Drug Detailing in Academic Medical Centers: Regulating for the Right Reasons, with the Right Evidence, at the Right Time

In this month's issue of AJOB, Huddle recapitulates several arguments against prohibiting drug detailing in academic medical centers. Briefly stated, Huddle argues that empiric data on the effects of detailing are not definitive, and that restricting interaction with sales representatives does not rise to the level of an “unqualified ethical imperative.” As a result, Huddle argues that professional ethics do not mandate an opposition to detailing, and that policy responses should be “prudential.” In recommending a blanket prohibition on detailing and framing the issue as one of professional ethics, Huddles believes the AAMC overstepped its bounds.

While Huddle outlines the reasoning for each of these arguments in more detail than is often articulated, the conclusions of each of these arguments are nonetheless incorrect.

Evidence

Debates over detailing often revolve around two key evidentiary questions. Does detailing affect physician behavior? If yes, does its impact create a net negative or positive effect for patients?

The answer to the first question is clearly “yes.” As Huddle notes, studies of detailing in the medical literature consist mostly of surveys employing physician self-report, which provide suggestive but inconclusive evidence that detailing impacts behavior. ^{1, 2} In contrast, studies from the economics literature provide convincing evidence that detailing positively impacts drug sales.³ This evidence, combined with the reasonable assumption that industry likely has a good reason for investing billions of dollars each year on drug detailing, strongly suggests that detailing achieves its intended effect of increasing the volume of prescriptions written by physicians for the higher cost, brand name products marketed by industry.

Does this effect create a net benefit or harm for patients? While it is possible that detailing may have some beneficial effects by educating physicians about new drugs, sales representatives often use education in a biased way to highlight the benefits of their products over others (for example, by discussing studies favorable to their drug and ignoring or downplaying other, less favorable research).⁴ In the absence of detailed expertise in the field - in which case education from sales representatives is unnecessary- the physician being detailed is hard-pressed to contextualize the information being presented, or even simply to distinguish true from false information. This may be particularly true in academic medical centers, where trainees have yet to develop the knowledge and experience base to inform their interpretation of new information. The alternative approach – a deep skepticism of the entire encounter and its overarching goal – begs the question of whether any useful education is occurring.

Industry documents and former sales representatives' narratives describe how the essential function of detailing is not transferring information to physicians, but building relationships and goodwill with them that ultimately translate into changes in prescribing behavior.^{5, 6} There are times when the incentives of industry and patient well-being may align, as exemplified by the role of industry in increasing use of ramipril following the HOPE trial.⁷ However, pharmaceutical companies have a fiduciary responsibility to their shareholders to maximize sales of their drugs, not to educate physicians about best practices. Thus, there are innumerable counterexamples where drug prescribing patterns fall contrary to guideline recommendations in a pattern that belies industry influence.

Adding to the argument is the neurobehavioral research cited in the AAMC report, which provides strong evidence for cognitive errors and self-deception in laboratory simulations involving conflicts of interest.⁸ This suggests that physicians – as humans – are unable to compensate for the influences on behavior that can arise from these conflicts. While this applies to all physicians, such mechanisms may be especially potent among trainees in academic medical centers, where sleep deprivation and the desire for creature comforts (such as a hot lunch provided by a visiting sales representative) may both lower defenses and create strong incentives for self-deception.

