Healthy Skepticism Library item: 19645

Warning: This library includes all items relevant to health product marketing that we are aware of regardless of quality. Often we do not agree with all or part of the contents.

Publication type: Journal Article

Avorn J, Choudhry NK
Controversies in Cardiovascular Medicine
American Heart Association 2011 Oct 11;
http://circ.ahajournals.org/content/121/20/2228.full

Abstract:

Funding for Medical Education: Maintaining a Healthy Separation From Industry

Full text:

Attitudes and policies about conflict of interest in medical education have evolved rapidly in the last several years. Temporally and conceptually, these changes have paralleled the rise of the relatively new field of behavioral economics in explaining commercial relationships. The traditional view of economics holds that in any transaction, all inputs and outcomes are simply the result of decisions by rational participants making choices that optimize their goals; however, the past decades have seen a reassessment of this “expected utility theory” with the growing awareness that unconscious biases and emotional factors profoundly affect behavior.1 The same insight has begun to influence how we think about medical education and the messages that are sent by and received from industry-sponsored speakers and programs. Because cardiology addresses the most common illnesses in developed societies and employs the most widely used and costly drugs and devices, these issues have been especially acute in this field.

Response by Harrington and Califf see p 2228
Unquestionably, the drug and device industries make important contributions to the development of products that reduce the morbidity and mortality of cardiovascular disease. Industry is also a major source of both funding for and collaboration in cardiology research, and it provides substantial support for professional cardiology societies, topics that are beyond the scope of this report. But in the realm of teaching, whether of medical students, house officers, or practitioners, there is growing concern that educational activities should not be supported or influenced by companies whose profitability and very existence depend on the sales volume of the products they make.2

Our position that industry should not directly support educational programs is not ideologically driven. Rather, it is based on the expectation that medical knowledge can be transferred most efficiently and with the least distortion if those responsible for this activity have minimal commercial interest in the topic being presented. After all, this assumption is no different from that underlying well-accepted policies at many journals and in other sectors of society. A person employed or funded by the maker of a drug or device is not permitted to author review articles or editorials on that topic at several journals, not because of the presumption of deceitfulness, but because of the concern over possibly unconscious bias that such relationships might induce. Many judicial and regulatory proceedings similarly require a person with commercial ties to a product to recuse himself or herself from deliberations related to that product. Medical education is an outlier in the extent to which we have traditionally ignored such conflicts in making decisions about funding and decision-maker participation.

Previous Section
Next Section
The Role of Industry-Funded Medical Education

Industry funding of medical education takes several forms, each of which in our view can pose important potential problems. The most direct form of industry-supported education is through funding for continuing medical education (CME) activities. Industry-funded CME providers include physician organizations, publication and education companies, medical schools, and hospitals, many of which derive substantial profit from these activities.3 According to the Accreditation Council for Continuing Medical Education, CME companies directly or jointly offered more than 100 000 CME activities accounting for more than 760 000 hours of education in 2008 alone.4 Support from pharmaceutical and medical device manufacturers for CME, which has quadrupled over the past decade, accounts for more than half of the $2.4 billion that is spent annually on CME.5 “Medical education companies” were established to act as intermediaries between physicians and industry, accepting funding from drug or device manufacturers and passing it along to physicians, medical schools, hospitals, or other organizations that offer lectures or educational programs. Although this arm’s-length “third-party” status is said to eliminate the manufacturer’s influence over the educational content of these programs, the fact that most medical education companies are heavily dependent on industry for their revenue raises questions about the adequacy of this separation. The world’s largest drug company, Pfizer, recently decided to stop funding educational programs through such intermediaries, recognizing that industry-supported programs should be labeled and acknowledged as such.6

There are other equally important but less visible forms of industry-funded education. “Speakers’ bureaus,” in which prominent physicians are paid to lecture about a company’s products, constitute another common form of industry involvement in teaching. Manufacturers may exert substantial influence over the content of these lectures and frequently provide the slides to be used; some physicians earn more than $50 000 per year from these talks alone. Other “key opinion leaders” (known in industry marketing language as KOLs) can earn up to $10 000 from a corporate sponsor for chairing a single educational symposium.

