Using structural equation modeling for network meta-analysis

Abstract

BackgroundNetwork meta-analysis overcomes the limitations of traditional pair-wise meta-analysis by incorporating all available evidence into a general statistical framework for simultaneous comparisons of several treatments. Currently, network meta-analyses are undertaken either within the Bayesian hierarchical linear models or frequentist generalized linear mixed models. Structural equation modeling (SEM) is a statistical method originally developed for modeling causal relations among observed and latent variables. As random effect is explicitly modeled as a latent variable in SEM, it is very flexible for analysts to specify complex random effect structure and to make linear and nonlinear constraints on parameters. The aim of this article is to show how to undertake a network meta-analysis within the statistical framework of SEM.MethodsWe used an example dataset to demonstrate the standard fixed and random effect network meta-analysis models can be easily implemented in SEM. It contains results of 26 studies that directly compared three treatment groups A, B and C for prevention of first bleeding in patients with liver cirrhosis. We also showed that a new approach to network meta-analysis based on the technique of unrestricted weighted least squares (UWLS) method can also be undertaken using SEM.ResultsFor both the fixed and random effect network meta-analysis, SEM yielded similar coefficients and confidence intervals to those reported in the previous literature. The point estimates of two UWLS models were identical to those in the fixed effect model but the confidence intervals were greater. This is consistent with results from the traditional pairwise meta-analyses. Comparing to UWLS model with common variance adjusted factor, UWLS model with unique variance adjusted factor has greater confidence intervals when the heterogeneity was larger in the pairwise comparison. The UWLS model with unique variance adjusted factor reflects the difference in heterogeneity within each comparison.ConclusionSEM provides a very flexible framework for univariate and multivariate meta-analysis, and its potential as a powerful tool for advanced meta-analysis is still to be explored.Electronic supplementary materialThe online version of this article (doi:10.1186/s12874-017-0390-9) contains supplementary material, which is available to authorized users.

Beverage and Food Fragrance Biotechnology, Novel Applications, Sensory and Sensor Techniques: An Overview

“…In 2018 Vilela and co-workers [191] applied an innovative multivariate analysis technique, structural equation modelling (SEM) for identifying wine sensory characteristics of monovarietal wine from the Portuguese Vinho Verde wine Demarcated Region. SEM is a statistical technique used to decrease the number of perceived variables (sensory descriptors) into a lesser number of latent (not observed) variables by examining the covariances between the observed variables [192,193]. This data analysis technique that can be viewed as a combination of factor analysis and multiple regression analysis had its origin at the beginning of the 20th century, from the important works of Charles Spearman [194], an English psychologist known to work in statistics, as a pioneer of factor analysis, and for Spearman's rank correlation coefficient.…”

Some reflections on combining meta‐analysis and structural equation modeling

“…SEM‐based meta‐analysis was later extended to multivariate meta‐analysis and 3‐level meta‐analysis . Recently, this approach has been extended to address the heterogeneity of effect sizes and to network meta‐analysis . SEM‐based meta‐analysis allows researchers to address many complicated research questions that are beyond the scope of a conventional meta‐analysis.…”

“…79 Recently, this approach has been extended to address the heterogeneity of effect sizes 80 and to network metaanalysis. 81 SEM-based meta-analysis allows researchers to address many complicated research questions that are beyond the scope of a conventional meta-analysis. For example, Shadish and others 26,82 were interested in testing causal models on study characteristics and several population effect sizes.…”

Evaluating network meta‐analysis and inconsistency using arm‐parameterized model in structural equation modeling

“…Cheung has shown how traditional and multivariate meta‐analysis can be carried out in SEM . For NMA, Tu and Wu recently demonstrated that the Lu‐Ades model can be implemented in SEM . They also extended the unrestricted weighted least squares (UWLS) model proposed in pairwise meta‐analysis into NMA .…”