The long recognized fact that human behavior is complex, not always rational, and driven by a number of factors—physical, psychological and economic—is receiving much renewed attention. Recent bestselling books, such as Predictably Irrational: The Hidden Forces That Shape Our Decisions, by Dan Ariely, and Thinking Fast and Slow, by the Nobel Laureate Daniel Kahneman, provide engaging examples of not only the complexity of human behavior, but often the irrationality of our thoughts and behaviors. Humans working in health care are certainly no exception. The desire to control or direct behaviors of providers, patients and institutions to improve the quality of medical care has increasingly centered on the power of incentives to drive behavioral changes. In current policy and medical practice, “aligning incentives” among stakeholders is the talk of the day.
For providers, desired behavioral changes include a range of activities such as stricter adherence to clinical practice guidelines and more consistent attendance at educational conferences. For patients, desired behavioral changes often focus on diet, exercise and participation in health screenings. The incentives to drive desired behaviors have taken the form of big policy ideas such as “pay for performance” for providers and “value-based” insurance premiums for patients. Other stakeholders in the health care industry – including pharmaceutical, medical device, and medical education companies – have also engaged in a variety of incentives, mostly financial, to influence the behavior of clinicians. Some obvious examples: paying for lunches to give informational talks about their products; providing small gifts such as coffee mugs, pens, calendars with brand reminders; providing free drug samples for distribution to patients; subsidizing medical education programs to lower costs for clinicians to attend; or paying clinicians directly for participating in educational programs or for enrolling patients in clinical trials.
Of course, health care professionals should operate under a behavioral and ethical code of professionalism, which, at its core, requires us to put the health and medical care interests of our patients above personal and institutional financial interests. In fact, the American Board of Internal Medicine Charter on Medical Professionalism correctly notes that one of the key responsibilities of medical professionalism is to maintain patient trust by managing conflicts of interests that inevitably arise when third parties explicitly seek to incentivize providers to direct behavior in specific ways. This is of particular importance given the pervasiveness of industry relationships with providers. In one study, 80% of physicians reported receiving food and beverages in their workplace paid by pharmaceutical companies; 78% reported receiving drug samples; 35% were reimbursed for costs of attending professional meetings or CME events, and 28% received payments for consulting, speaking, or enrolling patients in clinical trials.
Many clinicians, including this author, find it instinctively insulting that their administrative colleagues, policy makers or patients would believe that a relatively trivial financial incentive, such as the cost of a lunch at their workplace, or the gift of a pen or a coffee mug, would influence their behavior in ways that violate their professional obligations to their patients. But my reading of recent thoughtful research and scholarship in the area of what influences decision making and bonds of trust between individuals suggests that health care professionals in particular should be circumspect in our beliefs and opinions about the power of even small gifts and inducements to influence our behaviors.
However, I am also persuaded that we could over-react and not fully understand all of the unintended consequences of controlling too tightly all attempts to incentivize behaviors in the health care system.
On the cautionary side, a very interesting new book, Strings Attached: Untangling the Ethics of Incentives, by philosopher, political scientists and ethicist, Ruth Grant, warns that incentives are intended to “shift behavior from its usual paths” and, most importantly, often operate in an arena in which there is an imbalance of power between participants. Power asymmetries in health care are ubiquitous and should be under the careful scrutiny of ethicists. As was pointed out in a recent New York Times review of her book, Grant notes that one person’s incentive could be another person’s bribery or blackmail. Professor Grant proposes three standards to evaluate the ethical and moral legitimacy of incentives: the legitimacy of their purpose; the autonomy involved in choosing to accept an incentive, and the effect on the character of the parties involved. Like many of my colleagues in health care, I tend to inappropriately discount the power asymmetry in the doctor-patient relationship, and thus fail to recognize how even a small gift from a pharmaceutical company could be perceived by my patient as bribery, not as harmless inducement to participate in educational activities, or to learn more about their products.
Professors Carol Tavris and Elliot Aronson, in their book, Mistakes Were Made (But Not by Me), educate us about the powerful effects of cognitive dissonance in allowing us to self-justify errors and questionable acts that we commit. This theory provides the rationale for my feeling very confident, indeed certain, that I cannot be corrupted in my ethical obligations to my patients by accepting gifts from pharmaceutical representatives, because to admit such would be dissonant to my belief that I am a highly professional physician. Yet, this is discordant with evidence that physician-industry associations do influence prescribing behavior, which is the reason that companies spend $19 billion a year on these relationships. 
In fact, quite consistent with cognitive dissonance theory and the point of their book is the fact that although physicians in general deny that industry relationships have negative effects on patient care, they often are less convinced that the same is true of their physician colleagues!
A recent investigation in neuroscience and cognitive psychology into the science of influence and reciprocity provides many fascinating and yet troublesome observations concerning the ways in which subtle and unconscious factors influence decision making. Functional magnetic resonance (fMRI) imaging of the brain while two individuals play a game with economic consequences reveals selective activation of a specific brain region that anticipates the development of trusting behaviors between the individuals, and demonstrates that humans have a tendency to expect that favors given will be paid back.
Furthermore, the subjective value that persons place on objects such as a work of art correlates with activation of other specific brain regions and, most importantly, can be influenced by knowing who is sponsoring or funding the experiment, although art experts were able to insulate themselves against this judgment bias.  Other psychological and behavioral studies have found that all people cheat when they have a chance to do so; that we seek out information selectively and process information in a biased fashion, and that the psychological mechanisms underlying these behaviors are unconscious.
