Sharing information across patient subgroups to draw conclusions from sparse treatment networks.

Network meta-analysis (NMA) usually provides estimates of the relative effects with the highest possible precision. However, sparse networks with few available studies and limited direct evidence can arise, threatening the robustness and reliability of NMA estimates. In these cases, the limited amount of available information can hamper the formal evaluation of the underlying NMA assumptions of transitivity and consistency. In addition, NMA estimates from sparse networks are expected to be imprecise and possibly biased as they rely on large-sample approximations that are invalid in the absence of sufficient data. We propose a Bayesian framework that allows sharing of information between two networks that pertain to different population subgroups. Specifically, we use the results from a subgroup with a lot of direct evidence (a dense network) to construct informative priors for the relative effects in the target subgroup (a sparse network). This is a two-stage approach where at the first stage, we extrapolate the results of the dense network to those expected from the sparse network. This takes place by using a modified hierarchical NMA model where we add a location parameter that shifts the distribution of the relative effects to make them applicable to the target population. At the second stage, these extrapolated results are used as prior information for the sparse network. We illustrate our approach through a motivating example of psychiatric patients. Our approach results in more precise and robust estimates of the relative effects and can adequately inform clinical practice in presence of sparse networks.

Methodological review of NMA bias concepts provides groundwork for the development of a list of concepts for potential inclusion in a new risk of bias tool for network meta-analysis (RoB NMA Tool).

INTRODUCTION: Network meta-analyses (NMAs) have gained popularity and grown in number due to their ability to provide estimates of the comparative effectiveness of multiple treatments for the same condition. The aim of this study is to conduct a methodological review to compile a preliminary list of concepts related to bias in NMAs. METHODS AND ANALYSIS: We included papers that present items related to bias, reporting or methodological quality, papers assessing the quality of NMAs, or method papers. We searched MEDLINE, the Cochrane Library and unpublished literature (up to July 2020). We extracted items related to bias in NMAs. An item was excluded if it related to general systematic review quality or bias and was included in currently available tools such as ROBIS or AMSTAR 2. We reworded items, typically structured as questions, into concepts (i.e. general notions). RESULTS: One hundred eighty-one articles were assessed in full text and 58 were included. Of these articles, 12 were tools, checklists or journal standards; 13 were guidance documents for NMAs; 27 were studies related to bias or NMA methods; and 6 were papers assessing the quality of NMAs. These studies yielded 99 items of which the majority related to general systematic review quality and biases and were therefore excluded. The 22 items we included were reworded into concepts specific to bias in NMAs. CONCLUSIONS: A list of 22 concepts was included. This list is not intended to be used to assess biases in NMAs, but to inform the development of items to be included in our tool.

HIGHLIGHTS: • Our research aimed to develop a preliminary list of concepts related to bias with the goal of developing the first tool for assessing the risk of bias in the results and conclusions of a network meta-analysis (NMA).• We followed the methodology proposed by Whiting (2017) and Sanderson (2007) for creating systematically developed lists of quality items, as a first step in the development of a risk of bias tool for network meta-analysis (RoB NMA Tool).• We included items related to biases in NMAs and excluded items that are equally applicable to all systematic reviews as they are covered by other tools (e.g. ROBIS, AMSTAR 2).• Fifty-seven studies were included generating 99 items, which when screened, yielded 22 included items. These items were then reworded into concepts in preparation for a Delphi process for further vetting by external experts.• A limitation of our study is the challenge in retrieving methods studies as methods collections are not regularly updated.

A REML method for the evidence-splitting model in network meta-analysis.

