A Common Standard for

Conflict of Interest Disclosure

Merrill Goozner, Arthur Caplan, Jonathan Moreno,

Barnett S. Kramer, Thomas F. Babor, Wendy Cowles Husser

July 2008
A Common Standard for Conflict of Interest Disclosure

Merrill Goozner, Arthur Caplan, Jonathan Moreno, Barnett S. Kramer, Thomas F.

Babor, Wendy Cowles Husser

The reporting of conflicts of interest in science and medicine in the scientific

literature1 has come under increasing scrutiny in recent years. Failures to disclose
conflicts of interest have become front page news, and a major embarrassment to
publishers, editors, and professional societies.2

These failures to disclose relevant relationships have stemmed in part from a lack
of uniform definitions for conflicts of interest and confusion about what needs to be
reported. Academic investigators operate under varying institutional rules, and many are
unable to accurately describe their institutions’ policies.3 Science and biomedical journals
have a range of disclosure policies with differences in definitions of conflicts of interest,
reporting requirements, and promises to publish.4

In the face of heightened scrutiny, several publishers have moved in the past
several years to implement stricter conflict-of-interest disclosure and publication rules. 5 6
7
Organizations are paying greater attention to conflict-of-interest disclosure in the

1
Conflicts of interest most often refers to financial relationships, but can also involve personal,
professional, ideological, political or religious views, which are sometimes referred to as competing
interests. For the purpose of this paper, conflicts of interests refer to both financial and non-financial
relationships.
2
David Armstrong, “Amid Suits Over Mold, Experts Wear Two Hats,” Wall Street Journal, Jan. 9, 2007;
Christopher Rowland, “Medical Group Puts Stop to Talks on Drug Firm Ties,” Boston Globe, Dec. 27,
2006; David Armstrong, “Financial Ties to Industry Cloud Major Depression Study,” Wall Street Journal,
July 11, 2006; Reed Abelson, “Charities Tied to Doctors Get Drug Industry Gifts,” New York Times, June
27, 2006; Tinker Ready, “Divided Loyalties?” Washington Post, Feb. 7, 2006; Erika Check, “Journals
Scolded for Slack Disclosure Rules,” Nature, Jan. 18, 2006.
3
Boyd EA, Cho MK, Bero LA, “Financial Conflict-of-Interest Policies in Clinical Research: Issues for
Clinical Investigators,” Acad. Med. 2003;78:769-774.
4
Krimsky S, Rothenberg L.S. “Conflict of interest policies in science and medical journals: editorial
practices and author disclosures. Sci Eng Ethics. 2001;7:205-218.
5
Flanagin A, Fontanarosa PB, DeAngelis CD, “Update on JAMA’s Conflict of Interest Policy,” JAMA,
2006;296:220-21.
6
Graf C, Wager E, Bowman A, Fiack S, Scott-Lichter D, Robinson A; “Best Practice guidelines on
Publication Ethics: a Publisher’s Perspective,” Int J Clin Pract, 2007;61(Suppl. 152):1-26.
7
“Ethical Guidelines to Publication of Chemical Research,” American Chemical Society, January 2006.
context of redefining the rules of engagement between academic investigators and private
industry.8 The need for common standards in defining conflicts of interest has never been
greater.

Since the first conflict-of-interest disclosures appeared in the New England

Journal of Medicine in the early 1980s,9 the rationale for including this information in
published scientific articles has not changed. Scientific discourse depends on objectivity.
The conduct of science can be influenced by biases introduced by conflicts of interest,
whether they are financial, professional, intellectual, or fueled by academic competition.
The potential for bias is real, whether or not researchers believe those conflicts of interest
influence their conduct.

