Meta-epidemiology

Article information

Epidemiol Health. 2014;36.e2014019
Publication date (electronic) : 2014 September 25
doi : https://doi.org/10.4178/epih/e2014019
Department of Preventive Medicine, Jeju National University School of Medicine, JeJu, Korea
Correspondence: Jong-Myon Bae Department of Preventive Medicine, Jeju National University School of Medicine, 102 Jejudaehak-ro, Jeju 690-756, Korea Tel: +82-64-755-5567, Fax: +82-64-725-2593, E-mail: jmbae@jejunu.ac.kr
Received 2014 August 26; Accepted 2014 September 25.

Abstract

The concept of meta-epidemiology has been introduced with considering the methodological limitations of systematic review for intervention trials. The paradigm of meta-epidemiology has shifted from a statistical method into a new methodology to close gaps between evidence and practice. Main interest of meta-epidemiology is to control potential biases in previous quantitative systematic reviews and draw appropriate evidences for establishing evidence-base guidelines. Nowadays, the network meta-epidemiology was suggested in order to overcome some limitations of meta-epidemiology. To activate meta-epidemiologic studies, implementation of tools for risk of bias and reporting guidelines such as the Consolidated Standards for Reporting Trials (CONSORT) should be done.

INTRODUCTION

To establish the best decision-making processes in healthcare service, new fields of study including evidence-based medicine, evidence-based guidelines, and evidence-based health policy have emerged [1]. In addition, for academic cultivation, a new research methodology called systematic reviews (SR) was proposed, under which, the existing generated evidences are systematically collected and evaluated for synthesis into more valid and useful evidence [2-5].

In particular, the success of SR methodology in resolving the controversy over administering beta-blockers in myocardial infarction patients [6] has firmly established its use in published original articles (OA) related to randomized-controlled clinical trials (RCTs) investigating the efficacy of new medicinal or procedural interventions [7,8].

However, given the fact that the subject of SR is OA, some arguments have shown that, even in the development stage, SR methodology cannot overcome inherent limitations of OAs [4, 9-13]. Noble [5] summarized existing suggestions about the advantages and disadvantages of SRs. As one of the breakthroughs for overcoming these limitations, a new terminology, meta-epidemiology, was introduced [14]. In light of this trend, this study aimed to investigate the background of emergence, definition, purposes and research outcomes from its practical applications of meta-epidemiology.

INTRODUCTION TO META-EPIDEMIOLOGY

Emerging background of meta-epidemiology

While studies using SR methodology were being actively published, some of these studies with the same research hypotheses began to yield conflicting results; moreover, additional problems were identified in the SR reasoning owing to critical limitations inherent in the OA itself [15-19]. Problems associated with errors that can occur while performing the SR research process, include heterogeneity [20,21], publication bias [9,22-24], and outcome reporting bias [25-27], among others; however, studies have shown that fundamental problems in the methodology associated with conducting RCT research, such as allocation concealment or post-allocation patient blinding, made it difficult to provide a rationale for SR results [18,28-35].

On the basis of these findings, attempts have been made to evaluate the quality of the OA more strictly when conducting a SR and to apply the meta-analysis upon appraising the results of specific items [30,36-39]. In particular, as concept separation and establishment for allocation concealment and post-allocation blinding took place, entering the year 2000 [40,41], a serious movement was seen to confirm the validity of study results from SRs on RCT studies that did not reflect these in the research plan [18,27,30,37,42-45]. With this background, diverse methods, such as mega-regression [4,46-48], imputation [15], informative missing odds ratio [26,49], two statistical models [33], and others were attempted and the term “meta-epidemiology” was introduced [14,16,36,50].

