It's lunchtime. You are an intern on the medicine service. Your supervising resident comes by where you are charting and reminds you "There's a drug lunch today." You quickly finish your notes on this patient and head for the conference room where you find most of the other residents assembled, with a make-shift cafeteria line offering salad, and lasagna and soft drinks. You load up your plate and as you pick up the plastic flatware, you are greeted by the smiling face of the pharmaceutical manufacturer sales representative who provided the lunch. What do you say?
The preceding is a scenario almost every medical student, intern, resident and practicing physician will recognize. It is argued in this paper that what you say to the "drug rep" will influence your enjoyment of your meal, the cost of your meals across the year, your relationships with your fellow residents, your relationships with your patients, and the quality and cost of the medicine you practice.
Pharmaceutical manufacturers make money by manufacturing, and selling, pharmaceuticals. The prescription drug arrangement is unlike ordinary sales situations in that the person who makes the decision about what to buy is different from the person who pays the bill or uses the product. U.S. federal regulations limit marketing of prescription drugs directly to the public, so marketing activity is directed largely at physicians. The person making the decision of which drug to buy (the physician) has a fiduciary relationship to the person who is paying the bill (the patient): Above all, do no harm. (Or, in the era of managed care, perhaps the physician has a more complicated relationship with the third party paying the bill "for" the patient.) With this fiduciary relationship comes an obligation upon the physician to act in the patient's best interest. When the physician is courted by the drug rep, is this responsibility being carried out?
Most physicians view themselves as hard working people, who are acting in the patient's interest. This identity is usually important to them. They do not like the image of themselves as being bought by drug companies. They see themselves as independent thinkers. Yet lunches provided by drug com-panies, often with speakers paid for by drug companies, or indeed a talk by the drug rep, are commonplace in hospitals and clinics, as are evening meetings, dinners and receptions at medical conferences and continuing education conferences for practicing physicians, all sponsored and paid for by drug companies (Hodges 1995, Reeder, Dougherty & White 1993, Lurie, Rich, Simpson, Meyer, Schiedermayer, Goodman & McKinney 1990). Drug reps leave gifts in the office, for example, drug samples, pens, pads and "sticky notes" (Reeder, Dougherty & White 1993).
Given the identities they wish to enact, doctors would be expected to, and frequently do (Hodges 1995), say that the gifts from drug companies have no effect on their prescription practices. Indeed, they often experience the question as an insult to their integrity (personal observation at Indiana University Medical Center and University of Pennsylvania School of Medicine). To counter the argument that obviously the persuasion methods must be working or the drug companies wouldn't be spending millions of dollars on them, physicians may concede that others are influenced by these lunches and gifts, but they themselves are not (Hopper, Speece and Musial 1997).
How is it that a practice that would seem to outsiders to obviously influence physician behavior, and that research has shown does influence physician behavior (Chren & Landefeld 1994, Caudill, Johnson, Rich & McKinney 1996, Lurie, Rich, Simpson, Meyer, Schiedermayer, Goodman & McKinney 1990), continues and is defended by many physicians (Hodges 1995, Reeder, Dougherty & White 1993)? What allows physicians to sidestep the acknowledgment of this potential unwarranted influence? It could, of course, be frank dissimulation, driven by greed, and surely some of that exists. I argue below for a less malevolent explanation: That communicative behavior experienced as choosing to be polite, provides the opportunity for self-perception and cognitive dissonance to change attitudes toward pharmaceutical products.
I thanked him for lunch and told him I'd seen some good results from his product--which was true, though I didn't mention one bad experience. It would have been rude to just walk past him without saying anything, and then eat his food.
What presses one to be polite in this setting? There are several possibilities, and probably each makes a contribution. Unless situation-specific information dictates otherwise, strangers are to be responded to within the culture's politeness norms. An important aspect of this behavior, to be discussed in detail below, is that the social actor experiences polite behavior as freely chosen, although the fact that so few choose to do otherwise suggests that this choice is made under considerable constraint.