Those are interesting observations and have profound and troubling implications for the ways in which incentives can manipulate behavior in ways that are not obvious. On the other hand, in our zeal to overcome the biases and negative effects engendered by unrecognized conflicts of interest, we should not forget some of the positive effects facilitated by the web of provider-industry relationships so common in health care. These include, among others, the facilitating effects on drug and medical device development and approval. Increased awareness of drugs and devices through drug detailing efforts can in fact lead to improving care if these efforts lead to increasing the number of beneficial drugs prescribed for disorders that are currently undertreated.
Also, we should not be naïve to the consequences on patient care inflicted by well-intended but punitive policies to correct the negative effects of conflict of interest. It is wise to remember that Robert Merton originated the term “unintended consequences” to refer to the notion that interventions in complex highly adaptive systems (i.e., think health care) tends to create unanticipated and often undesirable outcomes. Merton noted that unanticipated problems are more likely when there is incomplete or erroneous analysis of a problem, or when overwhelming concern with immediate issues associated with a problem clouds the analysis of long term solutions.
One major potential unintended consequence of pursuing too restrictive conflict of interest policies in health care is that they can disqualify the very experts needed to provide advice and consultation. Consider the fact that more than one in three positions on the Food and Drug Administration advisory panels are vacant ( 218 vacant out of approximately 600 positions) because of concerns about conflicts of interest. This is the case even though a study evaluating the effects of conflict of interest on voting patterns in FDA advisory committees concluded that, “A weak relationship between certain types of conflicts and voting behaviors was detected, but excluding advisory committee members and voting consultants with conflicts would not have altered the overall vote outcome at any meeting studied.” 
How then can we respond to the legitimate concerns about bias and conflicts of interests? I suggest first we need to be creative and thoughtful about our approaches and have a healthy dose of skepticism when simplistic answers are proposed. This caution is brought home –in the recent commentary by Sunita Sah and colleagues – that even our mainstay of dealing with conflict of interest, the insistence on disclosure, has profound unintended and even perverse consequences. We can also structure relationship and networks in ways that minimize conflicts of interest. Furthermore, we can take advantage of other behavioral research that indicates that our propensity to cheat is moderated by just-in-time reminders of moral and ethical standards (see reference 8). We can go beyond simple disclosure of conflict of interest and put in place better patient education programs that make clear what conflicts of interests are present in their circumstances and their possible consequences to patient care. Finally, we need system changes to assure that second opinions are truly independent.
In our upcoming forum, “Health care Leadership: Navigating Ethics, Incentives and Conflicts of Interest,” we are asking questions such as: What unintended consequences follow from the transformation of patient/physician, patient/payer, physician/payer relationships into remunerative exchanges? In exchanges, increasingly laced with financial or reputational incentives, how is the primary character of the relationship changed? Are we inadvertently crowding out virtue, re-shaping “do what is right” into “do what is compensated?”
Let us know what you think about these complicated issues by adding your reflections in our Comments thread.
 ABIM Foundation. Medical professionalism in the new millennium: a physician charter. American Board of Internal Medicine Web site: http://www.abimfoundation.org/Professionalism/Physician Charter .aspx. Accessed October 1, 2012
 Campbell EG, Gruen RL, Mounford J et al. A national survey of physician-industry relationships. N Engl J Med 2007;356(17):1742-1750.
 Koehn N. When Life is a Bunch of Carrots. The New York Times, February 4, 2012. http://www.nytimes.com/2012/02/05/business/strings-attached-looks-at-incentives-and-ethics… Accessed October 1, 2012.
 Brennan TA, Rothman DJ, Blank L et al. Health Industry Practices that Create Conflicts of Interest: A Policy Proposal for Academic Medical Centers. JAMA 2006;295:429-33.
 Chren MM. Interactions between physicians and drug company representatives. AM J Med 1999;107:182-3.
 King-Casas B, Tomlin D, Anen C et al. Getting to know you: Reputation and trust in a two-person economic exchange. Science 2005;78-83.
 Kirk U, Harvey A, Montague PR. Domain expertise insulates aginst judgment bias by monetary favors through a modulation of ventromedial prefrontal cortex. PNAS 2011;188(25):10332-10336.
 Mazar N, Ariely D. Dishonesty in everyday life and its policy implications. J of Public Policy & Marketing 2006;25(1):1-21
 Dana, Jason; Weber, Roberto A.; and Kuang, Jason Xi, “Exploiting Moral Wiggle Room: Experiments Demonstrating an Illusory Preference for Fairness” (2005). Department of Social and Decision Sciences. Paper 8.
 Merton, Robert K. The Unanticipated Consequences of Purposive Social Action. American Sociological Review 1936; 1 (6): 895-904.
11 FDA Conflict of Interest Rules Means Fewer Experts on Advisory Panels.http://www.policymed.com/2011/05/fda-conflict-of-interest-rules-means-fewer-experts-on-advisory-panels.html, May 23, 2011
 Lurie P, Almeida CM, Stine N et al. Financial Conflict of Interest Disclosure and Voting Patterns at Food and Drug Administration Drug Advisory Committee Meetings. Author Affiliations: Public Citizen’s Health Research Group, Washington, DC (Drs Lurie, Almeida, and Wolfe and Mr N. Stine), and Department of Earth and Planetary Science, University of California, Berkeley (Mr A. Stine). JAMA. 2006;295(16):1921-1928.
 Loewenstein G, Sah S, Cain DM. The unintended consequences of conflict of interest disclosure. JAMA. 307(7):669-70, 2012 Feb 15