New horizons in evidence synthesis for older adults

Evidence synthesis, embedded within a systematic review of the literature, is a well-established approach for collating and combining all the relevant information on a particular research question. A robust synthesis can establish the evidence base, which underpins best practice guidance. Such endeavours are frequently used by policymakers and practitioners to inform their decision making. Traditionally, an evidence synthesis of interventions consisted of a meta-analysis of quantitative data comparing two treatment alternatives addressing a specific and focussed clinical question. However, as the methods in the field have evolved, especially in response to the increasingly complex healthcare questions, more advanced evidence synthesis techniques have been developed. These can deal with extended data structures considering more than two treatment alternatives (network meta-analysis) and complex multicomponent interventions. The array of questions capable of being answered has also increased with specific approaches being developed for different evidence types including diagnostic, prognostic and qualitative data. Furthermore, driven by a desire for increasingly up-to-date evidence summaries, living systematic reviews have emerged. All of these methods can potentially have a role in informing older adult healthcare decisions. The aim of this review is to increase awareness and uptake of the increasingly comprehensive array of newer synthesis methods available and highlight their utility for answering clinically relevant questions in the context of older adult research, giving examples of where such techniques have already been effectively applied within the field. Their strengths and limitations are discussed, and we suggest user-friendly software options to implement the methods described.

Data visualisation approaches for component network meta-analysis: visualising the data structure

BACKGROUND

Health and social care interventions are often complex and can be decomposed into multiple components. Multicomponent interventions are often evaluated in randomised controlled trials. Across trials, interventions often have components in common which are given alongside other components which differ across trials. Multicomponent interventions can be synthesised using component NMA (CNMA). CNMA is limited by the structure of the available evidence, but it is not always straightforward to visualise such complex evidence networks. The aim of this paper is to develop tools to visualise the structure of complex evidence networks to support CNMA.

METHODS

We performed a citation review of two key CNMA methods papers to identify existing published CNMA analyses and reviewed how they graphically represent intervention complexity and comparisons across trials. Building on identified shortcomings of existing visualisation approaches, we propose three approaches to standardise visualising the data structure and/or availability of data: CNMA-UpSet plot, CNMA heat map, CNMA-circle plot. We use a motivating example to illustrate these plots.

RESULTS

We identified 34 articles reporting CNMAs. A network diagram was the most common plot type used to visualise the data structure for CNMA (26/34 papers), but was unable to express the complex data structures and large number of components and potential combinations of components associated with CNMA. Therefore, we focused visualisation development around representing the data structure of a CNMA more completely. The CNMA-UpSet plot presents arm-level data and is suitable for networks with large numbers of components or combinations of components. Heat maps can be utilised to inform decisions about which pairwise interactions to consider for inclusion in a CNMA model. The CNMA-circle plot visualises the combinations of components which differ between trial arms and offers flexibility in presenting additional information such as the number of patients experiencing the outcome of interest in each arm.

CONCLUSIONS

As CNMA becomes more widely used for the evaluation of multicomponent interventions, the novel CNMA-specific visualisations presented in this paper, which improve on the limitations of existing visualisations, will be important to aid understanding of the complex data structure and facilitate interpretation of the CNMA results.

Pharmacological and electronic cigarette interventions for smoking cessation in adults: component network meta-analyses

BACKGROUND

Tobacco smoking is the leading preventable cause of death and disease worldwide. Stopping smoking can reduce this harm and many people would like to stop. There are a number of medicines licenced to help people quit globally, and e-cigarettes are used for this purpose in many countries. Typically treatments work by reducing cravings to smoke, thus aiding initial abstinence and preventing relapse. More information on comparative effects of these treatments is needed to inform treatment decisions and policies.

OBJECTIVES

To investigate the comparative benefits, harms and tolerability of different smoking cessation pharmacotherapies and e-cigarettes, when used to help people stop smoking tobacco.

SEARCH METHODS

We identified studies from recent updates of Cochrane Reviews investigating our interventions of interest. We updated the searches for each review using the Cochrane Tobacco Addiction Group (TAG) specialised register to 29 April 2022.

SELECTION CRITERIA

We included randomised controlled trials (RCTs), cluster-RCTs and factorial RCTs, which measured smoking cessation at six months or longer, recruited adults who smoked combustible cigarettes at enrolment (excluding pregnant people) and randomised them to approved pharmacotherapies and technologies used for smoking cessation worldwide (varenicline, cytisine, nortriptyline, bupropion, nicotine replacement therapy (NRT) and e-cigarettes) versus no pharmacological intervention, placebo (control) or another approved pharmacotherapy. Studies providing co-interventions (e.g. behavioural support) were eligible if the co-intervention was provided equally to study arms.

DATA COLLECTION AND ANALYSIS

We followed standard Cochrane methods for screening, data extraction and risk of bias (RoB) assessment (using the RoB 1 tool). Primary outcome measures were smoking cessation at six months or longer, and the number of people reporting serious adverse events (SAEs). We also measured withdrawals due to treatment. We used Bayesian component network meta-analyses (cNMA) to examine intervention type, delivery mode, dose, duration, timing in relation to quit day and tapering of nicotine dose, using odds ratios (OR) and 95% credibility intervals (CrIs). We calculated an effect estimate for combination NRT using an additive model. We evaluated the influence of population and study characteristics, provision of behavioural support and control arm rates using meta-regression. We evaluated certainty using GRADE.

MAIN RESULTS

Of our 332 eligible RCTs, 319 (835 study arms, 157,179 participants) provided sufficient data to be included in our cNMA. Of these, we judged 51 to be at low risk of bias overall, 104 at high risk and 164 at unclear risk, and 118 reported pharmaceutical or e-cigarette/tobacco industry funding. Removing studies at high risk of bias did not change our interpretation of the results. Benefits We found high-certainty evidence that nicotine e-cigarettes (OR 2.37, 95% CrI 1.73 to 3.24; 16 RCTs, 3828 participants), varenicline (OR 2.33, 95% CrI 2.02 to 2.68; 67 RCTs, 16,430 participants) and cytisine (OR 2.21, 95% CrI 1.66 to 2.97; 7 RCTs, 3848 participants) were associated with higher quit rates than control. In absolute terms, this might lead to an additional eight (95% CrI 4 to 13), eight (95% CrI 6 to 10) and seven additional quitters per 100 (95% CrI 4 to 12), respectively. These interventions appeared to be more effective than the other interventions apart from combination NRT (patch and a fast-acting form of NRT), which had a lower point estimate (calculated additive effect) but overlapping 95% CrIs (OR 1.93, 95% CrI 1.61 to 2.34). There was also high-certainty evidence that nicotine patch alone (OR 1.37, 95% CrI 1.20 to 1.56; 105 RCTs, 37,319 participants), fast-acting NRT alone (OR 1.41, 95% CrI 1.29 to 1.55; 120 RCTs, 31,756 participants) and bupropion (OR 1.43, 95% CrI 1.26 to 1.62; 71 RCTs, 14,759 participants) were more effective than control, resulting in two (95% CrI 1 to 3), three (95% CrI 2 to 3) and three (95% CrI 2 to 4) additional quitters per 100 respectively. Nortriptyline is probably associated with higher quit rates than control (OR 1.35, 95% CrI 1.02 to 1.81; 10 RCTs, 1290 participants; moderate-certainty evidence), resulting in two (CrI 0 to 5) additional quitters per 100. Non-nicotine/placebo e-cigarettes (OR 1.16, 95% CrI 0.74 to 1.80; 8 RCTs, 1094 participants; low-certainty evidence), equating to one additional quitter (95% CrI -2 to 5), had point estimates favouring the intervention over control, but CrIs encompassed the potential for no difference and harm. There was low-certainty evidence that tapering the dose of NRT prior to stopping treatment may improve effectiveness; however, 95% CrIs also incorporated the null (OR 1.14, 95% CrI 1.00 to 1.29; 111 RCTs, 33,156 participants). This might lead to an additional one quitter per 100 (95% CrI 0 to 2). Harms There were insufficient data to include nortriptyline and non-nicotine EC in the final SAE model. Overall rates of SAEs for the remaining treatments were low (average 3%). Low-certainty evidence did not show a clear difference in the number of people reporting SAEs for nicotine e-cigarettes, varenicline, cytisine or NRT when compared to no pharmacotherapy/e-cigarettes or placebo. Bupropion may slightly increase rates of SAEs, although the CrI also incorporated no difference (moderate certainty). In absolute terms bupropion may cause one more person in 100 to experience an SAE (95% CrI 0 to 2).

AUTHORS' CONCLUSIONS

The most effective interventions were nicotine e-cigarettes, varenicline and cytisine (all high certainty), as well as combination NRT (additive effect, certainty not rated). There was also high-certainty evidence for the effectiveness of nicotine patch, fast-acting NRT and bupropion. Less certain evidence of benefit was present for nortriptyline (moderate certainty), non-nicotine e-cigarettes and tapering of nicotine dose (both low certainty). There was moderate-certainty evidence that bupropion may slightly increase the frequency of SAEs, although there was also the possibility of no increased risk. There was no clear evidence that any other tested interventions increased SAEs. Overall, SAE data were sparse with very low numbers of SAEs, and so further evidence may change our interpretation and certainty. Future studies should report SAEs to strengthen certainty in this outcome. More head-to-head comparisons of the most effective interventions are needed, as are tests of combinations of these. Future work should unify data from behavioural and pharmacological interventions to inform approaches to combined support for smoking cessation.

MetaBayesDTA: codeless Bayesian meta-analysis of test accuracy, with or without a gold standard

BACKGROUND

The statistical models developed for meta-analysis of diagnostic test accuracy studies require specialised knowledge to implement. This is especially true since recent guidelines, such as those in Version 2 of the Cochrane Handbook of Systematic Reviews of Diagnostic Test Accuracy, advocate more sophisticated methods than previously. This paper describes a web-based application - MetaBayesDTA - that makes many advanced analysis methods in this area more accessible.

RESULTS

We created the app using R, the Shiny package and Stan. It allows for a broad array of analyses based on the bivariate model including extensions for subgroup analysis, meta-regression and comparative test accuracy evaluation. It also conducts analyses not assuming a perfect reference standard, including allowing for the use of different reference tests.

CONCLUSIONS

Due to its user-friendliness and broad array of features, MetaBayesDTA should appeal to researchers with varying levels of expertise. We anticipate that the application will encourage higher levels of uptake of more advanced methods, which ultimately should improve the quality of test accuracy reviews.

Non-opioid analgesics for the prevention of chronic postsurgical pain: a systematic review and network meta-analysis

BACKGROUND

Chronic postsurgical pain is common after surgery. Identification of non-opioid analgesics with potential for preventing chronic postsurgical pain is important, although trials are often underpowered. Network meta-analysis offers an opportunity to improve power and to identify the most promising therapy for clinical use and future studies.

METHODS

We conducted a PRISMA-NMA-compliant systematic review and network meta-analysis of randomised controlled trials of non-opioid analgesics for chronic postsurgical pain. Outcomes included incidence and severity of chronic postsurgical pain, serious adverse events, and chronic opioid use.

RESULTS

We included 132 randomised controlled trials with 23 902 participants. In order of efficacy, i.v. lidocaine (odds ratio [OR] 0.32; 95% credible interval [CrI] 0.17-0.58), ketamine (OR 0.64; 95% CrI 0.44-0.92), gabapentinoids (OR 0.67; 95% CrI 0.47-0.92), and possibly dexmedetomidine (OR 0.36; 95% CrI 0.12-1.00) reduced the incidence of chronic postsurgical pain at ≤6 months. There was little available evidence for chronic postsurgical pain at >6 months, combinations agents, chronic opioid use, and serious adverse events. Variable baseline risk was identified as a potential violation to the network meta-analysis transitivity assumption, so results are reported from a fixed value of this, with analgesics more effective at higher baseline risk. The confidence in these findings was low because of problems with risk of bias and imprecision.

CONCLUSIONS

Lidocaine (most effective), ketamine, and gabapentinoids could be effective in reducing chronic postsurgical pain ≤6 months although confidence is low. Moreover, variable baseline risk might violate transitivity in network meta-analysis of analgesics; this recommends use of our methods in future network meta-analyses.

SYSTEMATIC REVIEW PROTOCOL

PROSPERO CRD42021269642.

Origami plot: a novel multivariate data visualization tool that improves radar chart

OBJECTIVES

We propose the origami plot, which maintains the original functionality of a radar chart and avoids potential misuse of its connected regions, with newly added features to better assist multicriteria decision-making.

STUDY DESIGN AND SETTING

Built upon a radar chart, the origami plot adds additional auxiliary axes and points such that the area of the connected region of all dots is invariant to the ordering of axes. As such, it enables ranking different individuals by the overall performance for multicriteria decision-making while maintaining the intuitive visual appeal of the radar chart. We develop extensions of the origami plot, including the weighted origami plot, which allows reweighting of each attribute to define the overall performance, and the pairwise origami plot, which highlights comparisons between two individuals.

RESULTS

We illustrate the different versions of origami plots using the hospital compare database developed by the Centers for Medicare & Medicaid Services (CMS). The plot shows individual hospital's performance on mortality, readmission, complication, and infection, as well as patient experience and timely and effective care, as well as their overall performance across these metrics. The weighted origami plot allows weighing the attributes differently when some are more important than others. We illustrate the potential use of the pairwise origami plot in electronic health records (EHR) system to monitor five clinical measures (body mass index [BMI]), fasting glucose level, blood pressure, triglycerides, and low-density lipoprotein ([LDL] cholesterol) of a patient across multiple hospital visits.

CONCLUSION

The origami plot is a useful visualization tool to assist multicriteria decision making. It improves radar charts by avoiding potential misuse of the connected regions. It has several new features and allows flexible customization.

A multifaceted graphical display, including treatment ranking, was developed to aid interpretation of network meta-analysis

OBJECTIVE

Network meta-analysis (NMA) is becoming a popular statistical tool for analysing a network of evidence comparing more than two interventions. A particular advantage of NMA over pairwise meta-analysis is its ability to simultaneously compare multiple interventions including comparisons not previously trialled together, permitting intervention hierarchies to be created. Our aim was to develop a novel graphical display to aid interpretation of NMA to clinicians and decision-makers that incorporates ranking of interventions.

STUDY DESIGN AND SETTING

Current literature was searched, scrutinised, and provided direction for developing the novel graphical display. Ranking results were often found to be misinterpreted when presented alone and, to aid interpretation and effective communication to inform optimal decision-making, need to be displayed alongside other important aspects of the analysis including the evidence networks and relative intervention effect estimates.

RESULTS

Two new ranking visualisations were developed - the Litmus Rank-O-Gram, and the Radial SUCRA plot - and embedded within a novel multi-panel graphical display programmed within the MetaInsight app, with user feedback gained.