Checking for possible inconsistency between direct and indirect evidence is an important task in network meta-analysis. Recently, an evidence-splitting (ES) model has been proposed, that allows separating direct and indirect evidence in a network and hence assessing inconsistency. A salient feature of this model is that the variance for heterogeneity appears in both the mean and the variance structure. Thus, full maximum likelihood (ML) has been proposed for estimating the parameters of this model. Maximum likelihood is known to yield biased variance component estimates in linear mixed models, and this problem is expected to also affect the ES model. The purpose of the present paper, therefore, is to propose a method based on residual (or restricted) maximum likelihood (REML). Our simulation shows that this new method is quite competitive to methods based on full ML in terms of bias and mean squared error. In addition, some limitations of the ES model are discussed. While this model splits direct and indirect evidence, it is not a plausible model for the cause of inconsistency.

Comparison of two distinct arrival and treatment programs for bovine respiratory disease in high-risk feeder cattle entering a feedlot.

Antimicrobial metaphylaxis of high-risk cattle entering the feedlot is a common management strategy implemented against bovine respiratory disease (BRD). Typically, following a prescribed postmetaphylactic interval (PMI), animals displaying clinical signs of BRD are pulled from the feedlot pen and treated with antimicrobials when treatment criteria are met. The objective of this study was to compare 2 distinct sequential BRD treatment protocols each consisting of a metaphylactic antimicrobial plus 2 potential subsequent as-needed treatment antimicrobials. Heifers at high-risk for BRD (n = 1000; initial BW = 229 kg ± 1.6) purchased from sale barns in the southeastern U.S. were transported to a contract research feedlot in Nebraska and randomly assigned to 1 of 2 experimental groups (10 blocks of 100 animals each; 50 per treatment group). Experimental groups consisted of: (1) tulathromycin metaphylaxis (2.5 mg/kg) followed by ceftiofur crystalline free acid (6.6 mg/kg) and danofloxacin (8 mg/kg) for subsequent first and second as-needed BRD treatment, respectively (TCD) or (2) tildipirosin metaphylaxis (4 mg/kg) followed by florfenicol-flunixin meglumine (40 mg/kg florfenicol; 2.2 mg/kg flunixin meglumine) and enrofloxacin (12.5 mg/kg) for subsequent first and second as-needed BRD treatment, respectively (TFFE). Following expiration of the 7-d PMI, calves that showed signs of clinical BRD were pulled and examined to determine if treatment was necessary based on a clinical attitude score (CAS). Heifers with a CAS of 1 accompanied by ≥40°C rectal temperature, and all heifers with a CAS ≥ 2 regardless of rectal temperature, received the appropriate first-treatment antimicrobial. Upon relapse, following expiration of the post-treatment interval (PTI), heifers received the appropriate second-treatment antimicrobial. In the first 90 d, calves in the TFFE experimental group received more first-treatments than calves in the TCD experimental group (P = 0.054) and resulted in 50% greater mortality (P < 0.043) relative to the TCD heifers. From d 0 to closeout, first-treatment morbidity as well as mortality were greater in TFFE relative to TCD (P ≤ 0.032). Growth performance did not differ between treatments in the first 90 d; however, ADG was greater (P = 0.033) and G:F improved (P = 0.014) at closeout in TCD versus TFFE on a deads-in basis. Closeout economics revealed a $50.78/animal greater profit in the TCD experimental group relative to TFFE.

Analysis of adaptive platform trials using a network approach.

