For almost half a century, controversy has surrounded the presence and role of the pharmaceutical industry and its representatives in academic medical settings (1–7). Assertions of conflicts of interest created by industry largesse have become increasingly strident. At the heart of the conflict is the concern that pharmaceutical industry representatives bestow gifts to increase physicians’ use of new and expensive “on patent” drugs. Physicians’ choices to promote or prescribe these drugs may not be in the best interest of their patients if the new drugs have marginal or no benefit and cost more than older drugs.

There is mounting evidence that physician interactions with representatives of the pharmaceutical industry are frequent and often accompanied by gifts, yet the medical information provided may be inaccurate or commercially biased (2–18). Surveys of residents, including psychiatric trainees, conducted before 2002 (2, 19–21) reported that most opposed restrictions on interactions with pharmaceutical representatives and believed that accepting promotional gifts did not influence their prescribing behavior. In rebuttal, Schwartz et al. (8) reported that psychiatric residents preferentially prescribe certain medications in newly admitted clinic patients soon after pharmaceutical company sales visits.

Several professional organizations have recommended limitations on gifts, drug samples, and educational programs funded by industry (22, 23). The American Association of Directors of Psychiatric Residency Training (AADPRT) taskforce on the relationship between the pharmaceutical industry and psychiatry recommended that psychiatry departments develop clear policies, but provided no specific guidelines around interactions (24). More recently, Brennan et al. (25) proposed that academic medical centers take the lead in the “elimination” of physician-industry conflicts of interest and recommended banning all gifts regardless of value and encouraging industry contribution to a central repository at the academic institution, which could then disperse funds to programs for various activities. Some academic institutions have restricted gift-giving or pharmaceutical representative presence, and a few have banned these activities altogether (26–29). However, 62% of academic medical center departmental chairs receive compensation from industry (30). The degree to which the debate about the presence and influence of the pharmaceutical industry in academic psychiatry has influenced current opinions of psychiatrists is unknown; the most recent survey of psychiatric residents we were able to find was completed 7 years ago (21).

Over the last 4 years, psychiatric residents at Oregon Health and Science University (OHSU) have participated in educational sessions to explore the potential effect of interactions with the pharmaceutical industry on prescribing. Some residents and faculty across departments at OHSU have voiced dissatisfaction about the lack of a formal, campus-wide policy limiting interactions with the pharmaceutical industry. In 2006, a group of residents presented a plan to the Dean of the OHSU School of Medicine that would effectively ban the presence of the pharmaceutical industry. The Dean then convened an ad hoc task force to study the relationship between industry and the school, its departments, faculty, residents, and students. The task force developed an anonymous online survey to measure attitudes toward and relationships with the pharmaceutical industry and its representatives. This article reports on OHSU psychiatric residents, fellows, and faculty attitudes about and involvement with the pharmaceutical industry as determined from the survey.

The survey included 46 items. Categories of postgraduate year (PGY) or faculty rank were collapsed to a resident or faculty category to protect anonymity. In the first part of the survey, 10 attitudinal questions developed by McKinney et al. (31) addressed the degree to which respondents perceived positive and negative influences of industry, their preparedness to manage relationships with industry, and the value of industry presence in the academic setting. McKinney et al. (31) ensured construct validity by having individuals with expertise in questionnaire design, education, and pharmaceutical detailing review the questionnaire. Items were rated on a 5-point scale from strongly agree to strongly disagree. For analyses, these categories were collapsed to agree, disagree, and neutral/uncertain/missing. In the second part of the survey, behavioral questions measured, for the previous year, the number of pharmaceutical company conferences attended and gifts received. In the third part of the survey, participants were asked open-ended questions about the types of support that provided the greatest benefit for educational activities, whether or not relationships with industry compromised objectivity, and the adverse effects of industry withdrawal in educational settings. The survey was sent to all 49 faculty psychiatrists with OHSU e-mail accounts and 40 psychiatric residents and fellows, and it was re-sent twice to nonresponders over several months in 2007. The OHSU Institutional Review Board approved analysis of the data for publication.

The percentage of faculty and trainees endorsing a particular view (neutral/uncertain, disagree, agree) was compared using Fisher’s exact test. Secondary analyses, using an exact unconditional test (32), compared faculty and trainees with no opinion to those with a definite stance (combined disagree/agree) and then compared those who agreed to those who disagreed. McNemar’s test was used to compare correlated proportions. The number of gifts accepted was summarized by median and interquartile range (IQR); differences between groups were assessed using a randomization test (33). Analyses were performed using R version 2.6.1 (34); version 1.3 of the EXTSIG program was used to carry out the exact unconditional test of two proportions (35). The constant comparative method was used to categorize open-ended comments across broad categories of themes (36).