The sales representative, or “detailer,” is a major source of drug information for many physicians. This is especially true for new products, for which there may be little or no information available in the medical literature.7 On average, cardiologists meet with pharmaceutical sales representatives 9 times per month.8 Unfortunately, many of these salespeople may have limited scientific training and are paid on a commission basis, depending on how much of their company’s products are prescribed by the clinicians they target. There is also growing evidence of companies’ use of such communication to persuade physicians to prescribe products without a given Food and Drug Administration indication and, more importantly, without adequate evidence of efficacy or safety. Several of the nation’s largest drug companies have in recent years paid enormous sums in legal settlements mandated by state attorneys general for such off-label marketing of Neurontin (gabapentin, Pfizer: $430 million settlement),9 olanzapine (Zyprexa, Lilly: $1.3 billion settlement), and valdecoxib (Bextra, Pfizer: $2.3 billion settlement).

Previous Section
Next Section
The Impact of Industry-Funded Medical Education

These developments demonstrate that the lure of multibillion-dollar sales from blockbuster drugs and devices can distort the accuracy of information provided by manufacturers to physicians, even when no laws are broken. There is a growing literature in the social and behavioral sciences documenting how judgments and actions can be shaped by subtle factors; some of these findings as they relate to clinical information were presented in a 2007 publication of the Association of American Medical Colleges.10

Manufacturer-sponsored CME is associated with increases in prescription rates of the sponsor’s medication,11 perhaps because of the added attention paid to it during these sessions.12 In one study, the positive effects of the sponsor’s drug were mentioned 2.5 to 3 times more often than those of competitor’s drugs.13 Interactions with pharmaceutical representatives increase the likelihood that a physician requests to add the company’s drug to hospital formularies by more than 300%, and the receipt of honoraria to discuss a company’s new drug is associated with even larger effects on formulary requests.14 Interactions with detailers have been linked to more frequent prescribing of the marketed drug,15 as well as prescribing that is more expensive16 and less evidence-based.17

Several specific cases highlight our concerns. Rofecoxib (Vioxx) was one of the most widely used drugs in the United States until its withdrawal. Its capacity to nearly double the risk of myocardial infarction and stroke18 made it one of the most important causes of iatrogenic cardiovascular disease in modern history. Newly available data and company documents discovered through litigation documented that its manufacturer (Merck) skewed the depiction of cardiac risk in its presentation of clinical trial data19 and deployed programs to train its sales representatives to avoid direct responses to physicians who asked about the drug’s potential for cardiac toxicity.20 Other reports have documented the efforts of GlaxoSmithKline, the manufacturer of rosiglitazone (Avandia), to minimize or suppress early information that the drug could cause congestive heart failure and myocardial infarction.21 Cardiovascular device makers, too, have been found to have performed poorly in uncovering and disseminating information about potentially dangerous adverse effects of their products.22 Accounts such as these, relating to information provided by large, well-known companies about their widely used products, support some measure of doubt about the advisability of having manufacturers participate, even indirectly, in the delivery of education about their products.

Other reasons for concern about industry involvement in education in cardiology and other areas of medicine are based on worries about the balance presented among competing therapeutic and diagnostic choices, beyond the accuracy of specific facts. For example, in the 1990s, when calcium channel blockers were among the best-selling drugs in the world, their manufacturers were unlikely to sponsor CME lectures or conferences highlighting the advantages of off-patent thiazides for the management of hypertension, even though this was the first-line treatment recommended by the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure23 and it cost less than one twentieth the price of a brand-name calcium channel blocker. Currently, because individual branded statins (atorvastatin, Lipitor) and antiplatelet agents (clopidogrel, Plavix) are the first- and second-highest grossing drugs in the world, it is also unlikely that industry-sponsored talks would focus on the greater affordability of a generic statin or aspirin, respectively, even in patients for whom there is no trial-based evidence that the more costly drug is superior.

