A matrix‐based method of moments for fitting the multivariate random effects model for meta‐analysis and meta‐regression

Jackson, Dan; White, Ian R.; Riley, Richard D

doi:10.1002/bimj.201200152

Cited by 87 publications

(134 citation statements)

References 16 publications

Supporting

Mentioning

129

Contrasting

Order By: Relevance

“…Next, as an indirect comparison, we performed a network meta-analysis. The random effects network meta-analysis was performed using the multivariate meta-analysis (mvmeta) routine in the statistical software STATA 13 (StataCorp LLC, College Station, TX, USA) [14,15], and the results of direct and indirect comparisons were integrated. Furthermore, we examined treatment hierarchy using the surface under the cumulative ranking curve (SUCRA).…”

Section: Discussionmentioning

confidence: 99%

Effects of sulfonylurea treatment on blood plasminogen activator inhibitor-1 levels in patients with type 2 diabetes mellitus: A network meta-analysis

Ida¹,

Murata²,

Kaneko³

2018

J Diab Res Clin Met

View full text Add to dashboard Cite

Background: To compare the effects ofthree types of sulfonylureas (glibenclamide, gliclazide, and glimepiride) on blood plasminogen activator inhibitor-1 levels in patients with type2 diabetes mellitus, a network meta-analysis of randomized controlled trialswas performed. Methods: A literature search using MEDLINE, the Cochrane Central Register of Controlled Trials, and ClinicalTrials.gov was conducted. Randomized controlled trialsin which the effects of sulfonylureas on blood plasminogen activator inhibitor-1 levels in patients with type 2 diabetes mellitus were evaluated were includes. Outcome assessment included standardized mean differences and 95% confidence intervals. Results: Twelve randomized controlled trials (1,050 subjects) met the inclusion criteria and were included in the network meta-analysis. No significant difference was observedin blood plasminogen activator inhibitor-1 levels after using a placebo compared with those after using glibenclamide, gliclazide, and glimepiride. Blood plasminogen activator inhibitor-1levels were significantly lower after using gliclazide than after using glimepiride (standardized mean difference: -0.52; 95% confidence interval: -0.99%-0.44%). However, no significant difference was observed in blood plasminogen activator inhibitor-1 levels after using glibenclamide compared with those after using gliclazide and glime piride. Conclusions: Regarding the use of sulfonylureas for treating patients with type 2 diabetes mellitus, gliclazide may be preferable because of low blood plasminogen activator inhibitor-1 levels after its use. However, few studies have been published on the use of gliclazide, and the quality of these studies has been generally poor; thus, the results of this study should be interpreted with caution.

show abstract

Section: Discussionmentioning

confidence: 99%

Effects of sulfonylurea treatment on blood plasminogen activator inhibitor-1 levels in patients with type 2 diabetes mellitus: A network meta-analysis

Ida¹,

Murata²,

Kaneko³

2018

J Diab Res Clin Met

View full text Add to dashboard Cite

show abstract

“…In addition, a test of homogeneity does not shed light what the cause of heterogeneity might be. A valuable extension of random effects meta-analysis is meta-regression [14–15, 23–24], a meta-analysis that relates the size of the effect to one or more characteristics of the studies involved. Meta-regression can be useful for exploring sources of heterogeneity and for offering important insights as to the nature of interventions, of populations, or both.…”

Section: Meta-analysis In Clinical Trialsmentioning

confidence: 99%

Meta-analysis in clinical trials revisited

DerSimonian

Laird

2015

Contemporary Clinical Trials

2,000

1,368

View full text Add to dashboard Cite

In this paper, we revisit a 1986 article we published in this Journal, Meta-Analysis in Clinical Trials, where we introduced a random-effect model to summarize the evidence about treatment efficacy from a number of related clinical trials. Because of its simplicity and ease of implementation, our approach has been widely used (with more than 12,000 citations to date) and the “DerSimonian and Laird method” is now often referred to as the ‘standard approach’ or a ‘popular’ method for meta-analysis in medical and clinical research. The method is especially useful for providing an overall effect estimate and for characterizing the heterogeneity of effects across a series of studies. Here, we review the background that led to the original 1986 article, briefly describe the random-effects approach for meta-analysis, explore its use in various settings and trends over time and recommend a refinement to the method using a robust variance estimator for testing overall effect. We conclude with a discussion of repurposing the method for Big Data meta-analysis and Genome Wide Association Studies for studying the importance of genetic variants in complex diseases.

show abstract

“…In this random effects model, the true logit sensitivities and logit specificities are assumed to be bivariate normally distributed with common mean values, μ A and μ B , and between‐study variance‐covariance matrix, Σ , where

τ_{A}^{2}

and

τ_{B}^{2}

are the between‐study variances and τ AB = ρ·τ A ·τ B is the covariance between the logit sensitivity and specificity across studies ( ρ is their between‐study correlation). This model can be fitted using, for example, the multivariate method of moments procedure or restricted maximum likelihood to give the summary estimates,

{\overset{truê}{μ}}_{A}

and

{\overset{truê}{μ}}_{B}

, and the estimated between‐study variance matrix,

\overset{Σ}{true}

.…”

Section: Meta‐analysis Models For Diagnostic Test Accuracymentioning

confidence: 99%

Guidance for deriving and presenting percentage study weights in meta‐analysis of test accuracy studies

Burke

Ensor

Snell

et al. 2018

Research Synthesis Methods

Self Cite

View full text Add to dashboard Cite

Percentage study weights in meta-analysis reveal the contribution of each study toward the overall summary results and are especially important when some studies are considered outliers or at high risk of bias. In meta-analyses of test accuracy reviews, such as a bivariate meta-analysis of sensitivity and specificity, the percentage study weights are not currently derived. Rather, the focus is on representing the precision of study estimates on receiver operating characteristic plots by scaling the points relative to the study sample size or to their standard error. In this article, we recommend that researchers should also provide the percentage study weights directly, and we propose a method to derive them based on a decomposition of Fisher information matrix. This method also generalises to a bivariate meta-regression so that percentage study weights can also be derived for estimates of study-level modifiers of test accuracy. Application is made to two meta-analyses examining test accuracy: one of ear temperature for diagnosis of fever in children and the other of positron emission tomography for diagnosis of Alzheimer's disease. These highlight that the percentage study weights provide important information that is otherwise hidden if the presentation only focuses on precision based on sample size or standard errors. Software code is provided for Stata, and we suggest that our proposed percentage weights should be routinely added on forest and receiver operating characteristic plots for sensitivity and specificity, to provide transparency of the contribution of each study toward the results. This has implications for the PRISMA-diagnostic test accuracy guidelines that are currently being produced.

show abstract

A matrix‐based method of moments for fitting the multivariate random effects model for meta‐analysis and meta‐regression

Cited by 87 publications

References 16 publications

Effects of sulfonylurea treatment on blood plasminogen activator inhibitor-1 levels in patients with type 2 diabetes mellitus: A network meta-analysis

Effects of sulfonylurea treatment on blood plasminogen activator inhibitor-1 levels in patients with type 2 diabetes mellitus: A network meta-analysis

Meta-analysis in clinical trials revisited

Guidance for deriving and presenting percentage study weights in meta‐analysis of test accuracy studies

Contact Info

Product

Resources

About