INTRODUCTION
Corporate influence is the most pressing issue in public health. It cuts across many of our most intractable public health problems—from obesity to the opioid epidemic. Countless companies are making financial contributions to—and developing close relationships with—a variety of public health actors, including government agencies, research universities, academic medical centers, professional societies, and patient advocacy organizations.Footnote 1 Corporations often fund medical research and participate in public health interventions intended to address the very public health challenges they are creating or exacerbating. This has been well-documented in a variety of sectors: pharmaceuticals (including opioids), food and soda, alcohol, tobacco, and vaping products.Footnote 2 How we view relationships with industry, including how these relationships are framed in ethical discourse, shapes our legal and policy responses to them. In recent years, fueled in part by the opioid epidemic, the ethical framing of industry relationships has begun to evolve in significant ways. But legal and policy responses have not yet caught up.
Systemic analyses of corporate influence reveal the ways in which corporate strategies have evolved during the last half-century. Several scholars have drawn “webs” of corporate influence in public health,Footnote 3 described the “commercial determinants of health,”Footnote 4 and characterized industry-related conflicts of interest as “public health hazards.”Footnote 5 Some scholars have focused on the pharmaceutical sector. For example, Sergio Sismondo has characterized drug company strategies as “assemblage marketing”—in which doctors, researchers, and patient advocacy groups (among others) are recruited as “channel marketing partners.”Footnote 6 Others have identified strategies of influence that cut across a variety of sectors (from oil and gas to opioids)—which one might expect, given that companies in different sectors have been employing the services of the same public relations, management consultant, and crisis management firms.Footnote 7 These systemic analyses make clear that the paradigm response to corporate influence—the disclosure of financial conflicts of interest—while necessary, is not sufficient for addressing and neutralizing corporate strategies of influence in public health.
Some scholars have, of course, called for interventions beyond disclosure.Footnote 8 At least one of these calls comes from a former policymaker—Joshua Sharfstein who previously served as Secretary of Health for Maryland, principal deputy commissioner of the U.S. Food and Drug Administration (“FDA”), and the public health commissioner of Baltimore. He describes a “trend toward accepting disclosure and transparency of financial conflicts of interest as sufficient,” and he argues that a “more serious discussion of what roles should be protected from conflicts of interest is overdue,” and that, “[f]or certain critical educational, regulatory, and policy development functions, it may be necessary to prohibit financial conflicts of interest entirely.”Footnote 9 Even more recently, a group of journal editors and scholars of conflicts of interest published an article in the BMJ, launching a new series on conflicts of interest and calling for a shift from transparency to independence.Footnote 10 But current policymakers have yet to respond to such calls.
In order to address the systemic strategies of influence employed by a variety of sectors of industry in public health contexts, policymakers need to develop comprehensive counterstrategies. To support that claim, I outline here a very brief history of corporate influence in public health, which puts the current pervasive influence in context. I also review the main legal and policy responses, and describe the limitations of the current paradigm. In particular, I address the limitations of characterizing corporate influence as financial conflicts of interest that merely need to be disclosed. I further claim that corporate influence can be illuminated but not eliminated by transparency and, in particular, by the disclosure of financial conflicts of interest. Finally, I highlight the kinds of legal and policy responses that will be necessary to effectively address the pervasive corporate strategies of influence that we are currently experiencing.
I. A (VERY BRIEF) HISTORY OF CORPORATE INFLUENCE IN PUBLIC HEALTH
For as long as corporations have existed, they have attempted to influence consumers. Their survival depends on persuading customers to buy their goods and services. But many of the corporate strategies of influence that we know today go far beyond efforts to sway consumers. Such strategies were largely developed and refined in the latter half of the twentieth century. As I cannot fully chart this development in a few paragraphs, I will touch on some notable highlights instead. By many accounts, the corporate strategies of influence we know today were first developed in the 1950s by the tobacco industry, which was trying to cast doubt on the growing body of evidence demonstrating the individual and public health hazards of tobacco.Footnote 11 These companies recognized that comprehensive strategies were necessary, and that they would require building relationships with government, the academy, and civil society.Footnote 12 The origins of so-called “key opinion leaders,” physicians hired by drug companies to deliver scripted promotions of new pharmaceuticals to their peers, have also been traced back to this time period.Footnote 13
Developing alongside tobacco and drug company strategies were food industry strategies. The sugar industry sponsored a research program in the 1960s and 1970s to cast doubt on the health hazards of sucrose and identify fat as the dietary “culprit” in coronary heart disease.Footnote 14 Industry funding of studies on the relationship between diet and blood cholesterol also began in the 1970s—and continues to this date, with much of that funding now coming from the egg industry.Footnote 15 Food industry strategies were not confined to funding favorable research. For example, one major company sought to persuade “mothers” to buy its margarine, instead of butter, by building relationships with and influencing (among others) health professionals and consumer groups.Footnote 16
Corporate influence in medicine began to garner the attention of a few physicians and scholars by the early to mid-1980s. In 1980, Arnold Relman, then Editor of the New England Journal of Medicine, reminded physicians that medicine and public health have not always been flush with corporate funding.Footnote 17 He warned readers about the new “medical-industrial complex,” an echo of the concerns President Eisenhower expressed two decades earlier about the “military-industrial complex.”Footnote 18 Relman later also expressed concern that, in medicine, relationships with drug and device manufacturers had become “more pervasive, complex, and problematic.”Footnote 19