Moreover, the positive association between conflicts of interest and the outcomes
of research has been documented many times in many fields. Virtually all of this research
has focused on financial conflicts of interest.10 The causation of the relationship is the
subject of much speculation. Frequently cited factors include protocol construction,
sample selection bias, and selective interpretation. Because these potential causes of bias
have never been studied systematically, controlling for bias through publication criteria is
nearly impossible, although at least one medical journal, the Journal of the American
Medical Association, took steps in that direction when it required that the author
primarily responsible for “the integrity of the data and the accuracy of the data analysis . .
. not be employed by any commercial funding source.” The editors also expressed a
preference for studies where “data collection and data management are conducted
independently of the study sponsor and with additional monitoring and oversight, such as
under the auspices of an independent data and safety monitoring committee.”11

But that is the exception. The general consensus among editors is that disclosure
of financial conflicts of interest – at the outset of the submission process, during the
editing process, and, finally, to the readers – is the most practical available tool for
managing conflicts that can influence or even appear to influence scientific research.

But is disclosure a sufficient remedy for managing conflicts of interest in all

situations? The evolution of existing policies over the years shows there have been
sharply diverging views about what is allowed for authors of reviews, for instance, or
editorials. In 1990, editors at the New England Journal of Medicine adopted a policy
limiting the right of authors with conflicts of interest to write reviews and editorials. The
policy was reaffirmed in 1996 when readers complained that several prominent reviewers

8
Brockway LM, Furcht LT; “Conflicts of interest in biomedical research – the FASEB guidelines,” FASEB
Journal;20:2435-38.
9
Relman AS, “Dealing with conflicts of interest,” NEJM. 1984; May 3; 310(18):1182-3.
10
See, for instance, Bekelman JE, et al, “Scope and impact of financial conflicts of interest in biomedical
research,” JAMA. 2003; Jan 22-29;289(4):454-65. Review; Ridker PM, Torres J, “Reported outcomes in
major cardiovascular clinical trials funded by for-profit and not-for-profit organizations: 2000-2005,”
JAMA. 2006; May 17;295(19):2270-74; Lesser et al, “Relationship between funding source and conclusion
among nutrition-related scientific articles,” PLoS Medicine. 2007;
11
Fontanarosa PB, Flanagin A, DeAngelis CD, “Reporting conflicts of interest, financial aspects of
research, and role of sponsors in funded studies,” JAMA. 2005; 294(1):110-11.
had not disclosed relevant conflicts of interest, and an internal investigation revealed the
policy was frequently ignored.12 In 2002, a different set of editors amended the policy to
allow editorials and reviews to be written by authors with conflicts of interest below
$10,000 a year as long as they were fully disclosed. The editors explained that they didn’t
want to disqualify highly qualified people from offering their opinions in the pages of the
journal.13

The argument in favor of the older policy rests on a desire to reserve a high
standard of independence and objectivity for reviews that claim to represent the state of
the science. Prohibiting reviewers from having conflicts of interest protects journals
against the perception of bias influencing the analyses that appear in their pages. In that
sense, evidence reviews written by individuals or groups have a lot in common with
medical groups that author clinical practice guidelines. Both review the available
evidence, assign each component a weight based on its quality, and derive a conclusion
based on the overall weight of evidence. When the conclusions are controversial and the
authors have financial ties to affected parties, allegations that conflicts of interest
influenced the analysis are inevitable.14 While the model policy described in this paper
does not include this restriction, journals may choose to adopt this strategy – insisting
that writers of reviews and editorials adhere to a strict no-conflict-of-interest policy – as
one way of ensuring high ethical standards for their journal.

Those journals that choose to adopt this standard will not close themselves off
from including the opinions of their fields’ leading thinkers if they have financial ties to
industry. In addition to reviews and editorials, many journals allow signed commentaries
to appear in their pages. Such commentaries, as long as they are accompanied by
appropriate disclosure statements, would still be allowed under such a policy. Nor should
it make it difficult for editors to find reviewers or editorialists. The talent pool is broad
enough to accommodate such a policy. For instance, leading institutions in the field of
evidence-based medicine have adopted strict rules limiting the use of conflicted scientists
on its clinical practice guideline-writing committees. The U.S. Preventive Services Task
Force at the Agency for Healthcare Research and Quality excludes scientists from
specific discussions if they have extensive ties to companies with a stake in the
outcome.15 The Office of Medical Applications of Research within the National Institutes
of Health, which impanels scientists to develop “state of the science” statements on
controversial or underutilized medical technologies, prohibits scientists with relevant
conflicts of interest from serving on its panels.16