A second contribution to politeness in this setting is the norm of reciprocity: The drug rep is being polite, even generous, and one is therefore likely to reciprocate (Gouldner 1960, cited in Burgoon 1994). As with the politeness norm, behavior that follows the norm of reciprocity is likely to be experienced as freely chosen, while again the existence of the norm suggests that this choice is constrained.
But is this behavior merely politeness? As typically conceptualized by scholars, polite behavior is behavior which serves to soften face threats (Brown & Levinson 1978). Overall in the drug lunch encounter, the face threat is to the physician whose behavior is being shaped toward prescribing the drug rep's product. The communicative situation of the resident physician facing the drug rep at the end of the lunch line is, therefore, not a prototypic politeness situation. No request is being made that would threaten the autonomy of the drug rep; no critique or rejection is being offered that would threaten the competence or inclusion face of the drug rep (Lim & Bowers 1991).
This seeming failure of politeness theory to apply to the drug lunch situation comes in part from a too narrow conceptualization of politeness and of face. In this limited conceptualization, face and face threats are only seen evaluatively and the content is lost. In its larger meaning, one's face, as outlined for example in the seminal work of Goffman (1959), is the character one claims to be. The presentation of self is not merely about looking good or even looking independent. One presents oneself as a certain kind of person who should be treated in a certain kind of way. Moreover, the character one presents oneself as being not only lays claim to moral rights to be treated in such-and-such a way, but it provides the stuff around which interactions are organized. In the case of a physician, for example, a role-consistent compliment such as "I trust your judgment about these things; you always seem to turn out right" is likely to be appreciated, while a role-inconsistent compliment such as "It's great that you make the time to spend with your kids" is likely to be greeted with ambivalence at best.
In the drug lunch setting, the identity of the resident physician is as a hard-working person who doesn't have time to go out for lunch, who makes little money and appreciates the free food, and who recognizes the importance of medical education. In order to maintain this identity, the drug lunch situation must be defined primarily as one of convenience and education, not one of influence and marketing. An alternative explanation, then, of the "polite" conversation with the drug rep is that it is about maintaining face, but not about maintaining the face of the other whose face one is threatening. Rather, the polite behavior serves to maintain the face of the actor. In this sense, both the physician and the drug rep are part of a team effort to maintain this definition of the situation, from which each benefits in different ways. This is not unlike the teamwork suggested by Emerson's (1994, cited in Griffin 1994) description of the cooperative and motivated teamwork involved in maintaining the definition of a gynecological examination as a routine medical procedure. The explanation of the polite conversation with the drug rep as serving to maintain the definition of the situation as educational, also explains why just saying "thanks" is not enough to avoid being rude.
This teamwork extends beyond the resident-rep dyad: Other residents also want to maintain the illusion of education without venality. If one resident is curt with the rep, other residents are likely to bring social sanctions to bear and coax the errant resident back in line. This group pressure also springs, of course, from the desire to have the drug rep return--especially if it's a good lunch.
It is likely that all of the phenomena discussed in the last section--politeness norms, reciprocity, maintaining the definition of the situation as educational, and group pressure--play a role in shaping the conversation between the resident and the drug rep. A polite conversation with the rep is likely to leave one to enjoy lunch free of guilt, and to enjoy the team bond with other residents. And enjoying the setting will allow one to return and thus save money over buying one's own lunch. Whatever the constraints that influence it, however, "being polite" in this context is likely to involve discussing positive experiences with and good feelings about the drug rep's product. These discussions serve many purposes for the physician as social actor and the drug rep as persuader. Key consequences from the physician's side are (1) increasing the salience of the positive experiences with the product, thus increasing availability of this information in the future, (2) providing the opportunity for the drug rep to deliver company-approved information concerning the products , (3) establishing a friendly relationship between the physician and the rep, and (4) serving as an object of self-perception for the physician.