CONCLUSION

This display was designed to improve the reporting, and facilitate a holistic understanding, of NMA results. We believe uptake of the display would lead to better understanding of complex results and improve future decision-making.

Double-counting of populations in evidence synthesis in public health: a call for awareness and future methodological development

Background

There is a growing interest in the inclusion of real-world and observational studies in evidence synthesis such as meta-analysis and network meta-analysis in public health. While this approach offers great epidemiological opportunities, use of such studies often introduce a significant issue of double-counting of participants and databases in a single analysis. Therefore, this study aims to introduce and illustrate the nuances of double-counting of individuals in evidence synthesis including real-world and observational data with a focus on public health.

Methods

The issues associated with double-counting of individuals in evidence synthesis are highlighted with a number of case studies. Further, double-counting of information in varying scenarios is discussed with potential solutions highlighted.

Results

Use of studies of real-world data and/or established cohort studies, for example studies evaluating the effectiveness of therapies using health record data, often introduce a significant issue of double-counting of individuals and databases. This refers to the inclusion of the same individuals multiple times in a single analysis. Double-counting can occur in a number of manners, such as, when multiple studies utilise the same database, when there is overlapping timeframes of analysis or common treatment arms across studies. Some common practices to address this include synthesis of data only from peer-reviewed studies, utilising the study that provides the greatest information (e.g. largest, newest, greater outcomes reported) or analysing outcomes at different time points.

Conclusions

While common practices currently used can mitigate some of the impact of double-counting of participants in evidence synthesis including real-world and observational studies, there is a clear need for methodological and guideline development to address this increasingly significant issue.

Meta-analysis of dichotomous and ordinal tests with an imperfect gold standard

Standard methods for the meta-analysis of medical tests, without assuming a gold standard, are limited to dichotomous data. Multivariate probit models are used to analyse correlated dichotomous data, and can be extended to model ordinal data. Within the context of an imperfect gold standard, they have previously been used for the analysis of dichotomous and ordinal test data from a single study, and for the meta-analysis of dichotomous tests. However, they have not previously been used for the meta-analysis of ordinal tests. In this article, we developed a Bayesian multivariate probit latent class model for the simultaneous meta-analysis of ordinal and dichotomous tests without assuming a gold standard, which also allows one to obtain summary estimates of joint test accuracy. We fitted the models using the software Stan, which uses a state-of-the-art Hamiltonian Monte Carlo algorithm, and we applied the models to a dataset in which studies evaluated the accuracy of tests, and test combinations, for deep vein thrombosis. We demonstrate the issues with dichotomising ordinal test accuracy data in the presence of an imperfect gold standard, before applying and comparing several variations of our proposed model which do not require the data to be dichotomised. The models proposed will allow researchers to more appropriately meta-analyse ordinal and dichotomous tests without a gold standard, potentially leading to less biased estimates of test accuracy. This may lead to a better understanding of which tests, and test combinations, should be used for any given medical condition.

Impact of program characteristics on weight loss in adult behavioral weight management interventions: systematic review and component network meta-analysis

OBJECTIVE

Behavioral weight management programs (BWMPs) for adults lead to greater weight loss at 12 months than minimal-intervention control treatments. However, there is considerable heterogeneity in the content of BWMPs and outcomes of treatment. This study assessed the contribution of individual components of BWMPs, using Bayesian component network meta-analysis.

METHODS

Randomized controlled trials of BWMPs in adults were identified (latest search: December 2019) and arms coded for presence or absence of 29 intervention components grouped by type, content, provider, mode of delivery, and intensity.

RESULTS

A total of 169 studies (41 judged at high risk of bias) were included in the main analysis. Six components had effect estimates indicating clinically significant benefit and credible intervals (CrIs) excluding no difference: change in diet (mean difference [MD] = -1.84 kg, 95% CrI: -2.91 to -0.80); offering partial (MD = -2.12 kg, 95% CrI: -3.39 to -0.89) or total meal replacements (MD = -2.63 kg, 95% CrI: -4.58 to -0.73); delivery by a psychologist/counselor (MD = -1.45 kg, 95% CrI: -2.81 to -0.06) or dietitian (MD = -1.31 kg, 95% CrI: -2.40 to -0.24); and home setting (MD = -1.05 kg, 95% CrI: -2.02 to -0.09).

CONCLUSIONS

Future program development should consider including these components; other approaches continue to warrant evaluation of effectiveness.

Behavioural programmes for cigarette smoking cessation: investigating interactions between behavioural, motivational and delivery components in a systematic review and component network meta-analysis

AIMS

To investigate the comparative and combined effectiveness of four types of components of behavioural interventions for cigarette smoking cessation: behavioural (e.g. counselling), motivational (e.g. focus on reasons to quit), delivery mode (e.g. phone) and provider (e.g. nurse).

DESIGN

Systematic review and component network meta-analysis of randomised controlled trials identified from Cochrane reviews. Interventions included behavioural interventions for smoking cessation (including all non-pharmacological interventions, e.g. counselling, exercise, hypnotherapy, self-help materials), compared with another behavioural intervention or no support. Building on a 2021 review (CD013229), we conducted three analyses, investigating: comparative effectiveness of the components, whether models that allowed interactions between components gave different results to models assuming additivity, and predicted effect estimates for combined effects of components that had showed promise but where there were few trials.

SETTING

Community and health-care settings.

PARTICIPANTS

Adults who smoke tobacco.

MEASUREMENTS

Smoking cessation at ≥6 months, preferring sustained, biochemically validated outcomes where available.

FINDINGS

Three hundred and twelve trials (250 563 participants) were included. Fifty were at high risk of bias using Cochrane risk of bias tool, V1 (ROB1); excluding these studies did not change findings. Head-to-head comparisons of components suggested that support via text message (SMS) compared with telephone (OR 1.48, 95% CrI 1.13-1.94) or print materials (OR 1.44, 95% CrI 1.14-1.83) was more effective, and individual delivery was less effective than delivery as part of a group (OR 0.78, 95% CrI 0.64-0.95). There was no conclusive evidence of synergistic or antagonistic interactions when combining components that were commonly used together. Adding multiple components that are commonly used in behavioural counselling suggested clinically relevant and statistically conclusive evidence of benefit. Components with the largest effects that could be combined, but rarely have been, were estimated to increase the odds of quitting between two and threefold. For example, financial incentives delivered via SMS, with tailoring and a focus on how to quit, had an estimated OR of 2.94 (95% CrI 1.91-4.52).

CONCLUSIONS

Among the components of behavioural support for smoking cessation, behavioural counselling and guaranteed financial incentives are associated with the greatest success. Incorporating additional components associated with effectiveness may further increase benefit, with delivery via text message showing particular promise.

Anticoagulants for thrombosis prophylaxis in acutely ill patients admitted to hospital: systematic review and network meta-analysis

OBJECTIVE

To assess the benefits and harms of different types and doses of anticoagulant drugs for the prevention of venous thromboembolism in patients who are acutely ill and admitted to hospital.

DESIGN

Systematic review and network meta-analysis.

DATA SOURCES

Cochrane CENTRAL, PubMed/Medline, Embase, Web of Science, clinical trial registries, and national health authority databases. The search was last updated on 16 November 2021.

ELIGIBILITY CRITERIA FOR SELECTING STUDIES

Published and unpublished randomised controlled trials that evaluated low or intermediate dose low-molecular-weight heparin, low or intermediate dose unfractionated heparin, direct oral anticoagulants, pentasaccharides, placebo, or no intervention for the prevention of venous thromboembolism in acutely ill adult patients in hospital.

MAIN OUTCOME MEASURES

Random effects, bayesian network meta-analyses used four co-primary outcomes: all cause mortality, symptomatic venous thromboembolism, major bleeding, and serious adverse events at or closest timing to 90 days. Risk of bias was also assessed using the Cochrane risk-of-bias 2.0 tool. The quality of evidence was graded using the Confidence in Network Meta-Analysis framework.

RESULTS

44 randomised controlled trials that randomly assigned 90 095 participants were included in the main analysis. Evidence of low to moderate quality suggested none of the interventions reduced all cause mortality compared with placebo. Pentasaccharides (odds ratio 0.32, 95% credible interval 0.08 to 1.07), intermediate dose low-molecular-weight heparin (0.66, 0.46 to 0.93), direct oral anticoagulants (0.68, 0.33 to 1.34), and intermediate dose unfractionated heparin (0.71, 0.43 to 1.19) were most likely to reduce symptomatic venous thromboembolism (very low to low quality evidence). Intermediate dose unfractionated heparin (2.63, 1.00 to 6.21) and direct oral anticoagulants (2.31, 0.82 to 6.47) were most likely to increase major bleeding (low to moderate quality evidence). No conclusive differences were noted between interventions regarding serious adverse events (very low to low quality evidence). When compared with no intervention instead of placebo, all active interventions did more favourably with regard to risk of venous thromboembolism and mortality, and less favourably with regard to risk of major bleeding. The results were robust in prespecified sensitivity and subgroup analyses.

CONCLUSIONS

Low-molecular-weight heparin in an intermediate dose appears to confer the best balance of benefits and harms for prevention of venous thromboembolism. Unfractionated heparin, in particular the intermediate dose, and direct oral anticoagulants had the least favourable profile. A systematic discrepancy was noted in intervention effects that depended on whether placebo or no intervention was the reference treatment. Main limitations of this study include the quality of the evidence, which was generally low to moderate due to imprecision and within-study bias, and statistical inconsistency, which was addressed post hoc.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO CRD42020173088.

Challenges of modelling approaches for network meta-analysis of time-to-event outcomes in the presence of non-proportional hazards to aid decision making: Application to a melanoma network

BACKGROUND

Synthesis of clinical effectiveness from multiple trials is a well-established component of decision-making. Time-to-event outcomes are often synthesised using the Cox proportional hazards model assuming a constant hazard ratio over time. However, with an increasing proportion of trials reporting treatment effects where hazard ratios vary over time and with differing lengths of follow-up across trials, alternative synthesis methods are needed.

OBJECTIVES

To compare and contrast five modelling approaches for synthesis of time-to-event outcomes and provide guidance on key considerations for choosing between the modelling approaches.

METHODS

The Cox proportional hazards model and five other methods of estimating treatment effects from time-to-event outcomes, which relax the proportional hazards assumption, were applied to a network of melanoma trials reporting overall survival: restricted mean survival time, generalised gamma, piecewise exponential, fractional polynomial and Royston-Parmar models.

RESULTS

All models fitted the melanoma network acceptably well. However, there were important differences in extrapolations of the survival curve and interpretability of the modelling constraints demonstrating the potential for different conclusions from different modelling approaches.

CONCLUSION

The restricted mean survival time, generalised gamma, piecewise exponential, fractional polynomial and Royston-Parmar models can accommodate non-proportional hazards and differing lengths of trial follow-up within a network meta-analysis of time-to-event outcomes. We recommend that model choice is informed using available and relevant prior knowledge, model transparency, graphically comparing survival curves alongside observed data to aid consideration of the reliability of the survival estimates, and consideration of how the treatment effect estimates can be incorporated within a decision model.

Feasibility study for interactive reporting of network meta-analysis: experiences from the development of the MetaInsight COVID-19 app for stakeholder exploration, re-analysis and sensitivity analysis from living systematic reviews

BACKGROUND

Network meta-analysis (NMA) has been increasingly adopted worldwide by Cochrane reviews, guideline developers and decision-making bodies to identify optimal treatment choices. However, NMA results are often produced statically, not allowing stakeholders to 'dig deeper' and interrogate with their own judgement. Additionally, amid the COVID-19 pandemic, unnecessary or duplicated reviews have been proposed which analyse from the same pool of evidence. We developed the 'MetaInsight COVID-19' app as a prototype for an interactive platform to eliminate such duplicated efforts, by empowering users to freely analyse the data and improve scientific transparency.

METHODS

MetaInsight COVID-19 ( https://crsu.shinyapps.io/metainsightcovid/ ) was developed to conduct NMA with the evolving evidence on treatments for COVID-19. It was updated weekly between 19th May - 19th Oct 2020, incorporating new evidence identified from a living systematic review.

RESULTS

The app includes embedded functions to facilitate study selection based on study characteristics, and displays the synthesised results in real time. It allows both frequentist and Bayesian NMA to be conducted as well as consistency and heterogeneity assessments. A demonstration of the app is provided and experiences of building such a platform are discussed.

CONCLUSIONS

MetaInsight COVID-19 allows users to take control of the evidence synthesis using the analytic approach they deem appropriate to ascertain how robust findings are to alternative analysis strategies and study inclusion criteria. It is hoped that this app will help avoid many of the duplicated efforts when reviewing and synthesising the COVID-19 evidence, and, in addition, establish the desirability of an open platform format such as this for interactive data interrogation, visualisation, and reporting for any traditional or 'living' NMA.

Non-pharmacological interventions for preventing delirium in hospitalised non-ICU patients

BACKGROUND

Delirium is an acute neuropsychological disorder that is common in hospitalised patients. It can be distressing to patients and carers and it is associated with serious adverse outcomes. Treatment options for established delirium are limited and so prevention of delirium is desirable. Non-pharmacological interventions are thought to be important in delirium prevention.  OBJECTIVES: To assess the effectiveness of non-pharmacological interventions designed to prevent delirium in hospitalised patients outside intensive care units (ICU).

SEARCH METHODS

We searched ALOIS, the specialised register of the Cochrane Dementia and Cognitive Improvement Group, with additional searches conducted in MEDLINE, Embase, PsycINFO, CINAHL, LILACS, Web of Science Core Collection, ClinicalTrials.gov and the World Health Organization Portal/ICTRP to 16 September 2020. There were no language or date restrictions applied to the electronic searches, and no methodological filters were used to restrict the search.

SELECTION CRITERIA

We included randomised controlled trials (RCTs) of single and multicomponent non-pharmacological interventions for preventing delirium in hospitalised adults cared for outside intensive care or high dependency settings. We only included non-pharmacological interventions which were designed and implemented to prevent delirium.  DATA COLLECTION AND ANALYSIS: Two review authors independently examined titles and abstracts identified by the search for eligibility and extracted data from full-text articles. Any disagreements on eligibility and inclusion were resolved by consensus. We used standard Cochrane methodological procedures. The primary outcomes were: incidence of delirium; inpatient and later mortality; and new diagnosis of dementia. We included secondary and adverse outcomes as pre-specified in the review protocol. We used risk ratios (RRs) as measures of treatment effect for dichotomous outcomes and between-group mean differences for continuous outcomes. The certainty of the evidence was assessed using GRADE. A complementary exploratory analysis was undertaker using a Bayesian component network meta-analysis fixed-effect model to evaluate the comparative effectiveness of the individual components of multicomponent interventions and describe which components were most strongly associated with reducing the incidence of delirium.

