The education of medical students should be based on the best clinical information available, rather than on commercial interests. Previous research looking at university-wide conflict of interest (COI) policies used in Canadian medical schools has shown very poor regulation. An analysis of COI policies was undertaken to document the current policy environment in all 17 Canadian medical schools.
A web search was used to initially locate COI policies supplemented by additional information from the deans of each medical school. Strength of policies was rated on a scale of 0 to 2 in 12 categories and also on the presence of enforcement measures. For each school, we report scores for all 12 categories, enforcement measures, and summative scores.
COI policies received summative scores that ranged from 0 to 19, with 0 the lowest possible score obtainable and 24 the maximum. The highest mean scores per category were for disclosure and ghostwriting (0.9) and for gifts and scholarships (0.8).
This study provides the first comprehensive evaluation of all 17 Canadian medical school-specific COI policies. Our results suggest that the COI policy environment at Canadian medical schools is generally permissive. Policy development is a dynamic process. We therefore encourage all Canadian medical schools to develop restrictive COI policies to ensure that their medical students are educated based on the best clinical evidence available, free of industry biases and COI relationships that may influence the future medical thinking and prescribing practices of medical students in Canada once they graduate.
Citation: Shnier A, Lexchin J, Mintzes B, Jutel A, Holloway K (2013) Too Few, Too Weak: Conflict of Interest Policies at Canadian Medical Schools. PLoS ONE 8(7): e68633. doi:10.1371/journal.pone.0068633
Editor: K. Brad Wray, State University of New York, Oswego, United States of America
Received: March 20, 2013; Accepted: June 5, 2013; Published: July 4, 2013
Copyright: © 2013 Shnier et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The authors have no support or funding to report.
Competing interests: In 2007 Joel Lexchin was a consultant to a law firm acting for Apotex Inc. In 2008 he was an expert witness for the Canadian federal government in its defence against a lawsuit challenging the ban on direct-to-consumer advertising. In 2010 he was an expert witness for a law firm representing the family of a plaintiff who allegedly died from an adverse reaction from a product made by Allergan. He is currently on the Management Board of Healthy Skepticism Inc. and is the Chair of the Health Action International - Europe Association Board. Barbara Mintzes is on the Health Action International - Europe Association Board. Joel Lexchin and Barbara Mintzes are Academic Editors at PLOS ONE. This does not alter the authors’ adherence to all the PLOS ONE policies on sharing data and materials. The other authors report no conflicts of interest.
Conflicts of interest with industry may occur in medical education in the classroom, in the conduct and reporting of research, at the bedside, and in the treatment of patients. The education of medical students should be based on the best clinical information available, unbiased by the commercial interests of industries marketing pharmaceutical or other health products. In many Canadian medical schools, students are taught by faculty who work in partnership with industry, e.g., receive research grants from companies, serve on companies’ speakers’ bureaus or advisory committees, or own shares in companies.  The financial relationships of faculty with industry may affect, or reasonably appear to affect, the integrity of their academic or publishing interests, professional medical opinions, and the information that they disseminate to medical students. ,  These relationships between medical faculty and industry represent conflicts of interest (COI) and compromise not only the public’s confidence and trust in medical researchers and universities, , ,  but also the potential for robust, evidence-based clinical education for medical students. .
When medical school faculty members have ties with, or financial interest in, pharmaceutical companies, they are more likely to report results that are favourable to the sponsoring companies.  Faculty with financial COI tend to publish significantly more, and at a higher rate, than faculty without industry relationships.  At the same time, these faculty members are also more likely to conduct lower quality, but more commercializable research, as compared with those who undertake independently funded research. ,  Quality of research is evaluated based on the following criteria: whether clinical trial data is selectively reported, the medication being tested in a trial is compared to one that is known to be inferior, inappropriate doses of a competitor drug are used in a trial, and the length of clinical trials is altered to produce data that is favorable to the sponsors’ drugs, among other methods. , .