But note the caveat in that last sentence. When does a potential conflict of interest
become relevant? Here, existing policies provide very little guidance. One journal

12
Angell M, Kassirer JP, “Editorials and conflicts of interest,” NEJM. 1996; 335(14):1055-6.
13
Drazen JM, Curfman GD, “Financial associations of authors,” NEJM. 2002; 346: 1901-2.
14
Daniel Coyne, “Influence of Industry on Renal Guideline Development,” Clin J Am Soc Nephrol 2:3-7,
2007.
15
Email communication from Mary Barton, Scientific Director, U.S. Preventive Services Task Force, Nov.
7, 2007.
16
See NIH Consensus Development Program (http://consensus.nih.gov/aboutCDP.htm#panel, accessed
October 1, 2007).
recommends that authors unsure if a particular financial arrangement is relevant “should
always err on the side of full disclosure.”17 This presumes that there are some
relationships that need not be disclosed, even to editors. A safer standard would require
that authors disclose all financial or other relationships on the forms submitted to editors,
and draft a statement to accompany the submission that reflects the author’s interpretation
of relevance. The editors will then have the information readily at hand to review the self-
reported statement and make adjustments if they think greater disclosure is warranted.

This safeguard still doesn’t provide guidance to editors or authors for determining
relevance, however. If an author has a financial relationship with a company on a subject
entirely apart from the subject of the article, is that relevant? If the financial relationship
is with a company that competes with the company whose product is being discussed in
the article, is that relevant? If the financial relationship with the competitor company is
on a subject entirely apart from the subject of the article, is that relevant? How about
patents and pending patents in an area related to the article subject, held either directly or
by the researcher’s institution? What if research funding comes from a non-profit group
or research institute that is funded by industry? What percent of a research institute’s
funds must come from industry before it is considered industry-funded for disclosure
purposes? How long does a financial relationship persist?

The safest course for avoiding even the appearance that authors or editors are
hiding a potentially relevant conflict of interest would be to err on the side of greatest
disclosure, while still recognizing that some relationships are clearly irrelevant. Any
financial relationship with a company with a direct stake in the contents of a submission,
whether or not that relationship is on that subject, ought to be disclosed under the theory
that one might hesitate to bite the hand that feeds one. The same is true for a relationship
with a firm that competes with another firm’s product that is the subject of the article,
even if that relationship isn’t on that subject. But one could imagine a scenario in today’s
corporate world where two firms compete on some products, but not on the product that
is the subject of the article. If the author’s financial relationship with the competitor firm
is on a different subject, too, then the article need not disclose this relationship, which
could be referred to as a conflict-of-interest twice removed.

Defining relevance for patenting disclosure is difficult. Clearly, a patent on a fuel

cell technology needn’t be disclosed in an article about receptors on a particular type of
cancer cells. But a pending patent identifying the potential medical use of a receptor on
one type of cancer cell should be disclosed in an article about a similar receptor that
happens to be on a different type of cancer cell. Arguments that this is not relevant will
ring hollow if later that patent is licensed to a company developing a drug that influences
both receptors. As commercialization has pushed further upstream in the scientific
process, the potential utility of patented inventions has become much more difficult to
predict. Scientific advances in one area may increase the value of patented technologies
in another area that at first blush years earlier may have seemed only marginally related.
A broad definition of relevance that errs on the side of greater disclosure for patents and
pending patents is warranted.
17
Fontanarosa et al.,
 In recent years, there has been a proliferation of trade associations, social
advocacy groups and non-profit research foundations in the U.S. and other countries.
They not only fund and conduct research, but may be involved in recruiting scientists to
engage in advocacy and promotional activities.1 Many of these groups receive grants
from private firms; some are entirely funded by industry. Yet the exact purpose of these
organizations are not readily apparent to the general public, scientists and editors through
their names, nor is their financial support from industry.18 What level of industrial
support qualifies one of these non-profits or advocacy groups as industry-funded? If a
grant is for a specific research or promotional project, then recipients of those funds
clearly have a conflict of interest for purposes of disclosure. And if a group is entirely
funded by industry, then any grant from that group, even if industry support is
unrestricted, should be considered a conflict of interest. But what if the group receives
only part of its funds in unrestricted grants from industry? We believe 50 percent or more
funding from industry is a reasonable standard for qualifying a group as industry-funded,
and whose financial support for a researcher or project should be disclosed as such.