Of 89 surveys sent, 58 (65%) were fully or partially completed. Of 40 faculty members who indicated academic rank, 21 were assistant professors, six were associate professors, and 13 were professors. Of the 17 trainees, 12 were residents and five were fellows. The response proportion was 91% for assistant professors (n=23), 55% for associate professors (n=11), 87% for professors (n=13) (84% for all faculty, n=49), 13% for PGY-1 residents (n=8), and 50% for PGY-2–PGY-6 residents (n=32) (43% for all residents, n=40). Seventy percent of men (n=41) and 52% of women (n=15) responded.

Views on the Pharmaceutical Industry

More than two-thirds of respondents (n=39) disagreed that pharmaceutical representatives have an important teaching role in the medical school (Table 1). Only 24% of faculty (n=10) and 18% of residents (n=3) believed that pharmaceutical representatives provide useful and accurate information on new drugs; 20% of faculty (n=8) and 12% of residents (n=2) agreed that representatives provided useful and accurate information on established drugs. Forty-one percent of faculty (n=17) and 53% of trainees (n=9) agreed that pharmaceutical representatives should be restricted from making presentations on campus. Respondents were divided on whether they had sufficient training about how to interact with pharmaceutical representatives; however, most respondents endorsed that pharmaceutical companies supported important conferences and speakers, with only three respondents, all faculty, disagreeing with this statement.

Fifty-four percent of faculty (n=22) and 12% of residents (n=2) agreed that they would maintain the same degree of contact if no gifts or food were distributed. The distribution of responses differed between faculty and residents on this measure (p=0.01, Fisher’s exact test). The percentage of trainees who were neutral/uncertain (41%, n=7) was not significantly different (p=0.063, exact unconditional test) from the percentage of faculty who were neutral/uncertain (17%, n=7). Among the 44 respondents with an opinion, the percentage of faculty who agreed with the statement was between 9 and 74 percentage points higher (95% confidence interval) than the percentage of trainees who agreed (65% faculty, n=22, versus 20% trainees, n=2; p=0.01, exact unconditional test). This suggests that gifts and food are more important bonds between residents and pharmaceutical representatives than between faculty and pharmaceutical representatives.

Faculty and trainees differed in their response to the statement “Discussion with pharmaceutical representatives does not have an impact on my prescribing behavior” (p=0.046, Fisher’s exact test). Follow-up tests showed that faculty and trainees were comparable with respect to having some opinion (66% faculty, n=27, versus 88% trainees, n=15; p=0.084, exact unconditional test). Among the 27 faculty and 15 trainees with a definite opinion, the percentage of faculty who agreed (56%, n=15) was marginally different from the percentage of trainees who agreed (27%, n=4, p=0.082, exact unconditional test). As such, a trend is suggested for faculty to feel less influenced by discussions with drug representatives than residents.

Fifty-three percent of residents (n=9) and 24% of faculty (n=10) disagreed with the statement “Accepting gifts or food from pharmaceutical representatives has no impact on my prescribing behavior,” and 65% of residents (n=11) and 59% of faculty (n=24) disagreed with the statement “Accepting food or gifts from pharmaceutical representatives has no impact on other OHSU physicians prescribing choices.” Among the 33 respondents with a definite opinion about these two statements (Table 1), six respondents, all faculty members, indicated that accepting gifts affected others’ behavior, but not their own. Eight respondents, including seven faculty members and one resident, indicated that gifts affected neither their behaviors nor others'. Nineteen respondents indicated that they did not agree with either proposition, including nine residents and 10 faculty members. Among the 33 respondents with a definite opinion, a significantly higher proportion of respondents believed that gifts did not affect their own prescribing behavior (42%, n=14), while only 24% (n=8) did not believe that the gifts affected other physicians’ behavior (p=0.03, McNemar’s exact test). Of note, the significance of this finding entirely depended on faculty responses, because no trainee responses between these two questions were discordant.

Acceptance of Gifts

Gifts accepted from pharmaceutical representatives fell into five main categories: small items such as pens with a company logo, books and educational materials, gift certificates, medical equipment, and drug samples (Table 2). Nine faculty and two trainees reported accepting no gifts of any kind, while 32 faculty and 15 trainees accepted at least one gift. The most commonly accepted gifts for both groups were pens, books, and drug samples. Among those who had accepted gifts, there were no differences between faculty and trainees, except that trainees were more likely than faculty to have accepted educational material (p=0.01). However, the median number of books/educational materials accepted by each group did not differ (p=1.0, randomization test).

More than 90% of respondents (n=52) attended at least one industry-sponsored event on campus in the previous year (Table 3). In contrast, 47% of residents (n=8) and 49% of faculty (n=20) had not attended any off-campus events in the previous year. There were no differences between faculty and trainees in median number of conferences attended on or off campus.

