Over recent years, organized medicine and the public at large have increasingly scrutinized how relationships between physicians and industry, predominantly the pharmaceutical industry (“Pharma”), affect all levels of medical education from medical students to practitioners. This commentary reviews how Pharma’s marketing efforts have affected medical education, how some academic programs help learners comprehend and cope with marketing biases, and further steps that the profession and its learners might undertake to better understand and deal with influences that marketing biases exert on the medical database.
We appreciate that Pharma provides many important and valuable contributions, enhancing public wellbeing and professional knowledge in multiple ways that might not otherwise occur. However, the marketing-cum-education activities that result from Pharma companies’ pursuits of market share and profits are concerning. These actions are especially prominent in the increasingly Darwinian world of “me too” drugs, where similarities in effectiveness and adverse effect profiles often outweigh differences. In these efforts, marketing interests and tactics color the information on which contemporary evidence-based medicine is presumably practiced.
For all intents and purposes, the debate about whether Pharma marketing biases affect medical education has been settled. A steady flow of detailed, analytic reporting has appeared in major journals, books written by distinguished former editors of the New England Journal of Medicine (1–3) and JAMA (4), among many others, and books written by highly regarded investigative reporters (5–7). Conflicts of interest among academic physicians with Pharma ties and Pharma’s influence on continuing medical education (CME) have been scrutinized by Senators Charles Grassley and Herbert Kohl. Their inquiries are premised on the likelihood that Pharma slants the dissemination of information concerning the safety and effectiveness of treatments and consequently adversely affects prescribing behavior and drives up the costs of health care (8, 9). Medical students have recognized that their faculty’s industry ties may bias what they are taught in key aspects of therapeutics, resulting in the American Medical Student Association developing national scorecards for grading medical schools on policies regarding faculty-industry ties. The 11 dimensions assessed in these scorecards include policies governing disclosures, transparency, and limits to gifts and honoraria for off-campus CME, purchasing, curriculum, consulting, medication samples, speaking, on-campus CME, and access of pharmaceutical sales representatives to physicians and trainees. As of June 2009, of 149 U.S. medical schools, nine received As (6%), 36 Bs (24%), 18 Cs (12%), 17 Ds (11%), and 35 Fs (23%), 23 because they failed to submit policies or declined to respond. Another 27 schools were awaiting assessment of their newly designed policies (10).
How Do Trainees and Clinicians Learn Therapeutics, and How Do Marketing Biases Affect These Processes?
The traditional “hydraulic” model of pedagogy suggests that we learn therapeutics by having teachers, supervisors, “influence leaders,” texts, and journals pour information into us. But it is obviously more complicated than that. To start, learners are not passive recipients. We eagerly seek out knowledge to reduce the morbidity and suffering of patients and to bolster our sense of therapeutic competence and mastery. Consequently, we seek better and faster, but not necessarily cheaper, ways to treat—as long as the costs of treatment do not come out of our pockets. Desperate for better tools and outcomes, we may sometimes minimize adverse aspects of new treatments and not think too much about costs. Many of us are easy sells.
That said, medical learners are beset by multiple sources of information. First, many academic institutions shield residents and medical students from pharmaceutical representatives while attending physicians and faculty may still have considerable direct contact with them and their academic variants known as scientific liaisons. In office practice, many clinicians receive most of their information about new and existing (proprietary) products from pharmaceutical representatives. Some clinicians have one to two visits per month, although these practices appear to be waning.
Next, medical learners are influenced by the popular media, professional advertising, and patients who have been “informed” by direct-to-consumer advertisements, clever product placements, and celebrity mentions in the media. Pharma utilizes infomercials passed on as legitimate news items on TV programs, and public groups advocating for specific disorders are heavily subsidized by Pharma’s generous contributions. The effectiveness of promoted treatments is often oversold.
Of course, much information comes from informal interactions with trusted authorities, peers, teachers, and key opinion leaders. We learn from lecturers in medical schools and speakers at grand rounds and CME. Many of these people have developed financial relationships with Pharma on speakers’ bureaus and advisory boards and as research collaborators (11). A recent Association of American Medical Colleges (AAMC) (12) survey reported that 54% of academically related CME activities in 2008 were industry funded.
