Correspondence
Sponsorship, Authorship, and Accountability
N Engl J Med 2002; 346:290-292January 24, 2002
- Article
To the Editor:
Two comments seem in order with respect to the revised “Uniform Requirements” document (Sept. 13 issue).1 First, no mention is made of the more than 100 for-profit medical communication companies.2 In addition to ghostwriting entire medical research articles,3,4 these third parties — funded wholly by drug and device manufacturers — form partnerships with investigators supported by industry sponsors to ensure that manuscripts submitted for publication have the proper spin.
Second, the document states that editors should publish information about potential conflicts of interest “if they believe it will be important to readers in judging the manuscript.” This raises an important question: Does withholding information about relevant financial conflicts from readers — for any reason — impair their ability to assess the potential for bias in the design and analysis of trials? I believe the answer is self-evident.
4 ReferencesLaurence Landow, M.D.
7620 Old Georgetown Rd., Bethesda, MD 20814
laurence.landow@verizon. net 1
Davidoff F ,DeAngelis CD ,Drazen JM , et al. Sponsorship, authorship, and accountability. N Engl J Med 2001;345:825-827
CrossRef | Web of Science | Medline2
Relman AS . Separating continuing medical education from pharmaceutical marketing. JAMA 2001;285:2009-2012
CrossRef | Web of Science | Medline3
Cullen DJ . Ghostwriting in scientific anesthesia journals. Anesthesiology 1997;87:195-196
CrossRef | Web of Science | Medline4
Rennie D ,Flanagin A . Authorship! Authorship! Guests, ghosts, grafters, and the two-sided coin. JAMA 1994;271:469-471
CrossRef | Web of Science | Medline
To the Editor:
The authors of the editorial “Sponsorship, Authorship, and Accountability” are right to be concerned about commercial control of clinical trials. They forthrightly declare that “a submitted manuscript is the intellectual property of its authors” but then quickly seem to lose their nerve, adding that commercial sponsors should not have “sole control of the data” and that “contracts should give the researchers a substantial say in trial design.” Such language, whatever its good intentions, concedes too much power to sponsors.
Authors should have full control of the data and total “say” over the conduct and interpretation of the trial. They may agree to share the raw data with their sponsors and allow them time to review the manuscript before submission, but if the work is to be credible, a minimal requirement is that the authors of a paper reporting a sponsored clinical trial have the same responsibility for the work as the authors of any other published research. Such an unequivocal statement is missing from the editorial and the appended “Uniform Requirements.”
Arnold S. Relman, M.D.
Channing Laboratory, Boston, MA 02115-5804To the Editor:
Academic researchers must have complete freedom to participate in and approve all aspects of industry-sponsored clinical trials, including any publication resulting from such a trial. We encourage all authors to abide by the new rules.
That having been said, the editorial has serious defects. Its biased tone slights the vital contributions of our industry to the clinical trial process. Throughout the editorial, the integrity of academic investigators is assumed, whereas the industry's integrity is constantly questioned. The editorial ignores the fact that the sponsors of trials usually do most of the actual work in clinical trials. There are many well-respected, highly ethical, and experienced clinicians working for industry who are not subject to the pressures that affect most academics, such as the need to obtain grants and secure tenure.
Given that our industry supports the need to ensure the independence of researchers, the editorial is unnecessarily antagonistic, which weakens its effect. A more balanced approach would have recognized the essential roles of both industry and investigators in developing safe and effective medicines for patients.
Alan F. Holmer
Pharmaceutical Research and Manufacturers of America, Washington, DC 20005To the Editor:
The editorial states that sponsors “must impose no impediment, direct or indirect, on the publication of the study's full results.” No mechanism is suggested for enforcing this requirement. I suggest that this be an enforceable requirement for approval by institutional review boards (IRBs).
