Introduction

The most important professional responsibility of physicians is serving patients' best interest. However, this primary interest is sometimes undermined by secondary interest. One of the most notable concerns about conflict of interest has been the relationship between physicians and the pharmaceutical industry. In several countries, pharmaceutical representatives (PRs) give physicians a variety of gifts, information about products, and financial support for educational events, all with the aim of promoting products [1]–[6]. A gift-giving relationship between physicians and the pharmaceutical industry is ethically problematic for the following reasons: gifts cost money [7]; acceptance of gifts possibly erodes the faith of a patient in his or her doctor [8], [9]; gifts establish the obligation to respond possibly resulting in influence on physicians' decisions in patient care [10], [11]. Some studies have reported associations between exposure to information directly from the pharmaceutical industry and higher frequency, lower quality, or higher cost of prescribing, and others have not found such associations [12]. There is little systematic research on the impact of financial support for continuing education for physicians, although there are some reports of potential bias in favor of the pharmaceutical industry [13]–[15]

Remedies to control conflict of interest range from individual discretion to collective regulations including legal action [16], [17]. In the past decade, guidelines and recommendations were developed by an expert panel, professional societies, accrediting bodies, the pharmaceutical industry associations and the government in the United States [18]–[22]. In Japan, physicians working for public hospitals are prohibited from accepting kickbacks [23]. The Code of Pharmaceutical Research and Manufacturers of America in 2009 was translated and adopted by the Japan Pharmaceutical Manufacturer Association in 2011 [24]. However, at present, no professional societies in Japan have issued guidelines on physicians' relationships with pharmaceutical representatives in the context of clinical practice and medical education. After introduction of those guidelines and regulations in the United States, physician-industry relationships were nationally surveyed in 2009 and still high proportion (84%) of respondents had some type of interactions with PRs, although that was lower than that in 2004 (94%) [25]. In a questionnaire survey of 515 obstetricians and gynecologists in the United States, having read the guidelines was not associated with less reliance on PRs [26] These results indicate that the guidelines and regulations are partly effective but limited in reducing the number of physicians who engage in relationships with the pharmaceutical industry. In a survey of US family medicine residents, those from restricted programs reported fewer interactions with PRs, perceived less benefit from interactions with PRs, and rated gifts as less appropriate [27]. The previous national survey of Japanese physicians also showed an association between presence of local rules and less physician interaction with PRs [5]. It was also shown that attending medical schools with restricting policy was associated with reduced prescribing of newly marketed psychotropic medications [28]. These findings indicate that institutional rules and regulations may be more effective than general guidelines and recommendations in affecting physicians' attitude and behavior toward PRs.

The effectiveness of educational interventions aimed at helping individual physicians judiciously manage the conflict of interest has been inconclusive [29], [30].

Some qualitative studies investigated the reasons why physicians continue to interact with PRs regardless of the ethical implications. Physicians believe that the information from PRs is useful and that they can adequately evaluate the information without negative influence [31], [32]. Another reason for meeting with PRs has been reported to be as a part of their social contracts [31]. In one qualitative study, physicians were categorized into three types (avoiders, ambivalent engagers, and confident engagers) according to their attitude toward relationships with PRs [33]. In that study, physicians determine how to interact with PRs according to the perceived benefits and risks of the relationship, which were differently weighed from person to person. These findings suggest most physicians regularly interact with PRs and various factors influence such interactions. However, they do not provide enough information about if, when, how and why their attitudes and behaviors change over time.

To better understand how and why physicians interact with PRs over time, we conducted in-depth individual interviews with physicians focusing on chronological changes in their attitude and behavior concerning relationships with PRs and factors affecting such changes.

Methods

Results

A total of nine physicians (eight men and one woman) were interviewed. Five physicians were specialized in general medicine, two in family medicine, one in respiratory medicine, and one in infectious disease. Their years since graduation ranged from 10 to 31. Six (#1, 3, 4, 5, 7, 8) out of nine participants had graduated more than 20 years earlier. Six physicians (#1, 2, 4, 5, 7, 9) reported having interacted with PRs throughout their career. After the initial period of exposure, physicians #3 and #8 had a period in which they had no interaction with PRs but resumed the relationship after their workplaces changed. Physician #6 had contact with PRs in his medical school but never saw PRs after he began practice because of local regulations.

Three stages of chronological change in attitude and behavior were derived from a collective narrative of 9 physicians: stage 1, early interaction (pre-contemplation); stage 2, change in attitude (contemplation); stage 3, change in behavior (action) (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Chronological changes in Japanese physicians' attitude and behavior toward interaction with pharmaceutical representatives.

https://doi.org/10.1371/journal.pone.0106586.t001

Discussion

In the early years of their career, the interactions with PRs were passive. They were unaware of the implications of such interactions. As their career progressed, they started to think more actively about how they should interact with PRs and their attitudes toward PRs changed over time. The attitudinal changes were influenced by a variety of factors. Two thirds of our participants often had difficulty in making their behaviors consistent with their attitudes.