MAIN RESULTS

We included 22 RCTs that recruited a total of 5718 adult participants. Fourteen trials compared a multicomponent delirium prevention intervention with usual care. Two trials compared liberal and restrictive blood transfusion thresholds. The remaining six trials each investigated a different non-pharmacological intervention. Incidence of delirium was reported in all studies.  Using the Cochrane risk of bias tool, we identified risks of bias in all included trials. All were at high risk of performance bias as participants and personnel were not blinded to the interventions. Nine trials were at high risk of detection bias due to lack of blinding of outcome assessors and three more were at unclear risk in this domain.  Pooled data showed that multi-component non-pharmacological interventions probably reduce the incidence of delirium compared to usual care (10.5% incidence in the intervention group, compared to 18.4% in the control group, risk ratio (RR) 0.57, 95% confidence interval (CI) 0.46 to 0.71, I² = 39%; 14 studies; 3693 participants; moderate-certainty evidence, downgraded due to risk of bias).  There may be little or no effect of multicomponent interventions on inpatient mortality compared to usual care (5.2% in the intervention group, compared to 4.5% in the control group, RR 1.17, 95% CI 0.79 to 1.74, I² = 15%; 10 studies; 2640 participants; low-certainty evidence downgraded due to inconsistency and imprecision).  No studies of multicomponent interventions reported data on new diagnoses of dementia.  Multicomponent interventions may result in a small reduction of around a day in the duration of a delirium episode (mean difference (MD) -0.93, 95% CI -2.01 to 0.14 days, I² = 65%; 351 participants; low-certainty evidence downgraded due to risk of bias and imprecision). The evidence is very uncertain about the effect of multicomponent interventions on delirium severity (standardised mean difference (SMD) -0.49, 95% CI -1.13 to 0.14, I²=64%; 147 participants; very low-certainty evidence downgraded due to risk of bias and serious imprecision). Multicomponent interventions may result in a reduction in hospital length of stay compared to usual care (MD -1.30 days, 95% CI -2.56 to -0.04 days, I²=91%; 3351 participants; low-certainty evidence downgraded due to risk of bias and inconsistency), but little to no difference in new care home admission at the time of hospital discharge (RR 0.77, 95% CI 0.55 to 1.07; 536 participants; low-certainty evidence downgraded due to risk of bias and imprecision). Reporting of other adverse outcomes was limited.  Our exploratory component network meta-analysis found that re-orientation (including use of familiar objects), cognitive stimulation and sleep hygiene were associated with reduced risk of incident delirium. Attention to nutrition and hydration, oxygenation, medication review, assessment of mood and bowel and bladder care were probably associated with a reduction in incident delirium but estimates included the possibility of no benefit or harm.  Reducing sensory deprivation, identification of infection, mobilisation and pain control all had summary estimates that suggested potential increases in delirium incidence, but the uncertainty in the estimates was substantial.  Evidence from two trials suggests that use of a liberal transfusion threshold over a restrictive transfusion threshold probably results in little to no difference in incident delirium (RR 0.92, 95% CI 0.62 to 1.36; I² = 9%; 294 participants; moderate-certainty evidence downgraded due to risk of bias).  Six other interventions were examined, but evidence for each was limited to single studies and we identified no evidence of delirium prevention.  AUTHORS' CONCLUSIONS: There is moderate-certainty evidence regarding the benefit of multicomponent non-pharmacological interventions for the prevention of delirium in hospitalised adults, estimated to reduce incidence by 43% compared to usual care. We found no evidence of an effect on mortality. There is emerging evidence that these interventions may reduce hospital length of stay, with a trend towards reduced delirium duration, although the effect on delirium severity remains uncertain. Further research should focus on implementation and detailed analysis of the components of the interventions to support more effective, tailored practice recommendations.

Effectiveness and safety of non-steroidal anti-inflammatory drugs and opioid treatment for knee and hip osteoarthritis: network meta-analysis

OBJECTIVE

To assess the effectiveness and safety of different preparations and doses of non-steroidal anti-inflammatory drugs (NSAIDs), opioids, and paracetamol for knee and hip osteoarthritis pain and physical function to enable effective and safe use of these drugs at their lowest possible dose.

DESIGN

Systematic review and network meta-analysis of randomised trials.

DATA SOURCES

Cochrane Central Register of Controlled Trials (CENTRAL), Medline, Embase, regulatory agency websites, and ClinicalTrials.gov from inception to 28 June 2021.

ELIGIBILITY CRITERIA FOR SELECTING STUDIES

Randomised trials published in English with ≥100 patients per group that evaluated NSAIDs, opioids, or paracetamol (acetaminophen) to treat osteoarthritis.

OUTCOMES AND MEASURES

The prespecified primary outcome was pain. Physical function and safety outcomes were also assessed.

REVIEW METHODS

Two reviewers independently extracted outcomes data and evaluated the risk of bias of included trials. Bayesian random effects models were used for network meta-analysis of all analyses. Effect estimates are comparisons between active treatments and oral placebo.

RESULTS

192 trials comprising 102 829 participants examined 90 different active preparations or doses (68 for NSAIDs, 19 for opioids, and three for paracetamol). Five oral preparations (diclofenac 150 mg/day, etoricoxib 60 and 90 mg/day, and rofecoxib 25 and 50 mg/day) had ≥99% probability of more pronounced treatment effects than the minimal clinically relevant reduction in pain. Topical diclofenac (70-81 and 140-160 mg/day) had ≥92.3% probability, and all opioids had ≤53% probability of more pronounced treatment effects than the minimal clinically relevant reduction in pain. 18.5%, 0%, and 83.3% of the oral NSAIDs, topical NSAIDs, and opioids, respectively, had an increased risk of dropouts due to adverse events. 29.8%, 0%, and 89.5% of oral NSAIDs, topical NSAIDs, and opioids, respectively, had an increased risk of any adverse event. Oxymorphone 80 mg/day had the highest risk of dropouts due to adverse events (51%) and any adverse event (88%).

CONCLUSIONS

Etoricoxib 60 mg/day and diclofenac 150 mg/day seem to be the most effective oral NSAIDs for pain and function in patients with osteoarthritis. However, these treatments are probably not appropriate for patients with comorbidities or for long term use because of the slight increase in the risk of adverse events. Additionally, an increased risk of dropping out due to adverse events was found for diclofenac 150 mg/day. Topical diclofenac 70-81 mg/day seems to be effective and generally safer because of reduced systemic exposure and lower dose, and should be considered as first line pharmacological treatment for knee osteoarthritis. The clinical benefit of opioid treatment, regardless of preparation or dose, does not outweigh the harm it might cause in patients with osteoarthritis.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO number CRD42020213656.

Interrater agreement in dementia diagnosis: A systematic review and meta-analysis

OBJECTIVES

Dementia remains a clinical diagnosis with a degree of subjective assessment and potential for interrater disagreement. We described interrater agreement of clinical dementia diagnosis for various diagnostic criteria.

METHODS

We conducted a PROSPERO-registered (CRD42020168245) systematic review and meta-analysis. We searched multiple cross-disciplinary databases from inception until April 2020 for relevant papers, extracted data and described study quality in duplicate. Study quality was assessed using the Guidelines for Reporting Reliability and Agreement Studies. We used random-effects models to obtain summary estimates of interrater agreement using kappa and, where possible, Gwet's AC1/2 coefficients.

RESULTS

We found 7577 titles and 22 eligible studies. Meta-analysis was possible for all-cause dementia using the Diagnostic and Statistical Manual of Mental Disorders third edition revised (DSM-III-R) criteria (kappa = 0.66, 95% CI = [0.53,0.78]), Alzheimer's disease using the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's disease and Related Disorders Association (NINCDS-ADRDA) criteria (kappa = 0.71, 95% CI = [0.65,0.77] for presence/absence and AC2 = 0.61, 95% CI = [0.53,0.70] when distinguishing probable/possible cases), and vascular dementia using the International Classification of Diseases version 10 (ICD-10) criteria kappa = 0.79 (95% CI = [0.70,0.87]). Data was more limited for other criteria and dementia types. AC1/2 coefficients generally indicated higher agreement. One study was rated as high quality.

CONCLUSIONS

Diagnostic criteria for clinical dementia may have good but imperfect agreement. This has important implications for clinical practice and research studies, which frequently assume these criteria are perfect tests, such as diagnostic test accuracy studies frequently conducted for biomarkers and neuropsychological tests, and for trials where incident dementia is the outcome.

Non-pharmacological interventions for preventing delirium in hospitalised non-ICU patients

BACKGROUND

Delirium is an acute neuropsychological disorder that is common in hospitalised patients. It can be distressing to patients and carers and it is associated with serious adverse outcomes. Treatment options for established delirium are limited and so prevention of delirium is desirable. Non-pharmacological interventions are thought to be important in delirium prevention.  OBJECTIVES: To assess the effectiveness of non-pharmacological interventions designed to prevent delirium in hospitalised patients outside intensive care units (ICU).

SEARCH METHODS

We searched ALOIS, the specialised register of the Cochrane Dementia and Cognitive Improvement Group, with additional searches conducted in MEDLINE, Embase, PsycINFO, CINAHL, LILACS, Web of Science Core Collection, ClinicalTrials.gov and the World Health Organization Portal/ICTRP to 16 September 2020. There were no language or date restrictions applied to the electronic searches, and no methodological filters were used to restrict the search.

SELECTION CRITERIA

We included randomised controlled trials (RCTs) of single and multicomponent non-pharmacological interventions for preventing delirium in hospitalised adults cared for outside intensive care or high dependency settings. We only included non-pharmacological interventions which were designed and implemented to prevent delirium.  DATA COLLECTION AND ANALYSIS: Two review authors independently examined titles and abstracts identified by the search for eligibility and extracted data from full-text articles. Any disagreements on eligibility and inclusion were resolved by consensus. We used standard Cochrane methodological procedures. The primary outcomes were: incidence of delirium; inpatient and later mortality; and new diagnosis of dementia. We included secondary and adverse outcomes as pre-specified in the review protocol. We used risk ratios (RRs) as measures of treatment effect for dichotomous outcomes and between-group mean differences for continuous outcomes. The certainty of the evidence was assessed using GRADE. A complementary exploratory analysis was undertaker using a Bayesian component network meta-analysis fixed-effect model to evaluate the comparative effectiveness of the individual components of multicomponent interventions and describe which components were most strongly associated with reducing the incidence of delirium.

MAIN RESULTS

We included 22 RCTs that recruited a total of 5718 adult participants. Fourteen trials compared a multicomponent delirium prevention intervention with usual care. Two trials compared liberal and restrictive blood transfusion thresholds. The remaining six trials each investigated a different non-pharmacological intervention. Incidence of delirium was reported in all studies.  Using the Cochrane risk of bias tool, we identified risks of bias in all included trials. All were at high risk of performance bias as participants and personnel were not blinded to the interventions. Nine trials were at high risk of detection bias due to lack of blinding of outcome assessors and three more were at unclear risk in this domain.  Pooled data showed that multi-component non-pharmacological interventions probably reduce the incidence of delirium compared to usual care (10.5% incidence in the intervention group, compared to 18.4% in the control group, risk ratio (RR) 0.57, 95% confidence interval (CI) 0.46 to 0.71, I2 = 39%; 14 studies; 3693 participants; moderate-certainty evidence, downgraded due to risk of bias).  There may be little or no effect of multicomponent interventions on inpatient mortality compared to usual care (5.2% in the intervention group, compared to 4.5% in the control group, RR 1.17, 95% CI 0.79 to 1.74, I2 = 15%; 10 studies; 2640 participants; low-certainty evidence downgraded due to inconsistency and imprecision).  No studies of multicomponent interventions reported data on new diagnoses of dementia.  Multicomponent interventions may result in a small reduction of around a day in the duration of a delirium episode (mean difference (MD) -0.93, 95% CI -2.01 to 0.14 days, I2 = 65%; 351 participants; low-certainty evidence downgraded due to risk of bias and imprecision). The evidence is very uncertain about the effect of multicomponent interventions on delirium severity (standardised mean difference (SMD) -0.49, 95% CI -1.13 to 0.14, I2=64%; 147 participants; very low-certainty evidence downgraded due to risk of bias and serious imprecision). Multicomponent interventions may result in a reduction in hospital length of stay compared to usual care (MD -1.30 days, 95% CI -2.56 to -0.04 days, I2=91%; 3351 participants; low-certainty evidence downgraded due to risk of bias and inconsistency), but little to no difference in new care home admission at the time of hospital discharge (RR 0.77, 95% CI 0.55 to 1.07; 536 participants; low-certainty evidence downgraded due to risk of bias and imprecision). Reporting of other adverse outcomes was limited.  Our exploratory component network meta-analysis found that re-orientation (including use of familiar objects), cognitive stimulation and sleep hygiene were associated with reduced risk of incident delirium. Attention to nutrition and hydration, oxygenation, medication review, assessment of mood and bowel and bladder care were probably associated with a reduction in incident delirium but estimates included the possibility of no benefit or harm.  Reducing sensory deprivation, identification of infection, mobilisation and pain control all had summary estimates that suggested potential increases in delirium incidence, but the uncertainty in the estimates was substantial.  Evidence from two trials suggests that use of a liberal transfusion threshold over a restrictive transfusion threshold probably results in little to no difference in incident delirium (RR 0.92, 95% CI 0.62 to 1.36; I2 = 9%; 294 participants; moderate-certainty evidence downgraded due to risk of bias).  Six other interventions were examined, but evidence for each was limited to single studies and we identified no evidence of delirium prevention.  AUTHORS' CONCLUSIONS: There is moderate-certainty evidence regarding the benefit of multicomponent non-pharmacological interventions for the prevention of delirium in hospitalised adults, estimated to reduce incidence by 43% compared to usual care. We found no evidence of an effect on mortality. There is emerging evidence that these interventions may reduce hospital length of stay, with a trend towards reduced delirium duration, although the effect on delirium severity remains uncertain. Further research should focus on implementation and detailed analysis of the components of the interventions to support more effective, tailored practice recommendations.

Nutritional supplementation for nonalcohol-related fatty liver disease: a network meta-analysis

BACKGROUND

The prevalence of non-alcohol-related fatty liver disease (NAFLD) varies between 19% and 33% in different populations. NAFLD decreases life expectancy and increases risks of liver cirrhosis, hepatocellular carcinoma, and the requirement for liver transplantation. Uncertainty surrounds relative benefits and harms of various nutritional supplements in NAFLD. Currently no nutritional supplement is recommended for people with NAFLD.