Some argue that industry-sponsored education is no worse than industry-sponsored medical journals, in which advertisements for products appear alongside peer-reviewed articles. However, in most good journals, it is quite clear what is an advertisement and what is a scientific article, so that physicians can evaluate each kind of information in its own context. Relevant exceptions include the purchase of company-dominated sponsored “supplements” linked to otherwise peer-reviewed journals and the recent disclosure of the creation and control by 1 manufacturer of an entire journal.24

Other Kinds of Conflict of Interest in Medical Education

Defenders of corporate sponsorship of medical education programs often object that numerous participants in education can have potential conflicts of various kinds. Many medical school faculty members may have strongly held beliefs about particular kinds of treatment. Some may receive partial salary support through their institutions from industry-funded grants for work on clinical trials or other research. Faculty members or their institutions, such as medical schools or teaching hospitals, may also hold shares of patents and thereby derive financial benefit from the use of particular drugs, devices, or tests. These issues warrant careful attention, but the focus of the present discussion is on a more direct kind of influence: The transmission of knowledge that is funded by companies whose business is based on sales of a product that is the subject of that communication.

The funding of an educational activity by a corporate entity whose primary goal, by definition, must be the promotion of product sales and the enhancement of shareholder value represents a kind of conflict that is different in nature from the biases introduced by faculty with strongly held beliefs or who receive industry support for the conduct of research in the context of a full-time university salary. These other potential sources of bias can be important. But in an appropriately structured university setting, they should be mediated by institutional policies and values that can serve as a check on undue influence. Although such mediation is sometimes deficient, these deficiencies do not justify the glib statement that “everyone has conflicts” as a reason to ignore the problems of direct industry support of education.

Spurred by student activism, our own institution, Harvard Medical School, in 2008 adopted a new policy that attempts to address some of these forms of potential conflict. In all presentations to students, faculty members are now required to disclose at the start of their lecture any commercial relationships that may have bearing on the topic about which they are teaching, so that trainees can be aware of relevant external relationships. This policy simply extends to pre-MD education the same arrangements that are becoming universal in CME.

Another aspect of the potentially blurred line between commercial and scientific communication is the relationship between industry influence and another aspect of physician education, the creation of clinical recommendations and practice guidelines.25 Several instances of preventable iatrogenic cardiovascular disease illustrate this problem. The use of postmenopausal estrogen therapy was widely believed to help prevent ischemic cardiovascular events, until well-controlled randomized trials revealed that such a protective effect did not occur in most patients and that estrogen-replacement regimens could in fact increase the risk of cardiovascular disease.26,27 Although the incorrect belief in a protective effect was due in part to observational studies that did not adequately control for important differences between estrogen users and nonusers, evidence revealed more recently in litigation makes it clear that the drugs’ manufacturers also aggressively shaped clinical and public opinion through programs of ghost writing in the medical literature28 and public relations campaigns.29

Formal clinical practice guidelines represent a more concrete way in which improper influence by industry stakeholders can shape messages in ways that may not faithfully reflect the existing evidence base. Choudhry et al25 found that 59% of practice guideline developers had a financial relationship with a company that made a product addressed by those guidelines. This potential problem was identified as a major concern in 2 current reports from the Institute of Medicine30,31 and is also being addressed by a new committee formed by the Institute of Medicine in 2009.

Influence by commercial stakeholders can help determine the design of clinical trials that might help shape updated guidelines, as well as distort the way physicians practice. For example, in 2006, the National Kidney Foundation produced guidelines for anemia management in patients with renal disease; the manufacturer of erythropoietin was the principal sponsor of the guideline development process, and most of the participants and leaders of the initiative were consultants to that company, its competitor, or both.32 Those guidelines recommended use of erythropoietin at levels high enough to achieve hemoglobin goals in excess of those warranted by the available evidence. Later the same year, such aggressive use of erythropoietin was found in a randomized trial to significantly increase the risk of a composite outcome of death, myocardial infarction, congestive heart failure, and stroke.33 The Food and Drug Administration added a black-box warning to the drug’s label, and the National Kidney Foundation later modified its practice guideline.

