Elsevier

The Spine Journal

Volume 15, Issue 5, 1 May 2015, Pages 1133-1140
The Spine Journal

Technical Report
Spine device clinical trials: design and sponsorship

https://doi.org/10.1016/j.spinee.2015.01.027Get rights and content

Abstract

Background context

Multicenter prospective randomized clinical trials represent the best evidence to support the safety and effectiveness of medical devices. Industry sponsorship of multicenter clinical trials is purported to lead to bias.

Purpose

To determine what proportion of spine device-related trials are industry-sponsored and the effect of industry sponsorship on trial design.

Study design

Analysis of data from a publicly available clinical trials database.

Methods

Clinical trials of spine devices registered on ClinicalTrials.gov, a publicly accessible trial database, were evaluated in terms of design, number and location of study centers, and sample size. The relationship between trial design characteristics and study sponsorship was evaluated using logistic regression and general linear models.

Results

One thousand six hundred thrity-eight studies were retrieved from ClinicalTrials.gov using the search term “spine.” Of the 367 trials that focused on spine surgery, 200 (54.5%) specifically studied devices for spine surgery and 167 (45.5%) focused on other issues related to spine surgery. Compared with nondevice trials, device trials were far more likely to be sponsored by the industry (74% vs. 22.2%, odds ratio (OR) 9.9 [95% confidence interval 6.1–16.3]). Industry-sponsored device trials were more likely multicenter (80% vs. 29%, OR 9.8 [4.8–21.1]) and had approximately four times as many participating study centers (p<.0001) and larger sample sizes. There were very few US-based multicenter randomized trials of spine devices not sponsored by the industry.

Conclusions

Most device-related spine research is industry-sponsored. Multicenter trials are more likely to be industry-sponsored. These findings suggest that previously published studies showing larger effect sizes in industry-sponsored vs. nonindustry-sponsored studies may be biased as a result of failure to take into account the marked differences in design and purpose.

Introduction

The number and variety of devices for spine surgery are increasing. Both novel devices and devices representing incremental improvements in preexisting technology are commonly the subject of clinical trials aiming to establish or confirm their safety and effectiveness. Although single-arm trials may suffice in some situations, the gold standard for proving safety and effectiveness is a randomized clinical trial (RCT). An even higher standard is the multicenter RCT (MRCT). Multicenter RCTs play an important role in surgical devices for several reasons. First, multiple centers may be required to achieve adequate trial sample sizes in reasonable time frames. Perhaps, more importantly, MRCTs show that the surgical intervention can be performed adequately by a reasonably large number of physicians, not just a single physician.

Multicenter RCTs are difficult, time-consuming, and expensive, often requiring a team of trained personnel, as opposed to a single coordinating investigator, to implement, oversee, and report the study. Human research carries a liability risk that an individual physician or hospital may not want or be able to assume [1]. Additionally, public sector funding for clinical trials has significantly declined over the past two decades, increasing the competition for research dollars [2]. The time, financial burden, and liability risk required to conduct multicenter clinical trials markedly limits the ability of an individual physician to develop a medical device without outside support [1]. Not surprisingly, many trials are sponsored by the industry.

Several recent articles [3], [4], [5], including a Cochrane summary [6], have concluded that trials sponsored by the industry are more likely to report positive results, suggesting that these trials are biased. However, none of the authors of these comparisons took into account the component studies' interventions, designs, or purposes. In the scientific spirit of reproducibility of results through replication, device-related spine trials could be repeated in the nonindustry setting, allowing a direct comparison of the potential effect of industry sponsorship. As noted above, several factors make nonindustry trial replication difficult.

The aim of this study was to determine the type and frequency of clinical trial sponsorship and determine if the sponsor type systematically affects the scope and design of device-related spine surgery clinical trials.

Section snippets

Materials and methods

No publicly available database provides a complete picture of currently existing clinical trials. For many reasons, Clinicaltrials.gov is a reasonable proxy. ClinicalTrials.gov is a federally funded and publicly available database run by the US National Institutes of Health and is part of the World Health Organization International Clinical Trials Registry Program. Listing a trial in a publicly accessible trials database is a United States federal requirement for premarket studies. Moreover,

Results

A search of ClinicalTrials.gov using the term “spine” yielded 1,638 records. Of these, 531 were not specifically about spine pathology, 222 focused on osteoporosis treatment or prevention, and 465 involved the spine but were not surgical in nature. Of the remaining 420 studies, 52 were suspended, terminated, or withdrawn, leaving 367 studies for analysis. Of the 367 trials that focused on spine surgery, 200 (54.5%) specifically involved devices for the surgical treatment of spine-related

Discussion

Published metaanalyses that purported to compare effect sizes as a function of industry sponsorship have assumed that, other than the sponsor type, the component trials have similar design and purpose. To date, this assumption has not been validated. The present study sought to evaluate the impact of study sponsorship on the scope and design of ongoing device-related spine surgery trials. Analysis of data from ClinicalTrials.gov showed that most spine surgery-related device trials are sponsored

Conclusions

Most spine surgery device-related clinical trials are industry-sponsored; industry-sponsored device-related spine surgery trials are more likely to be multicenter and have larger sample sizes. Very few nonindustry-sponsored MRCTs are conducted in the United States. The design, intervention, and purpose of industry-sponsored trials differ from those of nonindustry-sponsored trials, potentially rendering invalid any prior estimation of bias in published effect sizes associated with industry

References (15)

  • A. Brown et al.

    Association of industry sponsorship to published outcomes in gastrointestinal clinical research

    Clin Gastroenterol Hepatol

    (2006)
  • A.P. White et al.

    Counterpoint: physician-industry relationships can be ethically established, and conflicts of interest can be ethically managed

    Spine

    (2007)
  • J.D. Zuckerman et al.

    Conflict of interest in orthopaedic research

    J Bone Joint Surg Am

    (2004)
  • M. Bhandari et al.

    Association between industry funding and statistically significant pro-industry findings in medical and surgical randomized trials

    CMAJ

    (2004)
  • R.V. Shah et al.

    Industry support and correlation to study outcome for papers published in Spine

    Spine

    (2005)
  • S.N. Khan et al.

    The roles of funding source, clinical trial outcome, and quality of reporting in orthopedic surgery literature

    Am J Orthop (Belle Mead NJ)

    (2008)
  • A. Lundh et al.

    Industry sponsorship and research outcome

There are more references available in the full text version of this article.

Cited by (19)

  • Industry payments to spine surgeons from 2014 to 2019: trends and comparison of payments to spine surgeons versus all physicians

    2022, Spine Journal
    Citation Excerpt :

    Linear trends in subcategories of general payments from 2014 to 2019 did not significantly differ for top eight and bottom eight companies except in education payments (p=.01), which declined substantially for the top eight companies (Fig. 3b, Supplemental Table 2). Interactions between healthcare companies and physicians are a common and essential part of research and advancements in healthcare [24]. The financial scope of these relationships ranges from the cost of meals and travel to consulting and speaker fees for individual physicians to research grants to both physicians and academic institutions.

View all citing articles on Scopus

FDA device/drug status: Not applicable.

Author disclosures: DJC: Nothing to disclose. RAC: Nothing to disclose.

The authors are employees of SI-BONE, Inc. SI-BONE, Inc., sponsored this research. Two of the trials retrieved in the ClinicalTrials.gov data set that comprises the focus of the analysis reported herein were sponsored by SI-BONE, Inc. All databases, analysis code, and analyses are available on request. No funding was received in support of this work.

View full text