Another issue that must be addressed by guidelines is how far back to look in
determining conflicts of interest. Clearly, there is a statute of limitations on a financial
relationship. Some journals and organizations have adopted a standard that says the
relationship should be current. In other words, if the check was written yesterday for
work conducted in the past, the conflict of interest no longer exists for disclosure
purposes. This has been rejected by the vast majority of journals with disclosure policies,
whose look-back periods range from one to five years. Given the long lead time between
the onset of scientific work and publication, a three-year look back period is a reasonable
minimum standard.

Some journals and organizations with conflict-of-interest disclosure policies also

set dollar levels below which conflicts need not be declared. This is probably unwise for
disclosures by authors to editors because this could wind up hiding potential conflicts of
interest that, while small, are still relevant. In no case should a relevant conflict of interest
remain undisclosed to readers.

Finally, there is the question of whether to report non-financial competing

interests such as academic rivalries or political and religious beliefs. There is a general
consensus that while they do exist and can influence research outcomes, they are hard to
identify and difficult to describe for purposes of disclosure. Many researchers are
unaware of their own biases or are unconcerned about their potential impact on research
outcomes, believing that the scientific method will be self-correcting even though many
experiments are too expensive, complicated, or narrowly conceived to be confirmed
through the traditional safeguard of replicating results. Yet in recognizing that some
effort to disclose non-financial conflicts as a source of potential bias is warranted, there is
as yet no objective basis for establishing a standard. The relationship between financial

18
Typical examples include the Center for Medicine in the Public Interest, which receives funds from the
pharmaceutical industry, and the Foundation for Lung Cancer Detection, Prevention and Treatment, a non-
profit affiliated with Weill Cornell Medical College and funded by Liggett & Myers, a tobacco firm.
conflicts of interest and research results is well documented. Similar research about non-
financial conflicts of interest doesn’t exist. Therefore, a voluntary approach to disclosure
of relevant non-financial conflicts of interest has been outlined in the model policy.

Ensuring that the disclosed information is complete and accurate ultimately

depends on the good will of the investigators, reviewers and editors who voluntarily
submit this information. A common understanding of what constitutes a conflict of
interest and when it should be disclosed would facilitate voluntary compliance with
journal policies. This model policy represents one attempt to arrive at that common
understanding.
 MODEL CONFLICT OF INTEREST DISCLOSURE GUIDELINE
FOR SCIENTIFIC AND MEDICAL JOURNALS

For Authors:

All manuscripts for articles, original research reports, editorials, comments,

reviews, book reviews and letters that are submitted to the journal must be accompanied
by a conflict of interest disclosure statement, or a declaration by the authors that they
have no conflicts of interest to declare. All articles that are published in the journal will
be accompanied by a conflict-of-interest disclosure statement, or a statement that the
authors have replied that they have no conflicts of interest to declare. If the journal prints
unsigned editorials, they should not be written by anyone with a conflict of interest.

To facilitate this policy, all authors must privately disclose to the editors of the
journal at the time of their submission ALL potential conflicts of interest. These include
all financial and non-financial interests and relationships (see below for definitions),
direct employment with a private sector entity (whether full or part-time), and service on
private sector and non-profit boards and advisory panels, whether paid or unpaid. Authors
should also disclose to editors any conflict of interest that may have influenced either the
conduct or the presentation of research, including but not limited to close relationships
with those who might be helped or hurt by the publication, academic interests and
rivalries, and any personal, religious or political convictions relevant to the topic at hand.