If positive information about the drug is made differentially available (Kahneman & Tversky 1973), then that information is likely to be retrieved when prescribing decisions are being made, thus increasing the likelihood that the product will be prescribed. If the physician listens to the company information--especially, if s/he continues with polite responses like "good point" and "that's an interesting study" and "what about the situation where..."--this information, too, will become differentially available and has been found to contain inaccuracies that favor the company products (Ziegler, Lew & Singer 1995, Lexchin 1997, Hemminki 1977).
Having a friendly relationship with the drug rep makes one want to "toss some business that way." For example, one resident physician (A. G., name withheld for privacy reasons) who had arranged for a certain drug rep to provide that day's meal and who vehemently denied any undue influence on physicians from accepting this meal, nevertheless, concluded her formal thanking of the rep by saying "So, remember who brought us this great meal." If accepting the meal is not supposed to influence prescribing, exactly when is one supposed to remember who fed one? One study (Sandberg 1997) found that drug reps who paid attention to medical students in these one-on-one encounters thereby engendered goodwill toward both the rep and the message, and that this was more effective than giving books to medical students. If this student-rep bond results in a less critical attitude on the part of the student toward the information source, then attitude change is likely to increase.
The drug rep is probably aware of all of these effects, and may consciously manipulate them by being intentionally friendly, by paying attention to medical students and interns, by encouraging talk of positive experience with the product, and even by encouraging negative feedback which may induce guilt or provide an opportunity to provide counter information. Some reps are exquisitely skillful at this.
But the resident physician or medical student is probably less aware of ways in which the situation serves to shape their communicative behavior. This is, of course, just the situation said to produce cognitive dissonance and attitude change (Festinger 1957). In the now-famous Festinger and Carlsmith (1959) $1/$20 experiment, subjects who had engaged in an extremely boring task were offered either $1 or $20 to lie to an incoming subject about the boringness of the task. The $1 subjects subsequently reported they felt the experiment was more interesting than the $20 subjects. Cognitive dissonance theory explains the attitude change in the $1 subjects as resulting from the inner discomfort of holding the belief that the task was boring and the observation of one's action as saying it was not. This inner discomfort or dissonance can be reduced by adjusting slightly the attitude in the direction of the behavior. For the $20 subjects, this inner dissonance was less because the large payment offered a ready explanation for the behavior.
The physician at the drug lunch is like the $1 subject. The physician engages in a "white lie" of mentioning to the rep positive experiences with the rep's product(s), but s/he does this for a small sum (a cafeteria style meal) and indeed for a meal that has already been "freely given" to him or her. The inducement to this behavior is therefore not salient. Cognitive dissonance would predict that in this setting the physician's attitude would change in the direction of the "lie". (1)
Not all physicians in this situation are likely to react in exactly the same way. This would seem to depend in part on the attitude the physician has toward the drug rep's products. If the physician has uniformly positive thoughts and feelings about these products, saying so to the rep is not a lie, and hence no attitude change would be expected. If the physician had uniformly negative thoughts and feelings about the products, s/he may not be able to think of anything positive to say and may dodge the rep; or s/he may produce a positive statement, but feel intense pressure to do so. This felt pressure makes the uniformly negative physician more like the Festinger and Carlsmith $20 subject and again no attitude change would be expected (though resentment of the pressure might adversely affect the physician-rep relationship and produce the opposite effect from those noted above when the rep pays attention to the medical student). Physicians with ambivalent and/or uncertain attitudes toward the rep's products--perhaps those early in their training--are more likely to be influenced by the drug lunch setup.
Bem (1967) has offered self-perception theory as an alternative explanation for phenomena explained by cognitive dissonance. Self-perception theory is based on the principle of "I guess I was hungrier than I thought; look how many pancakes I ate!" On this account, the physician at the drug lunch notices the positive things s/he has said to the drug rep and concludes "I must like Croxifin more than I thought."