Some scholars attribute corporate influence in research to the Bayh-Dole Act of 1980, which created an incentive for universities to commercialize the products of their research by allowing universities to retain title to inventions resulting from federally-funded research programs.Footnote 20
The 1980s and 1990s were also witness to a proliferation of disputes between academic researchers and the industry sponsors of their research.Footnote 21 In these high-profile cases, sponsors exerted pressure on researchers (often to behave in ways that were clearly unethical), and the researchers resisted. Such cases were described by some scholars as the “tip of the iceberg.”Footnote 22 But what lay beneath the surface was arguably more insidious—the systemic problem of subtle or tacit corporate influence and acquiescence on the part of many other industry-funded researchers.Footnote 23 The high-profile disputes also revealed a related problem—academic institutions failing to support researchers for fear of alienating corporate donors.Footnote 24
To document the developments in any one sector would require a book—or series of books. In Ghost-Managed Medicine (2018), for example, Sergio Sismondo provides evidence that the practice of “ghost writing” journal articles was viewed too narrowly when it began to attract attention in the 1990s and early 2000s. The term was usually understood to refer to the drug company practice of writing articles and then seeking academic authors to sign on to them, often with minimal input.Footnote 25 Sismondo's book makes clear that this practice has been part of drug companies' broader strategies of “ghost management.” These strategies entail planning the series of articles necessary for the approval and promotion of a new drug, finding academics willing to be listed as authors, and trying to place these articles in the best venues.Footnote 26
In her books, Food Politics (2007), Soda Politics (2015), and Unsavory Truth (2018), Marion Nestle has extensively documented the influence of food and soda companies on nutrition research and obesity policy.Footnote 27 In his books, Doubt is Their Product (2008) and The Triumpth of Doubt (2020), David Michaels describes corporate strategies of influence across a variety of industries, among them, tobacco, alcohol, opioids, oil and gas, and football—all designed to cast doubt on science demonstrating the harmful effects of industry products or practices.Footnote 28 Corporate strategies not only seek to influence consumers and markets; they also have in their sights government bodies, academic institutions, and civil society groups—as well as the individual members of these entities.Footnote 29 These strategies recognize that corporations need to exercise political power, as well as market power, to protect and promote their commercial interests.
The resulting practices have become pervasive. For example, one recent study found two-thirds of patient advocacy groups in the United States received industry funding.Footnote 30 Another study painted an even more troubling picture, finding that 83% of the 104 largest patient advocacy organizations receive funding from drug, device, and biotechnology companies.Footnote 31 A third recent study reported that two soda companies alone (Coca-Cola and Pepsi) funded close to 100 health-related civil society groups in the United States.Footnote 32 What companies call their “penetration” of civil society groups has become so pervasive that it is now difficult—in some cases, almost impossible—to find relevant groups that are independent of industry, and on which policymakers can rely for comments that genuinely reflect the views and interests of patients or the relevant publics.Footnote 33 Notably, an investigation into an FDA consultation process (a consultation mandated by statute) revealed that 93% of the participating patient advocacy groups had received funding from industry.Footnote 34 Another study in England identified that, in almost four out of every five occasions in which patient advocacy organizations contributed to an appraisal of a medical technology during 2015 and 2016, the organization in question had, in the same year, received funding from the manufacturer of the technology being assessed or a competitor technology.Footnote 35 If we make the charitable assumption that policymakers genuinely want access to the views of the relevant publics, untainted by industry, the pervasive nature of corporate influence is making it increasingly difficult to solicit and receive such input in policy and regulatory processes.
II. A (VERY BRIEF) HISTORY OF RESPONSES TO CORPORATE INFLUENCE IN PUBLIC HEALTH
Disclosure of financial relationships with companies—ordinarily framed as financial conflicts of interest—remains the most popular policy response to corporate influence.Footnote 36 As I explain more fully elsewhere, there are many avenues for corporate influence, and such influence can exist and prove highly problematic even when there are no financial conflicts of interest.Footnote 37 Here, I make the distinct, albeit related, claim that disclosure is necessary but not sufficient to address the systemic problem of corporate influence.
In medicine and public health, the conflict of interest definition most frequently employed is the one found in the Institute of Medicine's 2009 report on conflicts of interest in medical research, education, and practice: “a set of circumstances that creates a risk that professional judgment or actions regarding a primary interest will be unduly influenced by a secondary interest.”Footnote 38 Despite its broad acceptance, several scholars have raised legitimate concerns about this definition. Marc Rodwin cautions that the report's approach to conflicts of interest “neglects the actor's compromised loyalty to the party or mission she is supposed to serve” and that, by referring to conflicts between primary and secondary interests rather than conflicts between obligations and interests, this definition “diminishes the conflict's significance.”Footnote 39
My own concerns about the conflict of interest concept encompass, but also go beyond, definitional concerns. In short, corporate influence may result from relationships or practices that may not necessarily give rise to financial conflicts of interest. The conflicts of interest framing also tends to lead to an emphasis on individual conflicts and the neglect of institutional conflicts. In addition, the conflicts of interest concept can be exploited to import concerns about “intellectual” conflicts and detract attention from the systemic problem of corporate influence.Footnote 40 In this piece, however, I focus on the limitations and potential adverse effects of disclosure of conflicts of interest. To be clear, I am not arguing against disclosure. On the contrary, disclosure in particular and transparency in general should ordinarily be considered the minimum that is ethically required. But we should not allow disclosure to crowd out other legal and policy responses that may be more effective at addressing corporate influence, especially given these limitations and potential adverse effects.
