By Harriet Rosenberg and Adrienne Shnier
May 9th 2016
Part 1: Weak Policies and Ghostly Authorship
This blog is presented in interview form and is an amalgamation of an interview plus several subsequent conversations between Adrienne Shnier, a graduate student at York University and Harriet Rosenberg, a retired York University professor. Adrienne’s PhD thesis is titled Medical education and financial conflict of interest relationships with the pharmaceutical industry in Canada: An analysis of four areas of medical education and will be available online in August 2016.
HR: Your research is part of a reform movement within medicine that focuses on the influence and impact of conflicts of interest—especially financial conflicts of interest.
Your thesis research and publications have been about the education of Canadian physicians, broadly speaking, and the role that conflict of interest awareness and policies have played in shaping medical science and ultimately patient care. You argue that neo-liberal government policies, in Canada and around the world, over the last 3 decades have played an important role in transforming the practices and culture of medicine.
AS: Policies that are pro-business have contributed to an environment of commercialization of medical information and the protection of that information as proprietary. This has led to trends toward the privatization and commodification of science in ways that most people don’t realize.
For example, new for-profit research organizations have emerged in recent years. Relationships with these types of organizations are now embedded institutionally, and so the study of policies around conflict of interest (COI), especially, financial conflicts of interest (fCOIs) becomes very important. Pharmaceutical industry funding for research has expanded dramatically. For example, it is worthwhile noting that as of 2005 the pharmaceutical industry has provided 70% of total funding for clinical trials in the US (see here and here).
HR: The concept of conflict of interest, especially financial conflict of interest seems pretty straightforward. At a minimum, medical students, practicing physicians, and researchers should be aware of and avoid any potential for bias associated with taking money from the pharmaceutical industry. Yet the concept and associated policies have been fought over for decades. Those supporting close ties with industry argue that everyone has conflicts and that the term is itself disrespectful, pejorative and condescending.
They view those who want stringent conflict of interest policies as being sanctimonious moralizers. But those who want to separate medicine from the influence of the pharmaceutical industry point to a vast literature of revelations from investigations, lawsuits, legal documents, and academic articles and books that show how industry has manipulated clinical trials to exaggerate benefit and downplay harms.
AS: A solution to these debates about fCOIs, is to remove them from the personal, individualistic realm and have policies that apply to all rather than having to investigate and pursue case-by-case situations. Having clear, enforceable policies in the first place and then applying pre-specified penalties would help to protect the integrity of the profession.
HR: You looked at issues of fCOI within the Canadian context. There were four areas that came under the broad umbrella category of the education of doctors. You started with the education of medical students in your published 2013 paper.
AS: My co-authors and I looked at 17 Canadian medical schools to evaluate the presence of and stringency of fCOI policies.
HR: Your title says it all, “Too few, too weak.”
AS: We were concerned about the degree to which the schools protected their faculty and students against fCOI with industry. We were also curious about whether the potential for hidden fCOIs and commercial messages existed in medical school education — for example, did teaching faculty disclose their ties to industry? Were faculty permitted to have fCOIs with industry? Were medical students taught formally, as part of the curriculum, about how to identify fCOIs and behaviors and attitudes associated with engaging in them?
HR: You also asked about disclosures of financial links to industry in educational materials.
AS: The category of on-site education asked whether industry was permitted to provide direct financial support for educational activities, including Continuing Medical Education (CME). Approximately half of the schools did not address this category; whereas, approximately one-third of the schools had policies that allowed industry funding to be allocated to a specific topic, but that funding should be provided to the department rather than an individual. The remaining schools had policies that protected learners through prohibiting the provision of industry funds directly to educational activities, while still allowing unrestricted grants.
HR: Were freebies like lunches, pens and other trinkets examined for how they legitimated industry influence?
AS: No, but approximately half of the schools had policies on gifts, while the other half did not.
HR: What is the take-home message here?
AS: Most schools had weak policies. Commentators on our article in the press were surprised. After this study was published, some medical school administrators said they would tighten, or were in the process of tightening, their fCOI policies.