The notions of the range of acceptance and range of rejection from social judgment theory (Sherif, Sherif & Nebergall 1965), can be used to extend the reach of cognitive dissonance and self-perception by making them complementary rather than competing explanations (Fazio, Zanna and Cooper 1977, Rhodewalt 1986). The range of acceptance is the set of all those attitudinal statements that one finds plausibly true, while the range of rejection represents all those statements one finds untenable. When one observes oneself engaging in behavior within one's range of acceptance, those aspects of one's opinion consonant with the observed behavior will become salient, thus producing a priming effect, but no long-lasting change. When one observes oneself engaging in a behavior within one's range of rejection, however, dissonance is aroused and permanent attitude change typically occurs.
The overall attitude portrayed by the physician to the drug rep may be within the physician's range of acceptance, in which case a short-term priming effect would be expected. Or the overall attitude portrayed might be in the physician's range of rejection, and we would anticipate permanent attitude change. But this latter attitude change will only occur when the physician perceives his or her behavior to have been freely chosen. The drug rep who wants to produce maximal usage of his or her product will therefore want to encourage physicians to present their attitudes as sufficiently more positive than their range of acceptance, so as to induce dissonance, but not so far beyond the range of acceptance that the pressure to provide the positive portrayal becomes salient. Ironically, this analysis suggests that the more uncertain the physician is--that is, the wider the range of acceptance--the less permanent effect the positive portrayal to the drug rep will have.
"I would never change the drug I prescribe just because a drug rep gave me a pen."
"But if there are three plausible choices and one of them is on the pen you use all day, aren't you more likely to think of that one first and hence prescribe it?"
"Well, yes, that's possible; I hadn't thought of that."
Physicians, like other human beings, do not like to have their integrity questioned. Physicians, also like other human beings, like to have pleasant things: Free lunches, nice gifts, conferences with lots of free time in exotic places that can be written off their taxes as a business expense. Given the weight of evidence that suggests that accepting food and gifts from drug reps changes physician prescribing patterns (Chren & Landefeld 1994, Caudill, Johnson, Rich & McKinney 1996, Lurie, Rich, Simpson, Meyer, Schiedermayer, Goodman & McKinney 1990), how can an understanding of this interpersonal influence situation be used to counteract the corrupting influence of drug lunches and their brethren on the fiduciary relationship of physician to patient?
Several efforts have been made in this direction. One approach is to establish guidelines regulating contact between resident physicians and drug reps. It has been proposed, for example (Anonymous 1993) that for drug-company sponsored educational events, the residents and faculty should retain full control of the educational content, that educational materials should be reviewed before giving them to residents and that gifts and drug lunches not be allowed; these authors reported that half of the drug reps were satisfied with these arrangements, whereas the other half reacted negatively. Johnstone, Vanezuela and Sullivan (1995) encouraged discussion of legal and ethical issues concerning these contacts, banned drug lunches entirely, and adopted the American Medical Association guidelines on gifts. (See http://www.ama-assn.org/ med-sci/pra2/ethics.htm#Gifts) Despite the existence of guidelines at both the national and institutional level, even chief residents may not be aware of them (only 46% were aware in a study of emergency medicine residents by Reeder, Dougherty & White 1993). Guidelines, per se, however, do not appear to solve the problem, as Sergeant, Hodgetts, Godwin, Walker and McHenry (1996) found that only 23% of Canadian family practice residents studied had read the policy statement by the Canadian Medical Association which prohibits personal gifts (http://www.cma.ca/inside/policybase/1994/ 1%2D15.htm), and whether the resident had or had not read the policy was unrelated to whether the resident would attend a private dinner paid for by a drug rep (about 50% of residents agreed they would attend). More encouragingly, Brotzman and Mark (1993) compared family practice residents from programs that restricted contact with drug reps, to family practice residents in programs that did not. They found that residents in restricted programs saw fewer benefits from interacting with drug reps and were less likely to feel accepting gifts was appropriate. Although one cannot assess causality in a cross-sectional study, one possible explanation for this finding is that attitude follows behavior, just as is argued here occurs in the drug lunch situation.