We have seen a proliferation of conflicts of interest disclosure policies in medicine and public health during the last two decades. A study of 1,400 high-impact science journals found that, in 1997, a mere 15.7% of these journals had policies requiring author disclosure of conflicts of interest.Footnote 41 By one recent estimate, there are now in excess of 3,100 journals in medicine and health care that are listed by the International Committee of Medical Journal Editors (“ICMJE”) as following their policy requiring authors to disclose financial conflicts of interest.Footnote 42 A recent analysis of randomized controlled drug trials found disclosures of financial ties between the principal investigator and the manufacturer of the drug in 60% of the trials, and undisclosed ties in a further 8%.Footnote 43 Despite the proliferation of both disclosure policies and disclosure itself, there are still significant omissions in the scope, as well as the practice, of disclosure. Notably, two-thirds of U.S. medical journals have not published disclosure policies for editors, even though editor conflicts are widespread—an issue that only recently garnered significant attention in the literature.Footnote 44
Disclosure has also become the policy paradigm in medical education. In 2005, students at Harvard Medical School raised concerns when they discovered that their lecturers were not disclosing financial relationships with the manufacturers of the drugs about which they were lecturing.Footnote 45 Two years later, the American Medical Students Association launched its “scorecard,” giving grades (from A to F) to medical schools for the existence and content of their conflicts of interest policies.Footnote 46 After considerable pressure from medical students, Harvard Medical School subsequently introduced a policy requiring all lecturers to disclose and describe, when relevant, their conflicts of interest at the beginning of their lectures.Footnote 47
Over time, disclosure has become not only the paradigm institutional policy response, but also the dominant legal response. The Physican Payments Sunshine Act of 2010 (“PPSA”) (also known as section 6002 of the Affordable Care Act) requires drug and device manufacturers to disclose their payments to physicians and teaching hospitals.Footnote 48 Although information about these payments has been made available for several years on a searchable website, the vast majority of patients do not know about the website and, even when they do, they rarely visit it to determine whether their physician has accepted payments.Footnote 49 Despite this, lawmakers have continued to build on the disclosure provision. Section 6111 of the SUPPORT Act 2018—drafted in response to the opioid epidemic—expands companies' legal obligation to disclose payments: commencing in 2022, companies will also need to disclose payments they have made to physician assistants, nurse practitioners, clinical nurse specialists, certified registered nurse anesthetists, and certified nurse-midwives.Footnote 50 The section is optimistically entitled “Fighting the Opioid Crisis with Sunshine.”Footnote 51 But it will take far more than sunshine to address the opioid crisis and the many other public health challenges to which corporate influence has contributed.
To be clear, I do not oppose the expansion of the categories of payment recipient for which disclosure is required. In addition to those professionals listed in the SUPPORT Act, one might add several others:Footnote 52 for example, patient advocacy groups (so many of which receive funding from drug, device, and biotechnology companies) and a range of professionals giving nutrition or dietary advice (many of whom receive gifts from food and soda companies).Footnote 53 However, I am concerned about lawmakers' and policymakers' reliance upon disclosure—principally, or worse yet, completely—to address corporate influence. Recent evidence makes clear that compliance with disclosure and enforcement are often woefully inadequate. For example, one study found that one in four writers of practice guidelines had potentially relevant undisclosed ties to the pharmaceutical industry.Footnote 54 Even if there were perfect compliance with disclosure requirements, there are compelling arguments that disclosure alone is insufficient—not only for the writers of practice guidelines, but also more broadly in medicine and public health.Footnote 55
III. ON THE LIMITS OF DISCLOSURE OF CONFLICTS OF INTEREST
Before developing a theoretical critique of conflicts of interest disclosure, I briefly review the empirical research on disclosure. Much of this research has historically involved one-shot (rather than multiple-round) experiments—often using student and other non-professional participants who are engaged in role-play.Footnote 56 Some more recent evidence comes from experiments involving professionals (including health professionals) and from observational work in the field.Footnote 57 As the analysis below makes clear, the findings on the effects of disclosure are somewhat mixed.
The first challenge in discussing the research on disclosure is to be clear about the relevant parties. For scholars already familiar with the substantial behavioral science literature on disclosure, I will build on terminology often employed in that research while noting its limitations. Although I expand on these concerns below, they can be put briefly. First, the advisor/advisee frame evokes a dyad (or, at best, a triad), but, in reality, corporate influence is best viewed as a web of relations. Second, the terms do not readily conjure at least two kinds of relationships in which disclosure may also be problematic: intra-professional and institutional relationships.