OBJECTIVES

• To assess the benefits and harms of different nutritional supplements for treatment of NAFLD through a network meta-analysis • To generate rankings of different nutritional supplements according to their safety and efficacy SEARCH METHODS: We searched the Cochrane Central Register of Controlled Trials, MEDLINE, Embase, Science Citation Index Expanded, Conference Proceedings Citation Index-Science, the World Health Organization International Clinical Trials Registry Platform, and trials registers until February 2021 to identify randomised clinical trials in people with NAFLD.

SELECTION CRITERIA

We included only randomised clinical trials (irrespective of language, blinding, or status) for people with NAFLD, irrespective of method of diagnosis, age and diabetic status of participants, or presence of non-alcoholic steatohepatitis (NASH). We excluded randomised clinical trials in which participants had previously undergone liver transplantation.

DATA COLLECTION AND ANALYSIS

We performed a network meta-analysis with OpenBUGS using Bayesian methods whenever possible and calculated differences in treatments using hazard ratios (HRs), odds ratios (ORs), and rate ratios with 95% credible intervals (CrIs) based on an available-case analysis, according to National Institute of Health and Care Excellence Decision Support Unit guidance.

MAIN RESULTS

We included in the review a total of 202 randomised clinical trials (14,200 participants). Nineteen trials were at low risk of bias. A total of 32 different interventions were compared in these trials. A total of 115 trials (7732 participants) were included in one or more comparisons. The remaining trials did not report any of the outcomes of interest for this review. Follow-up ranged from 1 month to 28 months. The follow-up period in trials that reported clinical outcomes was 2 months to 28 months. During this follow-up period, clinical events related to NAFLD such as mortality, liver cirrhosis, liver decompensation, liver transplantation, hepatocellular carcinoma, and liver-related mortality were sparse. We did not calculate effect estimates for mortality because of sparse data (zero events for at least one of the groups in the trial). None of the trials reported that they measured overall health-related quality of life using a validated scale. The evidence is very uncertain about effects of interventions on serious adverse events (number of people or number of events). We are very uncertain about effects on adverse events of most of the supplements that we investigated, as the evidence is of very low certainty. However, people taking PUFA (polyunsaturated fatty acid) may be more likely to experience an adverse event than those not receiving an active intervention (network meta-analysis results: OR 4.44, 95% CrI 2.40 to 8.48; low-certainty evidence; 4 trials, 203 participants; direct evidence: OR 4.43, 95% CrI 2.43 to 8.42). People who take other supplements (a category that includes nutritional supplements other than vitamins, fatty acids, phospholipids, and antioxidants) had higher numbers of adverse events than those not receiving an active intervention (network meta-analysis: rate ratio 1.73, 95% CrI 1.26 to 2.41; 6 trials, 291 participants; direct evidence: rate ratio 1.72, 95% CrI 1.25 to 2.40; low-certainty evidence). Data were sparse (zero events in all groups in the trial) for liver transplantation, liver decompensation, and hepatocellular carcinoma. So, we did not perform formal analysis for these outcomes. The evidence is very uncertain about effects of other antioxidants (antioxidants other than vitamins) compared to no active intervention on liver cirrhosis (HR 1.68, 95% CrI 0.23 to 15.10; 1 trial, 99 participants; very low-certainty evidence). The evidence is very uncertain about effects of interventions in any of the remaining comparisons, or data were sparse (with zero events in at least one of the groups), precluding formal calculations of effect estimates. Data were probably because of the very short follow-up period (2 months to 28 months). It takes follow-up of 8 to 28 years to detect differences in mortality between people with NAFLD and the general population. Therefore, it is unlikely that differences in clinical outcomes are noted in trials providing less than 5 to 10 years of follow-up.

AUTHORS' CONCLUSIONS

The evidence indicates considerable uncertainty about effects of nutritional supplementation compared to no additional intervention on all clinical outcomes for people with non-alcohol-related fatty liver disease. Accordingly, high-quality randomised comparative clinical trials with adequate follow-up are needed. We propose registry-based randomised clinical trials or cohort multiple randomised clinical trials (study design in which multiple interventions are trialed within large longitudinal cohorts of patients to gain efficiencies and align trials more closely to standard clinical practice) comparing interventions such as vitamin E, prebiotics/probiotics/synbiotics, PUFAs, and no nutritional supplementation. The reason for the choice of interventions is the impact of these interventions on indirect outcomes, which may translate to clinical benefit. Outcomes in such trials should be mortality, health-related quality of life, decompensated liver cirrhosis, liver transplantation, and resource utilisation measures including costs of intervention and decreased healthcare utilisation after minimum follow-up of 8 years (to find meaningful differences in clinically important outcomes).

Lifestyle modifications for nonalcohol-related fatty liver disease: a network meta-analysis

BACKGROUND

The prevalence of nonalcohol-related fatty liver disease (NAFLD) varies between 19% and 33% in different populations. NAFLD decreases life expectancy and increases the risks of liver cirrhosis, hepatocellular carcinoma, and requirement for liver transplantation. There is uncertainty surrounding the relative benefits and harms of various lifestyle interventions for people with NAFLD.

OBJECTIVES

To assess the comparative benefits and harms of different lifestyle interventions in the treatment of NAFLD through a network meta-analysis, and to generate rankings of the different lifestyle interventions according to their safety and efficacy.

SEARCH METHODS

We searched CENTRAL, MEDLINE, Embase, Science Citation Index Expanded, Conference Proceedings Citation Index - Science, World Health Organization International Clinical Trials Registry Platform, and trials registers until February 2021 to identify randomised clinical trials in people with NAFLD.

SELECTION CRITERIA

We included only randomised clinical trials (irrespective of language, blinding, or status) in people with NAFLD, whatever the method of diagnosis, age, and diabetic status of participants, or presence of non-alcoholic steatohepatitis (NASH). We excluded randomised clinical trials in which participants had previously undergone liver transplantation.

DATA COLLECTION AND ANALYSIS

We planned to perform a network meta-analysis with OpenBUGS using Bayesian methods and to calculate the differences in treatments using hazard ratios (HRs), odds ratios (ORs), and rate ratios (RaRs) with 95% credible intervals (CrIs) based on an available-participant analysis, according to National Institute of Health and Care Excellence Decision Support Unit guidance. However, the data were too sparse for the clinical outcomes. We therefore performed only direct comparisons (head-to-head comparisons) with OpenBUGS using Bayesian methods.

MAIN RESULTS

We included a total of 59 randomised clinical trials (3631 participants) in the review. All but two trials were at high risk of bias. A total of 33 different interventions, ranging from advice to supervised exercise and special diets, or a combination of these and no additional intervention were compared in these trials. The reference treatment was no active intervention. Twenty-eight trials (1942 participants) were included in one or more comparisons. The follow-up ranged from 1 month to 24 months. The remaining trials did not report any of the outcomes of interest for this review. The follow-up period in the trials that reported clinical outcomes was 2 months to 24 months. During this short follow-up period, clinical events related to NAFLD such as mortality, liver cirrhosis, liver decompensation, liver transplantation, hepatocellular carcinoma, and liver-related mortality were sparse. This is probably because of the very short follow-up periods. It takes a follow-up of 8 years to 28 years to detect differences in mortality between people with NAFLD and the general population. It is therefore unlikely that differences by clinical outcomes will be noted in trials with less than 5 years to 10 years of follow-up. In one trial, one participant developed an adverse event. There were no adverse events in any of the remaining participants in this trial, or in any of the remaining trials, which seemed to be directly related to the intervention.

AUTHORS' CONCLUSIONS

The evidence indicates considerable uncertainty about the effects of the lifestyle interventions compared with no additional intervention (to general public health advice) on any of the clinical outcomes after a short follow-up period of 2 months to 24 months in people with nonalcohol-related fatty liver disease. Accordingly, high-quality randomised clinical trials with adequate follow-up are needed. We propose registry-based randomised clinical trials or cohort multiple randomised clinical trials (a study design in which multiple interventions are trialed within large longitudinal cohorts of participants to gain efficiencies and align trials more closely to standard clinical practice), comparing aerobic exercise and dietary advice versus standard of care (exercise and dietary advice received as part of national health promotion). The reason for the choice of aerobic exercise and dietary advice is the impact of these interventions on indirect outcomes which may translate to clinical benefit. The outcomes in such trials should be mortality, health-related quality of life, decompensated liver cirrhosis, liver transplantation, and resource use measures including costs of intervention and decreased healthcare use after a minimum follow-up of eight years, to find meaningful differences in the clinically important outcomes.

Treatment for bleeding oesophageal varices in people with decompensated liver cirrhosis: a network meta-analysis

BACKGROUND

Approximately 40% to 95% of people with liver cirrhosis have oesophageal varices. About 15% to 20% of oesophageal varices bleed within about one to three years after diagnosis. Several different treatments are available, including, among others, endoscopic sclerotherapy, variceal band ligation, somatostatin analogues, vasopressin analogues, and balloon tamponade. However, there is uncertainty surrounding the individual and relative benefits and harms of these treatments.

OBJECTIVES

To compare the benefits and harms of different initial treatments for variceal bleeding from oesophageal varices in adults with decompensated liver cirrhosis, through a network meta-analysis; and to generate rankings of the different treatments for acute bleeding oesophageal varices, according to their benefits and harms.

SEARCH METHODS

We searched CENTRAL, MEDLINE, Embase, Science Citation Index Expanded, World Health Organization International Clinical Trials Registry Platform, and trials registers until 17 December 2019, to identify randomised clinical trials (RCTs) in people with cirrhosis and acute bleeding from oesophageal varices.

SELECTION CRITERIA

We included only RCTs (irrespective of language, blinding, or status) in adults with cirrhosis and acutely bleeding oesophageal varices. We excluded RCTs in which participants had bleeding only from gastric varices, those who failed previous treatment (refractory bleeding), those in whom initial haemostasis was achieved before inclusion into the trial, and those who had previously undergone liver transplantation.

DATA COLLECTION AND ANALYSIS

We performed a network meta-analysis with OpenBUGS software, using Bayesian methods, and calculated the differences in treatments using odds ratios (OR) and rate ratios with 95% credible intervals (CrI) based on an available-case analysis, according to National Institute of Health and Care Excellence Decision Support Unit guidance. We performed also the direct comparisons from RCTs using the same codes and the same technical details.

MAIN RESULTS

We included a total of 52 RCTs (4580 participants) in the review. Forty-eight trials (4042 participants) were included in one or more comparisons in the review. The trials that provided the information included people with cirrhosis due to varied aetiologies and those with and without a previous history of bleeding. We included outcomes assessed up to six weeks. All trials were at high risk of bias. A total of 19 interventions were compared in the trials (sclerotherapy, somatostatin analogues, vasopressin analogues, sclerotherapy plus somatostatin analogues, variceal band ligation, balloon tamponade, somatostatin analogues plus variceal band ligation, nitrates plus vasopressin analogues, no active intervention, sclerotherapy plus variceal band ligation, balloon tamponade plus sclerotherapy, balloon tamponade plus somatostatin analogues, balloon tamponade plus vasopressin analogues, variceal band ligation plus vasopressin analogues, balloon tamponade plus nitrates plus vasopressin analogues, balloon tamponade plus variceal band ligation, portocaval shunt, sclerotherapy plus transjugular intrahepatic portosystemic shunt (TIPS), and sclerotherapy plus vasopressin analogues). We have reported the effect estimates for the primary and secondary outcomes when there was evidence of differences between the interventions against the reference treatment of sclerotherapy, but reported the other results of the primary and secondary outcomes versus the reference treatment of sclerotherapy without the effect estimates when there was no evidence of differences in order to provide a concise summary of the results. Overall, 15.8% of the trial participants who received the reference treatment of sclerotherapy (chosen because this was the commonest treatment compared in the trials) died during the follow-up periods, which ranged from three days to six weeks. Based on moderate-certainty evidence, somatostatin analogues alone had higher mortality than sclerotherapy (OR 1.57, 95% CrI 1.04 to 2.41; network estimate; direct comparison: 4 trials; 353 participants) and vasopressin analogues alone had higher mortality than sclerotherapy (OR 1.70, 95% CrI 1.13 to 2.62; network estimate; direct comparison: 2 trials; 438 participants). None of the trials reported health-related quality of life. Based on low-certainty evidence, a higher proportion of people receiving balloon tamponade plus sclerotherapy had more serious adverse events than those receiving only sclerotherapy (OR 4.23, 95% CrI 1.22 to 17.80; direct estimate; 1 RCT; 60 participants). Based on moderate-certainty evidence, people receiving vasopressin analogues alone and those receiving variceal band ligation had fewer adverse events than those receiving only sclerotherapy (rate ratio 0.59, 95% CrI 0.35 to 0.96; network estimate; direct comparison: 1 RCT; 219 participants; and rate ratio 0.40, 95% CrI 0.21 to 0.74; network estimate; direct comparison: 1 RCT; 77 participants; respectively). Based on low-certainty evidence, the proportion of people who developed symptomatic rebleed was smaller in people who received sclerotherapy plus somatostatin analogues than those receiving only sclerotherapy (OR 0.21, 95% CrI 0.03 to 0.94; direct estimate; 1 RCT; 105 participants). The evidence suggests considerable uncertainty about the effect of the interventions in the remaining comparisons where sclerotherapy was the control intervention.

AUTHORS' CONCLUSIONS

Based on moderate-certainty evidence, somatostatin analogues alone and vasopressin analogues alone (with supportive therapy) probably result in increased mortality, compared to endoscopic sclerotherapy. Based on moderate-certainty evidence, vasopressin analogues alone and band ligation alone probably result in fewer adverse events compared to endoscopic sclerotherapy. Based on low-certainty evidence, balloon tamponade plus sclerotherapy may result in large increases in serious adverse events compared to sclerotherapy. Based on low-certainty evidence, sclerotherapy plus somatostatin analogues may result in large decreases in symptomatic rebleed compared to sclerotherapy. In the remaining comparisons, the evidence indicates considerable uncertainty about the effects of the interventions, compared to sclerotherapy.

Primary prevention of variceal bleeding in people with oesophageal varices due to liver cirrhosis: a network meta-analysis

BACKGROUND

Approximately 40% to 95% of people with cirrhosis have oesophageal varices. About 15% to 20% of oesophageal varices bleed in about one to three years. There are several different treatments to prevent bleeding, including: beta-blockers, endoscopic sclerotherapy, and variceal band ligation. However, there is uncertainty surrounding their individual and relative benefits and harms.

OBJECTIVES

To compare the benefits and harms of different treatments for prevention of first variceal bleeding from oesophageal varices in adults with liver cirrhosis through a network meta-analysis and to generate rankings of the different treatments for prevention of first variceal bleeding from oesophageal varices according to their safety and efficacy.