All the participants had interaction with PRs in the same manner as their superiors did in their early years and began to actively think about their interactions with PRs thereafter. This process is consistent with the findings of the qualitative report of how residents learn in the clinical workplace, in which residents were shown to act first and after that interpret the meaning of the activities [39]. This learning process can also be explained by the theory of situated leaning, in which learners enter the community of practice at the periphery and they move towards fuller participation by absorbing and being absorbed into the culture of practice [40]. As their career progresses, they started to have their own attitude toward relationships with PRs. In our study, their attitudes proved to be not constant but changing due to a variety of factors, including work environment, role models, views of patient and the public, acquisition of skills in information seeking and EBM, and learning about the concepts of professionalism and conflict of interest.

Transitions into teaching positions provoked ethical consideration about physician-PR relationships in some participants, which is consistent with the results of recent surveys. In a survey of residents and faculty members at a US medical school, faculty members was shown to be more concerned about direct contact between residents and PRs than residents themselves [41]. In the survey of physicians across the specialties, attending physicians had a more critical attitude toward receiving gifts from PRs than non-attending physicians [42]. We speculate that attending physicians may have a more critical attitude toward the interactions because they believe that residents and medical students are more vulnerable to influence from such interactions than themselves.

Some became hesitant to form relationships with PRs when they recognized views of patients and the public on the interaction. They were probably afraid of losing patients' and the public's trust. Patients' trust in physicians was shown to be affected by gifts from the industry to physicians in recent surveys [8], [9]. On the other hand, most physicians in the US agreed to the statement,” physicians should put the patient's welfare above the physician's financial interest [43].” Thus, announcing the results of the patient surveys to physicians is expected to promote their critical attitude toward the relationships. In 1993, to conform to the revised rules of Japan Fair Trade Commission, Japanese Pharmaceutical Manufacturers Association established the first version of Promotion Code for Prescription Drugs which restricted industry support to physicians unrelated to the products to 100,000 Japanese Yen (US $800.00) per institution [23]. The Code substantially changed the way the pharmaceutical representatives interact with physicians. It is likely that such changes in the society made some physicians feel that the views of the public toward physician-PR relationships became more critical.

Some participants indicated that fear of negative influence from gifts on their own prescribing behaviors affected their attitude toward relationships with PRs. The survey of psychiatry residents, intern and clerks in the University of Toronto showed that there was a positive correlation between the number of promotional items received and the belief that discussions with representatives have no impact on prescribing behavior [44]. In other words, enhancing perceived influence by interaction with PRs will possibly reduce the interaction. Also, evidence of the influence of promotion on physicians' behavior will be an important and effective component of teaching physician-PR relationships.

However, physicians' attitudinal changes affected by a variety of events did not necessarily lead to behavioral changes. Participants often legitimatized and rationalized their continuing interactions with PRs even when they understood that such interactions might have unfavorable effects on patient care. This is compatible with the results in recent qualitative studies, which showed that physicians used rationalizations to resolve their cognitive dissonance [31], [45]. The other factor contributing to the inconsistency between attitude and behavior is considered to be conformity to social or cultural norms [31], [32]. “Social norm” is defined as the implicit rules for acceptable behaviors, values, and beliefs in social psychology. We often get along with a group even if we don't believe they are right and behave as if we accept them, which is explained as public compliance without private acceptance under normative pressure [46]. According to the results of our study, they seemed to know their group's behavior was ethically problematic but they were pressured to conform to the group's norms. Inversely, using the tendency to conform to social norm can be effective in promoting beneficial behavior [46].

One of nine physicians (#6) worked in the hospital where he was not allowed to have any interaction with PRs. He was the only “avoider” among the participants. The likely reason why he was an “avoider” is early abstention and being kept in a work environment without physician-PR relationship. This case supports the notion that creating new social norms that disapprove physician-PR relationships may help individual physicians to avoid such relationships.

Our study has limitations. First, participants were hardly typical of Japanese physicians because they all attended the symposium of conflict of interest with strong interest, which may limit transferability of the findings. However, we believe that having some knowledge about the concept of conflict of interest helped them to reflect on their attitude and behavior. Second, social desirability bias and post hoc rationalization may have concealed some negative aspects of their attitude and behavior. Third, the accuracy of our participants' stories especially with regard to changes in their attitudes and behaviors may not correct, because their stories were constructed backwards based on their memories at the time of the interviews. It is possible that participants would have revealed different stories if we had repeated interviews over time. Fourth, coders' bias in the data analysis in qualitative studies cannot be totally avoided. We attempted to minimize such bias by having the whole process from coding to theorization visualized on Excel file and having it reviewed by three authors.

In conclusion, physicians' attitude toward relationships with PRs changed over time and a variety of factors affected such change. Paying attention to these factors and creating new social norms may be both necessary to produce change in behavior consistent with change in attitude.

Acknowledgments

We thank Takashi Otani for his advice on qualitative research methodology. We also thank Brian K. Purdue for proofreading the manuscript.