BACKGROUND: Adaptive platform trials allow randomized controlled comparisons of multiple treatments using a common infrastructure and the flexibility to adapt key design features during the study. Nonetheless, they have been criticized due to the potential for time trends in the underlying risk level of the population. Such time trends lead to confounding between design features and risk level, which may introduce bias favoring one or more treatments. This is particularly true when experimental treatments are not all randomized during the same time period as the control, leading to the potential for bias from non-concurrent controls. METHODS: Two analysis methods addressing this bias are stratification and adjustment. Stratification uses only comparisons between treatment cohorts randomized during identical time periods and does not use non-concurrent randomizations. Adjustment uses a modeled analysis including time period adjustment, allowing all data to be used, even from periods without concurrent randomization. We show that these competing approaches may be embedded in a common framework using network meta-analysis principles. We interpret the stages between adaptations in a platform trial as separate fixed design trials. This allows platform trials to be viewed as networks of direct randomized comparisons and indirect non-randomized comparisons. Network meta-analysis methodology can be re-purposed to aggregate the total information from a platform trial and to transparently decompose this total information into direct randomized evidence and indirect non-randomized evidence. This allows sensitivity to indirect information to be assessed and the two analysis methods to be clearly compared. RESULTS: Simulations of platform trials were analyzed using a network approach implemented in the netmeta package in R. The results demonstrated bias of unadjusted methods in the presence of time trends in risk level. Adjustment and stratification were both unbiased when direct evidence and indirect evidence were consistent. Network tests of inconsistency may be used to diagnose inconsistency when it exists. In an illustrative network analysis of one of the treatment comparisons from the STAMPEDE platform trial in metastatic prostate cancer, indirect comparisons using non-concurrent controls were inconsistent with the information from direct randomized comparisons. This supports the primary analysis approach of STAMPEDE, which used only direct randomized comparisons. CONCLUSION: Network meta-analysis provides a natural methodology for analyzing the network of direct and indirect treatment comparisons from a platform trial. Such analyses provide transparent separation of direct and indirect evidence, allowing assessment of the impact of non-concurrent controls. We recommend time-stratified analysis of concurrently controlled comparisons for primary analyses, with time-adjusted analyses incorporating non-concurrent controls reserved for secondary analyses. However, regardless of which methodology is used, a network analysis provides a useful supplement to the primary analysis.

Reacciones al estrés y aspectos psicoterapéuticos / Reactions to Stress and Psychotherapeutic Aspects

Resumen El periodismo científico de este artículo versa sobre una excepción habitual en el ambiente especializado, acerca de diversos trastornos mentales (TM) asociados al estrés y su comorbilidad, inadecuado enfrentamiento por el paciente/familia, su condición genético-epigenética y el deficiente manejo brindado de algunos profesionales de la salud. El desdén del manejo psicoterapéutico (prejuicio-impreparación) dificulta la atención adecuada de los casos. Se insta a desarrollar apego maternofilial apropiado, medidas para el logro de la salud mental social y programas asistenciales para atender los casos.

Abstract The scientific journalism of this article deals with a common exception in the specialized environment about various mental disorders associated with stress and its comorbidity, inadequate confrontation by the patient/family, its genetic-epigenetic condition and the deficient management provided by some health professionals. The disdain of psychotherapeutic management (prejudice- lack of preparation) hinders the adequate attention of cases. It is of the utmost importance to develop proper maternal and filial attachment, adequate measures for the achievement of social mental health and assistance programs to attend the cases.

Debaryomyces hansenii, Stenotrophomonas rhizophila, and Ulvan as Biocontrol Agents of Fruit Rot Disease in Muskmelon (Cucumis melo L.).

The indiscriminate use of synthetic fungicides has led to negative impact to human health and to the environment. Thus, we investigated the effects of postharvest biocontrol treatment with Debaryomyces hansenii, Stenotrophomonas rhizophila, and a polysaccharide ulvan on fruit rot disease, storability, and antioxidant enzyme activity in muskmelon (Cucumis melo L. var. reticulatus). Each fruit was treated with (1) 1 × 106 cells mL-1 of D. hansenii, (2) 1 × 108 CFU mL-1 of S. rhizophila, (3) 5 g L-1 of ulvan, (4) 1 × 106 cells mL-1 of D. hansenii + 1 × 108 CFU mL-1 of S. rhizophila, (5) 1 × 108 CFU mL-1 of S. rhizophila + 5 g L-1 of ulvan, (6) 1 × 106 cells mL-1 of D. hansenii + 1 × 108 CFU mL-1 of S. rhizophila + 5 g L-1 of ulvan, (7) 1000 ppm of benomyl or sterile water (control). The fruits were air-dried for 2 h, and stored at 27 °C ± 1 °C and 85-90% relative humidity. The fruit rot disease was determined by estimating the disease incidence (%) and lesion diameter (mm), and the adhesion capacity of the biocontrol agents was observed via electron microscopy. Phytopathogen inoculation time before and after adding biocontrol agents were also recorded. Furthermore, the storability quality, weight loss (%), firmness (N), total soluble solids (%), and pH were quantified. The antioxidant enzymes including catalase, peroxidase, superoxide dismutase, and phenylalanine ammonium lyase were determined. In conclusion, the mixed treatment containing D. hansenii, S. rhizophila, and ulvan delayed fruit rot disease, preserved fruit quality, and increased antioxidant activity. The combined treatment is a promising and effective biological control method to promote the shelf life of harvested muskmelon.