A second approach is to educate physicians to make them better consumers of pharmaceutical information. Some progress may result merely from an open discussion about what guidelines should be. Shaughnessy, Slawson, and Bennett (1995) describe a formal curriculum developed for family practice residents to teach them to evaluate presentations by drug reps. They found that residents rapidly learned to identify logical fallacies and misleading sales techniques, and indeed that drug reps were generally happy with an educational identity. They argue for the physician as information manager, rather than merely as information repository. A similar intervention by Hopper, Speece and Musial (1997) found that after the intervention, residents were more aware of potential unethical marketing practices and more willing to believe that other physicians' prescribing patterns could be influenced by accepting gifts.
One notes with a sad, but perhaps too-close-to-home, irony that it is other physicians' prescribing that is believed to be potentially affected by accepting gifts. The dialog at the beginning of this section suggests, in accord with common sense, that proposing a cognitive mechanism for the effects of drug company marketing, as is argued for here, is more likely to find acceptance by physicians concerning their own behavior than one that puts their integrity in question by suggesting a motivated effect. A face saving approach is more likely to provide an initial success that can perhaps lead to examination of a broader set of issues in the longer term.
It is not always possible to separate motivation and cognition, and indeed, if the motivation for the polite conversation with the drug rep is to maintain the definition of the situation as education rather than marketing, then it is this motivated cognition that produces the behavior that allows cognitive dissonance (or self-perception) to work and change the physician's attitude toward the rep's product. Thus, the attempt to not see the situation as marketing results in its being a successful marketing situation: A motivated behavior unleashes a cognitive bias.
Shaughnessy, Slawson, and Bennett (1995) argued that the physician should be an information manager, not an information repository. The present analysis suggests a further step is needed: It is not merely an understanding of marketing tactics, or good training in scientific assessment of data, that will provide a buffer for physicians in their effort to uphold professional principles. Physicians need to be trained in how their own minds work in connection with their communicative behaviors. If they could understand that polite behaviors are not freely chosen, but highly constrained by the situation, and that being polite both changes access to their own memories and opens those memories to slanted drug-company information, perhaps trust in doctors could be augmented and medical care quality could be raised and costs could be lowered (Caudill, Johnson, Rich & McKinney 1996).
Anonymous, (1993). Development of residency program guidelines for interaction with the pharmaceutical industry. Education Council, Residency Training Programme in Internal Medicine, Department of Medicine, McMaster University, Hamilton, Ont.(see comments). Canadian Medical Association Journal, 149, 405-408.
Bem, Daryl J. (1967). Self-perception: An alternative interpretation of cognitive dissonance phenomena. Psychological Review, 74, 183-200..
Brotzman, G. L., & Mark, D. H. (1993). The effect on resident attitudes of regulatory policies regarding pharmaceutical representative activities. Journal of General Internal Medicine, 8, 130-134.
Brown, Penelope, & Levinson, Stephen C. (1987). Politeness: Some universals in language use. Cambridge, UK: Cambridge University Press.
Burgoon, Judee K. (1994). Nonverbal signals. In M. L. Knapp, & G. R. Miller (Eds.), Handbook of interpersonal communication(2nd ed.) (pp. 229-285). Thousand Oaks, CA: Sage.
Caudill, T. S., Johnson, M. S., Rich, E. C., & McKinney, W. P. (1996). Physicians, pharmaceutical sales representatives, and the cost of prescribing. Archives of Family Medicine, 5, 201-206.
Chren, M. M., Landefeld, C. S., , , & , (1994). Physicians' behavior and their interactions with drug companies. A controlled study of physicians who requested additions to a hospital drug formulary(see comments in: JAMA 1994 Aug 3;272(5):355). Journal of the American Medical Association, 271, 684-689.
Emerson, Joan P. (1994) Behavior in private places: Sustaining definitions of reality in gynecological examinations. pp.189-202 in Peter Kollock & Jodi O'Brien (Eds.) The production of reality. Thousand Oaks, CA, Pine Forge
Fazio, Russell H., Zanna, Mark P., & Cooper, Joel (1977). Dissonance and self-perception: An integrative view of each theory's proper domain of application. Journal of Experimental Social Psychology, 13, 464-479.