SEARCH METHODS

We searched CENTRAL, MEDLINE, Embase, Science Citation Index Expanded, World Health Organization International Clinical Trials Registry Platform, and trials registers to December 2019 to identify randomised clinical trials in people with cirrhosis and oesophageal varices with no history of bleeding.

SELECTION CRITERIA

We included only randomised clinical trials (irrespective of language, blinding, or status) in adults with cirrhosis and oesophageal varices with no history of bleeding. We excluded randomised clinical trials in which participants had previous bleeding from oesophageal varices and those who had previously undergone liver transplantation or previously received prophylactic treatment for oesophageal varices.

DATA COLLECTION AND ANALYSIS

We performed a network meta-analysis with OpenBUGS using Bayesian methods and calculated the differences in treatments using hazard ratios (HR), odds ratios (OR), and rate ratios with 95% credible intervals (CrI) based on an available-case analysis, according to National Institute for Health and Care Excellence Decision Support Unit guidance. We performed the direct comparisons from randomised clinical trials using the same codes and the same technical details.

MAIN RESULTS

We included 66 randomised clinical trials (6653 participants) in the review. Sixty trials (6212 participants) provided data for one or more comparisons in the review. The trials that provided the information included people with cirrhosis due to varied aetiologies and those at high risk of bleeding from oesophageal varices. The follow-up in the trials that reported outcomes ranged from 6 months to 60 months. All but one of the trials were at high risk of bias. The interventions compared included beta-blockers, no active intervention, variceal band ligation, sclerotherapy, beta-blockers plus variceal band ligation, beta-blockers plus nitrates, nitrates, beta-blockers plus sclerotherapy, and portocaval shunt. Overall, 21.2% of participants who received non-selective beta-blockers ('beta-blockers') - the reference treatment (chosen because this was the most common treatment compared in the trials) - died during 8-month to 60-month follow-up. Based on low-certainty evidence, beta-blockers, variceal band ligation, sclerotherapy, and beta-blockers plus nitrates all had lower mortality versus no active intervention (beta-blockers: HR 0.49, 95% CrI 0.36 to 0.67; direct comparison HR: 0.59, 95% CrI 0.42 to 0.83; 10 trials, 1200 participants; variceal band ligation: HR 0.51, 95% CrI 0.35 to 0.74; direct comparison HR 0.49, 95% CrI 0.12 to 2.14; 3 trials, 355 participants; sclerotherapy: HR 0.66, 95% CrI 0.51 to 0.85; direct comparison HR 0.61, 95% CrI 0.41 to 0.90; 18 trials, 1666 participants; beta-blockers plus nitrates: HR 0.41, 95% CrI 0.20 to 0.85; no direct comparison). No trials reported health-related quality of life. Based on low-certainty evidence, variceal band ligation had a higher number of serious adverse events (number of events) than beta-blockers (rate ratio 10.49, 95% CrI 2.83 to 60.64; 1 trial, 168 participants). Based on low-certainty evidence, beta-blockers plus nitrates had a higher number of 'any adverse events (number of participants)' than beta-blockers alone (OR 3.41, 95% CrI 1.11 to 11.28; 1 trial, 57 participants). Based on low-certainty evidence, adverse events (number of events) were higher in sclerotherapy than in beta-blockers (rate ratio 2.49, 95% CrI 1.53 to 4.22; direct comparison rate ratio 2.47, 95% CrI 1.27 to 5.06; 2 trials, 90 participants), and in beta-blockers plus variceal band ligation than in beta-blockers (direct comparison rate ratio 1.72, 95% CrI 1.08 to 2.76; 1 trial, 140 participants). Based on low-certainty evidence, any variceal bleed was lower in beta-blockers plus variceal band ligation than in beta-blockers (direct comparison HR 0.21, 95% CrI 0.04 to 0.71; 1 trial, 173 participants). Based on low-certainty evidence, any variceal bleed was higher in nitrates than beta-blockers (direct comparison HR 6.40, 95% CrI 1.58 to 47.42; 1 trial, 52 participants). The evidence indicates considerable uncertainty about the effect of the interventions in the remaining comparisons.

AUTHORS' CONCLUSIONS

Based on low-certainty evidence, beta-blockers, variceal band ligation, sclerotherapy, and beta-blockers plus nitrates may decrease mortality compared to no intervention in people with high-risk oesophageal varices in people with cirrhosis and no previous history of bleeding. Based on low-certainty evidence, variceal band ligation may result in a higher number of serious adverse events than beta-blockers. The evidence indicates considerable uncertainty about the effect of beta-blockers versus variceal band ligation on variceal bleeding. The evidence also indicates considerable uncertainty about the effect of the interventions in most of the remaining comparisons.

Secondary prevention of variceal bleeding in adults with previous oesophageal variceal bleeding due to decompensated liver cirrhosis: a network meta-analysis

BACKGROUND

Approximately 40% to 95% of people with cirrhosis have oesophageal varices. About 15% to 20% of oesophageal varices bleed in about one to three years of diagnosis. Several different treatments are available, which include endoscopic sclerotherapy, variceal band ligation, beta-blockers, transjugular intrahepatic portosystemic shunt (TIPS), and surgical portocaval shunts, among others. However, there is uncertainty surrounding their individual and relative benefits and harms.

OBJECTIVES

To compare the benefits and harms of different initial treatments for secondary prevention of variceal bleeding in adults with previous oesophageal variceal bleeding due to decompensated liver cirrhosis through a network meta-analysis and to generate rankings of the different treatments for secondary prevention according to their safety and efficacy.

SEARCH METHODS

We searched CENTRAL, MEDLINE, Embase, Science Citation Index Expanded, World Health Organization International Clinical Trials Registry Platform, and trials registers until December 2019 to identify randomised clinical trials in people with cirrhosis and a previous history of bleeding from oesophageal varices.

SELECTION CRITERIA

We included only randomised clinical trials (irrespective of language, blinding, or status) in adults with cirrhosis and previous history of bleeding from oesophageal varices. We excluded randomised clinical trials in which participants had no previous history of bleeding from oesophageal varices, previous history of bleeding only from gastric varices, those who failed previous treatment (refractory bleeding), those who had acute bleeding at the time of treatment, and those who had previously undergone liver transplantation.

DATA COLLECTION AND ANALYSIS

We performed a network meta-analysis with OpenBUGS using Bayesian methods and calculated the differences in treatments using hazard ratios (HR), odds ratios (OR) and rate ratios with 95% credible intervals (CrI) based on an available-case analysis, according to National Institute of Health and Care Excellence Decision Support Unit guidance.

MAIN RESULTS

We included a total of 48 randomised clinical trials (3526 participants) in the review. Forty-six trials (3442 participants) were included in one or more comparisons. The trials that provided the information included people with cirrhosis due to varied aetiologies. The follow-up ranged from two months to 61 months. All the trials were at high risk of bias. A total of 12 interventions were compared in these trials (sclerotherapy, beta-blockers, variceal band ligation, beta-blockers plus sclerotherapy, no active intervention, TIPS (transjugular intrahepatic portosystemic shunt), beta-blockers plus nitrates, portocaval shunt, sclerotherapy plus variceal band ligation, beta-blockers plus nitrates plus variceal band ligation, beta-blockers plus variceal band ligation, sclerotherapy plus nitrates). Overall, 22.5% of the trial participants who received the reference treatment (chosen because this was the commonest treatment compared in the trials) of sclerotherapy died during the follow-up period ranging from two months to 61 months. There was considerable uncertainty in the effects of interventions on mortality. Accordingly, none of the interventions showed superiority over another. None of the trials reported health-related quality of life. Based on low-certainty evidence, variceal band ligation may result in fewer serious adverse events (number of people) than sclerotherapy (OR 0.19; 95% CrI 0.06 to 0.54; 1 trial; 100 participants). Based on low or very low-certainty evidence, the adverse events (number of participants) and adverse events (number of events) may be different across many comparisons; however, these differences are due to very small trials at high risk of bias showing large differences in some comparisons leading to many differences despite absence of direct evidence. Based on low-certainty evidence, TIPS may result in large decrease in symptomatic rebleed than variceal band ligation (HR 0.12; 95% CrI 0.03 to 0.41; 1 trial; 58 participants). Based on moderate-certainty evidence, any variceal rebleed was probably lower in sclerotherapy than in no active intervention (HR 0.62; 95% CrI 0.35 to 0.99, direct comparison HR 0.66; 95% CrI 0.11 to 3.13; 3 trials; 296 participants), beta-blockers plus sclerotherapy than sclerotherapy alone (HR 0.60; 95% CrI 0.37 to 0.95; direct comparison HR 0.50; 95% CrI 0.07 to 2.96; 4 trials; 231 participants); TIPS than sclerotherapy (HR 0.18; 95% CrI 0.08 to 0.38; direct comparison HR 0.22; 95% CrI 0.01 to 7.51; 2 trials; 109 participants), and in portocaval shunt than sclerotherapy (HR 0.21; 95% CrI 0.05 to 0.77; no direct comparison) groups. Based on low-certainty evidence, beta-blockers alone and TIPS might result in more, other compensation, events than sclerotherapy (rate ratio 2.37; 95% CrI 1.35 to 4.67; 1 trial; 65 participants and rate ratio 2.30; 95% CrI 1.20 to 4.65; 2 trials; 109 participants; low-certainty evidence). The evidence indicates considerable uncertainty about the effect of the interventions including those related to beta-blockers plus variceal band ligation in the remaining comparisons.

AUTHORS' CONCLUSIONS

The evidence indicates considerable uncertainty about the effect of the interventions on mortality. Variceal band ligation might result in fewer serious adverse events than sclerotherapy. TIPS might result in a large decrease in symptomatic rebleed than variceal band ligation. Sclerotherapy probably results in fewer 'any' variceal rebleeding than no active intervention. Beta-blockers plus sclerotherapy and TIPS probably result in fewer 'any' variceal rebleeding than sclerotherapy. Beta-blockers alone and TIPS might result in more other compensation events than sclerotherapy. The evidence indicates considerable uncertainty about the effect of the interventions in the remaining comparisons. Accordingly, high-quality randomised comparative clinical trials are needed.

Methodologies for systematic reviews with meta-analysis of randomised clinical trials in pain, anaesthesia, and perioperative medicine

Systematic reviews and meta-analyses (SRMAs) are increasing in popularity, but should they be used to inform clinical decision-making in anaesthesia? We present evidence that the certainty of evidence from SRMAs in anaesthesia (and in general) may be unacceptably low because of risks of bias exaggerating treatment effects, unexplained heterogeneity reducing certainty in estimates, random errors, and widespread prevalence of publication bias. We also present the latest methodological advances to help improve the certainty of evidence from SRMAs. The target audience includes both review authors and practising clinicians to help with SRMA appraisal. Issues discussed include minimising risks of bias from included trials, trial sequential analysis to reduce random error, updated methods for presenting effect estimates, and novel publication bias tests for commonly used outcome measures. These methods can help to reduce spurious conclusions on clinical significance, explain statistical heterogeneity, and reduce false positives when evaluating small-study effects. By reducing concerns in these domains of Grading of Recommendations, Assessment, Development and Evaluation, it should help improve the certainty of evidence from SRMAs used for decision-making in anaesthesia, pain, and perioperative medicine.

Behavioural interventions for smoking cessation: an overview and network meta-analysis

BACKGROUND

Smoking is a leading cause of disease and death worldwide. In people who smoke, quitting smoking can reverse much of the damage. Many people use behavioural interventions to help them quit smoking; these interventions can vary substantially in their content and effectiveness.

OBJECTIVES

To summarise the evidence from Cochrane Reviews that assessed the effect of behavioural interventions designed to support smoking cessation attempts and to conduct a network meta-analysis to determine how modes of delivery; person delivering the intervention; and the nature, focus, and intensity of behavioural interventions for smoking cessation influence the likelihood of achieving abstinence six months after attempting to stop smoking; and whether the effects of behavioural interventions depend upon other characteristics, including population, setting, and the provision of pharmacotherapy. To summarise the availability and principal findings of economic evaluations of behavioural interventions for smoking cessation, in terms of comparative costs and cost-effectiveness, in the form of a brief economic commentary.

METHODS

This work comprises two main elements. 1. We conducted a Cochrane Overview of reviews following standard Cochrane methods. We identified Cochrane Reviews of behavioural interventions (including all non-pharmacological interventions, e.g. counselling, exercise, hypnotherapy, self-help materials) for smoking cessation by searching the Cochrane Library in July 2020. We evaluated the methodological quality of reviews using AMSTAR 2 and synthesised data from the reviews narratively. 2. We used the included reviews to identify randomised controlled trials of behavioural interventions for smoking cessation compared with other behavioural interventions or no intervention for smoking cessation. To be included, studies had to include adult smokers and measure smoking abstinence at six months or longer. Screening, data extraction, and risk of bias assessment followed standard Cochrane methods. We synthesised data using Bayesian component network meta-analysis (CNMA), examining the effects of 38 different components compared to minimal intervention. Components included behavioural and motivational elements, intervention providers, delivery modes, nature, focus, and intensity of the behavioural intervention. We used component network meta-regression (CNMR) to evaluate the influence of population characteristics, provision of pharmacotherapy, and intervention intensity on the component effects. We evaluated certainty of the evidence using GRADE domains. We assumed an additive effect for individual components.

MAIN RESULTS

We included 33 Cochrane Reviews, from which 312 randomised controlled trials, representing 250,563 participants and 845 distinct study arms, met the criteria for inclusion in our component network meta-analysis. This represented 437 different combinations of components. Of the 33 reviews, confidence in review findings was high in four reviews and moderate in nine reviews, as measured by the AMSTAR 2 critical appraisal tool. The remaining 20 reviews were low or critically low due to one or more critical weaknesses, most commonly inadequate investigation or discussion (or both) of the impact of publication bias. Of note, the critical weaknesses identified did not affect the searching, screening, or data extraction elements of the review process, which have direct bearing on our CNMA. Of the included studies, 125/312 were at low risk of bias overall, 50 were at high risk of bias, and the remainder were at unclear risk. Analyses from the contributing reviews and from our CNMA showed behavioural interventions for smoking cessation can increase quit rates, but effectiveness varies on characteristics of the support provided. There was high-certainty evidence of benefit for the provision of counselling (odds ratio (OR) 1.44, 95% credibility interval (CrI) 1.22 to 1.70, 194 studies, n = 72,273) and guaranteed financial incentives (OR 1.46, 95% CrI 1.15 to 1.85, 19 studies, n = 8877). Evidence of benefit remained when removing studies at high risk of bias. These findings were consistent with pair-wise meta-analyses from contributing reviews. There was moderate-certainty evidence of benefit for interventions delivered via text message (downgraded due to unexplained statistical heterogeneity in pair-wise comparison), and for the following components where point estimates suggested benefit but CrIs incorporated no clinically significant difference: individual tailoring; intervention content including motivational components; intervention content focused on how to quit. The remaining intervention components had low-to very low-certainty evidence, with the main issues being imprecision and risk of bias. There was no evidence to suggest an increase in harms in groups receiving behavioural support for smoking cessation. Intervention effects were not changed by adjusting for population characteristics, but data were limited. Increasing intensity of behavioural support, as measured through the number of contacts, duration of each contact, and programme length, had point estimates associated with modestly increased chances of quitting, but CrIs included no difference. The effect of behavioural support for smoking cessation appeared slightly less pronounced when people were already receiving smoking cessation pharmacotherapies.