Previous Section
Next Section
Opinions From the Leaders of Medicine

Over the last 3 years, 3 influential mainstream organizations have issued major reports recommending the separation of industry from all forms of medical education. In 2007, the Macy Foundation, which has been a major force in medical education for decades, examined the role of industry support in CME activities. It concluded that “[p]harmaceutical and medical device companies and healthcare professionals have inherently conflicting interests in [CME]. Commercial entities have a legitimate obligation to enhance shareholder value by promoting sales of their products, whereas healthcare professionals have a moral obligation to improve patient/public health without concern for the sale of products.”5 The report noted that “[n]o amount of strengthening of the ‘firewall’ between commercial entities and the content and processes of [CME] can eliminate the potential for bias.”5 The Macy Foundation report concluded that accredited organizations providing CME “should not accept any commercial support from pharmaceutical or medical device companies” and that the “[f]aculty of academic health centers should not serve on speakers’ bureaus or as paid spokespersons for pharmaceutical or device manufacturers.”

In 2008, the Association of American Medical Colleges released its own report on industry funding of CME. The recommendations were formulated by a broad-based committee that, along with academics and consumers, included the chief executive officers of Pfizer, Eli Lilly & Company, Medtronic, and Amgen and was chaired by the former chief executive officer of Merck. It determined that the potential of commercial sources to influence the objectivity of academic teaching “engenders public skepticism, not only in the commitment of medical schools and teaching hospitals to their primary public purpose, but also in the commitment of academia and industry together to promote the public’s interest by fostering the most effective, evidence-based medical care possible.”34 On the basis of this finding, the Association of American Medical Colleges recommended that industry funds should not be provided directly to any CME course; rather, any such support should be coordinated through a centralized office in a given institution, and “[i]n no case should the medical center permit CME funding directly to individual faculty.”34 It also proposed that “academic medical centers should strongly discourage participation by their faculty in industry-sponsored speakers bureaus.”34

In 2009, the Institute of Medicine issued a report entitled “Conflict of Interest in Medical Research, Education and Practice.” In its section on education, the Institute of Medicine recommended that “faculty should not participate in speakers bureaus and similar promotional activities in which they either present content directly controlled by industry or formulate their remarks to win favor and continued speaking fees,”31 and it called for a “broad-based consensus development process to propose a new system of funding accredited continuing medical education that is free of industry influence, enhances public trust in the integrity of the system, and provides high-quality education.”31

As recently as 2006, an article in JAMA by a group of leaders from academic medical centers35 that called for an end to direct industry sponsorship of CME drew heated rebuttals.36 But since then, several major academic medical centers have established tougher new rules banning commercial relationships in several areas of training. Often, these initiatives came in response to concerns raised by medical students individually or through the American Medical Student Association, which has taken an active lead in this area. In 2008, Stanford Medical School announced that it would no longer accept course-specific industry funding for CME, requiring that any such support be given instead to a pooled school-wide education fund.37 Although initial skepticism greeted the notion that any drug company would provide general educational support without linking it to a particular therapeutic area or faculty member, earlier this year Stanford announced a $3 million open-ended educational grant from Pfizer, with use of the funds to be totally determined by the medical school.38 The University of Pittsburgh Schools of the Health Sciences have recently adopted a similar policy.39 The Sloan-Kettering Cancer Institute has gone further and banned industry support in any form for CME.40

Although it is too early to discern the effects of these changes, there is widespread belief that despite the Pfizer-Stanford grant, the volume of funding that companies will be willing to commit to such general educational pools will be substantially less than they are now spending on more tightly controlled product-linked programming. This would be compatible with the explanation that such funding (often dispensed via the marketing department) is primarily aimed at increasing product sales, rather than as a form of general corporate “good citizenship.” The coming years will determine whether this concern is borne out.