Huddle is correct that these neurobehavioral studies have not been tested in real-life settings involving physician-industry relationships. However, demanding perfect, unequivocal evidence as the standard on which to base ethical discourse and policy is often unrealistic ; if applied uniformly, such a standard would paralyze clinical medicine, ethical decision-making, and policy interventions. When the totality of evidence is considered, Bayesian reasoning strongly suggests that drug detailing has a net negative effect on the care of patients. These negative effects can take a variety of forms, whether it be (1) prescribing an expensive new drug for a condition in which an older, less expensive agent has a better efficacy and safety record as well as lower cost to the patient; (2) inappropriately treating a condition with drugs where lifestyle modifications or non-drug alternatives are preferable; (3) blurring the line between pathology and normal human functioning to create an expanded market for drug use (so-called “disease mongering”); and (4) diminishing public trust in the medical profession by fostering perceptions of collusion between physicians and the pharmaceutical industry. Publicity surrounding the aggressive promotion and subsequent harms of Vioxx (rofecoxib), Bextra (valdecoxib), Zyprexa (olanzapine), and other drugs highlight some of the more dramatic of these ill effects. However, the potential health and economic harms that arise from detailing of other drugs compound on a daily basis even in the absence of news headlines.

Professionalism

The second key flaw in Huddle's argument concerns his interpretation of the imperatives of professionalism. As outlined in AAMC's policy statement, a cornerstone of professionalism is physicians' obligation to apply independent, objective judgment to act in the best interest of patients.⁹ Many physicians believe that interactions with industry, such as accepting gifts, do not influence their judgment. As noted by one physician quoted in the AAMC'S report, “You don't really think that I would let a pizza lunch influence my decision-making process for my patients, do you?” However, neurobehavioral research demonstrates that people have less control over their cognitive processes than we would like to believe.⁸ Acts of benevolence, even if subtle, can generate behavior that reciprocates the generosity even without conscious intent. People are inclined to cheat a little while still considering themselves honest and ethical, and to not look for things about themselves that they don't want to learn. These cognitive processes don't distinguish ethical from unethical people, but are a reality of typical human experience.

It may be true that some physicians can engage industry without being influenced; i.e. by unconsciously reciprocating or otherwise considering commercial interests in addition to the well-being of their patients. The problem is that it is difficult to know how susceptible a given physician is to influence, not only for external observers but for the physician herself. Since strong data suggests that physicians can be unconsciously influenced, and that we have limited ability to identify a priori who is influenced and to what degree, it is prudent to avoid the conflict of interest altogether. This is consistent with the purpose of professional standards, which exist to support the aims of medicine (e.g., to improve the health and well-being of patients) rather than to make heavy-handed judgments about moral vs. immoral behavior.¹⁰ In a similar vein, academic medical centers play an important role not just in the teaching of trainees with respect to the technical aspects of medicine, but in modeling standards of behavior that the next generation of physicians will carry forward into their professional lives.

Policy

Finally, Huddle is wrong in concluding that the policy recommendations of AAMC are not “prudential.” Reasonable people may – and do - disagree about the appropriate policy response to drug detailing, just as reasonable people disagree about all types of regulation and legislation. As Huddle properly notes, part of the debate over detailing relates not only to interpretation of the evidence but to ideological differences, corresponding with disagreements over the balance between regulation and physician autonomy. However, waiting for near-unanimous consent on this or any issue is unwarranted. Policies in medicine have an important role in improving practice, rather than simply codifying existing norms for which there is overwhelming agreement. In addition, the national investment in medical training and its attendant academic infrastructures in part reflect a larger social contract, with responsibilities to improve access to care and the public's health. Insofar as drug detailing raises the costs of our health care system by unnecessarily driving the use of expensive drugs, policy interventions such as the AAMC's prohibition have a legitimate role, with particular poignancy for the current debate about health care reform.

A central role of professional organizations is to help establish best practices and a forward-thinking vision to improve the structure and practice of medicine and in doing so to improve the well-being of patients. As such, the AAMC's recommendations are a thoughtful and appropriate response to the reality of pharmaceutical marketing and will positively impact the practice of medicine and set a leadership example for academic medicine that the rest of the country can and should follow. These recommendations combine the right evidence, the right reasons, and the right time.

This is a commentary on article Huddle TS. The pitfalls of deducing ethics from behavioral economics: why the Association of American Medical Colleges is wrong about pharmaceutical detailing. Am J Bioeth. 2010;10(1):1-8.