Definition and purpose of meta-epidemiology

According to Zhang [14], the term “meta-epidemiology” first appeared in published literature in 1997, in an editorial review by Naylor [50]. In 2002, Sterne et al. [16] attempted to make its meaning more explicit by referring to it as a “statistical method” for examining the influence of qualitative problems in RCTs. However, it is now in the process of being recognized as another epidemiological research methodology that controls meta-confounders, similar to traditional epidemiological research methodology that controls confounding variables [14,30,51]. Here, the difference from traditional epidemiology is that the subjects of traditional epidemiological studies are individuals, whereas those of meta-epidemiological studies are OAs that published the results of RCTs performed.

Thus, meta-epidemiology is based on the combination of two concepts: epidemiology and meta-analysis. To fit the purposes of these two concepts, meta-epidemiology strives to achieve the following: (1) to describe the distribution of research evidence for a specific question; (2) to examine heterogeneity and associated risk factors; and (3) to control bias across studies and summarize research evidence as appropriate [14].

Meta-epidemiology in the literature

Journals that have been published to date with the word “meta-epidemiology” in the journal title or research methodology section have been organized by year as shown in the Appendix 1. Published results with earnest applications of meta-epidemiology are seen after 2008, whereas initial studies were performed to control the influence of allocation concealment and post-allocation blinding [30,32]. As an example, Wood et al. [30] took 1,346 clinical trial papers, which were subjects in 146 published meta-analyses, and divided them, based on the existence or non-existence of allocation concealment and post-allocation blinding, and then re-analyzed them. It was shown that when these items were not properly followed, the subjective evaluations of their effects were exaggerated. More recently, a trend of applying potential meta-confounders, such as genotype [52], study design [36,53], and the number of participants [54], can be seen.

Meta-meta-epidemiology and network meta-epidemiology

As seen in the aforementioned background of meta-epidemiology, if meta-epidemiology was developed to control diverse SR results, then the word “meta-meta-epidemiology” can be proposed for diverse results from meta-epidemiology [55].

Nonetheless, meta-epidemiology has a few limitations [47,55, 56]. First, the study results that allow analysis are dichotomous and cannot handle continuous outcomes; second, with the reduced number of journals as study subjects, statistical power is limited; and third, indirect comparisons cannot be applied. With the goal of overcoming these limitations, Chaimani et al. [47] proposed the term “network meta-epidemiology.” Concurrently, the term “mixed treatment comparison meta-analysis” was introduced to emphasize the point about making (in)direct comparisons when multiple intervention types are introduced [7,46]. To execute this, the development of research conducting tools [57,58], Copas parametric model [59,60], graphs presented [61], and published items [62], are currently under way.

For easily distinguishing the concepts of meta-meta-epidemiology and network meta-epidemiology, derived from meta-epidemiology as seen above, Trinquart et al. [55] presented a mutual comparison table.

CONCLUSIONS AND SUGGESTIONS

Valid SR results must be present to develop good clinical diagnostic guidelines, which would ultimately contribute to the improvement of overall healthcare. This is the background and aim of the meta-epidemiology emergence. To obtain valid SR results, the key challenge is to improve the quality of RCTs that are the subject of analysis by SR [36,63]. To this end, some suggestions are being made.

First, development and distribution of standardized quality assessment tools is needed to accurately assess the risk of error occurrence [64]. Cochrane Collaboration has proposed a tool called risk of bias [65-67]; more active meta-epidemiologic studies are needed using such tools to investigate what influences are imposed on SR reasoning [36].

Second, there is a need to more clearly organize the concepts behind the terminologies used in RCT quality assessment [68]. This is because concepts such as allocation concealment and post-allocation blinding must be revised and disseminated in a unified manner to researchers, as well as existing research methodology textbooks.

Third, in order to accurately verify and interpret the RCT study results in SRs, reports must be made without overlooking any of the designated items [35]. Since there are some reports indicating that following the reporting guideline in The Consolidated Standards for Reporting Trials (CONSORT) improves the quality of journals [69,70], RCT researches would need to obey this guideline.

Fourth, there is an international movement to have the study registered in an open venue prior to conducting the study to prevent overlooking study results, because cases of conflicting between initial plans and final results were surfaced [71]. As there is also the advantage of reducing publication bias [24], there is a need to accept this wholeheartedly in accordance with the international trend.

