PurposeTo understand if greater team patient census is associated with differences in how teaching rounds are conducted, in terms of time with patients, content of rounds, persons involved and the nature of discussions.MethodsOver three months of a summer period, trained medical student research assistants (RAs) shadowed teaching rounds on our general medical inpatient services. Teams generally consist of an attending physician, a supervising resident, two first year residents (interns) and/or fourth-year medical students (acting interns), and two third year medical students. Teaching rounds occur most mornings, with the entire team going to the patient's room, presentation of the patient by the student or intern generally outside the patient's room, with the entire team then going into the patient's room for interaction with the patient. The RAs coded teaching rounds in 21/2 minute increments as to content (patient care, other teaching (i.e. mini-lectures, etc.) and administration), persons involved (attending, resident, interns, students, patient), and nature of discussions (open discussions with most members of the team, dyad interactions, lectures/presentations by a single individual, or dead time, such as walking). In addition, team patient census was collected and noted each day. Analysis was with Pearson's correlations.ResultsData were available for 96 teaching rounds. The mean patient census was 8.2, SD 2.7, range 2-14. Time on rounds was conducted in an open group fashion 44% of the time, 32% time in dyad fashion, 9% a solo individual, and 14% dead or not coded. Attendings were involved in the interactions 72% of the time, residents 41%, interns 37%, and students 23%. Greater patient census was associated with greater time with patients (r = .44, p < .0001), greater attending involvement (r = .34, p = .0008), greater resident involvement (r = .32, p = .002), but not intern or student involvement. Greater patient census was associated with rounds conducted in open formats (r = .35, p = .0004) and dyad format (r = .27, p = .007).ConclusionThe nature of teaching rounds changes with greater patient census. More time is spent with patients, but at the expense of attendings and residents dominating rounds, at the expense of the interns and students. Future studies should investigate how patient census influences teaching quality, both for students and for residents.

primary care (72% vs. 79%, p < .01), diabetes/endocrinology (71% vs. 82%, p < .01), eye (73% vs. 74%, p = .76), and podiatry/foot-care (71% vs.74%, p = .15). The homeless had marginally significantly lower odds of failing to have 4 diabetes-related Primary care visits (odds ratio [OR] 0.8 [95% confidence interval (CI) 0.6-1.0]). Drug abuse was associated with greater likelihood of failure to have 4 diabetes-related primary care visits for homeless (OR 2.0 [95%CI 1.2-3.7]), but not housed subjects (1.0 [0.7-1.5]) (interaction p = .04). Alcohol abuse and mental illness were associated with slightly lower likelihood of failure to have 4 diabetes-related primary care visits (0.

Introduction: Active engagement in research and scholarly output is challenging for pharmacists following residency training. Objectives: To determine the current structure of institutional support for pharmacy research, describe barriers to increasing staff research support and scholarly output, and benchmark best practices.