Given that the principal concern underpinning much of the social science research is whether the receipt of payments may create bias in the advice of professional advisors (e.g., financial advisors advising clients, or physicians advising patients), studies often employ the term “advisor.”Footnote 58 In many of these studies (as I have noted), however, the advisors are not actual professional advisors but students engaged in role-play—constituting what social scientists characterize as “convenience samples.”Footnote 59 When using the term “advisor,” we should keep in mind not only the paradigmatic cases of physicians and other health professionals, but also of scientists and other academics. These experts often act in advisory roles—for example, giving congressional testimony, serving on advisory boards, and drafting guidelines (whether clinical practice guidelines, dietary guidelines, or some other kind of guidance).Footnote 60 In addition, they engage in other scholarly activities, including conducting research and reporting their findings in scholarly journals. Although this would not ordinarily be considered providing advice, we should clearly be concerned about the impact of scientists' financial relationships with the industry sponsors of their research—especially given the ways in which such relationships influence not only research findings and the interpretation of those findings, but also the very questions that scientists explore and ignore.Footnote 61
In addition to individuals, when thinking about the “advisor,” we should keep institutions in mind, whether or not those institutions are engaged in activities that would ordinarily be considered advisory. Consider, for example, patient advocacy organizations making representations in regulatory or policy-making processes. There is a growing body of evidence that such groups frequently participate in policy processes without disclosing their financial relationships with industry.Footnote 62 Other relevant institutions include health professional associations, teaching hospitals/academic medical centers, and research universities. Even when they are not engaged in activities that might ordinarily be considered advisory, these entities are engaging in activities on which others rely, and we should be concerned about the impact of their financial relationships on those activities. Reliance on so-called “advisors” may be either specific or general. As an example of the latter, we, as a society, depend on universities to generate unbiased and credible research.Footnote 63
Similarly, we should also think about “advisees” broadly. We miss relationships of concern if we focus only on individual patients receiving advice from their doctors. Just as researchers are “advisors” of concern, we should also conceptualize advisee to include those who rely on research and recognize that some of these individuals may also be professionals. Two examples of this would be the peer reviewer of an article submitted to a scholary journal,Footnote 64 and a physician reading and considering whether to rely upon a study published in that journal.Footnote 65 But, like advisors, advisees may also be institutional. So, for example, when the Centers for Disease Control and Prevention (“CDC”) received representations from various patient advocacy organizations that did not disclose their industry funding when opposing the CDC's draft guidelines on opioid prescribing,Footnote 66 the CDC was effectively an advisee with regard to those representations. But the CDC is also, of course, an advisor in the drafting, revising, and issuing of those guidelines. So these categories are not mutually exclusive.
The third term I would add to this model is my own: influencer. This would include, for example, drug and device manufacturers and food and soda companies. It would also encompass their trade associations. The literature on disclosure tends to look only at payments as a form of influence.Footnote 67 But it is important to recognize that influence may also result from “in-kind” contributions—whether goods (such as free drug samples), services (such as displaying the government's MyPlate icon on food packaging), or anything else of value to the recipient. Moreover, it is important to recognize that these influencers, which we might consider originating influencers, create webs of influence that turn many of the advisors and advisees discussed above into co-influencers. Both points are notable given that the studies on disclosure tend to focus only on originating influencers and on financial payments as a means of exercising that influence.
What dyad and triad-based models of disclosure also omit is that additional parties may be affected by the influence and advice.Footnote 68 When the influencer is an opioid company, and the advisor is a health professional association or some other advisory body producing opioid prescribing guidelines for doctors, the advisees are doctors. The missing party in such models is patients, who are clearly affected. Lawyers would call them third parties—although in the case of a triad of advisor, advisee, and influencer, that term does not fit well. And, as those third parties—whether patients, communities, or populations—are often the very people whose interests we should be most concerned about protecting, an additional concern is that the term third party may threaten to relegate rather than prioritize their interests and wellbeing.Footnote 69
Finally, I do not refer to “disclosers” as a distinct category because disclosers of conflicts of interest—or the underlying gifts or relationships—may be influencers, advisors, third parties (a less troubling usage in this context), or some combination thereof. In the case of the PPSA, the obligation to disclose is placed on influencers, the drug and device companies.Footnote 70 In contrast, the guidelines of the ICMJE place the obligation to disclose on authors—advisors for the purpose of the schema in the disclosure literature.Footnote 71 But authors (and, for that matter, speakers) may also be asked to disclose payments to others with whom they have personal or professional relationships—for example, their domestic partners or the institutions that employ them.Footnote 72 Where the obligation falls can be significant: some professionals might avoid conflicts they are required to disclose themselves, but they would be less troubled by the same conflicts if only the influencer were required to disclose them (especially when no one else pays attention to the disclosure.) For this reason, we should be attentive to the identity of the discloser, both in experiments and in the real world.