COI relationships are present not only in the classroom, but also surface when industry provides resources to medical schools. Although corporate pharmaceutical funding for education may offer educational opportunities for students, these programs tend to provide students with industry-friendly information, which can compromise clinical judgment if it is at odds with the scientific evidence. For example, between 2002 and 2006, the pain management course for medical and other health science professional students held at University of Toronto was partly funded by grants from Purdue Pharma LP, the maker of OxyContin. As part of the course, a chronic pain management book that was funded and copyrighted by Purdue Pharma was distributed to the students by a lecturer who was external to University of Toronto and had financial ties to Purdue Pharma. Concerns were raised that some of the contents of the book were not consistent with the current best evidence for narcotic medication administration.  Without effective, stringent COI policies at medical schools to regulate such interactions between faculty, students, and industry, medical students are subject to direct or indirect interactions with industry, as well as industry resources, that have the potential to influence their future medical thinking and prescribing practices.
The implementation of COI policies has been effective in altering the future prescribing practices of medical residents. Epstein and colleagues conducted an analysis of the antidepressant prescribing practices of 1652 graduates from 162 psychiatric residency programs in the US before 2001 and after these programs adopted COI policies in 2008.  The authors found that residents who graduated before the introduction of COI policies in 2001 tended to prescribe less appropriately than 2008 graduates, where inappropriate prescribing was defined as prescribing heavily marketed and brand reformulated antidepressants (e.g., extended release products) at a higher rate. Furthermore, 2008 residents who graduated from programs with maximally restrictive COI policies prescribed these drugs significantly less often than 2008 graduates from programs with minimally restrictive COI policies.
The Association of Faculties of Medicine of Canada has voted to support the 2008 report by the Association of American Medical Colleges ,  to better manage and, when necessary, prohibit interactions between academics and industry that can create COI and undermine professionalism standards. Previous research on COI policies as applied to Canadian medical schools has shown very poor regulation. Mathieu and colleagues used the American Medical Students Association (AMSA) scorecard to analyze COI policies at Canadian universities that host medical schools.  They found that the university-wide policies were generally weak in the areas of faculty-industry relationships, samples, sales representatives, on-site and off-site training, industrial relationships, and educating students about COI.
However, the scope of Mathieu and colleagues’ study was limited because the authors only analyzed university-wide COI policies and not those specific to medical schools. Further, they omitted the Northern Ontario School of Medicine (NOSM) from their analysis. In addition, they did not contact the universities directly and only relied on institutional policies found via a web search. Finally, only a single coder evaluated the COI policies. To address these limitations, we undertook an analysis of COI policies at both the university and faculty levels to document the current COI policy environment in all 17 Canadian medical schools.
A list of all 17 medical schools (14 English language and 3 French language) in Canada was obtained from the web site of the Association of Faculties of Medicine of Canada (AFMC) <http://www.afmc.ca/faculties-e.php>. The web site of each of the schools was searched in late July 2011 for policies related to COI or documents interpreting policies using the terms “policy”, “policies”, “conflict-of-interest”, “conflicts-of-interest” and “COI” in English, and “politique” and “conflit d’intérêts” in French. The name of each policy and the latest of either the date of adoption or the date of the policy’s most recent review were recorded. After a preliminary list of policies for each school was assembled, an e-mail with the list of policies in English or French, as appropriate, was sent to each dean explaining the purpose of the study and requesting confirmation that this list contained the pertinent policies for the particular medical school. These emails also requested that the deans send us any additional policies we might have overlooked, or draft policies not yet in place. The deans were informed that we were only interested in publicly available policies and while respondents’ names would be confidential, the medical schools and their policies would be identified in any subsequent publication. Two reminder emails were sent at one-month intervals. We did not search for, request, or analyze policies from affiliated teaching hospitals.
