This blog post discusses the article "Changing Interactions Between Physician Trainees and the Pharmaceutical Industry” published in the February 27, 2013 electronic edition of the Journal of General Internal Medicine. The study was conducted by former Fellow in the Lab on Institutional Corruption Kirsten Austad, Lab affiliates Aaron Kesselheim M.D. J.D. and Eric Campbell Ph.D., and Jerry Avorn M.D. and other members of the Division of Pharmacoepidemiology and Pharmacoeconomics in the Department of Medicine at Brigham and Women’s Hospital and Harvard Medical School.
Within a week of its online-first publication the article received worldwide coverage, including The Boston Globe, Stuttgarter Zeitung, Popular Science, Pharmalot, HealthDay (in English and Spanish), and a host of medical and legal themed news portals and blogs.
It is well known that physicians have frequent interactions with the pharmaceutical and device industries. While some interactions are in the context of research collaborations, others are more promotional in nature, and may involve sponsored educational dinners or free product samples (Wazana, JAMA, 2000). While recent surveys suggest that the public is skeptical of physician-industry relationships (Consumer Reports survey, Aug 2010), many doctors find these promotional relationships to be useful and deny that they influence medical judgment. Physicians’ acceptance of industry promotion as a routine part of medical practice may have its roots in the fact that marketing interactions begin early in medical training.
Like all professions, medical school socializes students in to the role of doctor, and each stage of training presents unique opportunities for interactions with industry marketing representatives. First-year students are “pre-clinical,” spending their time learning the fundamentals of their profession in a classroom setting where they may receive lectures from professors who also serve on speakers’ bureaus. During third- and fourth-year they transition to immersion in the hospital environment where they begin to learn the practical aspects of patient care and may observe the daily interactions between their supervising physicians and industry representatives, including free meals at sponsored lunch talks. After graduation, trainees enter residency where they carry out the patient care responsibilities and may utilize industry representatives present in the clinical environment as resources to inform their prescribing.
In the past decade, many academic medical centers have implemented policies to shield trainees from promotional interactions with industry. These policies include banning free meals or other gifts, mandatory disclosure of teaching faculty members’ conflicts of interest, and formal curriculum time devoted to learning about professional ethics. However, there is little empirical evidence to guide development of these policies and perhaps as a result, policies vary widely between institutions and there is no consensus on where efforts would best be focused.
Our study aimed to systemically examine how pharmaceutical and device industry promotional representatives interact with trainees over the course of their medical education, and how trainees view the role of industry marketing in medical education. We surveyed a large, random sample of first-year (pre-clinical) medical students, fourth-year (clinical) medical students, and third-year residents, representing all 121 U.S. allopathic (M.D.) schools. Demographics of our respondents compared favorably to the survey of graduating medical students conducted by the Association of American Medical Colleges (AAMC), which boasted an 83% response rate in 2010 (AAMC website), confirming that our sample was indeed representative.
Our results were surprising: despite the significant changes over the past decade, 33% of first-year students, 57% fourth-year students, and 54% residents reported accepting a gift within the past six months. Though this level of exposure is reduced compared with a more limited study of third-year medical students published in 2005 (Sierles et al, JAMA), this demonstrates that gift-giving is still prevalent. Receipt of gifts was common despite the fact that a minority of trainees felt it was appropriate for medical students and residents to accept gifts of less than fifty dollars in value (by year of training: 26% vs. 23% vs. 35%). Observing mentors’ interactions with industry was also common, with 33% first-years, 59% fourth-years, and 53% of third-year residents reporting this occurrence. The frequency of trainees using industry for educational purposes also increased with training level across a variety of sources, including sponsored lectures, sales representatives, and promotional materials.
It is well-documented that most physicians believe they are less susceptible to influence from gifting than their colleagues (Wazana, JAMA, 2000). In our survey, all levels of training were more likely to report that their peers are influenced by accepting gifts from industry than they are swayed (52% vs. 33% for first-years, 46% vs. 36% for fourth-years, and 42% vs. 34% for third-year residents). Though this was not a longitudinal survey, we noticed a potential trend based on year of training for certain attitudes. For example, while 68% of first-year students reported that physician-industry interactions threaten the public’s trust in doctors, only 55% of four-year students and 46% of third-year residents agreed with this contention. There are two potential mechanisms for such changes in attitude. One possibility is that as trainees learn clinical medicine through observation and emulation of mentors, they also absorb the views of the role models around them. Alternatively, if students accept gifts that are prevalent in the environment, they may subconsciously adopt attitudes that resolve the cognitive dissonance. Other research (Sah and Loewenstein, 2010) has suggested that perceived self-sacrifice is a powerful justification for residents in accepting gifts. Thus, since residents’ work hours and financial strain due to loan payment are relatively worse than medical students’, this could also mediate the attitudinal changes we observed.