AUTHORS' CONCLUSIONS

Behavioural support for smoking cessation can increase quit rates at six months or longer, with no evidence that support increases harms. This is the case whether or not smoking cessation pharmacotherapy is also provided, but the effect is slightly more pronounced in the absence of pharmacotherapy. Evidence of benefit is strongest for the provision of any form of counselling, and guaranteed financial incentives. Evidence suggested possible benefit but the need of further studies to evaluate: individual tailoring; delivery via text message, email, and audio recording; delivery by lay health advisor; and intervention content with motivational components and a focus on how to quit. We identified 23 economic evaluations; evidence did not consistently suggest one type of behavioural intervention for smoking cessation was more cost-effective than another. Future reviews should fully consider publication bias. Tools to investigate publication bias and to evaluate certainty in CNMA are needed.

Quantitative Evidence Synthesis Methods for the Assessment of the Effectiveness of Treatment Sequences for Clinical and Economic Decision Making: A Review and Taxonomy of Simplifying Assumptions

Sequential use of alternative treatments for chronic conditions represents a complex intervention pathway; previous treatment and patient characteristics affect both the choice and effectiveness of subsequent treatments. This paper critically explores the methods for quantitative evidence synthesis of the effectiveness of sequential treatment options within a health technology assessment (HTA) or similar process. It covers methods for developing summary estimates of clinical effectiveness or the clinical inputs for the cost-effectiveness assessment and can encompass any disease condition. A comprehensive review of current approaches is presented, which considers meta-analytic methods for assessing the clinical effectiveness of treatment sequences and decision-analytic modelling approaches used to evaluate the effectiveness of treatment sequences. Estimating the effectiveness of a sequence of treatments is not straightforward or trivial and is severely hampered by the limitations of the evidence base. Randomised controlled trials (RCTs) of sequences were often absent or very limited. In the absence of sufficient RCTs of whole sequences, there is no single best way to evaluate treatment sequences; however, some approaches could be re-used or adapted, sharing ideas across different disease conditions. Each has advantages and disadvantages, and is influenced by the evidence available, extent of treatment sequences (number of treatment lines or permutations), and complexity of the decision problem. Due to the scarcity of data, modelling studies applied simplifying assumptions to data on discrete treatments. A taxonomy for all possible assumptions was developed, providing a unique resource to aid the critique of existing decision-analytic models.

Uptake of methodological advances for synthesis of continuous and time-to-event outcomes would maximize use of the evidence base

Objective

The objective of the study is to establish how often continuous and time-to-event outcomes are synthesized in health technology assessment (HTA), the statistical methods and software used in their analysis and how often evidence synthesis informs decision models.

Study Design and Setting

This is a review of National Institute of Health Research HTA reports, National Institute for Health and Care Excellence (NICE) technology appraisals, and NICE guidelines reporting quantitative meta-analysis or network meta-analysis of at least one continuous or time-to-event outcome published from April 01, 2018 to March 31, 2019.

Results

We identified 47 eligible articles. At least one continuous or time-to-event outcome was synthesized in 51% and 55% of articles, respectively. Evidence synthesis results informed decision models in two-thirds of articles. The review and expert knowledge identified five areas where methodology is available for improving the synthesis of continuous and time-to-event outcomes: i) outcomes reported on multiple scales, ii) reporting of multiple related outcomes, iii) appropriateness of the additive scale, iv) reporting of multiple time points, and v) nonproportional hazards. We identified three anticipated barriers to the uptake and implementation of these methods: i) statistical expertise, ii) software, and iii) reporting of trials.

Conclusion

Continuous and time-to-event outcomes are routinely reported in HTA. However, increased uptake of methodological advances could maximize the evidence base used to inform the decision making process.

Biological variation of cardiac troponins in chronic kidney disease

Background

Patients with chronic kidney disease often have increased plasma cardiac troponin concentration in the absence of myocardial infarction. Incidence of myocardial infarction is high in this population, and diagnosis, particularly of non ST-segment elevation myocardial infarction (NSTEMI), is challenging. Knowledge of biological variation aids understanding of serial cardiac troponin measurements and could improve interpretation in clinical practice. The National Academy of Clinical Biochemistry (NACB) recommended the use of a 20% reference change value in patients with kidney failure. The aim of this study was to calculate the biological variation of cardiac troponin I and cardiac troponin T in patients with moderate chronic kidney disease (glomerular filtration rate [GFR] 30–59 mL/min/1.73 m2).

Methods and results

Plasma samples were obtained from 20 patients (median GFR 43.0 mL/min/1.73 m2) once a week for four consecutive weeks. Cardiac troponin I (Abbott ARCHITECT® i2000SR, median 4.3 ng/L, upper 99th percentile of reference population 26.2 ng/L) and cardiac troponin T (Roche Cobas® e601, median 11.8 ng/L, upper 99th percentile of reference population 14 ng/L) were measured in duplicate using high-sensitivity assays. After outlier removal and log transformation, 18 patients’ data were subject to ANOVA, and within-subject (CVI), between-subject (CVG) and analytical (CVA) variation calculated. Variation for cardiac troponin I was 15.0%, 105.6%, 8.3%, respectively, and for cardiac troponin T 7.4%, 78.4%, 3.1%, respectively. Reference change values for increasing and decreasing troponin concentrations were +60%/–38% for cardiac troponin I and +25%/–20% for cardiac troponin T.

Conclusions

The observed reference change value for cardiac troponin T is broadly compatible with the NACB recommendation, but for cardiac troponin I, larger changes are required to define significant change. The incorporation of separate RCVs for cardiac troponin I and cardiac troponin T, and separate RCVs for rising and falling concentrations of cardiac troponin, should be considered when developing guidance for interpretation of sequential cardiac troponin measurements.

Treatment for ascites in adults with decompensated liver cirrhosis: a network meta-analysis

BACKGROUND

Approximately 20% of people with cirrhosis develop ascites. Several different treatments are available; including, among others, paracentesis plus fluid replacement, transjugular intrahepatic portosystemic shunts, aldosterone antagonists, and loop diuretics. However, there is uncertainty surrounding their relative efficacy.

OBJECTIVES

To compare the benefits and harms of different treatments for ascites in people with decompensated liver cirrhosis through a network meta-analysis and to generate rankings of the different treatments for ascites according to their safety and efficacy.

SEARCH METHODS

We searched CENTRAL, MEDLINE, Embase, Science Citation Index Expanded, World Health Organization International Clinical Trials Registry Platform, and trials registers until May 2019 to identify randomised clinical trials in people with cirrhosis and ascites.

SELECTION CRITERIA

We included only randomised clinical trials (irrespective of language, blinding, or status) in adults with cirrhosis and ascites. We excluded randomised clinical trials in which participants had previously undergone liver transplantation.

DATA COLLECTION AND ANALYSIS

We performed a network meta-analysis with OpenBUGS using Bayesian methods and calculated the odds ratio, rate ratio, and hazard ratio (HR) with 95% credible intervals (CrI) based on an available-case analysis, according to National Institute of Health and Care Excellence Decision Support Unit guidance.

MAIN RESULTS

We included a total of 49 randomised clinical trials (3521 participants) in the review. Forty-two trials (2870 participants) were included in one or more outcomes in the review. The trials that provided the information included people with cirrhosis due to varied aetiologies, without other features of decompensation, having mainly grade 3 (severe), recurrent, or refractory ascites. The follow-up in the trials ranged from 0.1 to 84 months. All the trials were at high risk of bias, and the overall certainty of evidence was low or very low. Approximately 36.8% of participants who received paracentesis plus fluid replacement (reference group, the current standard treatment) died within 11 months. There was no evidence of differences in mortality, adverse events, or liver transplantation in people receiving different interventions compared to paracentesis plus fluid replacement (very low-certainty evidence). Resolution of ascites at maximal follow-up was higher with transjugular intrahepatic portosystemic shunt (HR 9.44; 95% CrI 1.93 to 62.68) and adding aldosterone antagonists to paracentesis plus fluid replacement (HR 30.63; 95% CrI 5.06 to 692.98) compared to paracentesis plus fluid replacement (very low-certainty evidence). Aldosterone antagonists plus loop diuretics had a higher rate of other decompensation events such as hepatic encephalopathy, hepatorenal syndrome, and variceal bleeding compared to paracentesis plus fluid replacement (rate ratio 2.04; 95% CrI 1.37 to 3.10) (very low-certainty evidence). None of the trials using paracentesis plus fluid replacement reported health-related quality of life or symptomatic recovery from ascites.

FUNDING

the source of funding for four trials were industries which would benefit from the results of the study; 24 trials received no additional funding or were funded by neutral organisations; and the source of funding for the remaining 21 trials was unclear.

AUTHORS' CONCLUSIONS

Based on very low-certainty evidence, there is considerable uncertainty about whether interventions for ascites in people with decompensated liver cirrhosis decrease mortality, adverse events, or liver transplantation compared to paracentesis plus fluid replacement in people with decompensated liver cirrhosis and ascites. Based on very low-certainty evidence, transjugular intrahepatic portosystemic shunt and adding aldosterone antagonists to paracentesis plus fluid replacement may increase the resolution of ascites compared to paracentesis plus fluid replacement. Based on very low-certainty evidence, aldosterone antagonists plus loop diuretics may increase the decompensation rate compared to paracentesis plus fluid replacement.

Antibiotic prophylaxis to prevent spontaneous bacterial peritonitis in people with liver cirrhosis: a network meta-analysis

BACKGROUND

Approximately 2.5% of all hospitalisations in people with liver cirrhosis are for spontaneous bacterial peritonitis. Spontaneous bacterial peritonitis is associated with significant short-term mortality; therefore, it is important to prevent spontaneous bacterial peritonitis in people at high risk of developing it. Antibiotic prophylaxis forms the mainstay preventive method, but this has to be balanced against the development of drug-resistant spontaneous bacterial peritonitis, which is difficult to treat, and other adverse events. Several different prophylactic antibiotic treatments are available; however, there is uncertainty surrounding their relative efficacy and optimal combination.

OBJECTIVES

To compare the benefits and harms of different prophylactic antibiotic treatments for prevention of spontaneous bacterial peritonitis in people with liver cirrhosis using a network meta-analysis and to generate rankings of the different prophylactic antibiotic treatments according to their safety and efficacy.

SEARCH METHODS

We searched CENTRAL, MEDLINE, Embase, Science Citation Index Expanded, World Health Organization International Clinical Trials Registry Platform, and trials registers to November 2018 to identify randomised clinical trials in people with cirrhosis at risk of developing spontaneous bacterial peritonitis.

SELECTION CRITERIA

We included only randomised clinical trials (irrespective of language, blinding, or status) in adults with cirrhosis undergoing prophylactic treatment to prevent spontaneous bacterial peritonitis. We excluded randomised clinical trials in which participants had previously undergone liver transplantation, or were receiving antibiotics for treatment of spontaneous bacterial peritonitis or other purposes.

DATA COLLECTION AND ANALYSIS

We performed a network meta-analysis with OpenBUGS using Bayesian methods and calculated the odds ratio, rate ratio, and hazard ratio (HR) with 95% credible intervals (CrI) based on an available-case analysis, according to National Institute of Health and Care Excellence Decision Support Unit guidance.

MAIN RESULTS

We included 29 randomised clinical trials (3896 participants; nine antibiotic regimens (ciprofloxacin, neomycin, norfloxacin, norfloxacin plus neomycin, norfloxacin plus rifaximin, rifaximin, rufloxacin, sparfloxacin, sulfamethoxazole plus trimethoprim), and 'no active intervention' in the review. Twenty-three trials (2587 participants) were included in one or more outcomes in the review. The trials that provided the information included people with cirrhosis due to varied aetiologies, with or without other features of decompensation, having ascites with low protein or previous history of spontaneous bacterial peritonitis. The follow-up in the trials ranged from 1 to 12 months. Many of the trials were at high risk of bias, and the overall certainty of evidence was low or very low. Overall, approximately 10% of trial participants developed spontaneous bacterial peritonitis and 15% of trial participants died. There was no evidence of differences between any of the antibiotics and no intervention in terms of mortality (very low certainty) or number of serious adverse events (very low certainty). However, because of the wide CrIs, clinically important differences in these outcomes cannot be ruled out. None of the trials reported health-related quality of life or the proportion of people with serious adverse events. There was no evidence of differences between any of the antibiotics and no intervention in terms of proportion of people with 'any adverse events' (very low certainty), liver transplantation (very low certainty), or the proportion of people who developed spontaneous bacterial peritonitis (very low certainty). The number of 'any' adverse events per participant was fewer with norfloxacin (rate ratio 0.74, 95% CrI 0.59 to 0.94; 4 trials, 546 participants; low certainty) and sulfamethoxazole plus trimethoprim (rate ratio 0.19, 95% CrI 0.02 to 0.81; 1 trial, 60 participants; low certainty) versus no active intervention. There was no evidence of differences between the other antibiotics and no intervention in the number of 'any' adverse events per participant (very low certainty). There were fewer other decompensation events with rifaximin versus no active intervention (rate ratio 0.61, 65% CrI 0.46 to 0.80; 3 trials, 575 participants; low certainty) and norfloxacin plus neomycin (rate ratio 0.06, 95% CrI 0.00 to 0.33; 1 trial, 22 participants; low certainty). There was no evidence of differences between the other antibiotics and no intervention in the number of decompensations events per participant (very low certainty). None of the trials reported health-related quality of life or development of symptomatic spontaneous bacterial peritonitis. One would expect some correlation between the above outcomes, with interventions demonstrating effectiveness across several outcomes. This was not the case. The possible reasons for this include sparse data and selective reporting bias, which makes the results unreliable. Therefore, one cannot draw any conclusions from these inconsistent differences based on sparse data. There was no evidence of any differences in the subgroup analyses (performed when possible) based on whether the prophylaxis was primary or secondary.

FUNDING

the source of funding for five trials were organisations who would benefit from the results of the study; six trials received no additional funding or were funded by neutral organisations; and the source of funding for the remaining 18 trials was unclear.