Policies that were approved as of the end of September 2011 were analyzed. A grading system was modified from those that were already used by AMSA,  Chimonas and colleagues,  and Mason and Tattersall  for 12 different categories:
- gifts (including meals)
- consulting relationships (excluding scientific research and speaking)
- industry-funded speaking relationships and speakers’ bureaus
- honoraria, ghostwriting
- industry sales representatives
- on-site education activities
- compensation for travel or attendance at off-site lectures and meetings
- industry support for scholarships and funds for trainees
- medical school curriculum (or other documentation of educational objectives and course content)
AMSA uses a 0 to 3 scoring system where a score of 0 indicates that schools failed to respond to its request to send their policies. Since we initially identified policies using a web search, AMSA’s definition of what constituted a score of 0 was not relevant. Both AMSA and Mason and Tattersall regard a permissive policy as equivalent to the absence of a policy. We graded each category on a scale of 0 to 2, where 0 = no policy or permissive, 1 = moderate, and 2 = restrictive. (See Appendix S1 for the detailed scoring criteria for each individual category.) In addition, we scored enforcement measures: is it clear that a party is responsible for general oversight to ensure compliance and is it clear there are sanctions for noncompliance? Each of these enforcement measures was scored either “yes” or “no.” We did not attempt to identify if policies had been violated or to grade the severity of sanctions.
Scoring was done by two groups of two people, one for English (AS, KH) and one for French language schools (BM, AJ). Each person independently scored the policies and then compared results within their group. Disagreements were resolved through discussion. Once the scoring was completed, a follow-up e-mail in the appropriate language was sent to each dean. This email included the preliminary scoring for the medical school along with the policies that we used to obtain the score, an explanation of how each area was scored and a request that the dean review the scores for accuracy and notify us if he or she felt that a score was inaccurate. We also requested that the deans send us any new policies developed since the initial contact, but noted that the scores would be based on policies in place as of the end of September 2011. We asked the deans to respond within one month, and if we had not heard from them at that point, two further e-mail reminders were sent at one-month intervals.
After a response from the deans, the scores were reviewed by the original set of scorers, and a final set of scores was derived for each school. Similarly to Chimonas and colleagues,  we summed the scores in the first 12 individual categories for each school to come up with a summative score. Each category was weighted equally since each was identified as vital by a combination of the American Board of Internal Medicine-Institute on Medicine as a Profession (ABIM-IMAP), American Association of Medical Colleges (AAMC), and the Institute of Medicine (IOM).  We view this weighing system as also being applicable to the Canadian situation since the AFMC has endorsed the report by the AAMC on industry funding of medical education.  For the enforcement categories, the number of “yes” and “no” for each school was summed. We report scores for each category for each school, the summative scores for each school, and the mean for each category.
Since we collected only publicly available information about medical schools’ COI policies, the Human Participants Review Committee at York University, which approved this project, waived the requirement for informed consent from the deans of the schools that we contacted.
Of 17 medical schools contacted, 15 responded to the initial request for policies. Via web searches and responses from deans’ offices, we found a total of 50 policies and documents interpreting policies (collectively referred to as policies). Schools had as few as zero (NOSM) and as many as 8 relevant policies (University of British Columbia) per school. In addition, two deans sent us course outlines used in the teaching of COI to medical students. The dates of 16 policies were either not given (9) or were unclear (7). For the other 34 policies, seven were more than 10 years old (one dated back to 1976), while 12 were passed within two years of September 2011 (Table 1). Twenty-one policies were at the medical school level and the remainder (29) were university-wide.
Eleven schools responded when we asked them to review their initial scores, resulting in the revision of scores for five schools that provided policies that we had initially overlooked or that they had not initially sent us. In addition, two schools informed us that they had put in place new policies since our initial survey, while seven were in the process of developing or updating policies.
COI policies received summative scores that ranged from 0 (NOSM) to 19 (Western University, formerly University of Western Ontario), where 0 was the lowest score possible and 24 was the maximum score (Table 2). Twelve of the 17 schools scored less than 12/24 (50% of the maximum) and only one scored more than 18/24 (75% of maximum). Cumulative scores of 5 or less reflected ratings of mostly 0 (no policy or permissive) for each category, whereas cumulative scores of 8 or more reflected ratings of 1 or 2 (moderate or restrictive, respectfully) for most categories. The highest mean scores were assigned to disclosure and ghostwriting (0.9) and for gifts and scholarships (0.8). Policies on sampling received the lowest average score (0.2), followed by policies on sales representatives (0.3) and speaking and curriculum (0.4). No school had a restrictive policy that applied to samples. Of note, no category received a mean score of 1 or better (Table 3). Many COI policies with a rating less than 2 for disclosure failed to require disclosure of both past and present financial ties with industry on a publicly-available website and/or disclosure of any relationships to patients when this relationship may represent a COI.