Network Meta-Analysis.

There are often multiple potential interventions to treat a disease; therefore, we need a method for simultaneously comparing and ranking all of these available interventions. In contrast to pairwise meta-analysis, which allows for the comparison of one intervention to another based on head-to-head data from randomized trials, network meta-analysis (NMA) facilitates simultaneous comparison of the efficacy or safety of multiple interventions that may not have been directly compared in a randomized trial. NMAs help researchers study important and previously unanswerable questions, which have contributed to a rapid rise in the number of NMA publications in the biomedical literature. However, the conduct and interpretation of NMAs are more complex than pairwise meta-analyses: there are additional NMA model assumptions (i.e., network connectivity, homogeneity, transitivity, and consistency) and outputs (e.g., network plots and surface under the cumulative ranking curves [SUCRAs]). In this chapter, we will: (1) explore similarities and differences between pairwise and network meta-analysis; (2) explain the differences between direct, indirect, and mixed treatment comparisons; (3) describe how treatment effects are derived from NMA models; (4) discuss key criteria predicating completion of NMA; (5) interpret NMA outputs; (6) discuss areas of ongoing methodological research in NMA; (7) outline an approach to conducting a systematic review and NMA; (8) describe common problems that researchers encounter when conducting NMAs and potential solutions; and (9) outline an approach to critically appraising a systematic review and NMA.

Contrast-Based and Arm-Based Models for Network Meta-Analysis.

Network meta-analysis is used to synthesize evidence from a network of treatments. The models used in a network meta-analysis are more complex than those used for pairwise meta-analysis. Two types of models are available to undertake a network meta-analysis: contrast-based and arm-based models. Contrast-based models have been used in most published network meta-analyses. Arm-based models offer greater flexibility and handle treatments symmetrically, but risk compromising randomization. In this chapter, we (1) present the contrast-based and arm-based statistical models; (2) describe the theoretical differences between the models (noting when the estimates from the models are expected to diverge); (3) summarize the evidence comparing the two models from simulation studies and empirical investigations; and (4) provide a worked example applying the two models to a network using the R software package.

Improving Glycemic Control in Type 2 Diabetes Using Mobile Applications and e-Coaching: A Mixed Treatment Comparison Network Meta-Analysis.

OBJECTIVES: This study compared the effectiveness of glycemic control among usual care, care management using a mobile application (app), and management using an app with additional e-coaching for patients with type 2 diabetes mellitus (T2DM) using a mixed treatment comparison (MTC) network meta-analysis (NMA). METHODS: A systematic search for published randomized controlled trials (RCTs) was conducted, which included Pubmed, Web of Science, Cochrane Central Register of Controlled Trials, CINAL, Koreamed, KMbase, and ScienceOn, until October 2020. Among the 10,391 studies identified after removing duplicates, 14 RCTs were finally included in the MTC NMA. Data extraction and methodological quality assessment using version 2 of the Cochrane tool for assessing the risk-of-bias in randomized trials (RoB 2) was performed. The comparative efficacy was analyzed using the random-effects NMA based on a frequentist model by the intervention group and main outcome variables. RESULTS: At the 3-month follow-up after each intervention, a comparison of the P-scores revealed the app plus e-coaching intervention to be the most effective method for reducing the HbA1c level in a homogeneous gender ratio group (P-score 0.92). At the 6-month follow-up period, app intervention was the best in reducing the HbA1c level in the homogeneous gender ratio and under 60 years of age group (P-score 1.00). CONCLUSIONS: Based on MTC analysis using the data from published RCTs, mobile apps or apps with e-coaching interventions for T2DM patients were more effective in improving the HbA1c values, FBS, and hypoglycemia frequency than usual care. Nevertheless, further research will be needed to clarify the effects of adding e-coaching to the app. STUDY REGISTRATION: Research Registry UIN (reviewregistry780).