Finally, we constantly learn from esteemed peer-reviewed journals and authoritatively written textbooks, but how carefully do we assess the extent to which marketing bias creeps into these ordinarily trustworthy sources of evidence-based medicine? Because randomized controlled trials are often used as the gold standard for evidence-based medicine and the primary measure by which the Food and Drug Administration (FDA) decides on the safety and effectiveness of medications requesting approval for specific clinical indications, investigators have examined the extent and nature of industry-sponsored randomized controlled trials in psychiatry. One study (13) found that 239 published randomized controlled trials concerning psychiatric medications (60%) were industry sponsored. In another study (14), up to 62% of randomized controlled trials in general medicine were industry sponsored. Furthermore, results of industry-sponsored randomized controlled trials for antidepressants, for example, were five times more likely to be favorable for the sponsor’s products than those of randomized controlled trials funded by independent sources such as NIMH (12). Articles by Heres et al. (15), colorfully titled “Why olanzapine beats risperidone, risperidone beats quetiapine, and quetiapine beats olanzapine: an exploratory analysis of head-to-head comparison studies of second-generation antipsychotics,” and Safer (16), among others (17), illustrate the ways in which key elements in the initial clinical trial design can be selected to ensure in advance that the sponsor’s product will be favored (Table 1). Because clinical trials are very expensive and the primary intention of the sponsor is to ensure that its product will be advantaged, it behooves the sponsoring company to do whatever it can to make certain that its products will shine. As a result, information published in journal article abstracts and results and discussion sections are likely to emphasize favorable aspects of the trial and minimize harms. Because practicing clinicians mostly read journal article abstracts and are much less likely to read the methods and results sections, the take-away messages from these abstracts are often designed to highlight the main points that marketing teams hope to convey.
Depending on industry affiliations, other conflicts of interest, and the rigor with which specific authors read and report the literature, biased results may creep into review articles, meta-analyses, practice guidelines, and textbooks. Lauded textbooks carry a great deal of influence in educating medical students, residents, and practicing clinicians, but even highly respected textbooks occasionally include declarative statements and treatment recommendations based on uncritical acceptance of marketing biased results.
Educational Programs Preparing Students to Understand and Cope With Marketing Influences
Given that industry affects education in numerous ways, how can educators and students prepare themselves to discern and cope with marketing biases in therapeutic information? Psychiatric educators have grappled with these issues for some time (18). Using the e-mail list of the American Association of Directors of Psychiatric Residency Training (AADPRT), we conducted an informal survey of psychiatric educators regarding current innovative methods for teaching and preparing medical students and residents about industry marketing practices.
At the University of California Davis, workshops have been staged in which pharmacy students trained in the role of pharmaceutical representatives interactively role play with medical students. At Mayo, fourth-year medical student classes in pharmacology (related to clinical use rather than basic pharmacology) typically begin by analyzing drug company advertisements.
Using lectures, seminar series, and experiential learning, many departments described similar and overlapping approaches (e.g., University of Michigan, University of California at San Diego). Presentations woven into psychopharmacology courses or special teaching sessions typically cover topics such as comparisons of core values and ethical principles of the medical versus business professions. These sessions may include differing perspectives on customers versus patients; pharmaceutical industry economics, including costs related to drug development and marketing; analysis of drug company advertisements; reviews of drug company marketing strategies; and evidence from social psychology data that accepting even small gifts often influences physician practice patterns.
Experiential teaching methods at the University of Buffalo, Cambridge Hospital, Northwestern University, and Southwestern/University of Texas, among others, include such activities as grand rounds on Pharma prepared by residents; inviting pharmaceutical representatives to individually present their products for 10 minutes, immediately followed by critical analyses from faculty of the information; expanded versions of the previous strategy, in which panels of pharmaceutical representatives present as a group followed by residents questioning the pharmaceutical representatives and then critical analyses conducted by faculty of all the information presented after the pharmaceutical representatives leave; role plays in which postgraduate year 3 (PGY-3) residents act as pharmaceutical representatives to “detail” PGY-1s; and discussions of problem-based ethical scenarios, such as a physician speaking at an industry-sponsored dinner meeting and prescribing that company’s product.
At the University of Colorado at Denver, Feinstein et al. (19) have developed and implemented an elaborate evidence-based medicine (EBM) curriculum addressing many of these issues, including how to read randomized controlled trials, practice guidelines, and meta-analyses and how to identify clinical trial design biases. Elements of the 4-year course include “real-time” question formulation and literature searching and utilizing EBM in outpatient settings. Feinstein et al. report that other programs use current cases to conduct EBM daily ward teaching rounds and more elaborate weekly or biweekly EBM outpatient and ward teaching rounds; these programs also conduct EBM journal clubs with queries from the ward or clinics, an 8-week session on problems encountered in clinical practice, 2-week internal medicine EBM elective rotations, and EBM in the morning report.