Previous Section
Next Section
Options for the Alternative Funding of CME

We agree with the Macy Foundation, the Institute of Medicine, the Association of American Medical Colleges, and other groups that have called for ending or strictly limiting industry support of medical education at all levels. Because many clinicians, academic institutions, and professional organizations have become reliant on industry funding, this raises the question of where support for such programs would then come from.

The education of medical trainees and practitioners is a public good; there is ample justification for federal and state support of such education, especially for CME. Governments support other forms of higher education in which the life-and-death stakes are not as high and issues of potential bias are not as worrisome. One attractive means of publicly supported CME is “academic detailing,” initially developed by one of us (J.A.) in the 1980s.41,42 In this approach, medical school faculty working through a nonprofit organization with no ties to any pharmaceutical company perform a rigorous review of the evidence on optimal management of a given clinical topic (eg, hypercholesterolemia, or the evidence-based use of antiplatelet agents). That organization then condenses these syntheses into user-friendly, clinically relevant summaries of evidence-based recommendations. It then trains nurses, pharmacists, or physicians to present this information interactively in “educational outreach” to physicians in their own offices, in an encounter that is similar in some respects to the sales visits of drug company detailers but devoid of a commercial agenda. The Cochrane Collaborative has reviewed the growing literature evaluating this approach and concluded that it is an effective means of improving prescribing practices.43 Several states (including Pennsylvania, Massachusetts, New York, South Carolina, and Vermont, as well as the District of Columbia) have established publicly funded academic detailing programs to offer physicians such noncommercial evidence-based education on medication use. We participate in such programs, the materials for which are made available publicly at no cost through http://www.RxFacts.org. Senator Herb Kohl (Democrat, Wis) has introduced legislation that would provide federal funding for such programs on a wider scale, with the provision that those preparing or implementing such educational activities have no financial ties to the companies whose products are being discussed.

In addition, as more care is delivered through integrated health systems, such organizations will realize the clinical and economic value of investing in the continuing education of their practitioners. Third-party payers, including the federal government, should have similar incentives to pay for CME, as long as its content is determined by a truly independent group of professionals and not subject to influence by any public or private payer (which could introduce an opposite but equally problematic problem of potential bias). Entities responsible for healthcare financing and delivery are likely to realize eventually that if industry-sponsored education skews care toward more costly alternatives, supporting such programs themselves will be an important means of reducing unnecessary costs, as well as ensuring the quality of care.

Another obvious solution is for physicians to foot the bill for our own continuing education. Some clinicians may object that an end to company underwriting of CME programs will increase the cost they must pay for CME courses or lectures or require them to pay for their own food at mealtime talks. But despite the commoditization of so much of health care, medicine remains a profession; one of the hallmarks of professionalism is accepting the responsibility for one’s own lifelong learning. Professionals in other fields (such as lawyers or accountants) pay for the costs of their own continuing education, or the firms that employ them do so. Medicine is unusual among the professions in that so many of us expect the vendors with which we do business to underwrite the costs of our ongoing education. Those of us who offer and take CME lectures and courses will also need to accommodate ourselves to less lavish settings, gastronomy, and amenities to accompany our educational experiences.

Defenders of industry-funded medical education for physicians and of direct-to-consumer advertising of prescription drugs argue that even if there are concerns about balance or focus on costly products, company support leads to communication of messages that might otherwise be neglected; examples include the need to control cholesterol, or informing patients of conditions like depression, incontinence, or erectile dysfunction that they might otherwise not have discussed with their physicians. We believe that the wealthiest nation on earth should be able to provide neutral, evidence-based educational messages about medications to physicians and patients without having to accept the trade-off that industrial sponsorship of such education inevitably requires.

The United States currently spends more than $4 billion per year on direct-to-consumer “education” about drugs and 6 to 7 times that amount in industry-funded CME and promotion (including drug samples) for physicians, which are all tax-deductible business expenses. These are sound investments for the companies that make them, because they drive medication use toward the costliest drugs, even if these are no more effective than available products. The costs of this “free” information are thus ultimately borne by consumers and taxpayers. Industry sponsorship of such communication may appear to be cost-saving to the busy practitioner offered a free meal or a low-cost CME course, but as noted by a widely used World Wide Web site that covers industry involvement in medical education, there is no free lunch (see http://www.NoFreeLunch.org).