There is a very strong argument to be made that there is an ethical imperative to register1,2 and publish3 the full results of every clinical trial: patients who volunteer to be subjects for a trial do so with the understanding that they are helping to advance medical science. Since the research results must be made public in order to advance medical science, investigators make a commitment to the subjects — implied, even if not stated — to publish the full results of the research. This ethical issue makes enforcement of publication an appropriate matter for IRBs. I suggest that IRB approval for a clinical trial should be contingent on the commitment of the investigator to publish the results of the study within two years after the last observation is made. If the results are not published within this period, then the IRB should not approve any further studies by the investigator or the sponsor of the study until publication has occurred.
3 ReferencesMarcus M. Reidenberg, M.D.
Weill Medical College of Cornell University, New York, NY 10021
mmreid@med.cornell. edu 1
Horton R . Medical editors trial amnesty. Lancet 1997;350:756-756
CrossRef | Web of Science | Medline2
Reidenberg MM . Decreasing publication bias. Clin Pharmacol Ther 1998;63:1-3
CrossRef | Web of Science | Medline3
Reidenberg MM . Releasing the grip of big pharma. Lancet 2001;358:664-664
CrossRef | Web of Science | Medline
To the Editor:
On behalf of seven of the largest contract research organizations (CROs) — Covance, Icon Clinical Research, Inveresk Research Group, Kendle International, Parexel International, PPD Development, and Quintiles Transnational — I agree with the editors that all research must be conducted and reported objectively, dispassionately, and with the highest levels of scientific accuracy and integrity.
The perception that there is “head-to-head” competition between CROs and academic sites is mistaken and does not portray accurately the roles, objectives, and operations of CROs in the clinical research process. CROs work on behalf of a sponsor in a highly regulated environment1 to implement and manage a clinical trial according to the study protocol. They provide research services, including consultation in study design, facilitation of the recruitment of investigators and study patients, assurance of the protection of patients, assurance of the integrity of the data and the data analysis in order to maximize the quality of the research, and guidance through the complex regulatory environment. CROs do not sponsor clinical trials, do not own the data from trials, do not provide routine care for patients, and do not participate in agreements concerning publication rights and responsibilities, which are negotiated between sponsors and investigators. The contractual obligation of the CRO is to ensure the integrity of data and compliance with Food and Drug Administration and international regulations, not the presentation of specific results. The provision of such services enhances — rather than erodes — the quality and standards of clinical trials.
Both academic and community-based investigators participate in CRO-managed clinical investigations, and many of them participate in the development of study protocols. In CRO-managed studies, the investigator is neither our employee nor our customer but an integral partner in the research process. The breadth of our research spectrum encourages us to seek the best and the brightest physician-scientists in all clinical disciplines. We maintain that CROs contribute to high-quality clinical research by working in collaboration with — not competing against — clinical investigators at both academic medical centers and community-based clinics.
1 ReferencesDouglas Peddicord, Ph.D.
227 Massachusetts Ave., NE, Washington, DC 20002on Behalf of Covance, Icon Clinical Research, Inveresk Research Group, Kendle International, Parexel International, PPD Development, and Quintiles Transnational
To the Editor:
As a clinical investigator in an academic institution, I have experienced firsthand the problems with the integrity of the clinical-trials process and resulting publication that your journal and others are trying to prevent. In my experience, large, established pharmaceutical companies generally conduct themselves in an ethical and professional manner and are not the problem; rather, it is the smaller companies, with less experienced employees and more to lose from the publication of disappointing trial results, that are responsible for the worst abuses. Not only are investigators or institutions who try to negotiate with such companies to retain some degree of academic freedom unlikely to obtain funding for a trial, but also they and their institutions are subject to retribution. The following is an excerpt from recent correspondence between a pharmaceutical company and the legal department at my institution regarding a clinical trial in which we hoped to participate that I believe illustrates the problem: “Yours is one of the only institutions that has objected to our standard confidentiality clause. While I understand your position, I was frankly not overly pleased with our compromise and have asked [employee's name deleted] to convey to our clinical staff that I would prefer that [company's name deleted] not use your institutions for clinical trials in the future unless necessary.”