Festinger, Leon (1957). A theory of cognitive dissonance. Evanston, IL: Row, Peterson.
Festinger, Leon, & Carlsmith, James (1959). Cognitive consequences of forced compliance. Journal of Abnormal and Social Psycholgy, 58, 203-210.
Goffman, Erving (1959). The presentation of self in everyday life. New York: Anchor.
Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American Sociological Review, 25, 161-171.
Hemminki, E. (1977). Content analysis of drug-detailing by pharmaceutical representatives. Medical Education, 11, 10-215.
Hodges, B. (1995). Interactions with the pharmaceutical industry: experiences and attitudes of psychiatry residents, interns and clerks. Canadian Medical Association Journal, 153, 553-559.
Hopper, J. A., Speece, M. W., & Musial, J. L. (1997). Effects of an educational intervention on residents' knowledge and attitudes toward interactions with pharmaceutical representatives. Journal of General Internal Medicine, 12, 639-642.
Johnstone, R. E., Valenzuela, R. C., & Sullivan, D. (1995). Managing pharmaceutical sales activities in an academic anesthesiology department. Journal of Clinical Anesthesia, 7, 544-548.
Lexchin, J. (1997). What information do physicians receive from pharmaceutical representatives?. Canadian Family Physician, 43, 941-945.
Lim, Tae-Seop, & Bowers, John Waite (1991). Facework: Solidarity, approbation, and tact. Human Communication Research, 17, 415-450.
Lurie, N., Rich, E. C., Simpson, D. E., Meyer, J., Schiedermayer, D. L., Goodman, J. L., & McKinney, W. P. (1990). Pharmaceutical representatives in academic medical centers: interaction with faculty and housestaff. Journal of General Internal Medicine, 5, 240-243.
Reeder, M., Dougherty, J, & White, L. J. (1993). Pharmaceutical representatives and emergency medicine residents: A national survey. Annals of Emergency Medicine, 22, 1593-1596.
Rhodewalt, Frederick T. (1986). Self-presentation and the phenomenal self: On the stability and malleability of self-conceptions. In R. F. Baumeister (Ed.), Public self and private self (pp. 117-142). New York: Springer.
Sandberg, W. S., Carlos, R., Sandberg, E. H., & Roizen, M. F. (1997). The effect of educational gifts from pharmaceutical firms on medical students' recall of company names or products. Academic Medicine, 72, 916-918.
Sergeant, M. D., Hodgetts, P. G., Godwin, M., Walker, D. M., & McHenry, P. (1996). Interactions with the pharmaceutical industry: a survey of family medicine residents in Ontario. Canadian Medical Association Journal, 155, 1243-1248.
Shaughnessy, A. F., Slawson, D. C., & Bennett, J. H. (1995). Teaching information mastery: Evaluating information provided by pharmaceutical representatives(see comments). Family Medicine, 27, 581-585.
Sherif, Carolyn, Sherif, Muzafer, & Nebergall, Roger (1965). Attitude and attitude change: The social judgment-involvement approach. Philadelphia: W. B. Saunders.
Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging frequency and probability. Cognitive Psychology, 5, 207-232.
Ziegler, M. G., Lew, P., & Singer, B. C. (1995). The accuracy of drug
information from pharmaceutical sales representatives. Journal of the
American Medical Association, 273, 1296-1298.
1. Ironically, large gifts (say, an expensive dinner out or box seats at the ball
game) are seen as more problematic than small gifts (say, a pen) (Reeder,
Dougherty & White 1993), yet cognitive dissonance theory suggests that small gifts
may have larger and less conscious effects. Furthermore, as discussed below, small
and "relevant" gifts (pens, calendars, note pads) are often used many times a day
and usually contain the product name on them, thus greatly increasing the cognitive
accessibility of the product.