Overall, students support policies regulating interactions between the profession and the pharmaceutical industry. Between 87 and 94% of trainees agreed with mandatory disclosure of professors’ conflicts, though fewer (43-57%) felt that it was inappropriate for professors to have conflicts on a topic relevant to what they teach students. While most students felt that schools should ban the pharmaceutical industry from access to students in the pre-clinical learning environment (66-69.3%), slightly fewer felt that the same policy should apply to the clinical training sites (53-60.3%).
To explore how the learning environment influences trainees, we looked at whether responses were related to research intensity and strength of conflict of interest policy of the respondents’ medical schools. Amount of NIH funding to schools served as a surrogate for research intensity, and we found that trainees from medical schools with high research intensity were half as likely to have accepted a gift from the pharmaceutical industry in the preceding six months. This result may reflect the reality that many conflict of interest policies were crafted in response to concerns about safety of human subjects of clinical research, and thus research-focused academic medical centers were under more pressure to craft policies. Additionally, drug manufacturers may be more likely to cultivate relationships (facilitated by gifting) with trainees at less research-intensive schools who are more likely to become community practitioners.
To evaluate the impact of conflict of interest policy strength, we used each school’s AMSA PharmFree Scorecard grades from 2008 and 2010, which rates academic medical centers from A to F based on eleven areas (including gifts, disclosure, and sales representative access to clinical areas). Interestingly, we found no association between respondents’ medical school Scorecard grade and frequency of accepting gifts. Why might this be? Because the grade is a composite measure, it is possible that not all of the domains that contribute to the grade modulate students’ exposure to industry, or their perceptions of interactions. Notably, a recent study (King et al, BMJ, 2013) that found that graduates of medical schools with strong policies prohibiting industry gifts were less likely to prescribe heavily promoted psychotropic medications (vs. clinically appropriate, inexpensive generic alternatives), as compared to those who trained in environments without such rigorous gift restrictions.
Another possible explanation for the lack of association that we found between survey responses and medical schools’ PharmFree Scorecard score is that existence of a conflict of interest policy—even a strong one—is not enough. First, our data showed that a surprisingly low number of respondents (ranging from 10% of first-years to 24% of residents) reported understanding their schools’ conflict of interest policies, suggesting that more trainees need formal orientations to their institutions’ policies. Second, it is possible that school policies may be undermined by other aspects of the environment. According to Hafferty (Academic Medicine, 1994), the “hidden curriculum” denotes the norms and values learned through informal mechanisms such as off-handed comments made by attending physicians or observed behaviors of peers and mentors, and is thought to be an important contributor to development of professionalism. For example, a medical student may know that his school forbids accepting lunches from industry representatives but rationalize this behavior as appropriate and accept it himself after being told by a supervising resident “if you want to survive in the hospital environment, you take free food when you can get it.” Trainees in an environment where a policy prohibiting gifts exists but is not adhered to by all faculty or affiliated institutions could paradoxically become more likely to accept gifts. Our survey indicates that there are lapses in compliance: only 29% of first-year students, 59% of fourth-year students, and 52% of third-year residents felt that their faculty complied with the policy “very well.” This disparity between the formal policy and the implicit lessons taught via the hidden curriculum may mediate behaviors and attitudes.
The potential impact of the hidden curriculum is one of many questions arising from our data that merit further investigation. In one forthcoming project, we will consider how trainees learn about medications and explore their ability to correctly differentiate evidence-based treatments for common clinical scenarios. In the future, we also hope to conduct a follow-up study in which our survey respondents are re-contacted later in their professional development and results linked to their prescribing patterns. This longitudinal data will help us further examine how various aspects of the medical school learning environment affect physicians’ attitudes and behaviors relating to pharmaceutical and medical device industry promotion.