HR: Your research also includes the commercialization of medical journals and the concept of “honest publishing” in medical journals. The influence of industry on how clinical trials are conducted and written about in medical journals—including the highest profile medical journals— is an area of conflict of interest not very well known to the public.
People assume that medical science journals publish accounts of clinical trials that are conducted by academic researchers whose names appear on the title page and that these trials describe with care and honesty the potential benefits and harms of a drug —that they serve the public interest.
It is quite a shock to learn that things often don’t actually work that way and that leading physicians, including many medical journal editors themselves, warn of corruption in trials. These are people like Dr. Marcia Angell, former editor of the New England Journal of Medicine and long-time former editor of the British Medical Journal, Dr. Richard Smith among others, who wrote that the clinical trial was now the “marketing arm” of the pharmaceutical industry. A recent article by the redoubtable Stanford University Professor John Ioannidis argued that “evidence-based medicine has been hi-jacked” by commercial interests.
AS: It is also important to note that from the perspective of industry, data from clinical trials about the harms of their products can be considered to be commercially confidential information (CCI), which should not be released because it would give their competitors an advantage. In some cases, adverse event data is considered to be proprietary and, therefore, potentially inaccessible by the public see here).
HR: Some have called this the dominance of MBM—Marketing-Based-Medicine— over EBM, Evidence Based Medicine. In the now available documents from court cases we see “publication plans” with lists of numbered marketing messages to be inserted.
AS: The whole structure of clinical trials and journal publishing has been transformed by the creation of for-profit businesses that serve the pharmaceutical industry outside of the environment of academia.
There are companies called CROs—Contract Research Organizations—that are hired by pharmaceutical companies. Researchers have estimated that CROs are now responsible for conducting up to 75% of medical research (see here). The CROs are for-profit research businesses that conduct clinical trials—they have global reach—and recruit from around the world. They administer the trials, collect the data, and analyze the data. We know from documents from court cases that they are in close contact with the pharmaceutical companies who hired them throughout the process. Some of these documents that show the communications involving for-profit research and writing organizations are available on-line from the documents archive site called Industry Documents Library at the University of California San Francisco (see here and here).
These documents show memos and long-term marketing plans for about 5 year periods, and connections with other related businesses called Medical Education Communication Companies and Medical Writing Organizations. In this context, we have come to understand that the terms “medical writer” or “medical editor” indicate the involvement of a ghostwriter. Medical writers are hired through these organizations as contract workers to write up the results of trials, prepare conference posters and presentations, slide kits for physicians to present to colleagues, write letters to medical journal editors and develop a variety of “defense of product” communications when the product is challenged. There are important similarities between pharma, tobacco, and other industries in this regard.
These CROs and/or writing/editing businesses also seek out academic physicians as guest authors to put their names on the mastheads of trial write-ups and to be the face of the trial in medical journals. This practice has been revealed in legal transcripts and materials that judges may choose to unseal in court cases and are available on the Industry Documents Library. You can read the memos, the emails, the publication plans with proposed titles and slots for authors to-be-decided later (see here and here).
A recent 2016 publication by Cosgrove and colleagues shows that these practices continue into the present and contribute an additional dimension to research on financial conflicts of interest —those of medical journal editors themselves (see here).
HR: So this is why you use the term “honest publishing.” The work of these businesses —the CROs, the medical writers/editors, the medical education businesses are all associated with the pharmaceutical industry and seem to be generally hidden from the public. One could say “hidden in plain sight” because they are easily found on the internet and the medical editors even have their own professional organizations and do not see themselves in a negative way as ghostwriters even though their names do not usually appear in journal articles.
High profile medical journals around the world, including within Canada, have been grappling with this issue of “honest publishing” and the now complicated and unreliable term “author.”
What can you tell us about your research on “honest publishing” in Canada?
AS: There is very little written about this in the Canadian context; however, the issues seem to be the same for journals and for physicians, for medical students and for the public internationally because when articles are published in medical journals, they reach international physicians, medical students, and patient audiences.
HR: This leads to the question: How can you actually figure out what is going on with an article—even in the highest prestige journals—did the listed authors do the research and did they themselves write up the results?