Acknowledgements

This study was supported by the 2014 scientific promotion program funded by Jeju National University.

Notes

The author has no conflicts of interest to declare for this study.

SUPPLEMENTARY MATERIAL

Supplementary material is available at http://www.e-epih.org/.

References

1. Bae JM, Park BJ, Ahn YO. Perspectives of clinical epidemiology in Korea. J Korean Med Assoc 2013;56:718–723. (Korean).
2. Ioannidis JP, Schmid CH, Lau J. Meta-analysis in hematology and oncology. Hematol Oncol Clin North Am 2000;14:973–991.
3. Ahn HS, Kim HJ. An introduction to systematic review. J Korean Med Assoc 2014;57:49–59. (Korean).
4. Khoshdel A, Attia J, Carney SL. Basic concepts in meta-analysis: A primer for clinicians. Int J Clin Pract 2006;60:1287–1294.
5. Noble JH Jr. Meta-analysis: Methods, strengths, weaknesses, and political uses. J Lab Clin Med 2006;147:7–20.
6. Freemantle N, Cleland J, Young P, Mason J, Harrison J. beta Blockade after myocardial infarction: systematic review and meta regression analysis. BMJ 1999;318:1730–1737.
7. Salanti G, Higgins JP, Ades AE, Ioannidis JP. Evaluation of networks of randomized trials. Stat Methods Med Res 2008;17:279–301.
8. Bae JM. An overview of systematic reviews of diagnostic tests accuracy. Epidemiol Health 2014;36e2014016.
9. Egger M, Smith GD, Sterne JA. Uses and abuses of meta-analysis. Clin Med 2001;1:478–484.
10. Stevens KR, Ledbetter CA. Basics of evidence-based practice. Part 1: The nature of the evidenc. Semin Perioper Nurs 2000;9:91–97.
11. Michels KB. Quo vadis meta-analysis? A potentially dangerous tool if used without adequate rules. Important Adv Oncol 1992;:243–248.
12. Finckh A, Tramèr MR. Primer: strengths and weaknesses of meta-analysis. Nat Clin Pract Rheumatol 2008;4:146–152.
13. Morris RD. Meta-analysis in cancer epidemiology. Environ Health Perspect 1994;102 Suppl 8:61–66.
14. Zhang W. Meta-epidemiology: building the bridge from research evidence to clinical practice. Osteoarthritis Cartilage 2010;18 Suppl 2:S1.
15. Robertson C, Idris NR, Boyle P. Beyond classical meta-analysis: can inadequately reported studies be included? Drug Discov Today 2004;9:924–931.
16. Sterne JA, Jüni P, Schulz KF, Altman DG, Bartlett C, Egger M. Statistical methods for assessing the influence of study characteristics on treatment effects in ‘meta-epidemiological’ research. Stat Med 2002;21:1513–1524.
17. Olsen O, Gøtzsche PC. Cochrane review on screening for breast cancer with mammography. Lancet 2001;358:1340–1342.
18. Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA 1995;273:408–412.
19. Gluud LL. Bias in clinical intervention research. Am J Epidemiol 2006;163:493–501.
20. Ioannidis JP. Interpretation of tests of heterogeneity and bias in meta-analysis. J Eval Clin Pract 2008;14:951–957.
21. Turner RM, Davey J, Clarke MJ, Thompson SG, Higgins JP. Predicting the extent of heterogeneity in meta-analysis, using empirical data from the Cochrane Database of Systematic Reviews. Int J Epidemiol 2012;41:818–827.
22. 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 2009;MR000006.
23. Mahid SS, Qadan M, Hornung CA, Galandiuk S. Assessment of publication bias for the surgeon scientist. Br J Surg 2008;95:943–949.
24. Thornton A, Lee P. Publication bias in meta-analysis: its causes and consequences. J Clin Epidemiol 2000;53:207–16.
25. Kirkham JJ, Dwan KM, Altman DG, Gamble C, Dodd S, Smyth R, et al. The impact of outcome reporting bias in randomised controlled trials on a cohort of systematic reviews. BMJ 2010;340c365.
26. White IR, Welton NJ, Wood AM, Ades AE, Higgins JP. Allowing for uncertainty due to missing data in meta-analysis--part 2: hierarchical models. Stat Med 2008;27:728–745.
27. Savović J, Jones HE, Altman DG, Harris RJ, Jüni P, Pildal J, et al. Influence of reported study design characteristics on intervention effect estimates from randomized, controlled trials. Ann Intern Med 2012;157:429–438.