Fifteen of 17 schools had policies that identified a party responsible for enforcement of the policies (Table 2). Examples of responsible parties included “Department Head or equivalent” and “Department Chair, Dean or immediate supervisor.”
Eleven of 17 schools had policies that specified sanctions for noncompliance (Table 2). An example of such a policy from McGill University contains sanctions ranging from counselling of the individual involved all the way to termination for cause. Ten schools had policies that met requirements for both a specific party responsible for enforcement and specified sanctions for non-compliance.
The 17 Canadian medical schools received scores that ranged from 0 to 19 out of a possible maximum score of 24. The score of 0 was received by NOSM. This low score may reflect, in part, the fact that the school was only established in 2005. Western University received the highest score of 19. Of the 17 medical schools in Canada, over half (10) received summative scores of 5 or less out of 24, indicating that in most of the categories they had either no policy or a permissive policy. No single category managed to achieve an average score of 1 or more.
Fourteen (82%) of the schools received a rating of 0 (no policy or permissive policy) for samples. Samples have been shown to influence medical residents’ prescribing practices, with negative implications both for costs and prescribing appropriateness. Adair and Holmgren have shown that access to drug samples increases the likelihood that physicians will prescribe heavily advertised and more costly drugs as opposed to cheaper or over-the-counter drugs.  We also found that most medical faculties (70%) had permissive policies or no policy concerning faculty involvement in companies’ speakers’ bureaus. The United States (US) Institute of Medicine’s recent report on COI recommended banning such relationships  because speakers’ bureaus represent part of a company’s promotional activities and the content is often under the company’s control. .
Similarly, 70% of medical faculties had permissive or no policies concerning interactions with sales representatives. Sales representatives have been found to negatively influence prescribing practices, e.g., to lead to more frequent and expensive prescribing and poorer prescribing quality.  In a comparative study, recently graduated internists who had studied in a program that restricted contact with sales representatives were more critical of the information they provided and saw sales representatives less often than internists from a medical school without such restrictions.  Most schools (70%) also failed to cover conflicts of interest or drug promotion in the curriculum. This gap has important implications for students’ abilities to understand the context within which promotional activities occur and to weigh their own responses to ethical challenges that might arise.  Finally, nearly all schools had a party responsible for enforcing their policies (15/17) and the majority had sanctions for violations (10/17), but we do not have information on how often these sanctions are applied or how effective they are.
We found that COI policies were most stringent in the areas of disclosure, ghostwriting, gifts, (considered to be the easiest to prohibit ) and scholarships. These results parallel findings that AMSA obtained in its annual reviews of policies in US medical and osteopathic schools. Its 2012 analysis found that the policy areas that received the highest ratings were those that addressed scholarships, off-campus continuing medical education, purchasing, and gifts.  The importance of restricting gifts is emphasized in a review of COI policies at 14 American medical schools that found that exposure to a gift restriction policy during medical school was associated with reduced prescribing of two out of three newly introduced psychotropic medications. .
Our findings on ghostwriting are consistent with those of Chimonas and colleagues, even though their rating scale separated out no policy (score = 0) and permissive policies (score = 1).  They found that, although existing ghostwriting policies at American medical schools were among the most stringent of all of the policy areas, ghostwriting was also the most neglected policy area. Furthermore, other work has shown that meaningful sanctions for academic fraud are generally absent.  Because universities reward academic faculty for their publication records, limited enforcement can mean that faculty may find themselves complicit in ghostwriting activities, in spite of policies prohibiting them.