AS: This is difficult to know without complete transparency about the roles of individuals and research and writing companies in the publications process. In one of my thesis chapters, I provide a preliminary research tool that can be used to gauge the fCOI disclosures and transparency of individuals’ roles required in medical journals.
HR: So some articles are published and it may be the case that the listed authors did not even see the raw data. There is really no way to tell outside of what may eventually come out in lawsuits.
AS: Correct. We do not know if authors had access to the raw patient-level data or only the already-analyzed data from these companies.
HR: Sometimes in the acknowledgments of an article an “editor” is thanked for assistance and this may be a clue that a ghostwriter had significant involvement in the drafting and development of a manuscript. One of the risks in this situation is that because the medical writer’s employer has been hired by a drug company, this creates an environment that incentivizes the development of medical manuscripts that are generally favorable to the drug company.
AS: There is no way to know what the named authors have actually contributed, outside of the problematic ICMJE requirements for disclosure, onto which many journals have voluntarily subscribed….so the answer is, no.
HR: The ICMJE (International Committee of Medical Journal Editors) is a group of leading medical journal editors who got together to try to tighten rules around authorship given the scandals about ghost authorship and the lack of transparency about who was responsible for what parts of the trial write-up and what the role of industry was in all of this, including not only the text but the statistical analyses. This was brought to a head when over 1500 pages of documents were unsealed in relation to breast cancer lawsuits (against Wyeth, now Pfizer) by women who had taken Hormone Replacement drugs. The documents (available on line in the Industry Documents Library) revealed an entire backstage system of industry involvement in promoting the benefits of HRT and downplaying harms.
The medical journals now have tighter guidelines around authorship but it has been pointed out that there are many ways of getting around them.
Furthermore, write-ups in medical journals—regardless of who the authors are—are just a fraction of the story. The issue of data transparency and how the full record based on all the raw data gets condensed and possibly distorted in the published version is also very important. For example, Dr. Ben Goldacre has argued for transparency because the published record falls short. About 50% of trials are just not published. He and allies in leading medical journals like the BMJ have started an initiative called AllTrials to correct this.
HR: We will be looking at this more fully in our discussion with Dr. Elia Abi-Jaoude in a future blog.
You have developed a glossary of terms that comprise 50 items that help explain what has now become normative in medical journal publishing. Tell us about them.
AS: The glossary includes terms that describe various fCOI relationships and the involvement of industry in the research, data collection, data analysis, and writing processes. For example, do they disclose consulting for pharma, honoraria for speaking engagements for particular companies, do they (or family members) disclose owning stock, and do they appear on scientific advisory boards of pharmaceutical companies—just what are their ties to industry? Have the named authors had access to the raw data? Have the trials been registered in a clinical trial registry? Have any trials been suppressed? The answers to these, and other, questions at least allow the reader to put articles in medical journals into some context about their origins.
More difficult in journals is to find policies that address other aspects of industry influence. Do journals have policy language that addresses behind the scenes behavior? Do they have ways of detecting ghostwriting and preventing ghosted articles from being published? Can they verify if the names on the article are the people who did the work? Do they have penalties for breaches of policy?
Analysis of medical journal policies, themselves, were not a part of this study, but I hope to use the research tool to do this in a future study.
HR: How widespread is this problem of authorship?
AS: I have seen figures that indicate that between 9% and 49% of published medical articles are ghostwritten. At some phases in product promotion, it has been argued that 90% to 100% of articles during these key phases or within certain specialties are written by industry hired writers and not the academics whose names appear.
The industry-originated and planned research studies, data collection, analysis, interpretation, writing, and publication of these studies, are all the while being potentially informed by a larger promotional campaign for a drug. Collectively these coordinated undertakings have been called “ghost management” and are related directly to the commercial nature of the pharmaceutical industry.
Simply put: the meaning of authorship has changed.
End of Part 1
Angell, M. The truth about the drug companies. Random House Trade Paperbacks. 2005
Ford AR and Saibil, D. The push to prescribe. Women’s Press. 2009
Goldacre, B. Bad Pharma: How medicine is broken, and how we can fix it. 2013
Gotzsche, P. Deadly medicines and organized crime. CRC Press. 2013
Healy, D. Pharmageddon. University of California Press. 2013