects in favour of the intervention than unpublished or grey literature data (Ratio of ORs 1.15, 95% CI 1.04–1.28 and 1.34, 95% CI 1.09–1.66). In the remaining research projects pooled effect estimates and/or overall findings were not significantly changed by the inclusion of unpublished and/or grey literature data. The precision of the pooled estimate was increased with narrower 95% confidence interval. Conclusions Although we may anticipate that systematic reviews and meta-analyses not including unpublished or grey literature study results are likely to overestimate the treatment effects, current empirical research shows that this is only the case in a minority of reviews. Therefore, currently, a meta-analyst should particularly consider time, effort and costs when adding such data to their analysis. Future research is needed to identify which reviews may benefit most from including unpublished or grey data.

Background Network meta-analysis is used to compare three or more treatments for the same condition. Within a Bayesian framework, for each treatment the probability of being best, or, more general, the probability that it has a certain rank can be derived from the posterior distributions of all treatments. The treatments can then be ranked by the surface under the cumulative ranking curve (SUCRA). For comparing treatments in a network meta-analysis, we propose a frequentist analogue to SUCRA which we call P-score that works without resampling. Methods P-scores are based solely on the point estimates and standard errors of the frequentist network meta-analysis estimates under normality assumption and can easily be calculated as means of one-sided p-values. They measure the mean extent of certainty that a treatment is better than the competing treatments. Results Using case studies of network meta-analysis in diabetes and depression, we demonstrate that the numerical values of SUCRA and P-Score are nearly identical. Conclusions Ranking treatments in frequentist network meta-analysis works without resampling. Like the SUCRA values, P-scores induce a ranking of all treatments that mostly follows that of the point estimates, but takes precision into account. However, neither SUCRA nor P-score offer a major advantage compared to looking at credible or confidence intervals. Electronic supplementary material The online version of this article (doi:10.1186/s12874-015-0060-8) contains supplementary material, which is available to authorized users.

Preferred Reporting Items for Systematic review and Meta-Analysis Protocols (PRISMA-P) 2015 checklist: recommended items to address in a systematic review protocol.

Hintergrund und Fragestellung: Auf wissenschaftlichen Konferenzen vorgestellte Studien sind durch eine elektronische Literatursuche oft schwer zu identifizieren. Diese sogenannte ‚graue Literatur‘ wird deshalb oft nicht in systematische Reviews (SRs) eingeschlossen, was zu Disseminations-Bias[for full text, please go to the a.m. URL], Prävention zwischen Evidenz und Eminenz; 15. Jahrestagung des Deutschen Netzwerks Evidenzbasierte Medizin

„Cloud Computing“, das heißt in diesem Kontext die temporäre, stundenweise abgerechnete Anmietung von Rechenressourcen bei einem kommerziellen Provider. Dies erscheint auf den ersten Blick für viele Institute und Arbeitsgruppen, die nicht auf die Infrastruktur eines großen[for full text, please go to the a.m. URL], Mainz//2011; 56. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds), 6. Jahrestagung der Deutschen Gesellschaft für Epidemiologie (DGEpi)