28. Jüni P, Altman DG, Egger M. Systematic reviews in health care: Assessing the quality of controlled clinical trials. BMJ 2001;323:42–46.
29. Hróbjartsson A, Thomsen AS, Emanuelsson F, Tendal B, Hilden J, Boutron I, et al. Observer bias in randomised clinical trials with binary outcomes: systematic review of trials with both blinded and non-blinded outcome assessors. BMJ 2012;344e1119.
30. 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;336:601–605.
31. Nüesch E, Trelle S, Reichenbach S, Rutjes AW, Bürgi E, Scherer M, et al. The effects of excluding patients from the analysis in randomised controlled trials: meta-epidemiological study. BMJ 2009;339b3244.
32. Nüesch E, Reichenbach S, Trelle S, Rutjes AW, Liewald K, Sterchi R, et al. The importance of allocation concealment and patient blinding in osteoarthritis trials: a meta-epidemiologic study. Arthritis Rheum 2009;61:1633–1641.
33. 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;26:2745–2758.
34. Montori VM, Bhandari M, Devereaux PJ, Manns BJ, Ghali WA, Guyatt GH. In the dark: the reporting of blinding status in randomized controlled trials. J Clin Epidemiol 2002;55:787–790.
35. Kjaergard LL, Villumsen J, Gluud C. Reported methodologic quality and discrepancies between large and small randomized trials in meta-analyses. Ann Intern Med 2001;135:982–989.
36. Savović J, Jones H, Altman D, Harris R, Jűni P, Pildal J, et al. Influence of reported study design characteristics on intervention effect estimates from randomised controlled trials: combined analysis of meta-epidemiological studies. Health Technol Assess 2012;16:1–82.
37. Moher D, Pham B, Jones A, Cook DJ, Jadad AR, Moher M, et al. Does quality of reports of randomised trials affect estimates of intervention efficacy reported in meta-analyses? Lancet 1998;352:609–613.
38. Balk EM, Bonis PA, Moskowitz H, Schmid CH, Ioannidis JP, Wang C, et al. Correlation of quality measures with estimates of treatment effect in meta-analyses of randomized controlled trials. JAMA 2002;287:2973–2982.
39. Jackson R, Ameratunga S, Broad J, Connor J, Lethaby A, Robb G, et al. The GATE frame: critical appraisal with pictures. Evid Based Med 2006;11:35–38.
40. Altman DG, Schulz KF. Statistics notes: concealing treatment allocation in randomised trials. BMJ 2001;323:446–447.
41. Pildal J, Chan AW, Hróbjartsson A, Forfang E, Altman DG, Gøtzsche PC. Comparison of descriptions of allocation concealment in trial protocols and the published reports: cohort study. BMJ 2005;330:1049.
42. Pildal J, Hróbjartsson A, Jørgensen KJ, Hilden J, Altman DG, Gøtzsche PC. Impact of allocation concealment on conclusions drawn from meta-analyses of randomized trials. Int J Epidemiol 2007;36:847–857.
43. Schulz KF. Assessing allocation concealment and blinding in randomized controlled trials: why bother? ACP J Club 2000;132:A11–A12.
44. Hróbjartsson A, Pildal J, Chan AW, Haahr MT, Altman DG, Gøtzsche PC. Reporting on blinding in trial protocols and corresponding publications was often inadequate but rarely contradictory. J Clin Epidemiol 2009;62:967–973.
45. Egger M, Juni P, Bartlett C, Holenstein F, Sterne J. How important are comprehensive literature searches and the assessment of trial quality in systematic reviews? Empirical study. Health Technol Assess 2003;7:1–76.
46. Salanti G, Dias S, Welton NJ, Ades AE, Golfinopoulos V, Kyrgiou M, et al. Evaluating novel agent effects in multiple-treatments meta-regression. Stat Med 2010;29:2369–2383.
47. 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;42:1120–1131.
48. Rücker G, Carpenter JR, Schwarzer G. Detecting and adjusting for small-study effects in meta-analysis. Biom J 2011;53:351–368.
49. Lyles RH, Allen AS, Dana Flanders W, Kupper LL, Christensen DL. Inference for case-control studies when exposure status is both informatively missing and misclassified. Stat Med 2006;25:4065–4080.