A. Impact on Behavior of the Advisor
1. On the Advice
One of the central concerns about financial conflicts of interest arising from relationships with industry is the impact on the behavior of professionals—especially, biased advice. Several social scientists have explored whether disclosure of conflicts of interest might alleviate this impact.Footnote 73 The empirical data, while mixed, suggests that it would be a mistake to rely on disclosure for this effect. A central concern first raised by the social science research is that disclosure can lead to strategic exaggeration. In other words, bias in the advice may increase substantially when the advisor's conflict is disclosed.Footnote 74 This is even more worrying when the recipient of the disclosure does not sufficiently discount the advice in order to counteract the increased bias or does not have the opportunity or resources to seek independent advice. The advisor's bias may be inflated not only to anticipate discounting, but also in the belief that the advisee has been warned or put on notice about the conflict of interest.Footnote 75 Having disclosed their conflicts of interest, advisors may also feel at greater liberty to “gild the lily,” believing that they have put the advisee on notice—this is what some scholars call “moral licensing.”Footnote 76 A more recent series of studies, on the other hand, suggests that disclosure can make professional norms more salient, and increase the quality of advice given by the professional advisors.Footnote 77 But this research should not be read as a ringing endorsement of disclosure as an antidote to bias; notably, the author of the studies, Sunita Sah, expressly argues otherwise.Footnote 78 We should also keep in mind that, in Sah's studies, the advisors disclosed their own conflicts.Footnote 79 In contrast, the PPSA does not require physicians to disclose payments from industry; the obligation is instead imposed on drug and device companies.Footnote 80 Although other scholars found some evidence of a decrease in prescribing of brand drugs by salaried physicians' in the wake of Massachusetts' disclosure law (an effect the authors attribute to “social image” concerns), the study did not find such an effect in relation to physicians who received larger payments.Footnote 81 In addition, as scholars of disclosure have been aware for some time, physicians with financial interests in medical facilities tend to recommend more (or more expensive) tests even when they have been required to disclose those interests.Footnote 82
Other real-world evidence suggests that we should not rely on disclosure to improve the quality of science, or to address scientific bias associated with industry funding. A recent Cochrane Review found not only that industry-funded research produces more favorable outcomes for industry sponsors, but also that “industry sponsored studies had less concordance between results and conclusions compared with non-industry sponsored studies.”Footnote 83 So disclosure of conflicts of interest is clearly not improving the “quality” of biomedical science—as evidenced by the fact that the industry funding of the studies in the Cochrane Review was disclosed. Whether the context is medical practice, biomedical science, or public health, as Sah notes, the elimination of conflicts is “likely to have a much larger effect on improving advice quality” than disclosure.Footnote 84
2. On the Advisor's Receipt of Gifts/Payments
Some scholars have argued that the prospect of disclosing a conflict could discourage physicians from entering into relationships that are difficult to justify.Footnote 85 There is some evidence from lab experiments that the mandatory disclosure of conflicts of interest can lead to the avoidance of those conflicts, provided that such conflicts are perceived as avoidable.Footnote 86 In the field, however, the evidence is mixed.Footnote 87 In some cases, the implementation of a disclosure policy may be followed by an increase in conflicts.Footnote 88 One potential explanation for this phenomenon is social proof; in short, physicians decide what to do by looking at what others are doing.Footnote 89 When conflicts are pervasive and professionals observe that many of their peers have them, they may perceive these relationships as acceptable, even desirable. Notably, the PPSA, which places an obligation on drug and device companies—but not on physicians—to disclose the payments, does not appear to have had a significant impact on physicians' willingness to accept industry gifts.Footnote 90 The Open Payments website indicates that total annual payments between 2014 and 2018 have increased from $8.58 billion to $9.23 billion.Footnote 91 The total number of payments exceeded 11 million each year in 2014 to 2017, and dipped only slightly to 10.82 million for 2018.Footnote 92 More fine-grained analyses similarly suggest disclosure has not reduced physicians' willingness to accept gifts from industry. One recent study found that, on average, there was no significant change in monthly payments to 50,000 physicians in relation to 16 anti-diabetes brand drugs in the 15-month period following the PPSA coming into effect.Footnote 93 The authors concluded that “public pressure … if there is any, seems to be ineffective” at reducing physician payments.Footnote 94 They also observed that disclosure may be “especially ineffective towards more expensive drugs and heavier prescribers”Footnote 95—arguably, the cases of greatest concern.
B. Impact on Advisees
Much of the empirical literature explores the impact of disclosure of physicians' conflicts of interest on patients. But, as noted above, it is important to keep in mind that patients are not the only potential “advisees” in medicine and public health. Others include: health professionals reading conflicts of interest disclosures made by peers in scholarly journals;Footnote 96 public health agencies receiving representations (and disclosures, when made) from patient advocacy organizations; and physicians, patients, and others who read the reports of advisory committees (including committees drafting practice guidelines). The potential impacts and, in turn, the pitfalls of disclosure are likely to vary considerably depending on the context and audience.
The evidence on disclosure of conflicts of interest to patients raises significant concerns. One recent study notably concluded that the Open Payments Database, created pursuant to the PPSA, “has fallen well short of its aspiration to better inform patients of their physicians' industry relationships.”Footnote 97 The study found that, two years after the launch of the Open Payments website, only 13% of respondents knew that information about their physician's receipt of payments was available; worse still, only 3% knew whether their own doctor had received payments.Footnote 98 Another recent study found that mailing disclosure of physician conflicts directly to patients with their appointment letter was effective at informing only 57% of recipients of their physician's conflict of interest—disturbingly, 43% of respondents did not understand it, ignored it, or forgot it.Footnote 99 So approaches to disclosure, thus far, do not appear to be serving their intended function: to inform patients of their doctors' conflicts.