AUTHORS' CONCLUSIONS

Based on very low-certainty evidence, there is considerable uncertainty about whether antibiotic prophylaxis is beneficial, and if beneficial, which antibiotic prophylaxis is most beneficial in people with cirrhosis and ascites with low protein or history of spontaneous bacterial peritonitis. Future randomised clinical trials should be adequately powered, employ blinding, avoid postrandomisation dropouts (or perform intention-to-treat analysis), and use clinically important outcomes such as mortality, health-related quality of life, and decompensation events.

Induction immunosuppression in adults undergoing liver transplantation: a network meta-analysis

BACKGROUND

Liver transplantation is considered the definitive treatment for people with liver failure. As part of post-liver transplantation management, immunosuppression (suppressing the host immunity) is given to prevent graft rejections. Immunosuppressive drugs can be classified into those that are used for a short period during the beginning phase of immunosuppression (induction immunosuppression) and those that are used over the entire lifetime of the individual (maintenance immunosuppression), because it is widely believed that graft rejections are more common during the first few months after liver transplantation. Some drugs such as glucocorticosteroids may be used for both induction and maintenance immunosuppression because of their multiple modalities of action. There is considerable uncertainty as to whether induction immunosuppression is necessary and if so, the relative efficacy of different immunosuppressive agents.

OBJECTIVES

To assess the comparative benefits and harms of different induction immunosuppressive regimens in adults undergoing liver transplantation through a network meta-analysis and to generate rankings of the different induction immunosuppressive regimens according to their safety and efficacy.

SEARCH METHODS

We searched CENTRAL, MEDLINE, Embase, Science Citation Index Expanded, World Health Organization International Clinical Trials Registry Platform, and trials registers until July 2019 to identify randomised clinical trials in adults undergoing liver transplantation.

SELECTION CRITERIA

We included only randomised clinical trials (irrespective of language, blinding, or status) in adults undergoing liver transplantation. We excluded randomised clinical trials in which participants had multivisceral transplantation and those who already had graft rejections.

DATA COLLECTION AND ANALYSIS

We performed a network meta-analysis with OpenBUGS using Bayesian methods and calculated the odds ratio (OR), rate ratio, and hazard ratio (HR) with 95% credible intervals (CrIs) based on an available case analysis, according to National Institute of Health and Care Excellence Decision Support Unit guidance.

MAIN RESULTS

We included a total of 25 trials (3271 participants; 8 treatments) in the review. Twenty-three trials (3017 participants) were included in one or more outcomes in the review. The trials that provided the information included people undergoing primary liver transplantation for various indications and excluded those with HIV and those with renal impairment. The follow-up in the trials ranged from three to 76 months, with a median follow-up of 12 months among trials. All except one trial were at high risk of bias, and the overall certainty of evidence was very low. Overall, approximately 7.4% of people who received the standard regimen of glucocorticosteroid induction died and 12.2% developed graft failure. All-cause mortality and graft failure was lower with basiliximab compared with glucocorticosteroid induction: all-cause mortality (HR 0.53, 95% CrI 0.31 to 0.93; network estimate, based on 2 direct comparison trials (131 participants; low-certainty evidence)); and graft failure (HR 0.44, 95% CrI 0.28 to 0.70; direct estimate, based on 1 trial (47 participants; low-certainty evidence)). There was no evidence of differences in all-cause mortality and graft failure between other induction immunosuppressants and glucocorticosteroids in either the direct comparison or the network meta-analysis (very low-certainty evidence). There was also no evidence of differences in serious adverse events (proportion), serious adverse events (number), renal failure, any adverse events (proportion), any adverse events (number), liver retransplantation, graft rejections (any), or graft rejections (requiring treatment) between other induction immunosuppressants and glucocorticosteroids in either the direct comparison or the network meta-analysis (very low-certainty evidence). However, because of the wide CrIs, clinically important differences in these outcomes cannot be ruled out. None of the studies reported health-related quality of life.

FUNDING

the source of funding for 14 trials was drug companies who would benefit from the results of the study; two trials were funded by neutral organisations who have no vested interests in the results of the study; and the source of funding for the remaining nine trials was unclear.

AUTHORS' CONCLUSIONS

Based on low-certainty evidence, basiliximab induction may decrease mortality and graft failure compared to glucocorticosteroids induction in people undergoing liver transplantation. However, there is considerable uncertainty about this finding because this information is based on small trials at high risk of bias. The evidence is uncertain about the effects of different induction immunosuppressants on other clinical outcomes, including graft rejections. Future randomised clinical trials should be adequately powered, employ blinding, avoid post-randomisation dropouts (or perform intention-to-treat analysis), and use clinically important outcomes such as mortality, graft failure, and health-related quality of life.

A comparison of the statistical performance of different meta-analysis models for the synthesis of subgroup effects from randomized clinical trials

BACKGROUND

When investigating subgroup effects in meta-analysis, it is unclear whether accounting in meta-regression for between-trial variation in treatment effects, but not between-trial variation in treatment interaction effects when such effects are present, leads to biased estimates, coverage problems, or wrong standard errors, and whether the use of aggregate data (AD) or individual-patient-data (IPD) influences this assessment.

METHODS

Seven different models were compared in a simulation study. Models differed regarding the use of AD or IPD, whether they accounted for between-trial variation in interaction effects, and whether they minimized the risk of ecological fallacy.

RESULTS

Models that used IPD and that allowed for between-trial variation of the interaction effect had less bias, better coverage, and more accurate standard errors than models that used AD or ignored this variation. The main factor influencing the performance of models was whether they used IPD or AD. The model that used AD had a considerably worse performance than all models that used IPD, especially when a low number of trials was included in the analysis.

CONCLUSIONS

The results indicate that IPD models that allow for the between-trial variation in interaction effects should be given preference whenever investigating subgroup effects within a meta-analysis.

Antibiotic treatment for spontaneous bacterial peritonitis in people with decompensated liver cirrhosis: a network meta-analysis

BACKGROUND

Approximately 2.5% of all hospitalisations in people with cirrhosis are for spontaneous bacterial peritonitis (SBP). Antibiotics, in addition to supportive treatment (fluid and electrolyte balance, treatment of shock), form the mainstay treatments of SBP. Various antibiotics are available for the treatment of SBP, but there is uncertainty regarding the best antibiotic for SBP.

OBJECTIVES

To compare the benefits and harms of different antibiotic treatments for spontaneous bacterial peritonitis (SBP) in people with decompensated liver cirrhosis.

SEARCH METHODS

We searched CENTRAL, MEDLINE, Embase, Science Citation Index Expanded, World Health Organization International Clinical Trials Registry Platform, and trials registers until November 2018 to identify randomised clinical trials on people with cirrhosis and SBP.

SELECTION CRITERIA

We included only randomised clinical trials (irrespective of language, blinding, or publication status) in adults with cirrhosis and SBP. We excluded randomised clinical trials in which participants had previously undergone liver transplantation.

DATA COLLECTION AND ANALYSIS

Two review authors independently identified eligible trials and collected data. The outcomes for this review included mortality, serious adverse events, any adverse events, resolution of SBP, liver transplantation, and other decompensation events. We performed a network meta-analysis with OpenBUGS using Bayesian methods and calculated the odds ratio, rate ratio, and hazard ratio with 95% credible intervals (CrIs) based on an available-case analysis, according to the National Institute of Health and Care Excellence (NICE) Decision Support Unit guidance.

MAIN RESULTS

We included a total of 12 trials (1278 participants; 13 antibiotics) in the review. Ten trials (893 participants) were included in one or more outcomes in the review. The trials that provided the information included patients having cirrhosis with or without other features of decompensation of varied aetiologies. The follow-up in the trials ranged from one week to three months. All the trials were at high risk of bias. Only one trial was included under each comparison for most of the outcomes. Because of these reasons, there is very low certainty in all the results. The majority of the randomised clinical trials used third-generation cephalosporins, such as intravenous ceftriaxone, cefotaxime, or ciprofloxacin as one of the interventions.Overall, approximately 75% of trial participants recovered from SBP and 25% of people died within three months. There was no evidence of difference in any of the outcomes for which network meta-analysis was possible: mortality (9 trials; 653 participants), proportion of people with any adverse events (5 trials; 297 participants), resolution of SBP (as per standard definition, 9 trials; 873 participants), or other features of decompensation (6 trials; 535 participants). The effect estimates in the direct comparisons (when available) were very similar to those of network meta-analysis. For the comparisons where network meta-analysis was not possible, there was no evidence of difference in any of the outcomes (proportion of participants with serious adverse events, number of adverse events, and proportion of participants requiring liver transplantation). Due to the wide CrIs and the very low-certainty evidence for all the outcomes, significant benefits or harms of antibiotics are possible.None of the trials reported health-related quality of life, number of serious adverse events, or symptomatic recovery from SBP.

FUNDING

the source of funding for two trials were industrial organisations who would benefit from the results of the trial; the source of funding for the remaining 10 trials was unclear.

AUTHORS' CONCLUSIONS

Short-term mortality after SBP is about 25%. There is significant uncertainty about which antibiotic therapy is better in people with SBP.We need adequately powered randomised clinical trials, with adequate blinding, avoiding post-randomisation dropouts (or performing intention-to-treat analysis), and using clinically important outcomes, such as mortality, health-related quality of life, and adverse events.

Treatment for hepatorenal syndrome in people with decompensated liver cirrhosis: a network meta-analysis

BACKGROUND

Hepatorenal syndrome is defined as renal failure in people with cirrhosis in the absence of other causes. In addition to supportive treatment such as albumin to restore fluid balance, the other potential treatments include systemic vasoconstrictor drugs (such as vasopressin analogues or noradrenaline), renal vasodilator drugs (such as dopamine), transjugular intrahepatic portosystemic shunt (TIPS), and liver support with molecular adsorbent recirculating system (MARS). There is uncertainty over the best treatment regimen for hepatorenal syndrome.

OBJECTIVES

To compare the benefits and harms of different treatments for hepatorenal syndrome in people with decompensated liver cirrhosis.

SEARCH METHODS

We searched CENTRAL, MEDLINE, Embase, Science Citation Index Expanded, World Health Organization International Clinical Trials Registry Platform, and trial registers until December 2018 to identify randomised clinical trials on hepatorenal syndrome in people with cirrhosis.

SELECTION CRITERIA

We included only randomised clinical trials (irrespective of language, blinding, or publication status) in adults with cirrhosis and hepatorenal syndrome. We excluded randomised clinical trials in which participants had previously undergone liver transplantation.

DATA COLLECTION AND ANALYSIS

Two authors independently identified eligible trials and collected data. The outcomes for this review included mortality, serious adverse events, any adverse events, resolution of hepatorenal syndrome, liver transplantation, and other decompensation events. We performed a network meta-analysis with OpenBUGS using Bayesian methods and calculated the odds ratio (OR), rate ratio, hazard ratio (HR), and mean difference (MD) with 95% credible intervals (CrI) based on an available-case analysis, according to National Institute of Health and Care Excellence Decision Support Unit guidance.

MAIN RESULTS

We included a total of 25 trials (1263 participants; 12 interventions) in the review. Twenty-three trials (1185 participants) were included in one or more outcomes. All the trials were at high risk of bias, and all the evidence was of low or very low certainty. The trials included participants with liver cirrhosis of varied aetiologies as well as a mixture of type I hepatorenal syndrome only, type II hepatorenal syndrome only, or people with both type I and type II hepatorenal syndrome. Participant age ranged from 42 to 60 years, and the proportion of females ranged from 5.8% to 61.5% in the trials that reported this information. The follow-up in the trials ranged from one week to six months. Overall, 59% of participants died during this period and about 35% of participants recovered from hepatorenal syndrome. The most common interventions compared were albumin plus terlipressin, albumin plus noradrenaline, and albumin alone.There was no evidence of a difference in mortality (22 trials; 1153 participants) at maximal follow-up between the different interventions. None of the trials reported health-related quality of life. There was no evidence of differences in the proportion of people with serious adverse events (three trials; 428 participants), number of participants with serious adverse events per participant (two trials; 166 participants), proportion of participants with any adverse events (four trials; 402 participants), the proportion of people who underwent liver transplantation at maximal follow-up (four trials; 342 participants), or other features of decompensation at maximal follow-up (one trial; 466 participants). Five trials (293 participants) reported number of any adverse events, and five trials (219 participants) reported treatment costs. Albumin plus noradrenaline had fewer numbers of adverse events per participant (rate ratio 0.51, 95% CrI 0.28 to 0.87). Eighteen trials (1047 participants) reported recovery from hepatorenal syndrome (as per definition of hepatorenal syndrome). In terms of recovery from hepatorenal syndrome, in the direct comparisons, albumin plus midodrine plus octreotide and albumin plus octreotide had lower recovery from hepatorenal syndrome than albumin plus terlipressin (HR 0.04; 95% CrI 0.00 to 0.25 and HR 0.26, 95% CrI 0.07 to 0.80 respectively). There was no evidence of differences between the groups in any of the other direct comparisons. In the network meta-analysis, albumin and albumin plus midodrine plus octreotide had lower recovery from hepatorenal syndrome compared with albumin plus terlipressin.

FUNDING

two trials were funded by pharmaceutical companies; five trials were funded by parties who had no vested interest in the results of the trial; and 18 trials did not report the source of funding.

AUTHORS' CONCLUSIONS

Based on very low-certainty evidence, there is no evidence of benefit or harm of any of the interventions for hepatorenal syndrome with regards to the following outcomes: all-cause mortality, serious adverse events (proportion), number of serious adverse events per participant, any adverse events (proportion), liver transplantation, or other decompensation events. Low-certainty evidence suggests that albumin plus noradrenaline had fewer 'any adverse events per participant' than albumin plus terlipressin. Low- or very low-certainty evidence also found that albumin plus midodrine plus octreotide and albumin alone had lower recovery from hepatorenal syndrome compared with albumin plus terlipressin.Future randomised clinical trials should be adequately powered; employ blinding, avoid post-randomisation dropouts or planned cross-overs (or perform an intention-to-treat analysis); and report clinically important outcomes such as mortality, health-related quality of life, adverse events, and recovery from hepatorenal syndrome. Albumin plus noradrenaline and albumin plus terlipressin appear to be the interventions that should be compared in future trials.

Synthesizing existing evidence to design future trials: survey of methodologists from European institutions

Background

‘Conditional trial design’ is a framework for efficiently planning new clinical trials based on a network of relevant existing trials. The framework considers whether new trials are required and how the existing evidence can be used to answer the research question and plan future research. The potential of this approach has not been fully realized.

Methods

We conducted an online survey among trial statisticians, methodologists, and users of evidence synthesis research using referral sampling to capture opinions about the conditional trial design framework and current practices among clinical researchers. The questions included in the survey were related to the decision of whether a meta-analysis answers the research question, the optimal way to synthesize available evidence, which relates to the acceptability of network meta-analysis, and the use of evidence synthesis in the planning of new studies.

