Discordances originated by multiple meta-analyses on interventions for myocardial infarction: a systematic review

Abstract

To clarify the impact of multiple (covering the same population,intervention, control, and outcomes) systematic reviews (SRs) on interventions for myocardial infarction (MI). STUDY DESIGN AND SETTING: Clinical Evidence (BMJ Group) sections and related search strategies regarding MI were used to identify multiple SRs published between 1997 and 2007. Multiple SRs were classified as discordant if they featured conflicting results or interpretation of them.RESULTS: Thirty-six SRs (23.5% of 153 on the treatment or prevention of MI) were classified as multiple and grouped in 16 clusters [ie, at least two SRs with the same PICO (population, condition/disease, intervention, control) and at least one common outcome] exploring angioplasty, angiotensin-converting enzyme inhibitors,anticoagulants, antiplatelets, beta-blockers, and stents. Complete agreement on statistically significant differences between interventions was found in 7 of 10 clusters with a shared composite outcome. Agreement was reduced when single outcomes were considered. Despite substantial variation and limited agreement in reporting of major outcomes, SRs agreed in their conclusions on the superiority of either the intervention or control in 14 of 16 clusters. Sources of minor discrepancies were found in terms of study and outcome selection, subgroup analyses, and interpretation of findings. CONCLUSION: Multiple SRs agreed in their qualitative conclusions but not on reporting and on analyses of hard outcomes. Discordance on significance of treatment effects was due to a combination of variation in design with inclusion of different studies and lack of precision for single hard outcomes compared with a composite outcome. Such inconsistencies among SRs could potentially slow the translation of SRs' results to clinical and public health decision making and suggest the need for a broader methodological and clinical agreement on their design

Interventions for Age-Related Macular Degeneration

“…Since the time the searches were conducted for the current study, Cochrane authors have updated an earlier review of anti-VEGF effectiveness and also have published a review comparing the systemic safety of ranibizumab versus bevacizumab. 102,103 Unlike other research that has found duplication of systematic reviews on the same topic to be wasteful 101,104 or lead to discordant findings, 105,106 we conclude that sequential systematic reviews that at first glance appear to cover similar topics instead may represent evolution in the research question with increased clinical experience and serve as an indication of a rapidly developing field.…”

“…Some reasons for differing conclusions are understandable, for example when the studies synthesized in systematic reviews were conducted at dissimilar time periods or included different types of study designs. 8 But sometimes differing conclusions can be ascribed to use of systematic review methods that are potentially subject to bias. 9 …”

Topic duplication and research waste at the ‘Overviews of systematic reviews’ level: Survey of overlapping overviews

“…We would hope that overviews on the same topic would nd the same results and come to similar conclusions, but this is often not the case with systematic reviews [12,[35][36][37]. Potential discrepancies can cause endless debates (e.g.…”

“…Overviews may alternatively aim to answer narrow, focused clinical questions, and to identify and explore reasons for variation in the results, interpretation, or conclusions of systematic reviews analysis [9][10][11][12][13][14][15][16][17][18].…”

Meta-Analysis of the Association of HLA-DRB1 with Anti-Citrullinated Protein Antibody-Positive and Anti-Citrullinated Protein Antibody-Negative Rheumatoid Arthritis

“…The summary odds ratios (ORs) with 95% confidence intervals (CIs) were calculated to indicate the association of the different HLA-DRB1 alleles with the development of different ACPA-status RA. The statistical heterogeneity among the studies was assessed by a chi-square-based Q statistic test and the I 2 statistic (I 2 = 0% -50%, no or moderate heterogeneity; I 2 > 50%, significant heterogeneity) [15] [16]. If the homogeneity was significant (P > 0.10 for the Q test or I2% > 50%), the random-effect model was used.…”