In the paper, authors should include a draft statement that discloses all relevant
conflicts of interest and affiliations. Relevance for financial conflicts of interest with
private firms is defined as a relationship of any value with a firm with a stake in the
subject of the manuscript, or its competitors. Relevance for patents is defined as any
invention or pending invention connected in any way to the subject. Since relevance is
often in the eye of the beholder, you should err on the side of full disclosure in drafting
the disclosure statement. Editors will check your draft against the private financial
disclosure statement, and initiate discussions toward possible adjustments if necessary.

What to report: Any financial relationship from the past three years (dating from
the month of submission) of any size should be disclosed. These potential conflicts of
interest include:
 Direct employment, either full or part-time;
 Grants and research funding (but not grants to your institution or others within
your institution on which you did not work); this includes grants from trade
associations and non-profits substantially (50 percent or more) funded by
private-sector firms;
 Consultancies;
 Travel grants, speaking fees, writing fees and other honoraria;
 Paid expert testimony for one side in an adversarial proceeding (this does not
include testimony as a factual witness in a civil or criminal case);
  Patents granted, pending and applications, whether or not generating royalties;
 Stock ownership, investment in related “sector” funds, or stock options,
including those of immediate family members but excluding diversified mutual
funds and investment trusts; and
 Membership on private sector scientific or other advisory boards, whether paid
or unpaid.

In addition, any current negotiations regarding future employment or current job

offers, either full- or part-time, must be disclosed.

In disclosing these financial arrangements to editors, authors can include dollar

amounts even though they will not be printed in the journal. Editors may choose to
exclude this information from publication, but in no case should an editor or author
consider an arrangement irrelevant based on its size alone.

Non-Financial Conflicts of Interest: Authors may have strongly-held views about

the article being submitted for publication. Authors should consider disclosing and
editors may choose to print any affiliations or expressions of these views that may be
relevant. These may be personal, political, or intellectual and may include any expression
of strongly held views relevant to the subject of the submission. Such disclosures may be
original, or they make reference to opinions previously expressed in books or
monographs, op-eds or public comments, or to sworn testimony before or lobbying of
legislators or legislative bodies. Disclosable non-financial conflicts of interest would also
include membership or affiliation with non-governmental organizations that have an
interest in the submission.

For Editors, Articles Editors, and Peer Reviewers:

As a general rule, journals should require that all senior editorial personnel
(editors in chief, managing editors, full-time assistant editors) avoid all financial
relationships that might constitute a conflict of interest. Editorial managers should also
avoid personal, political, or intellectual entanglements, organizational or otherwise, that
could be construed as establishing a particular bias that might influence one’s judgments.

The well-established principle in editorial management is that there should be a

firewall separating editorial decision making from publishers, advertisers, circulation
managers and anyone connected with the business side of the journal’s operations.
Editors who maintain financial ties with companies or institutions that have an interest in
the content of a journal undermine the editorial independence that is crucial to a journal’s
credibility.

Scientific publishing is a highly competitive enterprise. The number of

manuscripts rejected far exceeds the number accepted at most journals. Requiring editors
to remain conflict-of-interest free assures rejected authors that bias or potential bias did
not enter into the editorial decision to reject their manuscript. Similarly, readers have the
right to know that the choice of articles, choice of reviewers, and choice of editorial or
commentary writers was made by senior editorial managers whose judgment was not
influenced by financial conflicts of interest or ideological bias.

If a journal editor or senior editorial manager has a conflict of interest, it must be

disclosed in each edition of the journal near the masthead and be found easily in the
online edition. Such notice should also be included on the author guidelines or
information page for electronic submissions. The same rules that govern author
disclosures should govern editor disclosures.