Results

In total, 76 researchers completed the survey. Two out of three survey participants (65%) were willing to possibly or definitely consider using evidence synthesis to design a future clinical trial and around half of the participants would give priority to such a trial design. The median rating of the frequency of using such a trial design was 0.41 on a scale from 0 (never) to 1 (always). Major barriers to adopting conditional trial design include the current regulatory paradigm and the policies of funding agencies and sponsors.

Conclusions

Participants reported moderate interest in using evidence synthesis methods in the design of future trials. They indicated that a major paradigm shift is required before the use of network meta-analysis is regularly employed in the design of trials.

Electronic supplementary material

The online version of this article (10.1186/s13063-019-3449-6) contains supplementary material, which is available to authorized users.

The Value of Further Research: The Added Value of Individual-Participant Level Data

Judgements based on average cost-effectiveness estimates may disguise significant heterogeneity in net health outcomes. Decisions about coverage of new interventions are often more efficient when they consider between-patient heterogeneity, which is usually operationalized as different selections for different subgroups. While most model-based cost-effectiveness studies are populated with aggregated-level sub-group estimates, individual-level data are recognized as the best source of evidence to produce unbiased and efficient estimates to explore this heterogeneity. This paper extends a previously published framework to assesses the added value of having access to individual-level data, compared to using aggregate-level data only, in the absence/presence of mutually exclusive population subgroups. Supported by a case study on the cost-effectiveness of interventions to increase uptake of smoke-alarms, the extended framework provided a quantification of the benefits foregone of not using individual-level data, pointed to the optimal number of subgroups and where further research should be undertaken. Although not indicating changes in reimbursement decisions, results showed that irrespective of using aggregate or individual-level data, no substantial additional gains are obtained if more than two subgroups are taken into account. However, depending on the evidence type used, different subgroups are revealed as warranting larger research funds. The use of individual-level data, rather than aggregate, may however influence not only the extent to which an appropriate understanding of existing heterogeneity is attained, but, more importantly, it may shape approval decisions for particular population subgroups and judgements of future research.

Development of an interactive web-based tool to conduct and interrogate meta-analysis of diagnostic test accuracy studies: MetaDTA

Background

Recommended statistical methods for meta-analysis of diagnostic test accuracy studies require relatively complex bivariate statistical models which can be a barrier for non-statisticians. A further barrier exists in the software options available for fitting such models. Software accessible to non-statisticians, such as RevMan, does not support the fitting of bivariate models thus users must seek statistical support to use R, Stata or SAS. Recent advances in web technologies make analysis tool creation much simpler than previously. As well as accessibility, online tools can allow tailored interactivity not found in other packages allowing multiple perspectives of data to be displayed and information to be tailored to the user’s preference from a simple interface. We set out to: (i) Develop a freely available web-based “point and click” interactive tool which allows users to input their DTA study data and conduct meta-analyses for DTA reviews, including sensitivity analyses. (ii) Illustrate the features and benefits of the interactive application using an existing DTA meta-analysis for detecting dementia.

Methods

To create our online freely available interactive application we used the existing R packages lme4 and Shiny to analyse the data and create an interactive user interface respectively.

Results

MetaDTA, an interactive online application was created for conducting meta-analysis of DTA studies. The user interface was designed to be easy to navigate having different tabs for different functions. Features include the ability for users to enter their own data, customise plots, incorporate quality assessment results and quickly conduct sensitivity analyses. All plots produced can be exported as either .png or .pdf files to be included in report documents. All tables can be exported as .csv files.

Conclusions

MetaDTA, is a freely available interactive online application which meta-analyses DTA studies, plots the summary ROC curve, incorporates quality assessment results and allows for sensitivity analyses to be conducted in a timely manner. Due to the rich feature-set and user-friendliness of the software it should appeal to a wide audience including those without specialist statistical knowledge. We encourage others to create similar applications for specialist analysis methods to encourage broader uptake which in-turn could improve research quality.

AD-8 for detection of dementia across a variety of healthcare settings

BACKGROUND

Dementia assessment often involves initial screening, using a brief tool, followed by more detailed assessment where required. The AD-8 is a short questionnaire, completed by a suitable 'informant' who knows the person well. AD-8 is designed to assess change in functional performance secondary to cognitive change.

OBJECTIVES

To determine the diagnostic accuracy of the informant-based AD-8 questionnaire, in detection of all-cause (undifferentiated) dementia in adults. Where data were available, we described the following: the diagnostic accuracy of the AD-8 at various predefined threshold scores; the diagnostic accuracy of the AD-8 for each healthcare setting and the effects of heterogeneity on the reported diagnostic accuracy of the AD-8.

SEARCH METHODS

We searched the following sources on 27 May 2014, with an update to 7 June 2018: ALOIS (Cochrane Dementia and Cognitive Improvement Group), MEDLINE (Ovid SP), Embase (Ovid SP), PsycINFO (Ovid SP), BIOSIS Previews (Thomson Reuters Web of Science), Web of Science Core Collection (includes Conference Proceedings Citation Index) (Thomson Reuters Web of Science), CINAHL (EBSCOhost) and LILACS (BIREME). We checked reference lists of relevant studies and reviews, used searches of known relevant studies in PubMed to track related articles, and contacted research groups conducting work on the AD-8 to try to find additional studies. We developed a sensitive search strategy and used standardised database subject headings as appropriate. Foreign language publications were translated.

SELECTION CRITERIA

We selected those studies which included the AD-8 to assess for the presence of dementia and where dementia diagnosis was confirmed with clinical assessment. We only included those studies where the AD-8 was used as an informant assessment. We made no exclusions in relation to healthcare setting, language of AD-8 or the AD-8 score used to define a 'test positive' case.

DATA COLLECTION AND ANALYSIS

We screened all titles generated by electronic database searches, and reviewed abstracts of potentially relevant studies. Two independent assessors checked full papers for eligibility and extracted data. We extracted data into two-by-two tables to allow calculation of accuracy metrics for individual studies. We then created summary estimates of sensitivity, specificity and likelihood ratios using the bivariate approach and plotting results in receiver operating characteristic (ROC) space. We determined quality assessment (risk of bias and applicability) using the QUADAS-2 tool.

MAIN RESULTS

From 36 papers describing AD-8 test accuracy, we included 10 papers. We utilised data from nine papers with 4045 individuals, 1107 of whom (27%) had a clinical diagnosis of dementia. Pooled analysis of seven studies, using an AD-8 informant cut-off score of two, indicated that sensitivity was 0.92 (95% confidence interval (CI) 0.86 to 0.96); specificity was 0.64 (95% CI 0.39 to 0.82); the positive likelihood ratio was 2.53 (95% CI 1.38 to 4.64); and the negative likelihood ratio was 0.12 (95% CI 0.07 to 0.21). Pooled analysis of five studies, using an AD-8 informant cut-off score of three, indicated that sensitivity was 0.91 (95% CI 0.80 to 0.96); specificity was 0.76 (95% CI 0.57 to 0.89); the positive likelihood ratio was 3.86 (95% CI 2.03 to 7.34); and the negative likelihood ratio was 0.12 (95% CI 0.06 to 0.24).Four studies were conducted in community settings; four were in secondary care (one in the acute hospital); and one study was in primary care. The AD-8 has a higher relative sensitivity (1.11, 95% CI 1.02 to 1.21), but lower relative specificity (0.51, 95% CI 0.23 to 1.09) in secondary care compared to community care settings.There was heterogeneity across the included studies. Dementia prevalence rate varied from 12% to 90% of included participants. The tool was also used in various different languages. Among all the included studies there was evidence of risk of bias. Issues included the selection of participants, conduct of index test, and flow of assessment procedures.

AUTHORS' CONCLUSIONS

The high sensitivity of the AD-8 suggests it can be used to identify adults who may benefit from further specialist assessment and diagnosis, but is not a diagnostic test in itself. This pattern of high sensitivity and lower specificity is often suited to a screening test. Test accuracy varies by setting, however data in primary care and acute hospital settings are limited. This review identified significant heterogeneity and risk of bias, which may affect the validity of its summary findings.

Use of a random effects meta-analysis in the design and analysis of a new clinical trial

In designing a randomized controlled trial, it has been argued that trialists should consider existing evidence about the likely intervention effect. One approach is to form a prior distribution for the intervention effect based on a meta‐analysis of previous studies and then power the trial on its ability to affect the posterior distribution in a Bayesian analysis. Alternatively, methods have been proposed to calculate the power of the trial to influence the “pooled” estimate in an updated meta‐analysis. These two approaches can give very different results if the existing evidence is heterogeneous, summarised using a random effects meta‐analysis. We argue that the random effects mean will rarely represent the trialist's target parameter, and so, it will rarely be appropriate to power a trial based on its impact upon the random effects mean. Furthermore, the random effects mean will not generally provide an appropriate prior distribution. More appropriate alternatives include the predictive distribution and shrinkage estimate for the most similar study. Consideration of the impact of the trial on the entire random effects distribution might sometimes be appropriate. We describe how beliefs about likely sources of heterogeneity have implications for how the previous evidence should be used and can have a profound impact on the expected power of the new trial. We conclude that the likely causes of heterogeneity among existing studies need careful consideration. In the absence of explanations for heterogeneity, we suggest using the predictive distribution from the meta‐analysis as the basis for a prior distribution for the intervention effect.

Planning a future randomized clinical trial based on a network of relevant past trials

Background

The important role of network meta-analysis of randomized clinical trials in health technology assessment and guideline development is increasingly recognized. This approach has the potential to obtain conclusive results earlier than with new standalone trials or conventional, pairwise meta-analyses.

Methods

Network meta-analyses can also be used to plan future trials. We introduce a four-step framework that aims to identify the optimal design for a new trial that will update the existing evidence while minimizing the required sample size. The new trial designed within this framework does not need to include all competing interventions and comparisons of interest and can contribute direct and indirect evidence to the updated network meta-analysis. We present the method by virtually planning a new trial to compare biologics in rheumatoid arthritis and a new trial to compare two drugs for relapsing-remitting multiple sclerosis.

Results

A trial design based on updating the evidence from a network meta-analysis of relevant previous trials may require a considerably smaller sample size to reach the same conclusion compared with a trial designed and analyzed in isolation. Challenges of the approach include the complexity of the methodology and the need for a coherent network meta-analysis of previous trials with little heterogeneity.

Conclusions

When used judiciously, conditional trial design could significantly reduce the required resources for a new study and prevent experimentation with an unnecessarily large number of participants.

Electronic supplementary material

The online version of this article (10.1186/s13063-018-2740-2) contains supplementary material, which is available to authorized users.

Network meta-analysis of diagnostic test accuracy studies identifies and ranks the optimal diagnostic tests and thresholds for health care policy and decision-making

OBJECTIVES

Network meta-analyses (NMA) have extensively been used to compare the effectiveness of multiple interventions for health care policy and decision-making. However, methods for evaluating the performance of multiple diagnostic tests are less established. In a decision-making context, we are often interested in comparing and ranking the performance of multiple diagnostic tests, at varying levels of test thresholds, in one simultaneous analysis.

STUDY DESIGN AND SETTING

Motivated by an example of cognitive impairment diagnosis following stroke, we synthesized data from 13 studies assessing the efficiency of two diagnostic tests: Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA), at two test thresholds: MMSE <25/30 and <27/30, and MoCA <22/30 and <26/30. Using Markov chain Monte Carlo (MCMC) methods, we fitted a bivariate network meta-analysis model incorporating constraints on increasing test threshold, and accounting for the correlations between multiple test accuracy measures from the same study.

RESULTS

We developed and successfully fitted a model comparing multiple tests/threshold combinations while imposing threshold constraints. Using this model, we found that MoCA at threshold <26/30 appeared to have the best true positive rate, whereas MMSE at threshold <25/30 appeared to have the best true negative rate.

CONCLUSION

The combined analysis of multiple tests at multiple thresholds allowed for more rigorous comparisons between competing diagnostics tests for decision making.

Influence of El Niño on atmospheric CO2 over the tropical Pacific Ocean: Findings from NASA's OCO-2 mission

Spaceborne observations of carbon dioxide (CO₂) from the Orbiting Carbon Observatory-2 are used to characterize the response of tropical atmospheric CO₂ concentrations to the strong El Niño event of 2015-2016. Although correlations between the growth rate of atmospheric CO₂ concentrations and the El Niño-Southern Oscillation are well known, the magnitude of the correlation and the timing of the responses of oceanic and terrestrial carbon cycle remain poorly constrained in space and time. We used space-based CO₂ observations to confirm that the tropical Pacific Ocean does play an early and important role in modulating the changes in atmospheric CO₂ concentrations during El Niño events-a phenomenon inferred but not previously observed because of insufficient high-density, broad-scale CO₂ observations over the tropics.

Glycemic Control During Continuous Subcutaneous Insulin Infusion Versus Multiple Daily Insulin Injections in Type 2 Diabetes: Individual Patient Data Meta-analysis and Meta-regression of Randomized Controlled Trials

OBJECTIVE

To compare glycemic control during continuous subcutaneous insulin infusion (CSII) and multiple daily insulin injections (MDI) in people with type 2 diabetes to identify patient characteristics that determine those best treated by CSII.

RESEARCH DESIGN AND METHODS

Randomized controlled trials were selected comparing HbA_1c during CSII versus MDI in people with type 2 diabetes. Data sources included Cochrane database and Ovid Medline. We explored patient-level determinants of final HbA_1c level and insulin dose using Bayesian meta-regression models of individual patient data and summary effects using two-step meta-analysis. Hypoglycemia data were unavailable.

RESULTS

Five trials were identified, with 287 patients randomized to receive MDI and 303 to receive CSII. Baseline HbA_1c was the best determinant of final HbA_1c: HbA_1c difference (%) = 1.575 - (0.216 [95% credible interval 0.371-0.043] × baseline HbA_1c) for all trials, but with largest effect in the trial with prerandomization optimization of control. Baseline insulin dose was best predictor of final insulin dose: insulin dose difference (units/kg) = 0.1245 - (0.382 [0.510-0.254] × baseline insulin dose). Overall HbA_1c difference was -0.40% (-0.86 to 0.05 [-4.4 mmol/mol (-9.4 to 0.6)]). Overall insulin dose was reduced by -0.25 units/kg (-0.31 to -0.19) (26% reduction on CSII), and by -24.0 units/day (-30.6 to -17.5). Mean weight did not differ between treatments (0.08 kg [-0.33 to 0.48]).

CONCLUSIONS

CSII achieves better glycemic control than MDI in people with poorly controlled type 2 diabetes, with ∼26% reduction in insulin requirements and no weight change. The best effect is in those worst controlled and with the highest insulin dose at baseline.