Background: Although pharmaceutical sales representatives provide physicians with information on new products, these encounters have rarely been studied in practice settings. We examined these interactions among practicing internists and assessed whether prior residency policies limiting pharmaceutical sales representative access affected the subsequent behavior of practitioners.

Methods: We conducted a mail survey of the internal medicine staffs of a medical school hospital and two affiliated community hospitals. A second request was sent to nonresponders. After the second mailing, a random sample of nonresponders was compared with a similar sample of respondents. Multivariate odds ratios (OR) and 95% confidence intervals (CI) were estimated with logistic regression.

Results: Of the 346 (40%) internists who responded, 22% were women and 60% were trained in university hospitals. There were no differences in gender, subspecialization, or type of training when survey responders and nonresponders were compared. Two hundred eighty-seven (83%) physicians had met with pharmaceutical sales representatives within the previous year, of whom 248 (86%) had received drug samples. Having had a policy that limited access to pharmaceutical sales representatives during residency did not affect the subsequent likelihood of seeing these representatives (P = 0.20) or accepting samples in practice (P = 0.99). Those describing themselves as busy practitioners were significantly less likely to abstain from meeting pharmaceutical sales representatives (OR = 0.2, 95% CI: 0.1 to 0.6, P <0.001). Those with very frequent contacts (>10 times/month) were virtually all busy practitioners.

Conclusions: Encounters between physicians and pharmaceutical sales representatives are common in internal medicine practice, especially in busy offices. Policies designed to limit pharmaceutical sales representative access during residency do not appear to affect the subsequent likelihood of meeting with pharmaceutical sales representatives or accepting samples.