A similar study of Australian medical schools found that their COI policies were even weaker than those at Canadian schools. Eleven out of 15 schools received less than 50% of the maximum possible number of points and only one barely exceeded 66%. All schools either had no policies or had policies that were unlikely to have a substantial effect on behavior in the areas of on- and off-campus educational activities. Lastly, policies on consulting relationships and disclosure had mean scores below 50%. .
Our study, in conjunction with the ongoing AMSA survey, the analyses of the US schools by Chimonas and colleagues, and the results from the Australian schools, clearly establishes that the poor control of COI at medical schools is not confined to a single country, but is an issue that needs to be addressed at both national and international levels. One effort to engage medical students in these issues has come from a collaboration between the World Health Organization and Health Action International that has resulted in a manual to teach medical students about pharmaceutical promotion . The manual is available in English, French, Russian and Spanish, and has been distributed across a wide range of countries.
This study has some limitations. Two schools did not respond to our initial request for any policies that we might have missed in our web search. Six medical schools failed to review our ratings despite repeat requests; their input could have validated, or alternatively, contradicted our findings. Furthermore, only medical schools’ COI policies were within the scope of our study, so we did not consider the policies of affiliated teaching hospitals (e.g., on samples or sales representatives). Hospitals may have had more restrictive policies, but this is unlikely based on previous research. .
Policy development is a dynamic process, and some Canadian medical schools have introduced new policies since September 2011, while others continue to revise their policies. It is important for medical schools to continue to develop and improve their COI policies to mitigate institution-industry relationships and to address the ways in which those relationships may affect the information that is taught to, and the attitudes of, medical students. Policies must also continue to develop, especially since the role of industry within universities continues to evolve. .
Practices that were once entrenched into medical culture, including the receipt of gifts, food, and drug samples, in addition to faculty consulting and speaking engagements with industry,  should no longer play direct or indirect roles in the education of medical students. Student-industry interactions can influence students’ education.  Students who have more contact with industry tend to have more favorable attitudes towards these types of interactions.  It has been reported that students who receive gifts from industry feel obliged to rely on industry representatives for information on medications. .
More stringent policies are not the only answer for helping to ensure medical education is free from faculty COI, but such policies have been shown to limit the acceptability of promotional items. ,  Medical schools across Canada are encouraged to achieve the most effective and stringent policies to regulate industry relations with both faculty and students.
Grading System for Categories in Policies.
Conceived and designed the experiments: JL. Performed the experiments: AS JL BM AJ KH. Analyzed the data: AS JL BM AJ KH. Wrote the paper: AS JL BM AJ KH.
- 1. Hébert P, MacDonald N, Flegel K, Stanbrook MB (2010) Competing interests and undergraduate medical education: time for transparency. CMAJ 182: 1279.
- 2. Cho MK, Shohara R, Schissel A, Rennie D (2000) Policies on faculty conflicts of interest at US universities. JAMA 284: 2203–2208.
- 3. Ehringhaus SH, Weissman JS, Sears JL, Goold SD, Feibelmann S, et al. (2008) Responses of medical schools to institutional conflicts of interest. JAMA 299: 665–671.
- 4. Association of American Universities Task Force on Research Accountability (2001) Report on individual and institutional financial conflict of interest. Washington, DC: Association of American Universities.
- 5. Grande D, Shea J, Armstrong K (2011) Pharmaceutical industry gifts to physicians: patient beliefs and trust in physicians and the health care system. Journal of General Internal Medicine 27: 1760–1763.
- 6. Busing N (2008) Canadian faculties of medicine not in denial. CMAJ 183: 463.
- 7. Zinner DE, Bjankovic D, Clarridge B, Blumenthal D, Campbell EG (2009) Participation of academic scientists in relationships with industry. Health Affairs 28: 1814–1825.
- 8. Downie J, Herder M (2007) Reflections on the commercialization of research conducted in public institutions in Canada. McGill Health Law Publication 1: 23–44.
- 9. Ubelacker S (2010) Med school pain course revised over concerns about possible pharma influence. Macleans.ca.