Concepts and emerging issues of network meta-analysis.

Most diseases have more than two interventions or treatment methods, and the application of network meta-analysis (NMA) studies to compare and evaluate the superiority of each intervention or treatment method is increasing. Understanding the concepts and processes of systematic reviews and meta-analyses is essential to understanding NMA. As with systematic reviews and meta-analyses, NMA involves specifying the topic, searching for and selecting all related studies, and extracting data from the selected studies. To evaluate the effects of each treatment, NMA compares and analyzes three or more interventions or treatment methods using both direct and indirect evidence. There is a possibility of several biases when performing NMA. Therefore, key assumptions like similarity, transitivity, and consistency should be satisfied when performing NMA. Among these key assumptions, consistency can be evaluated and quantified by statistical tests. This review aims to introduce the concepts of NMA, analysis methods, and interpretation and presentation of the results of NMA. It also briefly introduces the emerging issues in NMA, including methods for evaluation of consistency.

An evidence-splitting approach to evaluation of direct-indirect evidence inconsistency in network meta-analysis.

Network meta-analysis (NMA) compares the efficacy and harm between several treatments by combining direct and indirect evidence. The validity of NMA requires that all available evidence form a coherent network. Failure to meet such requirement is known as inconsistency. The most popular approach to inconsistency detection is to compare the direct and indirect evidence for each treatment contrast. Although several models have been proposed to evaluate direct-indirect evidence inconsistency, there is no comprehensive study on the implications of how these models separate direct from indirect evidence. The main objective of this study is to show that evidence is not properly split into direct and indirect evidence in current inconsistency models, and to propose a novel approach to inconsistency evaluation based on the principle of independence between direct and indirect evidence. We further demonstrated that current models for direct-indirect evidence inconsistency can potentially lead to misleading conclusions in inconsistency detection and NMA quality appraisal, while our proposed evidence-splitting model satisfies the principle of independence when splitting the direct from indirect evidence in multi-arm trials. Moreover, we showed that all these direct-indirect evidence inconsistency models differ in how the weight of the inconsistency parameter is split between the treatments of interest, yet only the evidence-splitting model assigns satisfying weights. Finally, we demonstrated how the evidence-splitting model can be implemented within the structural equation modeling framework. The evidence-splitting model may be a valuable tool to assess the inconsistency within NMA and evaluate the quality of its evidence.

Efficacy and Safety of Anti-Vascular Endothelial Growth Factor Monotherapies for Neovascular Age-Related Macular Degeneration: A Mixed Treatment Comparison.