Second, if patients receive the disclosure, do they understand the implications of the conflict? And, if so, what impacts does it have on their perceptions and beliefs? There is mixed evidence on the impact of conflicts of interest disclosure on patient trust. In a study at the Cleveland Clinic, the disclosure of physicians' conflicts of interest in the appointment letter did not significantly impact patient trust in the physician or hospital.Footnote 100 However, another recent study found that public disclosure of payments to physicians was associated not only with decreased trust in the medical profession, but also with decreased trust in one's own physician—irrespective of whether the respondent knew that their physician had received payments.Footnote 101 In contrast, a further group of studies alarmingly suggests that disclosure can increase trust in an advisor's expertise and, in turn, can enhance their persuasiveness—especially when the advisee does not deliberate about the disclosure or has not been given the opportunity to do so.Footnote 102
Patients may not know what to do with the disclosure and may simply ignore it.Footnote 103 But if patients understand the disclosure and its implications for medical practice, and they are able to assimilate this information into their decision-making process, a third and related question remains: are they willing and able to act on it? There is some evidence that patients receiving disclosures from their physicians may experience distrust and, at the same time, feel pressure to follow their physician's advice or recommendations. This may be due to their fear of signaling distrust to their doctor—what some scholars call “insinuation anxiety.”Footnote 104
Even absent insinuation anxiety, the question remains: do patients have the financial and temporal resources to act on the information? The participants in the Cleveland Clinic study did not change their appointments when they received the conflicts of interest disclosure with their appointment letter.Footnote 105 This is hardly surprising given the time costs of canceling their appointment and finding an alternative physician. The outcome might have been different, however, if patients had been given the information at the time of choosing the physician—if someone had explained to them in person the impact of conflicts of interest on medical advice, and if they had been offered an appointment with another physician with similar experience and expertise who did not have financial conflicts of interest. As some scholars of disclosure have observed, it is arguably perverse to place the burden of interpreting and responding to disclosure on patients, especially given their general lack of knowledge of the social science research on bias that would make clear why the relationships being disclosed are problematic.Footnote 106 Moreover, if disclosure is the sole, or primary, measure to address physician conflicts of interest, this imposes the burden of addressing a systemic problem on individuals who are not only ill-equipped to address it, but are also being harmed by it—and, for compelling evidence of harm, we need look no further than the opioid crisis.Footnote 107
As noted above, “advisees” may not only be patients. A recent study explored the impact of disclosures of authors' financial conflicts of interest on peer reviewers for a medical journal.Footnote 108 The study found that the disclosures had no impact on the reviewers' mean ratings of manuscript quality.Footnote 109 But the authors of the study rightly flagged that it is not clear how reviewers should address such a disclosure: “If a disclosure reveals an important [conflict of interest], should the reviewer recommend rejection? Apply special scrutiny to methods? Consult additional experts? No consensus exists about how reviewers should respond to … disclosures; nor, hence, is any direction provided about how to do so.”Footnote 110 Although the authors suggest that the editor could enlist a “highly qualified methodologist” or another reviewer to evaluate manuscripts with serious conflicts of interest, the authors conclude that—at least in its current form—“disclosure does not provide the long desired tool for solving the problems created by conflicts.”Footnote 111
C. Impact on the Influencer
Disclosure of conflicts of interest may have an impact on the corporate entity trying to exercise influence. In theory, disclosure could result in increased giving. One company may wish to outbid its competitors. Alternatively, it could encourage corporations to be more strategic in their giving and “fine-tune” their marketing efforts.Footnote 112 In some cases, disclosure might lead to the withdrawal of funding. Notably, in the wake of the U.S. Senate report on opioid industry funding of several patient advocacy groups and professional societies related to pain management, some opioid companies withdrew funding.Footnote 113 Companies may have done so because they were concerned that these entities would now be perceived as “front groups” for industry,Footnote 114 and that the influence of these groups—and, in turn, their own indirect influence through these groups—would be diminished as a result of the exposure and resulting criticism. Corporate funding of these kinds of groups, however, remains pervasive.Footnote 115 And there is no evidence that the pharmaceutical industry more broadly has pulled away from interactions with civil society groups as a form of influence. There is also little evidence to suggest that the mandated disclosure of industry payments to physicians has had a chilling effect on pharmaceutical companies' willingness to make gifts. As mentioned above, the Open Payments database reveals that gifts have not diminished in the last four years: roughly $9 billion across 11 million transactions.Footnote 116
IV. MOVING BEYOND DISCLOSURE
Although the empirical research on disclosure is extremely valuable, it is important to go beyond this analysis and consider the ways in which disclosure does not address systemic and institutional problems. When all or most practitioners or researchers working in a particular field receive money from drug companies, it may be difficult or impossible to find truly independent advice. At an institutional level, corporate influence may be even more problematic. I have already argued that corporate funding of patient advocacy groups undermines the ability of policymakers to have access to the views and interests of patients that are truly independent of the views of industry actors operating in the same space.Footnote 117 But when the authors of guidelines—whether clinical practice guidelines, dietary guidelines, or some other form of professional guidance—have financial conflicts of interest, this is especially problematic and disclosure is clearly inadequate.Footnote 118 Surely, what professionals, patients, and society need are guidelines from which the risk of bias associated with industry relationships has been eliminated. It is also problematic when the entity being influenced is one of a small number of institutions (one of only two or three so-called “centers of excellence” in a particular field) or, perhaps, the only entity that can perform the functions that we as a society rely on it to perform—such as a public health agency at the local, national, or international level. As influence becomes more pervasive, it may lead to the “exhaustion” of independent advisors and advice, and of science and scientists. It becomes almost impossible to find individuals and institutions without financial relationships, and to find research, practice, education, and policy that has not been influenced by such relationships. In addition, the pervasiveness of influence may increase the risks of potential adverse effects of disclosure. When disclosure reveals that relationships with industry are pervasive, it may reinforce the view among the relevant individuals or institutions that these relationships are unavoidable and, therefore, acceptable.Footnote 119 Even when relationships are pervasive, certain kinds of relationships with industry may still be characterized as scarce or competitive resources, and they may be perceived as not merely acceptable but desirable.Footnote 120
Several scholars and commentators have emphasized that transparency should only be the first step.Footnote 121 For example, De Jong and Steinbrook argue: “The next steps are to move from greater transparency to reforms that … diminish the incentives for industry payments to health care professionals in the first place.”Footnote 122 To this, I would add: we need measures that address the incentives not only for companies to make payments, but also for professionals to accept them; and, beyond removing incentives, we need to consider when prohibitions are warranted. A number of other scholars argue that, while disclosure is generally “a good thing,” it is “no panacea.”Footnote 123 Koch et al. argue that transparency with regard to physicians' relationships with pharmaceutical companies should be seen as “a stepping-stone rather than an end in itself.”Footnote 124 Bachynski and Goldberg describe disclosure as an “ethical floor.”Footnote 125 That metaphor vividly captures the idea that disclosure is necessary but not sufficient. One major concern is that, after disclosure has been implemented, the ethical elevator will remain stuck on the first floor!