50. Naylor CD. Meta-analysis and the meta-epidemiology of clinical research. BMJ 1997;315:617–619.
51. 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;316:311–312.
52. Valdes AM, Arden NK, Tamm A, Kisand K, Doherty S, Pola E, et al. A meta-analysis of interleukin-6 promoter polymorphisms on risk of hip and knee osteoarthritis. Osteoarthritis Cartilage 2010;18:699–704.
53. Tzoulaki I, Siontis KC, Ioannidis JP. Prognostic effect size of cardiovascular biomarkers in datasets from observational studies versus randomised trials: meta-epidemiology study. BMJ 2011;343d6829.
54. Dechartres A, Trinquart L, Boutron I, Ravaud P. Influence of trial sample size on treatment effect estimates: meta-epidemiological study. BMJ 2013;346f2304.
55. Trinquart L, Dechartres A, Ravaud P. Commentary: Meta-epidemiology, meta-meta-epidemiology or network meta-epidemiology? Int J Epidemiol 2013;42:1131–1133.
56. Veroniki AA, Vasiliadis HS, Higgins JP, Salanti G. Evaluation of inconsistency in networks of interventions. Int J Epidemiol 2013;42:332–345.
57. Salanti G, Del Giovane C, Chaimani A, Caldwell DM, Higgins JP. Evaluating the quality of evidence from a network meta-analysis. PLoS One 2014;9e99682.
58. Hutton B, Salanti G, Chaimani A, Caldwell DM, Schmid C, Thorlund K, et al. The quality of reporting methods and results in network meta-analyses: an overview of reviews and suggestions for improvement. PLoS One 2014;9e92508.
59. Mavridis D, Sutton A, Cipriani A, Salanti G. A fully Bayesian application of the Copas selection model for publication bias extended to network meta-analysis. Stat Med 2013;32:51–66.
60. Salanti G, Marinho V, Higgins JP. A case study of multiple-treatments meta-analysis demonstrates that covariates should be considered. J Clin Epidemiol 2009;62:857–864.
61. Chaimani A, Higgins JP, Mavridis D, Spyridonos P, Salanti G. Graphical tools for network meta-analysis in STATA. PLoS One 2013;8e76654.
62. Bafeta A, Trinquart L, Seror R, Ravaud P. Analysis of the systematic reviews process in reports of network meta-analyses: methodological systematic review. BMJ 2013;347f3675.
63. De Vito C, Manzoli L, Marzuillo C, Anastasi D, Boccia A, Villari P. A systematic review evaluating the potential for bias and the methodological quality of meta-analyses in vaccinology. Vaccine 2007;25:8794–8806.
64. Moja LP, Telaro E, D’Amico R, Moschetti I, Coe L, Liberati A. Assessment of methodological quality of primary studies by systematic reviews: results of the metaquality cross sectional study. BMJ 2005;330:1053.
65. Hartling L, Ospina M, Liang Y, Dryden DM, Hooton N, Krebs Seida J, et al. Risk of bias versus quality assessment of randomised controlled trials: cross sectional study. BMJ 2009;339b4012.
66. Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ 2011;343d5928.
67. Schouten LM, Hulscher ME, van Everdingen JJ, Huijsman R, Grol RP. Evidence for the impact of quality improvement collaboratives: systematic review. BMJ 2008;336:1491–1494.
68. Devereaux PJ, Manns BJ, Ghali WA, Quan H, Lacchetti C, Montori VM, et al. Physician interpretations and textbook definitions of blinding terminology in randomized controlled trials. JAMA 2001;285:2000–2003.
69. Augestad KM, Berntsen G, Lassen K, Bellika JG, Wootton R, Lindsetmo RO, et al. Standards for reporting randomized controlled trials in medical informatics: a systematic review of CONSORT adherence in RCTs on clinical decision support. J Am Med Inform Assoc 2012;19:13–21.
70. Hopewell S, Dutton S, Yu LM, Chan AW, Altman DG. The quality of reports of randomised trials in 2000 and 2006: comparative study of articles indexed in PubMed. BMJ 2010;340c723.
71. Hill CL, LaValley MP, Felson DT. Discrepancy between published report and actual conduct of randomized clinical trials. J Clin Epidemiol 2002;55:783–786.