Background: We aimed to evaluate the comparative efficacy and safety of anti-vascular endothelial growth factor (anti-VEGF) monotherapy to identify its utilization and prioritization in patients with neovascular age-related macular degeneration (nAMD). Methods: Eligible studies included randomized controlled trials comparing the recommended anti-VEGF agents (ranibizumab, bevacizumab, aflibercept, brolucizumab, and conbercept) under various therapeutic regimens. Outcomes of interest included the mean change in best-corrected visual acuity (BCVA), serious adverse events, the proportion of patients who gained ≥15 letters or lost <15 letters in BCVA, the mean change in central retinal thickness, and the number of injections within 12 months. Results: Twenty-seven trials including 10,484 participants and eighteen treatments were identified in the network meta-analysis. The aflibercept 2 mg bimonthly, ranibizumab 0.5 mg T&E, and brolucizumab 6 mg q12w/q8w regimens had better visual efficacy. Brolucizumab had absolute superiority in anatomical outcomes and a relative advantage of safety, as well as good performance of aflibercept 2 mg T&E. The proactive regimens had slightly better efficacy but a slightly increased number of injections versus the reactive regimen. Bevacizumab had a statistically non-significant trend toward a lower degree of efficacy and safety. Conclusion: The visual efficacy of four individual anti-VEGF drugs is comparable. Several statistically significant differences were observed considering special anti-VEGF regimens, suggesting that brolucizumab 6 mg q12w/q8w, aflibercept 2 mg bimonthly or T&E, and ranibizumab 0.5 mg T&E are the ideal anti-VEGF regimens for nAMD patients. In the current landscape, based on the premise of equivalent efficacy and safety, the optimal choice of anti-VEGF monotherapies seems mandatory to obtain maximal benefit.

Bayesian splines versus fractional polynomials in network meta-analysis.

BACKGROUND: Network meta-analysis (NMA) provides a powerful tool for the simultaneous evaluation of multiple treatments by combining evidence from different studies, allowing for direct and indirect comparisons between treatments. In recent years, NMA is becoming increasingly popular in the medical literature and underlying statistical methodologies are evolving both in the frequentist and Bayesian framework. Traditional NMA models are often based on the comparison of two treatment arms per study. These individual studies may measure outcomes at multiple time points that are not necessarily homogeneous across studies. METHODS: In this article we present a Bayesian model based on B-splines for the simultaneous analysis of outcomes across time points, that allows for indirect comparison of treatments across different longitudinal studies. RESULTS: We illustrate the proposed approach in simulations as well as on real data examples available in the literature and compare it with a model based on P-splines and one based on fractional polynomials, showing that our approach is flexible and overcomes the limitations of the latter. CONCLUSIONS: The proposed approach is computationally efficient and able to accommodate a large class of temporal treatment effect patterns, allowing for direct and indirect comparisons of widely varying shapes of longitudinal profiles.

A Systematic Review and Mixed Treatment Comparison of Pharmaceutical Interventions for Multiple Sclerosis.

BACKGROUND: Since 2010, 27 mixed-treatment comparisons (MTCs) of disease-modifying therapies (DMTs) for multiple sclerosis have been published. However, there has been continued evolution in the field of MTCs. Additionally, limitations in methodological approach and reporting transparency, even in the most recent publications, makes interpretation and comparison of existing studies difficult. OBJECTIVES: The objectives of this study are twofold: (1) to estimate the efficacy and safety of DMTs at European Commission-approved doses compared with placebo in adults with relapsing-remitting multiple sclerosis (RRMS) using MTC, and (2) to identify and address methodological challenges when performing MTC in RRMS, thereby creating a baseline for comparisons with future treatments. METHODS: Searches were completed in 14 databases, including MEDLINE, Embase, CENTRAL, CDSR and DARE, from inception to June 2018 to identify published or unpublished prospective, randomised controlled trials of all European Union-approved DMTs or DMTs expected to be approved in the near future in RRMS or rapidly-evolving severe RRMS. No language or date restrictions were applied. Studies were included in the MTC if they were judged to have sufficiently similar characteristics, based on the following: patient age; proportion of male participants; Expanded Disability Status Scale (EDSS) score; duration of disease; number of relapses prior to enrolment and proportion of previously treated patients. Background information from the included studies, as well as effect size and confidence intervals (where relevant) of defined outcomes were extracted. Reporting of the MTC was consistent with the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) network meta-analysis guidelines. RESULTS: In total, 33 studies were included in the MTC. Annualised relapse rate (ARR 28 trials) was significantly reduced in all treatments compared with placebo. Alemtuzumab had the highest probability (63%) of being the most effective treatment in terms of ARR compared with placebo (rate ratio [RR] 0.28, 95% credible interval [CrI] 0.21-0.38), followed by natalizumab (30% probability; RR 0.32, 95% CrI 0.23-0.43). The risk of 3- and 6-month confirmed disability progression (CDP3M, 13 trials; CDP6M, 14 trials) were similar; CDP6M was significantly reduced for alemtuzumab (hazard ratio [HR] 0.365; 95% CrI 0.165-0.725), ocrelizumab (HR 0.405, 95% CrI 0.188-0.853) and natalizumab (HR 0.459, 95% CrI 0.252-0.840) relative to placebo. There were no significant differences in the odds of serious adverse events (SAEs, 6 trials) between any treatment and placebo. The results of the MTC were limited by the lack of studies reporting direct comparisons between the included treatments and by heterogeneous reporting of key outcome data. CONCLUSIONS: Meta-analyses confirmed the benefit of all DMTs in terms of relapse rate compared with placebo with a comparable rate of SAEs for the DMTs that could be included in the network. The rigor and transparency of reporting in this study provide a benchmark for comparisons with future new agents.