Thank you for your efforts. I hope they help.
Elizabeth Loder, M.D.
Harvard Medical School, Boston, MA 02114
eloder@partners.org Author/Editor Response
The editorialists reply:
To the Editor: Dr. Landow points out that third parties, paid by sponsors, may be involved in the preparation of manuscripts. We reassert our original point that authorship implies both independence and accountability. We expect that all authors will provide, and take responsibility for, a dispassionate analysis of the data that they have collected. With respect to disclosure, each journal editor will need to make his or her own decision about what financial data are relevant; as we noted in the editorial, if there is any question, it makes sense to err on the side of disclosure.
The arguments of Dr. Relman and Mr. Holmer represent two poles in the debate over studies funded by industry and performed by academics. Academic investigators and industrial sponsors both contribute to the accrual of new medical knowledge. An appropriate balance is struck when a novel treatment can be tested in an environment in which all data are freely available to all investigators. When a report is prepared for publication, the primary goal must be the fair, honest, accurate, and complete dissemination of the accrued information. The system is best served when there is no attempt to present the results obtained in a more favorable light by selective reporting of the data.
Dr. Reidenberg raises the possibility of reporting the data from all clinical trials, regardless of the outcome. In many cases, publication in a journal is an endorsement of the data, but sometimes firm conclusions cannot be drawn. For example, data may not be published because the randomization process did not result in balanced treatment groups or because a differential-treatment trial that was designed to show the superiority of one treatment did not achieve this result but had insufficient power to show the equivalence of treatments or the inferiority of any treatment. In such circumstances, it may make sense to post the data analysis in a public electronic repository where interested parties could view it.
The CRO consortium argues that it performs an auxiliary function in clinical trials and that such organizations are not in competition with academic investigators. We did not mean to diminish the role of the CROs; these organizations provide substantial benefit with respect to the orderly accrual of complex sets of clinical research data. However, since CROs act as agents for the sponsor, it is possible that their choice of sites for the performance of research may not be based solely on the selection of “the best and brightest” but, rather, on the selection of those who are willing to accept restrictive terms. Indeed, the letter quoted by Dr. Loder demonstrates that sponsors seek to avoid investigators who will not accept restrictive terms in research contracts.
International Committee of Medical Journal Editors (ICMJE)
- Citing Articles (8)
Citing Articles
1
Xavier Bosch. (2011) Exorcising ghostwriting…. EMBO reports 12:6, 489-494
CrossRef2
Edward Abraham. (2009) Better infrastructure: Industry–academia partnerships—A marriage of convenience?. Critical Care Medicine 37:Supplement, S159-S164
CrossRef3
Dennis Paustenbach. (2008) Re: A rebuttal: Secret ties to industry and conflicting interests in cancer research, March, 2007; 50(3):227-233. American Journal of Industrial Medicine 51:9, 714-716
CrossRef4
Dave Davis. (2007) Continuing education, guideline implementation, and the emerging transdisciplinary field of knowledge translation. Journal of Continuing Education in the Health Professions 26:1, 5-12
CrossRef5
Jeffrey Peppercorn, Emily Blood, Eric Winer, Ann Partridge. (2007) Association between pharmaceutical involvement and outcomes in breast cancer clinical trials. Cancer 109:7, 1239-1246
CrossRef6
Richard A. Brand. (2003) Writing for Clinical Orthopaedics and Related Research. Clinical Orthopaedics and Related Research 413, 1-7
CrossRef7
Richard A. Brand, Joseph A. Buckwalter, Charlotte L. Talman, Daniel G. Happe. (2003) Industrial Support of Orthopaedic Research in the Academic Setting. Clinical Orthopaedics and Related Research 412, 45-53
CrossRef8
Marcus M. Reidenberg. (2002) Conflict of interest and medical publication. Science and Engineering Ethics 8:3, 455-457
CrossRef