For several years, the leading teams of social scientists exploring the effects of disclosure of conflicts of interest have expressed concern that disclosure “can often forestall more substantial institutional change.”Footnote 126 George Loewenstein, Daylian Cain, and Sunita Sah have continued to worry that a serious pitfall of disclosure is moral licensing: “the rationalization that, with disclosure, the profession has dispensed with its obligation to deal with conflicts of interest.”Footnote 127 For this reason, they also caution: “Care must be taken … to ensure that disclosure does not replace more effective measures, such as working harder to eliminate conflicts of interest in the first place.”Footnote 128 Sah similarly warns that “we must take care that once disclosure is implemented, it will not inoculate against more systemic change.”Footnote 129 She warns against viewing disclosure as an “all-encompassing solution” and argues that laws and policies that help eliminate conflicts of interest should be “the first priority.”Footnote 130 Kanter and Loewenstein also express concern that disclosure can “crowd out” more effective solutions.Footnote 131 These concerns have been echoed by bioethicists. For example, Tod Chambers alerts readers to “moral misdirection,” which occurs when disclosure of conflicts causes people to believe that the ethical issues raised by the disclosed relationship have been resolved.Footnote 132
A number of scholars have argued, in various ways and to various degrees, that we need separation from or the elimination of financial relationships with industry.Footnote 133 But policymakers have not yet acted on these appeals. In my view, the disclosure of conflicts of interest should be given a more modest role in the suite of legal and policy responses to corporate influence. In short, disclosure of conflicts of interest, together with a variety of other tools that might promote transparency, can help identify the nature and extent of corporate influence. But we cannot, and should not, rely on disclosure to eliminate that influence. In order to accord disclosure a more modest role, it is helpful to develop a temporal account of corporate influence and policy responses to it.
V. TOWARD A TEMPORAL ACCOUNT OF CORPORATE INFLUENCE AND RESPONSES
In other work, I have developed a framework for recognizing and addressing corporate influence in public health.Footnote 134 That framework emphasizes the ways in which corporations weave “webs of influence” with governmental and intergovernmental bodies, the academy, and civil society (including health professional associations and patient advocacy organizations) in order to protect and promote their commercial interests. I have demonstrated how these relationships create subtle reciprocity and influence that serve to frame public health problems and their solutions in ways that are most favorable to corporate sponsors, and that distort both research and policy agendas, as well as the resulting research and policies. In addition to the concept of “webs” of influence,Footnote 135 a number of other concepts have emerged in the academic literature that also provide systemic perspectives on corporate influence. They include characterizing corporate strategies as “playbooks,”Footnote 136 and analyzing the impact of commercial products or practices as commercial or corporate determinants of health.Footnote 137 Some scholars have also focused on the nature and types of corporate power that inform and underpin these strategies of influence.Footnote 138
I hope to add to these frameworks by articulating the importance of a temporal account of corporate influence in public health. Fleshing out such a framework in any one area of medicine or public health will require a much more detailed analysis than I am able to provide in this piece. But the purpose of the temporal account I propose is three-fold. First, the account should help scholars and policymakers recognize that the nature and extent of corporate influence in public health has evolved over time. Recognizing that things were not always this way may help address status quo bias or, worst still, a belief that the status quo is immutable.Footnote 139 Second, a temporal account should provide a framework for rethinking our legal and policy responses to corporate influence in public health—to take stock of where we are and where we might go. This framework can help us reassess what we can and cannot achieve with the current paradigm of conflicts disclosure, and enable us to develop other tools that can eliminate or mitigate corporate influence. Third, a temporal account can make us attentive to the adaptive practices of industry actors, as their strategies of influence will inevitably evolve in response to these tools. This, in turn, will allow lawmakers and policymakers to revisit, reassess, and refine their approaches.
A. Toward a Temporal Account of Corporate Influence
For our purposes, it is helpful to broadly categorize corporate influence in public health and the legal and policy responses to such influence in “waves.”Footnote 140 The first wave of influence might be characterized as “emergent influence.” In this phase, corporate actors develop and pilot marketing strategies intended to promote their products by influencing not only “end-users” but also a variety of other actors, including health professionals. Here, corporations begin to experiment with strategies designed to exercise not only market power but also political power. Companies recognize that these strategies are necessary to protect and promote their commercial interests. In this wave, there may be limited or no evidence of these strategies in the public domain. For example, the sugar industry's secret efforts half a century ago to shift critical attention from their own products to dietary fat were only exposed in the last five years.Footnote 141
In the second wave, which I call “expansive influence,” corporations refine their strategies for the exercise of political, as well as market power, and seek to apply them as widely as possible. In this phase, corporations use relationships with other bodies in government, the academy, and civil society in order to recruit new “partners.” So, for example, in the United Kingdom, Coca-Cola funded exercise initiatives in collaboration with a few local government bodies, before using those relationships as precedents for expanding the initiative and building similar relationships with a variety of other local governments.Footnote 142 It is during this phase that webs of relationships emerge, implicating institutions within government, the academy, and civil society. At this stage, evidence of influence is usually apparent. The full extent of the influence, however, may not be known, even when the existence of many relationships has been made public.