Continuing Medical Education
How can we engage in lifelong learning and obtain CME in which marketing biases play a lesser role? Several models are available. First, APA has developed a lifelong learning system around FOCUS, a CME journal that includes examinations and can provide 40 hours/year of category 1 CME. The current price is $292/year, placing this well within the reach of most practicing psychiatrists. In addition, every year numerous quality lectures are available from presentations at the APA annual meeting involving NIMH, NIDA, NIAAA, and other nonindustry sources. APA now makes available on DVD the best lectures from annual meetings. The Research Channel on cable TV offers scores of university-derived presentations suitable for academic CME. These presentations can easily be packaged with learning goals and objectives, CME questions, frequently asked questions, discussion guides, and other appurtenances necessary to qualify for CME accreditation for local audiences.
Further, rethinking exactly what constitutes and qualifies for category 1 CME credit may be in order. Ongoing peer supervision groups focusing on current patients seen in clinical practice combined with readings on pertinent practice-based research and journal clubs that follow quality guidelines are often more instructive and salient than CME presentations that may or may not relate to an individual’s practice. The ACCME and CME granting bodies might create better models for monitoring and accrediting such activities.
Academic departments of psychiatry and local district branches of APA can certainly survey their stakeholders regarding what they need to learn and can ask faculty to prepare scholarly updates for regularly scheduled grand rounds. District branches may do the same for local practitioners. A large number of “industry free” CME resources are also available for perusal (e.g., at http://pharmedout.org/pharmafree.htm).
Additional Activities to Prepare Students to Deal with Pharma Marketing Efforts
Current and future practitioners and educators who wish to learn how physicians are “targeted” for sales by Pharma and how experimental biases are used in clinical trial research designs can review the resources in Table 2. The goal of these pedagogical exercises is to enable learners to become increasingly discerning skeptics who can nevertheless judiciously retrieve useful information from marketing-influenced sources. Through such knowledge the practitioner may more carefully think about how to separate the actual potential added value of new products from hype.
Advanced Experiential Learning
Beyond seminars and journal clubs, several experiential and problem-based exercises also may help trainees and practitioners reconsider the effect of industry advertising on prescribing patterns. Data from New York State’s PSYKES study, in which participating clinics were informed about “polypharmacy” prescribing patterns involving antipsychotics, clearly demonstrate that simply providing participating clinics with system-level feedback can substantially improve clinicians’ prescribing patterns, even without additional incentives (M. Finnerty, personal communication and presentation to the Executive Committee on Practice Guidelines of APA, July 27, 2009).
Should comparative effectiveness of psychiatric treatments with regard to benefits, harms, and costs be embedded in routine practice and training? In clinical teaching and practice settings, newcomers might receive lists of the most commonly prescribed medications and their costs, and from then on they might be expected to estimate the ongoing costs of various medication options for their patients’ health plans, out-of-pocket expenses, and benefits and harms during initial and discharge treatment planning. As already occurs in many managed care and public settings, prescribing more expensive proprietary medications rather than less expensive generic medications might require evidence-based justification.
Consider this exercise: a practitioner is personally rewarded by receiving 10% of the cost savings accrued from prescribing less expensive, rather than more expensive, equally effective medications (according to the best available evidence). To what extent might such an incentive alter the prescribing habits of practitioners and trainees?
Reducing the Extent of Marketing-Driven Bias in Databases
Although we have focused primarily on biases related to Pharma marketing, industry biases in research design, in how findings are reported, and in advertising directed toward professionals and directly to consumers clearly pervade all types of business enterprises. Beyond Pharma, we have seen industry biases in research and advertising materials concerning medical devices pertinent to psychiatry, hospital and residential treatment centers, and, most recently, genetic testing programs that purport to provide accurate and clinically meaningful information regarding individual risk for major psychiatric disorders such as bipolar disorder. Most journals have been increasingly scrupulous about searching out industry biases and reporting objectively (20). However, even these sources have not always been able to achieve complete freedom from bias. Editors of Cochrane Reviews and practice guidelines have been increasingly aware of potential marketing biases and have expanded efforts to eliminate them. Despite scrupulous efforts to the contrary, textbooks and CME materials still occasionally propagate information slanted by industry marketing. Asking authors to simply disclose their industry ties is not enough. Pages of such disclosures already appear in the major texts and in CME programs, but these disclosures do not ensure that the information disseminated has been sufficiently vetted for the absence of biases. Some publications that offer synopses of current medical literature, such as The Medical Letter and Journal Watch, accept no advertising to reduce potential concerns about industry ties. Community standards for reviewing clinical trials and accepting their conclusions for texts and systematic reviews deserve to be upgraded. Editors should invariably reject papers from clinical trials in which marketing-related design biases are evident, and authors of systematic reviews and textbooks should reject results and conclusions from such publications. Thanks to continuing improvement in these efforts, we trust that treatment-related information available to trainees and practitioners in the future will reflect fewer marketing-driven slants than has sometimes been the case.