In dieser Arbeit werden die Eigenschaften statistischer Tests auf Biasin Meta-Analysen mit binären Zielgrößen im Rahmen von Simulationensowohl unter der Null-Hypothese als auch unter zwei verschiedenenAlternativen (mit moderatem und starkem Selektionsbias)untersucht. Insbesondere wird der Einfluß der Studiengröße derEinzelstudien, des zugrundeliegenden Therapieeffekts, der Ereignisratein der Kontrollgruppe und des Ausmaßes der Heterogenität zwischen denStudien auf die Ergebnisse der Tests auf Bias bewertet.In die Untersuchungen werden die beiden gebräuchlichsten Testverfahrenauf Bias in Meta-Analysen sowie ein neu vorgeschlagener Testeingeschlossen. Ein vorhandener Test basiert auf der Rangkorrelationzwischen einer standardisierten Form des geschätzten Therapieeffekt(z.B. log odds ratio) und der geschätzten Varianz des Therapieeffekts(z.B. geschätzte Varianz des log odds ratios). Der zweite Test basiertauf einer linearen Regression des standardisierten geschätztenTherapieeffekts auf die Inverse der Wurzel der geschätzten Varianz. Inbeiden Testverfahren wird insbesondere die geschätzte asymptotischeVarianz verwendet. Der neu vorgeschlagene Test auf Bias inMeta-Analysen stellt eine Modifikation des Rankkorrelationstestsdar. Anstelle des standardisierten geschätzten Therapieeffekts und dergeschätzten asymptotischen Varianz des Therapieeffekts werden Größenverwendet, die aus der nicht-zentralen hypergeometrischen Verteilungabgeleitet sind.Unter der Nullhypothese, daß kein Selektionsbias vorliegt, hält derneu vorgeschlagene Test, im Gegensatz zu den beiden existierendenTests, das Niveau in allen untersuchten Situationen ein. Alle Testshaben eine geringe Power zur Detektion von Selektionsbias in denmeisten praxisrelevanten Situationen., In this thesis, properties of statistical tests for bias inmeta-analysis are investigated via simulations both under the nullhypothesis of no selection bias and two different alternatives(moderate and strong selection bias). A total of four differentsimulation designs is used which differ only in the mechanism utilisedto generate sample sizes of individual trials; other factors (numberof trials per meta-analysis, true treatment effect, control eventrate, between-trial variance, amount of bias) are equal in alldesigns.Two commonly used tests, i.e. a rank correlation test and a linearregression test, as well as a newly proposed test are considered inthe simulations. The new test is a modification of the rankcorrelation test using different factors; the main difference consistsin the use of a different estimate for the within-trial variance. Inthe new test, the variance estimate is derived from the conditionaldistribution of the observed cell count given the marginal totalswhich follows a non-central hypergeometric distribution.A clear tendency of inflation of type I error rates with increasingtreatment effect is apparent for linear regression and rankcorrelation test in simulations with a large amount of sparse binarydata; results for the new test are much better in this situation. Thepower of all tests to detect selection bias is not sufficient in mostclinically important situations.

Background Statisticians investigate new methods in simulations to evaluate their properties for future real data applications. Results are often presented in a number of figures, e.g., Trellis plots. We had conducted a simulation study on six statistical methods for estimating the treatment effect in binary outcome meta-analyses, where selection bias (e.g., publication bias) was suspected because of apparent funnel plot asymmetry. We varied five simulation parameters: true treatment effect, extent of selection, event proportion in control group, heterogeneity parameter, and number of studies in meta-analysis. In combination, this yielded a total number of 768 scenarios. To present all results using Trellis plots, 12 figures were needed. Methods Choosing bias as criterion of interest, we present a ‘nested loop plot’, a diagram type that aims to have all simulation results in one plot. The idea was to bring all scenarios into a lexicographical order and arrange them consecutively on the horizontal axis of a plot, whereas the treatment effect estimate is presented on the vertical axis. Results The plot illustrates how parameters simultaneously influenced the estimate. It can be combined with a Trellis plot in a so-called hybrid plot. Nested loop plots may also be applied to other criteria such as the variance of estimation. Conclusion The nested loop plot, similar to a time series graph, summarizes all information about the results of a simulation study with respect to a chosen criterion in one picture and provides a suitable alternative or an addition to Trellis plots. Electronic supplementary material The online version of this article (doi:10.1186/1471-2288-14-129) contains supplementary material, which is available to authorized users.

Many healthcare interventions are complex, consisting of multiple, possibly interacting, components. Several methodological articles addressing complex interventions in the meta-analytical context have been published. We hereby provide an overview of methods used to evaluate the effects of complex interventions with meta-analytical models. We summarized the methodology, highlighted new developments, and described the benefits, drawbacks, and potential challenges of each identified method. We expect meta-analytical methods focusing on components of several multicomponent interventions to become increasingly popular due to recently developed, easy-to-use, software tools that can be used to conduct the relevant analyses. The different meta-analytical methods are illustrated through two examples comparing psychotherapies for panic disorder.