Conflicts of interest in randomized controlled trials reported in neurosurgical journals

Restricted access

OBJECTIVE

Randomized controlled trials (RCTs) form the basis of today’s evidence-based approach to medicine, and play a critical role in guidelines and the drug and device approval process. Conflicts of interest (COIs) are commonplace in medical research, but little is known about their influence. The authors aimed to evaluate the extent and influence of COIs in recent RCTs published in core neurosurgical journals using a cross-sectional analysis.

METHODS

Through review of 6 general neurosurgical journals, all interventional RCTs published from January 2009 to January 2019 were identified. Because it is difficult to objectively assess study outcome, the authors opted for a strict rating approach based on the statistical significance of unambiguously reported primary endpoints, and the reported statistical protocol.

RESULTS

A total of 129 RCTs met the inclusion criteria. During the study period, the Journal of Neurosurgery published the largest number of RCTs (n = 40, 31%). Any potential COI was disclosed by 57%, and a mean of 12% of authors had a personal COI. Nonfinancial industry involvement was reported in 10%, while 31% and 20% received external support and sponsoring, respectively. Study registration was reported by 56%, while 51% of studies were blinded. Registration showed an increasing trend from 17% to 76% (p < 0.001). The median randomized sample size was 92 (interquartile range 50–153), and 8% were designed to investigate noninferiority or equality. Sixty-three RCTs (49%) unambiguously reported a primary endpoint, of which 13% were composite primary endpoints. In 43%, study outcome was positive, which was associated with a noninferiority design (31% vs 3%, p = 0.007) and a composite primary endpoint (46% vs 9%, p = 0.002). Potential COIs were not significantly associated with study positivity (69% vs 59%, p = 0.433). In the multivariate analysis, only a composite primary endpoint remained predictive of a positive study outcome (odds ratio 6.34, 95% confidence interval 1.51–33.61, p = 0.017).

CONCLUSIONS

This analysis provides an overview of COIs and their potential influence on recent trials published in core neurosurgical journals. Reporting of primary endpoints, study registration, and uniform disclosure of COIs are crucial to ensure the quality of future neurosurgical randomized trials. COIs do not appear to significantly influence the outcome of randomized neurosurgical trials.

ABBREVIATIONS CI = confidence interval; COI = conflict of interest; CONSORT = Consolidated Standards of Reporting Trials; ICMJE = International Committee of Medical Journal Editors; IQR = interquartile range; OR = odds ratio; PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analyses; RCT = randomized controlled trial.

Downloadable materials

  • Supplemental Table 1 and Appendix 1 (ZIP 560 KB)

Article Information

Correspondence Victor E. Staartjes: Bergman Clinics, Naarden, The Netherlands. victor.staartjes@gmail.com.

INCLUDE WHEN CITING Published online August 16, 2019; DOI: 10.3171/2019.5.JNS183560.

Disclosures The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

© AANS, except where prohibited by US copyright law.

Headings

Figures

  • View in gallery

    Number of RCTs related to neurosurgery indexed in PubMed/MEDLINE per year between 1976 and 2019. Counts were obtained by searching for the 6 neurosurgical journals with the filter “randomized controlled trial” applied. Figure is available in color online only.

  • View in gallery

    PRISMA flowchart demonstrating the flow of studies throughout the searches and analysis. Figure is available in color online only.

  • View in gallery

    Association of any reported COI (A: 69% vs 59%, p = 0.433), the primary endpoint of this study, as well as nonfinancial industry involvement (B: 23% vs 6%, p = 0.067), a noninferiority or equality design (C: 31% vs 3%, p = 0.007), and a composite primary endpoint (D: 46% vs 9%, p = 0.002) with positive study outcome.

  • View in gallery

    Reported COIs (A: Z = −0.20, p = 0.840), the primary endpoint of this study, and study registration (B: Z = −5.12, p < 0.001) over time. While we observed no trend in reported COIs, the proportion of trials reporting registration increased from 17% in 2009% to 76% in 2018.

TrendMD

Metrics

Metrics

All Time Past Year Past 30 Days
Abstract Views 403 403 138
Full Text Views 85 85 24
PDF Downloads 40 40 10
EPUB Downloads 0 0 0

PubMed

Google Scholar