Português | English

ISSN (Print): 2359-4802 | ISSN (Online): 2359-5647

Edition: 29.4 - 13 Article(s)

Back to Sumary

Point of View http://www.dx.doi.org/10.5935/2359-4802.20160041

Understanding meta-epidemiological studies

Compreendendo os estudos de meta-epidemiologia

Leonardo Silva Roever Borges

Universidade Federal de Uberlândia, Uberlândia, MG - Brazil

Corresponding author

Leonardo Silva Roever Borges
Universidade Federal de Uberlândia - Departamento de Pesquisa Clínica
Rua Rafael Rinaldi, 431
Postal Code 38400384, Uberlândia, MG - Brazil
E-mail: leonardoroever@hotmail.com

Received in 2/2/2016
Accepted em 8/8/2016
Reviewed in 2/6/2016


The concept of meta-epidemiology has been introduced because of the methodological limitations of the systematic review of clinical trials of intervention. Meta-epidemiology has moved from a statistical method to a new methodology to close gaps between evidence and practice, controlling the potential biases in quantitative systematic review and drawing appropriate evidence to establish evidence-based guidelines. Network meta-epidemiology has been suggested to overcome some limitations of meta-epidemiology. This review aims to clarify the concept and major methods to conduct a meta-epidemiological study.

Keywords: Evidence-Based Practice / statistics & numeric data; Evidence-Based Medicine; Epidemiology.



Owing to the recent advances to overcome the limitations of systematic review (SR), 'meta-epidemiology' has been proposed as a new methodology aimed at investigating the conflicting results of a SR with the same hypothesis, as well as the problems inherent in the research process, such as heterogeneity, publication bias, allocation concealment or post-allocation patient blinding, which make it difficult to provide a rationale for the results of a SR and drawing of appropriate conclusions.1-2

The term 'meta-epidemiology' can be defined as a 'statistical method' to analyze the influence of qualitative problems in randomized clinical trials and their confounding variables. In randomized clinical trials, the topics of traditional epidemiological studies are the individuals, while the topics of meta-epidemiological studies are the original articles of randomized clinical trials and observational studies.3-5 Table 1 shows the characteristics of meta-epidemiological studies.



Meta-epidemiology is based on the combination of two concepts: epidemiology and metaanalysis. To adjust the purposes of those two concepts, meta-epidemiology strains to: (A) describe the distribution of the research evidence for a specific question; (B) examine heterogeneity and risk factors associated; and (C) control the biases between studies and summarize the research evidence. Considering such model, several methods, such as meta-regression, imputation, lack of informational odds ratio, double statistical models, have been tested, the term 'meta-epidemiology' being thus introduced.3,6,7 Meta-epidemiological studies analyze the articles of randomized clinical trials and observational studies, meta-meta-epidemiologic studies analyze the meta-epidemiologic studies, and network meta-epidemiology analyzes the metaanalyses of published randomized clinical trials, whose data were analyzed with a statistical method valid for indirect comparisons or network metaanalysis, also called multiple-treatment or mixed-treatment comparison metaanalysis. Table 2 shows the major characteristics of meta-epidemiological, meta-meta-epidemiological and network meta-epidemiological studies.3



Recently there was a trend towards the application of the potentials of confounding meta-variables, such as genotype, study design, number of participants, generation of allocation sequence, allocation, concealment, blinding, placebo-control vs. no treatment control, exclusion of patients, randomization, effect size, single-center vs. multicenter study, and experimental vs. observational study.8

Meta-epidemiological studies have limitations: study results allow for a dichotomous analysis and continuous results cannot be managed; if the number of study subjects is reduced, the statistical power is limited; and indirect comparisons cannot be applied. Aiming at overcoming such limitations, the term 'network meta-epidemiology' has been proposed to emphasize how to make direct comparisons when several types of interventions are assessed. Therefore, developing research tools, Copas parametric model, graphs presented and published items are paramount for their conduction.9

In a study assessing 31 metaanalyses on cardiovascular biomarkers (C-reactive protein, non-HDL-cholesterol, lipoprotein(a), post-load glucose, fibrinogen, B-type natriuretic peptide and troponins), the prognostic effect was significantly stronger in observational studies than in randomized clinical trials. Cardiovascular biomarkers often have less promising results in the evidence derived from randomized clinical trials than from observational studies.10



This topic is extremely new, generating new questions that fill the gaps in this type of investigation. In addition, this challenging topic requires new methodologies for science advance.



Conception and design of the research:Borges LSR. Acquisition of data: Borges LSR. Analysis and interpretation of the data: Borges LSR. Writing of the manuscript: Borges LSR. Critical revision of the manuscript for intellectual content: Borges LSR.

Potential Conflict of Interest

No potential conflict of interest relevant to this article was reported.

Sources of Funding

There were no external funding sources for this study.

Study Association

This study is not associated with any thesis or dissertation work.



1. Gluud LL. Bias in clinical intervention research. Am J Epidemiol. 2006;163(3):493-501.

2. Hopewell S, Loudon K, Clarke MJ, Oxman AD, Dickersin K. Publication bias in clinical trials due to statistical significance or direction of trial results. Cochrane Database Syst Rev. Jan 21;(1):MR000006.

3. Trinquart L, Dechartres A, Ravaud P. Commentary: meta-epidemiology, meta-meta-epidemiology or network meta-epidemiology? Int J Epidemiol. 2013;42(4):1131-3.

4. Bae ME. Meta-epidemiology. Epidemiol Health. 2014;36:e2014019.

5. Wood L, Egger M, Gluud LL, Schulz KF, Jüni P, Altman DG, et al. Empirical evidence of bias in treatment effect estimates in controlled trials with different interventions and outcomes: meta-epidemiological study. BMJ. 2008;336(7644):601-5.

6. Le Lorier J, Grégoire G. Meta-analysis and the meta-epidemiology of clinical research. Comments on paper by author of editorial were unwarranted. BMJ. 1998;316(7127):311-2.

7. Siersma V, Als-Nielsen B, Chen W, Hilden J, Gluud LL, Gluud C. Multivariable modelling for meta-epidemiological assessment of the association between trial quality and treatment effects estimated in randomized clinical trials. Stat Med. 2007;26(14):2745-58.

8. Kjaergard LL, Villumsen J, Gluud C. Reported methodologic quality and discrepancies between large and small randomized trials in metaanalyses. Ann Intern Med. 2001;135(11):982-9.

9. Chaimani A, Vasiliadis HS, Pandis N, Schmid CH, Welton NJ, Salanti G. Effects of study precision and risk of bias in networks of interventions: a network meta-epidemiological study. Int J Epidemiol. 2013;42(4):1120-31.

10. Tzoulaki I, Siontis KC, Ioannidis. Prognostic effect size of cardiovascular biomarkers in datasets from observational studies versus randomised trials: meta-epidemiology study. BMJ. 2011;7:343,d6829.