In the third wave, which I call “exhaustive influence,” corporate strategies have become so widespread that they have exhausted—or, at least, threaten to exhaust—individuals and entities in government, the academy, and civil society that are independent of industry influence. In this phase, corporations and trade associations work aggressively to maintain and expand their webs of influence, and to ensure that legal and policy responses do not undermine their influence. Sometimes, the numbers will speak for themselves. For example, when more than 80% of the largest patient advocacy groups receive industry funding,Footnote 143 we have good reason to be concerned that corporate influence among these entities is becoming exhaustive.Footnote 144 At other times, the numbers may not speak for themselves (or, worse still, may provide false assurance).Footnote 145 For this reason, we also need qualitative assessments to determine whether influence is exhaustive, or threatening to become so. But policymakers should not, of course, wait for the situation to become that dire before taking action to address corporate influence.
B. Toward a Temporal Account of Legal and Policy Responses to Corporate Influence
It is also helpful to consider legal and policy responses to corporate influence in analogous waves. In the first wave, “emergent responses,” the problem of corporate influence is primarily characterized as one of individual financial conflicts of interest, and laws and policies focus on the disclosure of those conflicts. Individuals who are exposed for failing to disclose relationships with the private sector are subject to naming and shaming. But institutional conflicts of interest and systemic perspectives on corporate influence are neglected. Where attention is paid to institutions, the focus is often narrow and ex post—pursuing accountability for one bad actor or a small group of bad actors. The first criminal proceedings against Purdue Pharma for the fraudulent marketing of OxyContin, a case that was settled in 2007, might be considered an example of that.Footnote 146
The thousands of cases currently pending against a variety of opioid manufacturers and distributorsFootnote 147 might be seen as evidence of expansive responses to corporate influence. But, as with the proliferation of disclosure policies, the effect of these approaches should not be overestimated. They are certainly responses to influence, but they cannot be relied upon to neutralize that influence. On the contrary, despite the inglorious history of opioid industry practices, government agencies continue to partner with drug companies in efforts to address the opioid epidemic, thereby creating further opportunities for corporate influence.Footnote 148 These partnerships are likely to promote highly-profitable pharmaceutical solutions to both pain management and addiction at the expense of other measures that also warrant exploration, among them, increased regulation of pharmaceuticals and drug company marketing practices.Footnote 149
In my view, the goal for policymakers and institutional leadership should be to develop counterstrategies of influence that are as ambitious as the corporate strategies of influence they seek to address. Although it may never be possible to achieve this entirely, the aspiration should be toward exhaustive responses, that is, a comprehensive suite of measures designed to eliminate and/or truly counteract corporate influence in public health. These measures must tackle the cumulative effects of the webs of relationships that corporations employ to exercise influence. This will require systemic and structural solutions to corporate influence—including “sequestration” or separation—as well as other reforms to keep corporate actors at arm's length.
The efficacy of these interventions should be periodically reassessed and appropriate adjustments made in order to address the adaptive practices of industry actors. Disclosure and other measures to promote transparency can facilitate such assessment by both researchers and policymakers. Attempts to measure corporate influence, the effects of corporate influence, and the efficacy of interventions designed to address corporate influence will be difficult.Footnote 150 So we should not wait for conclusive evidence before developing strategies to address corporate influence in public health. On the contrary, delaying interventions to tackle corporate influence pending further evidence would play into what some might characterize as a “playbook” strategy of corporate influence. We know enough about the pervasiveness of corporate influence in public health and the associated biases to take action now. That said, additional evidence regarding both the efficacy of legal and policy responses to corporate influence and the adaptive practices of industry actors in the wake of those responses can and should be used by lawmakers and policymakers in order to refine their approaches.
CONCLUSION
The earliest proponents of disclosure were aware of its limitations. In 1913, when Louis Brandeis famously wrote that “[s]unlight is said to be the best of disinfectants,” he was not arguing only for transparency.Footnote 151 While calling for disclosure of bankers' commissions or profits, he also contended that the banks' “sources of … undue power must be effectually stopped,”Footnote 152 and that some bankers' services should be entirely dispensed with. Brandeis, who acknowledged the sunlight metaphor was not his own, borrowed it from American Commonwealth (1888).Footnote 153 In that book, Viscount Bryce claimed that, when a “rankling sore in the body politic” is disclosed, it is “half destroyed.”Footnote 154 But that which is half destroyed may, of course, still fester or endure. Alas, the nuances of the claims by Brandeis and Bryce have been lost through careless repetition.
Lawmakers, policymakers, and institutional leaders in the academy and civil society need to develop counterstrategies that are as ambitious and pervasive as the corporate strategies of influence they seek to address. Disclosure alone cannot perform that function. By increasing transparency, policymakers may illuminate corporate influence but they cannot and should not rely on transparency to eliminate corporate influence. It is time for a new wave in public health law and policy that recognizes the limitations of disclosure of conflicts of interest; accords it a more modest role; and explores new approaches, including separation and other structural reforms to more effectively address the systemic problem of corporate influence.
He is extremely grateful to Michele Mekel, Marc Rodwin, and Sunita Sah for comments on an earlier draft. Please excuse any errors or omissions—final revisions to this piece were made during the COVID-19 pandemic.