Appendices

Appendix 1. Examples of meta-epidemiologic studies

A01. 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;336:601–605. A02. Zhang W, Robertson J, Jones AC, Dieppe PA, Doherty M. The placebo effect and its determinants in osteoarthritis: meta-analysis of randomised controlled trials. Ann Rheum Dis 2008;67:1716–1723. A03. Nüesch E, Reichenbach S, Trelle S, Rutjes AW, Liewald K, Sterchi R, et al. The importance of allocation concealment and patient blinding in osteoarthritis trials: a meta-epidemiologic study. Arthritis Rheum 2009;61:1633–1641. A04. Valdes AM, Arden NK, Tamm A, Kisand K, Doherty S, Pola E, et al. A meta-analysis of interleukin-6 promoter polymorphisms on risk of hip and knee osteoarthritis. Osteoarthritis Cartilage 2010;18:699–704. A05. Nüesch E, Trelle S, Reichenbach S, Rutjes AW, Bürgi E, Scherer M, et al. The effects of excluding patients from the analysis in randomised controlled trials: meta-epidemiological study. BMJ 2009;339b3244. A06. Oliver S, Bagnall AM, Thomas J, Shepherd J, Sowden A, White I, et al. Randomised controlled trials for policy interventions: a review of reviews and meta-regression. Health Technol Assess 2010;14:1–165. A07. Dechartres A, Boutron I, Trinquart L, Charles P, Ravaud P. Single-center trials show larger treatment effects than multicenter trials: evidence from a meta-epidemiologic study. Ann Intern Med 2011;155:39–51. A08. Herbison P, Hay-Smith J, Gillespie WJ. Different methods of allocation to groups in randomized trials are associated with different levels of bias. A meta-epidemiological study. J Clin Epidemiol 2011;64:1070–1075. A09. Tzoulaki I, Siontis KC, Ioannidis JP. Prognostic effect size of cardiovascular biomarkers in datasets from observational studies versus randomised trials: meta-epidemiology study. BMJ 2011;343d6829. A10. Savović J, Jones H, Altman D, Harris R, Jűni P, Pildal J, et al. Influence of reported study design characteristics on intervention effect estimates from randomised controlled trials: combined analysis of meta-epidemiological studies. Health Technol Assess 2012;16:1–82. A11. Dechartres A, Trinquart L, Boutron I, Ravaud P. Influence of trial sample size on treatment effect estimates: meta-epidemiological study. BMJ 2013;346f2304.

Article information Continued

Meta-confounders References
Concealment & blindness A01
Placebo control vs. untreated control A02
Concealment & blindness A03
Genetic polymorphism A04
Exclusion of patients A05
Randomization & effect size A06
Single center vs. multicenter A07
Concealment A08
Experimental vs. observational design A09
Study design A10
Sample size A11