Overview of Differences and Similarities of Published Mixed Treatment Comparisons on Pharmaceutical Interventions for Multiple Sclerosis.

INTRODUCTION: Mixed treatment comparisons (MTCs) are increasingly important in the assessment of the benefit-risk profile of pharmaceutical treatments for relapsing-remitting multiple sclerosis (RRMS). Interpretation of MTCs requires a clear understanding of the methods of analysis and population studied. The objectives of this work were to compare MTCs of pharmaceutical treatments for RRMS, including a detailed description of differences in populations, treatments assessed, methods used and findings; and to discuss key considerations when conducting an MTC. METHODS: Fourteen databases were searched until July 2019 to identify MTCs (published during or after 2010) in adults (at least 18 years of age) with RRMS or rapidly evolving severe RRMS treated with any form of pharmaceutical treatment. No language restriction was imposed. RESULTS: Twenty-seven MTCs assessing 21 treatments were identified. Comparison highlighted many differences in conduct and reporting between MTCs relating to the patient populations or treatments included, duration of follow-up and outcomes of interest measured. The lack of similarity between the MTCs leads to questions about variability in the robustness of analyses and makes comparisons between studies challenging. CONCLUSION: Given the importance of MTCs for healthcare decision-making, it is imperative that reporting of methods, results and assumptions is clear and transparent to allow accurate interpretation of findings. For MTCs to be relevant, the choice of outcome measures should reflect clinical practice. Combination of treatments or of outcomes measured at different points of time should be avoided, as should imputation without justification. Furthermore, all approved treatment options should be included and updates of MTCs should be conducted when data for new treatments are published.

Multivariate network meta-analysis incorporating class effects.

BACKGROUND: Network meta-analysis synthesises data from a number of clinical trials in order to assess the comparative efficacy of multiple healthcare interventions in similar patient populations. In situations where clinical trial data are heterogeneously reported i.e. data are missing for one or more outcomes of interest, synthesising such data can lead to disconnected networks of evidence, increased uncertainty, and potentially biased estimates which can have severe implications for decision-making. To overcome this issue, strength can be borrowed between outcomes of interest in multivariate network meta-analyses. Furthermore, in situations where there are relatively few trials informing each treatment comparison, there is a potential issue with the sparsity of data in the treatment networks, which can lead to substantial parameter uncertainty. A multivariate network meta-analysis approach can be further extended to borrow strength between interventions of the same class using hierarchical models. METHODS: We extend the trivariate network meta-analysis model to incorporate the exchangeability between treatment effects belonging to the same class of intervention to increase precision in treatment effect estimates. We further incorporate a missing data framework to estimate uncertainty in trials that did not report measures of variability in order to maximise the use of all available information for healthcare decision-making. The methods are applied to a motivating dataset in overactive bladder syndrome. The outcomes of interest were mean change from baseline in incontinence, voiding and urgency episodes. All models were fitted using Bayesian Markov Chain Monte Carlo (MCMC) methods in WinBUGS. RESULTS: All models (univariate, multivariate, and multivariate models incorporating class effects) produced similar point estimates for all treatment effects. Incorporating class effects in multivariate models often increased precision in treatment effect estimates. CONCLUSIONS: Multivariate network meta-analysis incorporating class effects allowed for the comparison of all interventions across all outcome measures to ameliorate the potential impact of outcome reporting bias, and further borrowed strength between interventions belonging to the same class of treatment to increase the precision in treatment effect estimates for healthcare policy and decision-making.

