More than two-thirds of the authors of National Comprehensive Cancer Network clinical practice guidelines reported at least one financial conflict of interest, according to research published in JAMA Oncology.
More than half of the guideline authors received a general payment of $1,000 or more.
Aaron P. Mitchell
“Clinical practice guidelines are highly influential within medicine, defining the standard of care for many ailments,” Aaron P. Mitchell, MD, oncology fellow at University of North Carolina Hospital, and colleagues wrote. “Ideally, they should be based on a comprehensive evaluation of available data and unbiased expert opinion. However, financial conflicts of interest between guideline authors and industry create the potential for undue influence of for-profit companies in medical practice.”
In a research letter — simultaneously published in JAMA Oncology — Vinay Prasad, MD, MPH, assistant professor of medicine at Oregon Health and Sciences University, and colleagues examined the impact of changes to the ASCO conflicts of interest disclosure policy.
Their results found that the change in policy to include all financial disclosures may lead to “information overload” at ASCO annual meetings.
“For conflict-of-interest disclosure to be meaningful, it must be able to be read and processed,” the researchers wrote.
Industry relationships among NCC guideline authors
Mitchell — along with Ethan M. Basch, MD, MSCr, director of the cancer outcomes research program and professor of hematology and oncology at UNC Chapel Hill School of Medicine, and Stacie B. Dusetzina, PhD, assistant professor at UNC Eshelman School of Pharmacy — accessed the Open Payments database to identify financial conflicts of interest of NCCN guideline authors in 2014.
The researchers included all physician members from the guideline committees for lung, breast, prostate and colorectal cancers (n = 125).
Study outcome measures included the proportion of guideline authors with financial conflicts of interest and the average compensation amount received per author from industry sources.
Eighty-six percent (n = 108) of guideline authors reported at least one financial conflict of interest, with a total value of $30,287,549.
Forty-seven percent (n = 59) of authors received research payments, comprising $29,036,127; eighty-four percent (n = 105) of authors received general payments, comprising $1,251,422.
Fifty-six percent (n = 70) of authors received a general payment of $1,000 or more, with a mean general payment of $10,011 (range, 0-106,859).
Research payments occurred less commonly but had a higher monetary value, with a mean research payment of $236,066 (range, 0-2,756,713).
Six percent (n = 8) of authors had financial conflicts of interests that exceeded the NCCN guideline policy on net maximum ($50,000) or single-company maximum ($20,000) payments.
The researchers acknowledged study limitations. Because the Open Payments database only captures information on physicians, the researchers did not have access to data on nonphysician authors. Further, the Open Payments database may include imprecise or inaccurate data.
“How best to manage the financial conflicts of interest in oncology practice and clinical practice guidelines remains an area of open discussion and research,” Mitchell and colleagues wrote. “More research is needed to determine which kinds of relationships are more likely to produce the unwanted consequence of physician bias to create rational, evidence-based policies that will allow for the participation of key clinical experts while managing real or perceived conflicts.”
Current disclosure protocols may not constitute meaningful disclosure of conflicts of interest, according to Prasad and colleagues.
ASCO requires the disclosure of financial conflicts of interest from presenters at its annual meeting, which are presented on a slide that precedes oral presentations.
A change to the conflict-of-interest policy that occurred between the 2014 and 2015 annual meetings now mandates that speakers disclose all conflicts of interest, whether or not they’re relevant.
Prasad and colleagues sought to determine whether this change would result in increased information overload — defined as the presentation of more information than can be managed in the time allotted to it — in conflict-of-interest slides.
The researchers reviewed 469 recorded presentations given by 458 speakers housed in the ASCO virtual meeting library, representing as many as five presentations in each oral and education session track. Researchers noted the length of time the conflict of interest slide was visible, and video player time stamping confirmed length timing.
The percentage of speakers with at least one conflict of interest increased from 124 (51.2%) in 2014 to 151 (69.9%) in 2015 (OR = 2.21; P < .001). The median number of conflicts of interest increased from one (interquartile range [IQR], 0-3) in 2014 to three (IQR, 0-8) in 2015 (P < .001).
Median visibility duration of conflict-of-interest slides decreased from 6 seconds (IQR, 4-8) in 2014 to 5 seconds (IQR, 2-9) in 2015 (P = .02). During that time, the median number of words per slide increased from seven to 13 (P < .001), and the median number of words per second increased from 1.5 to 3.2 (P < .001).
The number of conflicts of interest was comparable in oral and educational sessions in 2014; in 2015, oral sessions had a higher median number of disclosures (5.5 vs. 3; P = .004).
The researchers observed a significant uptake in information upload, from 14.9% of presentations in 2014 to 37.6% in 2015 (P < .001).
“Our findings raise concern about whether current disclosure protocols constitute meaningful disclosure,” Prasad and colleagues wrote. “The solution would be to extend the time of the disclosure slide so that no slide exceeds established limits of reading speed. Further investigation is needed on audience ability to process conflict of interest slides at professional meetings if disclosure is the major objective.”
The impact of disclosure
Increased industry–oncologist relationships have raised legitimate concerns about whether financial incentive correlates with undue influence, Ryan D. Nipp, MD, medical oncologist at Massachusetts General Hospital, and Beverly Moy, MD, MPH, clinical director of the breast oncology program at Massachusetts General Hospital and a HemOnc Today Editorial Board member, wrote in an accompanying editorial.
Ryan D. Nipp
“Both ASCO and the NCCN publicly report their financial conflict-of-interest policies, but consensus about how best to manage or prohibit potential financial conflicts of interest is lacking,” Nipp and Moy wrote. “In addition, self-disclosure alone may result in an underestimation of the true prevalence of financial conflicts of interest.”
The evolving nature of these relationships may precipitate a new conception of conflict-of-interest reporting.
“The need for comprehensible and meaningful financial conflict-of-interest policies among professional societies and universities is clear,” Nipp and Moy wrote. “Historically, the basis for most financial conflict-of-interest policies has relied solely on disclosure of relationships. In the modern era, in which public scrutiny of such relationships has never been higher, financial conflict-of-interest policies need to go beyond mere disclosure to also include management (ie, enforcement of actions, such as auditing or prohibition) of relationships that are a cause for concern.”
However, it is unwise to assume that financial conflicts of interest inherently represent unwarranted bias, according to Nipp and Moy.
“Although financial conflicts of interest may create a potential risk for compromised judgment among researchers and undue bias from outside commercial interests, there are also potential positive consequences associated with financial conflicts of interest,” Nipp and Moy wrote. “Collaboration between industry and academic researchers may facilitate the development of novel cancer therapies, a shared goal aligning the interests of industry, academia and patients.” – by Cameron Kelsall
Boothby A, et al. JAMA Oncol. 2016;doi:10.1001/jamaoncol.2016.2706.
Mitchell AP, et al. JAMA Oncol. 2016;doi:10.1001/jamaoncol.2016.2710.
Nipp RD and Moy B. JAMA Oncol. 2016;doi:10.1001/jamaoncol.2016.2726.
Disclosures: Mitchell, Prasad and the other researchers report no relevant financial disclosures. Moy reports that an immediate family member serves as a consultant for MOTUS GI Medical Technologies. Nipp reports no relevant financial disclosures.