Some journals publish the names of article editors in the article. Some do not.
Peer reviewers have traditionally remained anonymous. Whatever the journal’s policy,
peer reviewers and articles editors should follow the same rules as authors for disclosing
conflicts of interest. In other words, peer reviewers will send disclosure forms to the
editors, even though this information does not appear in the publication. This gives
editors the tools they need to ensure that peer review panels for individual papers are
fairly balanced.

Enforcement

Conflict of interest disclosure relies on the honor system. Editors do not have the
time, resources or inclination to serve as financial auditors or ideological arbiters.
Successful disclosure policies depend on the good will and integrity of editors, authors,
and peer reviewers. But even the highest quality journals with well-established conflict-
of-interest disclosure policies do not have universal compliance.19 This may be, in part,
due to a lack of understanding about the rules for disclosure, which this model policy is
meant to address.

Most cases of a failure to disclose relevant conflicts of interest arise because

someone brings it to the attention of the editors, either directly or through attention in the
press. In such cases, editors should investigate the allegation. If true, editors should
determine the cause of the failure to disclose a relevant conflict. The editors may wish to
establish a permanent panel that includes external parties to review difficult cases, or
consult their editorial board.

In all cases of failure to disclose a relevant conflict of interest, editors will publish an editor’s note
that becomes part of the permanent record of that article. In those rare cases where editors uncover a willful
desire to hide financial conflicts of interest, the editors will consider appropriate penalties, such as refusing
to allow that author to publish in the journal for a specified period of time. Some journals have adopted a
three-year ban for such instances.20

19
Goozner M, “Unrevealed: Non-Disclosure of Conflicts of Interest in Four Leading Medical and
Scientific Journals,” Center for Science in the Public Interest, July 2004 (available at
http://www.cspinet.org/new/pdf/unrevealed_final.pdf, accessed September 26, 2007).
20
Goehl TJ, “Embracing Scrutiny,” Environmental Health Perspectives, October 2004 (available at
http://www.ehponline.org/docs/2004/112-14/EHP112pa788PDF.pdf, accessed September 26, 2007).
Acknowledgements

The authors would like to thank the following for helping to underwrite the cost of
developing this model policy: Beldon Fund; Richard & Rhoda Goldman Fund; Josiah
Macy, Jr. Foundation; and Nathan Cummings Foundation.

Disclosure

Merrill Goozner directs the Integrity in Science project for the Center for Science in the
Public Interest, which lobbies Congress and regulatory agencies for laws requiring
greater disclosure of conflicts of interest in the science literature and on federal advisory
committees.

Authors:

Merrill Goozner, M.S.

Director, Integrity in Science Project
Center for Science in the Public Interest
1875 Connecticut Ave. NW, Suite 300
Washington, DC 20009
Tel: (202) 777-8374
Fax: 202-265-4954
E-mail: [email protected]

Arthur Caplan, Ph.D.

Emanuel and Robert Hart Professor of Bioethics
University of Pennsylvania
3401 Market Street, Suite 320
Philadelphia, PA 19104
Tel: (215) 898-7136
E-mail: [email protected]

Jonathan Moreno, Ph.D.

Professor, University of Pennsylvania
Department of History and Sociology of Science
303 Logan Hall, 249 S. 36th Street
Philadelphia, PA 19104-6304
Telephone: (215) 898-7629
E-mail: [email protected]

Barnett S. Kramer, M.D., M.P.H.

Editor-in-Chief, Journal of the National Cancer Institute
8120 Woodmont Ave., Suite 500
Bethesda, MD 20814
Telephone: (301) 496-1508
Email: [email protected]
Thomas F. Babor, M.D., M.P.H.
Associate Editor-in-Chief, Addiction
Department of Community Medicine and Health Care
University of Connecticut Health Center
263 Farmington Avenue, MC 6325
Farmington, CT 06030
Phone: (860) 679-5470
Email: [email protected]

Wendy Cowles Husser, M.A., M.P.A.

Executive Editor, Journal of the American College of Surgeons
633 N Saint Clair Street
Chicago, IL 60611
Tel: (312) 202-5306
Email: [email protected]