To the Editor: The potential impact of pharmaceutical industry promotional efforts on physician behavior has received growing attention both in the medical community and among the general public (1, 2). Recent research suggests that interactions between physicians and pharmaceutical companies affect physician requests to add medications to hospital formularies as well as prescribing behavior (3, 4).
One specific concern is the influence pharmaceutical company spending may have on resident physicians (5). Meetings between residents and pharmaceutical industry representatives are frequent and commonly take place in the form of company-sponsored conference lunches: over 88% of internal medical residency program directors report that their programs have pharmaceutical company sponsorship of educational conferences (6). These trends have compelled academic hospitals like Massachusetts General Hospital to limit access of pharmaceutical representatives to their physicians (7).
Because of these growing concerns in residency training, we looked at the association between pharmaceutical company sponsorship of conferences and patterns of antidepressant prescriptions at the adult outpatient clinic of our psychiatry residency program.
We first identified 25 medications that are used as antidepressants in the United States and the corresponding pharmaceutical company that manufactures each one. Patterns of use of these 25 antidepressants were then obtained from a database collected by the UCSF Coping with Depression program for clinical research purposes for the 2001—02 and 2002—03 academic years. The information included in the database was collected on clinical tracking forms that monitor all initiation, discontinuation, adjustment, or continuation of medications among adult outpatients. Data collection for each patient started with his or her initial visit and ended with the termination of treatment in the clinic.
This clinical information was then compared with records of pharmaceutical company sponsorship of conferences for residents working in the outpatient clinic of our department. During the 2001—02 and 2002—03 academic years, pharmaceutical companies sponsored three resident conferences per week, and sponsorship information was kept in administrative records. Using these records, we tabulated how often each company supported a resident conference over the course of each academic year. Because different sets of residents work in the outpatient clinic each year, the data from the different academic years were analyzed separately.
For the 2001—02 academic year, we found a strong positive correlation between pharmaceutical company sponsorship of resident conferences and the use of the corresponding antidepressant by patients in the adult outpatient clinic. The Pearson correlation coefficient was 0.87 (p<0.01). Since companies manufacturing generic antidepressants generally did not sponsor conferences, we conducted a separate analysis of only brand-name antidepressants and found a similarly strong positive correlation (0.83, p<0.01).
We also compared antidepressant use for patients who had never taken antidepressants before with the number of resident conferences sponsored by each pharmaceutical company and found a correlation of 0.74 (p<0.001). Medication decisions made for these patients should be less colored by previous medication trials or side effects and therefore should more clearly reflect the influence of pharmaceutical company sponsorship.
We found more modest but similar positive correlations using information gathered for the 2002—03 academic year. The correlation between the number of patients using each antidepressant and the percentage of company-sponsored conferences had a Pearson coefficient of 0.73 (p<0.01); analysis of only brand-name antidepressants had a Pearson coefficient of 0.60 (p<0.05). Patients with no prior antidepressant use had a Pearson coefficient of 0.66 (p<0.001).
These correlations are striking and raise serious concerns regarding how much influence pharmaceutical company presence at educational conferences has on resident physician decision-making. These correlations, however, do not capture the complex set of factors that affect how residents make prescribing decisions. Other potential influences on prescribing behavior include patient preferences, patient health insurance coverage of psychiatric medications, faculty supervisor preference of antidepressant, and differential efficacy and side effect profiles. Though controlling for these other factors would help us gauge more precisely the amount of influence pharmaceutical companies have on prescribing, limitations of the information available did not allow us to account for these variables in our analyses.
With this letter, we would like to bring attention to the importance of psychiatry residency training programs in establishing clearly defined boundaries in their relationships with pharmaceutical companies. Resident physicians are developing their notions of what are appropriate interactions with the drug industry. Residency training programs are therefore in an ideal position to model appropriate relationships with pharmaceutical companies and to ensure that these future clinicians make reasoned, evidence-based clinical decisions.
Some training programs, like our own, have addressed concerns about potential pharmaceutical company influence (or the appearance of influence) by establishing guidelines for pharmaceutical company representative interactions with residents (8, 9). In our department, company representatives are allowed to speak 5 minutes before the conferences and then are asked to leave. They are not allowed to distribute commercial materials within the conference area or to visit clinical areas without an appointment.
We recognize that the relationship between pharmaceutical companies and medicine is complex. Psychotropic medications have had a profound effect on the care of our patients, and drug companies have long supported many worthwhile research and scientific events at academic institutions. It is the responsibility of those of us involved in residency training to teach early career physicians to use these medications for the benefit of their patients while ensuring that their clinical decisions are impartial and sound.