Open-Ended Questions

Of those who completed any portion of the survey, the response rate for each of the open-ended questions varied between 82% (n=14) and 88% (n=15) for trainees and 59% (n=24) and 78% (n=32) for faculty. When asked about the types of industry support that provided the greatest benefit for and least amount of threat to educational activities, residents endorsed books (27% greatest benefit, n=4/15, 36% least threat, n=5/14) while faculty endorsed unrestricted educational grants and educational programs (46% greatest benefit, n=13/28, 38% least threat, n=9/24). Half of residents (n=7/14) viewed “detailing” (sales visits by pharmaceutical representatives) as the greatest threat to objectivity in educational activities while one-third of faculty (n=10/29) viewed biased speakers as the greatest threat. When asked which educational activities would be adversely affected by partial or complete termination of industry support, one-third of residents (n=5/15) expressed concern about resident activities, such as the annual resident retreat. Thirty-eight percent of faculty (n=12/32) indicated that national and local continuing medical education (CME) would be negatively affected by curtailing industry support. Twenty-nine percent of trainees and 15% of faculty (n=5/32) stated that there would be no adverse effects to either restricting or removing industry support.

Studies of psychiatric professionals’ attitudes toward and interactions with pharmaceutical representatives and industry have primarily focused on residents; we were unable to locate studies of psychiatry faculty. In 1995, Hodges et al. (20) examined psychiatric trainee experiences and attitudes at a large Canadian medical institution. Randall et al. (21) subsequently surveyed attitudes and interactions of U.S. psychiatric residents at an East Coast academic campus. Both studies used the same attitudinal questionnaire as in this study but revealed four interesting differences. First, the previous studies reported that between 56% and 65% of psychiatric residents, compared to 29% of OHSU trainees, agreed that accepting promotional items or gifts did not influence their own prescribing behavior (20, 21). Second, only one resident in our survey, compared to 29% of residents surveyed by Hodges et al. (20), believed that pharmaceutical representatives had an important teaching role. Third, more than half of psychiatric residents at OHSU, compared to 9%–10% in the previous studies (20, 21), believed that pharmaceutical representatives should be restricted from making presentations at the medical center, and many respondents’ open-ended responses further explained that detailing was a substantial threat to objectivity in educational activities. Finally, OHSU psychiatric residents (and faculty) appeared to attend fewer industry-sponsored events than the median of 10 drug lunches and one personal meeting per year for residents that were reported by Hodges et al. (20). These negative views may represent cultural differences in Oregon, effects of educational efforts to increase awareness of conflicts of interest at OHSU, strong leadership from a group of trainees on campus who worked to increase institutional separation from industry, and/or secular trends reflecting widespread criticism of academic-industry relationships. In conflict with these negative opinions, however, both faculty and residents supported pharmaceutical industry financial support of important educational events.

There were few differences of opinion between faculty and residents; however, the differences that did exist suggest that trainees perceive themselves as potentially more vulnerable to the influence of pharmaceutical representatives than do faculty members. Residents were more likely to indicate that food and gifts promoted interactions, and that discussions with drug representatives were somewhat more likely to influence their prescribing practices. This greater sense of vulnerability may, in part, explain why the impetus for change at OHSU came from the residency group directly to the Dean and bypassed departmental chairs and faculty.

Educators may acknowledge the risks of interactions with pharmaceutical representatives and industry; however, departments and graduate medical education are faced with the reality of growing financial challenges that are eased by pharmaceutical industry support. Pharmaceutical sponsorship of conferences is inversely correlated to the availability of departmental or hospital funds (37). Program directors believe that attendance at educational conferences would decrease if food were not provided (37). These concerns were pervasive in our study and conflicted with respondents’ negative views of the pharmaceutical industry.

There are important limitations to our study. Although our response rate was high for faculty and PGY 2–6 psychiatric residents, the overall number of participants was small. The results are from one academic medical center and may not represent participants of other institutions. Because there is increased scrutiny and criticism of the role of pharmaceutical industry in medical education, our results may represent a social bias, with respondents underreporting behaviors and not providing genuine responses to attitudinal questions despite anonymity.

In summary, unlike past resident cohorts, residents in recent years have trained during a time when major medical organizations have issued cautionary guidelines and position statements, and campuses have received media attention for banning pharmaceutical industry presence. Cautious comparisons to previous psychiatric resident surveys suggest the possibility that the recent OHSU cohort of residents is more aware of the risks of pharmaceutical industry interactions, is less trusting of information from pharmaceutical representatives, and supports increased restrictions on industry presence in academic medical centers. Despite their concerns, our residents and faculty are aware of the limited discretionary monies in academic centers and would not support complete discontinuation of all industry funds. These findings should be investigated nationally among psychiatric faculty and residents and the results considered in future revisions of AADPRT’s policy.

The authors appreciate the statistical consultation of M.R. Lasarev, M.S. At the time of submission, Drs. Misra and Ganzini reported no competing interests. Dr. Keepers is a site investigator for the Cyberonics A21 UNS clinical trial.