Physician Conflict of Interest Disclosures to Journals for Device-Industry Ties Fall Short

US physicians who have the most financial connections to the medical device industry don’t always fully disclose these relationships when publishing papers in journals, omitting potential conflicts of interest (CoI) relevant to the research more than one-third of the time.

The discrepancies came to light when researchers led by Kasra Ziai, MD (University of California, Irvine), dug into the 2015 Open Payments Database (OPD) to see how well self-disclosed conflicts for these top doctors matched up with the official record on general payments for food and beverage, consulting fees, honoraria, travel and lodging, etc. Their results, recently published online in JAMA Surgery, add to a growing body of literature showing a discordance between self-reported ties and reality.

Industry ties are common and aren’t necessarily a problem, Ziai and colleagues note. “Device manufacturers have a symbiotic relationship with clinicians,” they write. “Compensation is often appropriate for the work and intellectual capital provided to the companies can be an important source of revenue for a clinician or an investigator.”

Not only do many academic institutions restrict or require reporting of potential conflicts but journals also can provide a paper trail. “When evaluating published works, it is important to assess the integrity and academic credentials of the authors, who serve as putative experts. A relationship with a relevant manufacturer may increase the potential risk for bias in relevant studies,” the investigators observe, adding, “Inaccurate CoI statements can result in biased perception of the study results by readers, ranging from patients to healthcare professionals.”

Importantly, the International Committee of Medical Journal Editors (ICMJE) mandates that study authors disclose potential CoI related, whether directly or indirectly, to a publication within the past 3 years. The OPD, established in 2013 to house details gathered through the 2010 Sunshine Act, makes it possible to now see whether physicians are being fully transparent about links to industry.

Speaking with TCTMD, senior author Mehraneh D. Jafari, MD (University of California, Irvine), said that gaps in disclosure aren’t likely intentional. “I think it’s because the system is broken,” she commented. “Each journal has its own guidelines, and then each meeting has its own guidelines. And if you’re a prolific researcher it’s difficult to keep up with [these rules].” Moreover, Jafari added, the busiest physician investigators may not even be turning in these details themselves but relying on others to do so, which can lead to outdated or incomplete disclosure.

Mainly “it’s that they don’t know the rules. They don’t realize how stringent the true guidelines [for reporting] actually are and what true conflict of interest is,” she explained.

James N. Kirkpatrick, MD (University of Washington, Seattle), who is chair of his institution’s ethics committee, commented on the study for TCTMD, noting that it’s hard to say what explains the discrepancies in disclosure. “None of these studies can really investigate motives, and it’s probable that for the vast majority of people it’s simply an oversight.”

For example, he continued, “the time frames are not always aligned in terms of what is required and people’s disclosures do change over time, so it may not be adequately captured.” Or it may be that the companies reporting to the OPD have accidentally sent in inaccurate information, Kirkpatrick added.

While the new study isn’t unique in showing gaps, he said it’s still worthwhile to continue exploring these patterns. “This is something that continually has to be brought before the public. It’s not something that goes away, and therefore it shouldn’t go away in the press either,” Kirkpatrick stressed. “It’s something that everyone should be reminded to be really careful about, and in our busy lives . . . we tend to lose sight of the importance of this.”

An Argument for Full Disclosure

To better understand the situation, Ziai et al looked at 10 surgical and medical device manufacturers—Medtronic, Stryker, Intuitive Surgical, Covidien, Edwards Lifesciences, Ethicon, Olympus, WL Gore & Associates, LifeCell Corporation, and Baxter Healthcare—then extracted data on each company’s 10 most highly compensated physicians from 2015.

That year, these 100 physicians (88% men) altogether received $12.4 million in payments. The median amount was $95,993. Obstetricians and gynecologists led the pack, with 15 physicians in the group, followed by anesthesiologists, orthopedic surgeons, and general surgeons. Four vascular surgeons and five cardiothoracic surgeons were in the top 100, as were four physicians who worked in internal medicine/cardiovascular disease.

Searching PubMed for articles published by each of these individuals in 2016 turned up 412 papers, 225 (55%) of which were relevant to general payments listed in the OPD. However, only 84 of these 225 papers (37.3%) included all potentially relevant CoI.

Of the 64 physicians in the group who had authored papers, 86% failed to mention associated CoI in at least one publication. There was no difference in the validity of reporting between the half of physicians who were faculty at academic institutions and the half who were not. Nor did reporting vary by rank, though associate professors tended to receive higher mean payments than did either professors or assistant professors. Physicians’ h index did not seem to affect either their reporting or the amount of payments they received.

“A high level of inconsistency was found between self-declared CoI and the OPD among the physicians receiving the highest industry payments,” the researchers conclude. “Therefore, a policy of full disclosure for all publications, regardless of relevance, is proposed.”

It’s possible that differences in reporting policies across journals, meetings, and institutions may be “overwhelming and puzzling for authors,” they concede. “Another factor to consider is the authors’ perception of relevance of CoI in a study,” with less direct conflicts going unmentioned.

Full disclosure is the best way to come up with a very imperfect solution to a vexing problem. James N. Kirkpatrick

Full disclosure would get around these obstacles, Ziai and colleagues say, suggesting that “an unbiased third party should decide on the relevance of the disclosures.”

Greg D. Sacks, MD, PhD and O. Joe Hines, MD (University of California, Los Angeles), writing in an editorial accompanying the study, stress the importance of disclosure as a “cornerstone” of the safeguards in place to prevent undue influence. “Of all the biases threatening research validity, perhaps most pernicious is that introduced by a financial incentive exerting influence on a researcher. In some circumstances, remuneration from a company may support a researcher’s time used to perform a study, but unfortunately, a substantial body of literature consistently documents bias attributed to these financial payments, which are ubiquitous in medicine,” they write.

But while the results shown by Ziai et al are “alarming,” the editorialists say, “the relatively low number of reported conflicts may be in part due to the authors’ use of a strict criterion to determine whether a researcher disclosed his or her CoI.” Specifically, the researchers set the standard at reporting of all potential conflicts, not just those related to the manufacturer of the device being studied but also those related to competitors in the same field. This decision is the right one, Sacks and Hines agree, based on ICMJE standards.

“Although progress has been made with CoI, the current processes are inadequate, and a revision of CoI reporting procedures is essential. Given the enormous sums of money invested by industry into physician relationships, enforcing accurate disclosure of CoI is a critical step to ensuring the integrity and reputation of our fields,” they conclude.

Full disclosure will require cooperation among journals, Jafari noted. In the meantime, physicians can err on the side of reporting all financial ties, so that journal editors can make the call, as can the journal’s readers, she said.

Kirkpatrick agreed, though he pointed out that for some physicians with many conflicts, it may be hard to tease out what’s relevant and what’s not. Also challenging is that some people may be more—or less—susceptible to bias that others, even when they have the same disclosures. As such, “full disclosure is the best way to come up with a very imperfect solution to a vexing problem,” he advised.