- 10. Epstein AJ, Busch AB, Asch DA, Barry CL (2013) Does exposure to conflict of interest policies in psychiatry residency affect antidepressant prescribing? Medical Care 51: 199–203.
- 11. Association of American Medical Colleges (2008) Industry funding of medical education: report of an AAMC task force. Washington, D.C.: AAMC.
- 12. Mathieu G, Smith E, Potvin M-J, Williams-Jones B (2012) Conflict of interest policies at Canadian universities and medical schools. Some lessons from AMSA PharmFree Scorecard. Bioéthique Online 1.
- 13. AMSA: American Medical Student Association (2012) AMSA PharmFree Scorecard 2011–2012. AMSA.
- 14. Chimonas S, Patterson L, Raveis VH, Rothman DJ (2011) Managing conflicts of interest in clinical care: a national survey of policies at U.S. medical schools. Academic Medicine 86: 293–299.
- 15. Mason PR, Tattersall MHN (2011) Conflicts of interest: a review of institutional policy in Australian medical schools. MJA 194: 121–125.
- 16. Association of Faculties of Medicine of Canada (2008) Industry funding of medical education: anticipated questions and associated answers. Ottawa: AFMC.
- 17. Adair RF, Holmgren LR (2005) Do drug samples affect residents’ prescribing practices? A randomized trial. American Journal of Medicine 118: 881–884.
- 18. Institute of Medicine (2009) Conflict of interest in medical research, education, and practice. Washington DC: National Academies Press.
- 19. Steinbrook R (2009) Controlling conflicts of interest - proposals from the Institute of Medicine. New England Journal of Medicine 360: 2160–2163.
- 20. Spurling GK, Mansfield PR, Montgomery BD, Lexchin J, Doust J, et al. (2010) Information from pharmaceutical companies and the quality, quantity, and cost of physicians’ prescribing: a systematic review. PLoS Medicine 7: e1000352.
- 21. McCormick BB, Tomlinson G, Brill-Edwards P, Detsky AS (2001) Effect of restricting contact between pharmaceutical company representatives and internal medicine residents on posttraining attitudes and behavior. JAMA 286: 1994–1999.
- 22. Austad KE, Avorn J, Kesselheim AS (2011) Medical students’ exposure to and attitudes about the pharmaceutical industry: a systematic review. PLoS Medicine 8: e1001037.
- 23. SCCPD working group on industry relations (nd) SSCPD position paper on the role of industry in university-based CME/CPD.
- 24. King M, Essick C, Bearman P, Ross JS (2013) Medical school gift restriction policies and physician prescribing of newly marketed psychotropic medications: difference-in-difference analysis. BMJ 346: f264.
- 25. Stern S, Lemmens T (2011) Legal remedies for medical ghostwriting: imposing fraud liability on guest authors of ghostwritten articles. PLoS Medicine 8: e1001070.
- 26. World Health Organization, Health Action International (2010) Understanding and responding to pharmaceutical promotion: a practical guide. Geneva.
- 27. Naylor C (2002) Early Toronto experience with new standards for industry-sponsored clinical research: a progress report. CMAJ 166: 453–456.
- 28. Rothman DJ, Chimonas S (2010) Academic medical centers conflict of interest policies. JAMA 304: 2294–2295.
- 29. Grande D, Frosch DL, Perkins AW, Kahn BE (2009) Effect of exposure to small pharmaceutical promotional items on treatment preferences. Archives of Internal Medicine 169: 887–893.
- 30. Austad KE, Avorn J, Kesselheim AS (2011) Medical students’ exposure to and atttitudes about the pharmaceutical industry: a systematic review. PLoS Medicine 8: e1001037.
- 31. Industry funding working group (2011) Defining the relationship: an evidence based review and recommendations on the role of industry funding in medical schools. Canadian Federation of Medical Students.
- 32. Sierles F, Brodkey A, Clearly L, McCurdy FA, Mintz M, et al. (2005) Medical students’ exposure to and attitutdes about drug company interactions: a national survey. JAMA 294: 1034–1042.