Previous Section
Next Section
The Ongoing Development of Medical Practice and Education

The evolution of opinion on industry-supported medical education in cardiology and other branches of medicine mirrors the loss of naïveté we have seen before in the evolution of biomedical science. Before World War II, many clinical trials did not require randomization or control groups or blinding of investigators or patients, in the mistaken belief that patient selection, observer bias, or the placebo effect could never color the interpretation of whether a given treatment worked or not. “After all, we are astute scientific observers and would never let our own expectations or a patient’s self-report influence our judgment about how well a treatment works.” We now know better and understand that a trial that lacks these methodological safeguards can be fatally flawed, and its results would not be taken seriously.

In the mid 20th century, little attention was paid to ethical standards for research that involved human subjects or to the need to obtain informed consent for such studies. “After all, we are concerned, committed physicians and would never intentionally harm a research subject.” We now know better, and it is unthinkable (and impermissible) to conduct a trial without obtaining the consent of subjects and securing approval of the study design by an arms-length institutional review board.

Similarly, as recently as the 1980s, the medical profession generally ignored the potential influence of commercial ties on clinical research; rules about conflict of interest were vague, incomplete, or ignored: “After all, as researchers we would never allow the prospect of personal gain to distort the conduct and interpretation of our clinical studies.” We now know better, and most universities and academic medical centers enforce strict rules that prohibit those with financial interest in a given treatment from studying it in patients.

It is only in the last decade that parallel concerns have become common in considering the role of industry in medical education. Many physicians are still in a stage of denial similar to those that preceded other paradigm shifts: “After all, companies and their spokespeople would never shape an educational presentation just to favor a given product and intelligent physicians would not be influenced if they did.” We now know better, or should. As our understanding of these issues becomes more sophisticated, it is likely that in cardiology and other fields of medicine, industry-sponsored lectures, courses, and gourmet infomercials will eventually go the way of nonrandomized trials, research without informed consent, and clinical studies performed by investigators whose financial well-being depends on the results.

As in the earlier examples from medicine’s development, the issue is not venality or intentional deception. The problem is generally not that manufacturers or the educators they employ deceitfully manipulate educational content, despite some egregious examples to the contrary. As with the other problems of what we now call conflict of interest, the problem is simply that bias can so easily be introduced unintentionally. And because it is often unintentional, it is all the more difficult to detect and prevent.

The lessons of behavioral economics, as well as the literature on clinical decision making, provide ample evidence that although we all like to consider ourselves perfectly rational actors, what we believe and the decisions we make can be subtly but heavily influenced by bias, even if such bias was not intentional. Because the clinical stakes for our patients are so high, and because the healthcare system of the coming decade will have to expend its constrained resources in the most cost-effective manner, cardiology and the rest of medicine will need to move beyond having vendors provide or pay for the education of its practitioners. The transition will be challenging for a time, but our patients, the healthcare system, and we as professionals will be better off for it.

Previous Section
Next Section
Acknowledgments

Disclosures

Neither Dr Avorn nor Dr Choudhry accepts any personal compensation from any drug or device manufacturer. Drs Avorn and Choudhry are listed as coinvestigators on a grant application submitted to the Agency for Healthcare Research and Quality through the Brigham and Women’s Hospital that proposes adaptation of Agency for Healthcare Research and Quality comparative effectiveness reports for wider dissemination. Dr Avorn also provides pro bono consultative services on appropriate medication use to the Alosa Foundation, a nonprofit 501c3 educational organization with no relationship to any drug or device manufacturers; a family member serves pro bono in an administrative capacity, but neither Dr Avorn nor his family member receives a financial benefit of any kind from this work. Dr Choudhry also provides consultative services on appropriate medication use to the Alosa Foundation and is compensated for this work.