Elsevier

Journal of Clinical Epidemiology

Volume 70, February 2016, Pages 155-163
Journal of Clinical Epidemiology

Original Article
Meta-analyses with industry involvement are massively published and report no caveats for antidepressants

https://doi.org/10.1016/j.jclinepi.2015.08.021Get rights and content

Abstract

Objectives

To identify the impact of industry involvement in the publication and interpretation of meta-analyses of antidepressant trials in depression.

Study Design and Setting

Using MEDLINE, we identified all meta-analyses evaluating antidepressants for depression published in January 2007–March 2014. We extracted data pertaining to author affiliations, conflicts of interest, and whether the conclusion of the abstract included negative statements on whether the antidepressant(s) were effective or safe.

Results

We identified 185 eligible meta-analyses. Fifty-four meta-analyses (29%) had authors who were employees of the assessed drug manufacturer, and 147 (79%) had some industry link (sponsorship or authors who were industry employees and/or had conflicts of interest). Only 58 meta-analyses (31%) had negative statements in the concluding statement of the abstract. Meta-analyses including an author who were employees of the manufacturer of the assessed drug were 22-fold less likely to have negative statements about the drug than other meta-analyses [1/54 (2%) vs. 57/131 (44%); P < 0.001].

Conclusion

There is a massive production of meta-analyses of antidepressants for depression authored by or linked to the industry, and they almost never report any caveats about antidepressants in their abstracts. Our findings add a note of caution for meta-analyses with ties to the manufacturers of the assessed products.

Introduction

What is new?

Key findings

  1. No studies have systematically assessed the extent and implications of industry involvement in meta-analyses of trials evaluating antidepressants, one of the largest markets of pharmaceuticals, for the management of depression.

  2. Our study shows the massive presence of the industry in generating a factory of meta-analyses in this field [147 of 185 (79%) meta-analyses published in the last 7 years].

What this adds to what was known?
  1. Industry-authored meta-analyses almost never include any negative concluding statement in their abstracts about the antidepressants assessed.

What is the implication and what should change now?
  1. Caution is needed in interpreting meta-analyses with ties to the manufacturers of the assessed products.

Meta-analyses are often considered the highest level of evidence in the hierarchy of evidence-based medicine. The number of meta-analyses being conducted is growing rapidly [1]. As these data syntheses become more influential in shaping guidelines and clinical practice, the biopharmaceutical industry may have an enhanced interest in them [2]. There is some evidence that randomized trials and systematic reviews performed or sponsored by the industry may lead to more favorable conclusions than non–industry-related trials and systematic reviews [3], [4], [5], [6], [7]. Given that influential meta-analyses can be performed with more limited resources and faster than randomized trials, it is conceivable that the industry could easily help generate a large number of meta-analyses to support its products.

Antidepressants comprise one of the largest markets of pharmaceuticals, with several blockbuster drugs, and a market of $9.4 billion per year in the United States alone in 2013 [8]. A few meta-analyses of antidepressants published by the Food and Drug Administration and other non–industry-sponsored investigators reached extremely high visibility between 2005 and early 2008, causing heated debates about serious risks (e.g., suicidality [9], [10]) or questionable effectiveness in patients without severe depression [11]. However, there is no systematic study of industry-related meta-analyses in this field. Some important questions may be asked. Is there a large current meta-analysis literature of antidepressants under control or sponsorship by the industry? Do industry-related meta-analyses reach more favorable conclusions in their abstracts than other meta-analyses? To answer these questions, we aimed to examine systematically the extent and implications of industry involvement in meta-analyses of trials evaluating antidepressants for the management of depression published between 2007 and 2014.

Section snippets

Eligibility criteria

We considered all meta-analyses of randomized controlled trials evaluating drugs approved as antidepressants in patients with depressive conditions and published since 2007. Publications were eligible regardless of whether they used either group-level or individual-patient information.

Meta-analyses were defined as analyses that quantitatively combine results from multiple trials to produce a summary treatment effect. A thorough systematic review component was not required (e.g., pooled analyses

Results

Our search identified 1,111 citations. Three articles were unavailable [14], [15], [16], and we scrutinized 259 in full text. Of these, 74 were excluded (Fig. 1). Our evaluation thus consisted of 185 eligible meta-analyses (Table 1, Appendix at www.jclinepi.com).

After 2007, every year, 20–32 eligible meta-analyses were published. Most publications were published by corresponding authors from the United States (35%) and Europe (33%), assessed benefit (92%), evaluated patients with major

Discussion

Over the past 7 years, there have been nearly 200 published meta-analyses assessing the effectiveness of antidepressants for depression. Approximately 80% of these meta-analyses had a direct involvement from industry (sponsorship, industry authors, or authors with industry conflicts of interest), and one-third were written by industry employee authors. This represents a massive presence of the industry in generating a prolific production of meta-analyses in this field. Meta-analyses by industry

Acknowledgments

The authors thank Ms. Kelsey Vercammen for assisting with article retrieval; no compensation was received for completing this work.

Authors' contributions: S.E. and J.P.A.I. conceived and designed the study. S.E., S.B., A.A., and C.M. made substantial contributions to the acquisition, analysis, or interpretation of data for the work. S.E. drafted the manuscript, and S.B., A.A., C.M., and J.P.A.I. revised it critically for important intellectual content. All authors approved the final version to

References (32)

  • A.B. Haidich et al.