A simulation study to compare different estimation approaches for network meta-analysis and corresponding methods to evaluate the consistency assumption.

BACKGROUND: Network meta-analysis (NMA) is becoming increasingly popular in systematic reviews and health technology assessments. However, there is still ambiguity concerning the properties of the estimation approaches as well as for the methods to evaluate the consistency assumption. METHODS: We conducted a simulation study for networks with up to 5 interventions. We investigated the properties of different methods and give recommendations for practical application. We evaluated the performance of 3 different models for complex networks as well as corresponding global methods to evaluate the consistency assumption. The models are the frequentist graph-theoretical approach netmeta, the Bayesian mixed treatment comparisons (MTC) consistency model, and the MTC consistency model with stepwise removal of studies contributing to inconsistency identified in a leverage plot. RESULTS: We found that with a high degree of inconsistency none of the evaluated effect estimators produced reliable results, whereas with moderate or no inconsistency the estimator from the MTC consistency model and the netmeta estimator showed acceptable properties. We also saw a dependency on the amount of heterogeneity. Concerning the evaluated methods to evaluate the consistency assumption, none was shown to be suitable. CONCLUSIONS: Based on our results we recommend a pragmatic approach for practical application in NMA. The estimator from the netmeta approach or the estimator from the Bayesian MTC consistency model should be preferred. Since none of the methods to evaluate the consistency assumption showed satisfactory results, users should have a strong focus on the similarity as well as the homogeneity assumption.

The dark side of the force: Multiplicity issues in network meta-analysis and how to address them.

Standard models for network meta-analysis simultaneously estimate multiple relative treatment effects. In practice, after estimation, these multiple estimates usually pass through a formal or informal selection procedure, eg, when researchers draw conclusions about the effects of the best performing treatment in the network. In this paper, we present theoretical arguments as well as results from simulations to illustrate how such practices might lead to exaggerated and overconfident statements regarding relative treatment effects. We discuss how the issue can be addressed via multilevel Bayesian modelling, where treatment effects are modelled exchangeably, and hence estimates are shrunk away from large values. We present a set of alternative models for network meta-analysis, and we show in simulations that in several scenarios, such models perform better than the usual network meta-analysis model.

A comparison of arm-based and contrast-based models for network meta-analysis.

Differences between arm-based (AB) and contrast-based (CB) models for network meta-analysis (NMA) are controversial. We compare the CB model of Lu and Ades (2006), the AB model of Hong et al(2016), and two intermediate models, using hypothetical data and a selected real data set. Differences between models arise primarily from study intercepts being fixed effects in the Lu-Ades model but random effects in the Hong model, and we identify four key difference. (1) If study intercepts are fixed effects then only within-study information is used, but if they are random effects then between-study information is also used and can cause important bias. (2) Models with random study intercepts are suitable for deriving a wider range of estimands, eg, the marginal risk difference, when underlying risk is derived from the NMA data; but underlying risk is usually best derived from external data, and then models with fixed intercepts are equally good. (3) The Hong model allows treatment effects to be related to study intercepts, but the Lu-Ades model does not. (4) The Hong model is valid under a more relaxed missing data assumption, that arms (rather than contrasts) are missing at random, but this does not appear to reduce bias. We also describe an AB model with fixed study intercepts and a CB model with random study intercepts. We conclude that both AB and CB models are suitable for the analysis of NMA data, but using random study intercepts requires a strong rationale such as relating treatment effects to study intercepts.