    Most meta-analyses of drug interventions have narrow scopes and many focus on specific agents

    J Clin Epidemiol

    (2013)
  • D. Moher et al.

    Epidemiology and reporting characteristics of systematic reviews

    PLos Med

    (2007)
  • E. Stamatakis et al.

    Undue industry influences that distort healthcare research, strategy, expenditure and practice: a review

    Eur J Clin Invest

    (2013)
  • A.W. Jørgensen et al.

    Industry-supported meta-analyses compared with meta-analyses with non-profit or no support: differences in methodological quality and conclusions

    BMC Med Res Methodol

    (2008)
  • A.W. Jørgensen et al.

    Cochrane reviews compared with industry supported meta-analyses and other meta-analyses of the same drugs: systematic review

    BMJ

    (2006)
  • B. Djulbegovic et al.

    Treatment success in cancer: industry compared to publicly sponsored randomized controlled trials

    PLoS One

    (2013)
  • A. Lundh et al.

    Industry sponsorship and research outcome

    Cochrane Database Syst Rev

    (2012)
  • R.J. Kelly et al.

    Relationship between drug company funding and outcomes of clinical psychiatric research

    Psychol Med

    (2006)
  • U.S. Anti-Depressant Market - Old Fashion Branding Could Save The Day. Available at...
  • T.A. Hammad et al.

    Suicidality in pediatric patients treated with antidepressant drugs

    Arch Gen Psychiatry

    (2006)
  • Relationship between psychotropic drugs and pediatric suicidality: review and evaluation of clinical data. Available at...
  • I. Kirsch et al.

    Initial severity and antidepressant benefits: a meta-analysis of data submitted to the Food and Drug Administration

    PLos Med

    (2008)
  • S. Chimonas et al.

    From disclosure to transparency: the use of company payment data

    Arch Intern Med

    (2011)
  • D.R. Guay

    Vilazodone hydrochloride, a combined SSRI and 5-HT1A receptor agonist for major depressive disorder

    Consult Pharm

    (2012)
  • R. Fleurence et al.

    A systematic review of augmentation strategies for patients with major depressive disorder

    Psychopharmacol Bull

    (2009)
  • D.S. Robinson et al.

    Treatment effects of selegiline transdermal system on symptoms of major depressive disorder: a meta-analysis of short-term, placebo-controlled, efficacy trials

    Psychopharmacol Bull

    (2007)
  • Cited by (92)

    • The controversy on the treatment of left main coronary artery disease

      2022, Journal of Thoracic and Cardiovascular Surgery
    • The methodological quality was low and conclusions discordant for meta-analyses comparing proximal humerus fracture treatments: a meta-epidemiological study

      2022, Journal of Clinical Epidemiology
      Citation Excerpt :

      Assessments were conducted in duplicate by NS and PB with disagreements solved by consensus. In our protocol, we identified six methodological quality-related variables that have either been reported or hypothesized to affect the conclusions of meta-analyses: (1) overall AMSTAR2 rating, (2) critical AMSTAR2 items, (3) source of funding, (4) journal type, (5) inclusion criteria based on the reported PICO question (Population, Intervention, Comparator, Outcome), and (6) whether one of the authors was a content expert [21–30]. To determine the association between the quality-related variables and reported conclusions, we planned to use the Mann-Whitney test for variables with 2 groups (AMSTAR2 critical items 11, 13, and 15, and whether one of the authors was a content expert).

    • Redundant meta-analyses are common in genetic epidemiology

      2020, Journal of Clinical Epidemiology
      Citation Excerpt :

      It has been estimated that close to one-third of currently produced meta-analyses may be redundant and unnecessary [6]. Additional focused assessments of specific fields have shown a very high frequency of meta-analyses, for example, 185 meta-analyses of antidepressants for major depression [10], and high redundancy among network meta-analyses, including up to 28 published network meta-analyses for biologics [11]. For some topics, the number of published meta-analyses already far exceeds the number of primary studies [12,13].

    • Future of evidence ecosystem series: 1. Introduction Evidence synthesis ecosystem needs dramatic change

      2020, Journal of Clinical Epidemiology
      Citation Excerpt :

      There are also challenges beyond simply identifying new data—in particular, the effort required to introduce changing methods. This lack of prioritization and defined scope in the current landscape of evidence synthesis results in important overlaps and gaps [10,66–68]. The creation of PROSPERO is useful for understanding the landscape of research synthesis but insufficient because registration is not mandatory, and whether the registration was prospectively performed cannot be determined [69].

    View all citing articles on Scopus

    Conflict of interest: None.

    Funding: No funds were received for this study. S.E. is supported by MITACS Elevate and SickKids Restracomp Postdoctoral Fellowship Awards, S.B. by an Ontario Graduate Scholarship and Social Sciences and Humanities Research Council Doctoral Scholarship, C.M. by a graduate scholarship from the Ontario Mental Health Association, and the work of JPAI is supported by an unrestricted gift from Sue and Bob O'Donnell. Sponsors providing individual financial support to authors did not have a role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. S.E. and J.P.A.I. had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    View full text