Abstract
IMPORTANCE: In large academic centers, medical residents work on multiple clinical floors with transient interactions with nursing colleagues. Although teamwork is critical in delivering high-quality medical care, little research has evaluated the effect of interprofessional familiarity on inpatient team performance. OBJECTIVE: To determine the effectiveness of increased familiarity between medical residents and nurses on team performance, psychological safety, and communication. DESIGN, SETTING, AND PARTICIPANTS: A 12-month randomized clinical trial in an inpatient general medical service at a large academic medical center was completed from June 25, 2019, to June 24, 2020. Participants included 33 postgraduate year (PGY)-1 residents in an internal medicine residency program and 91 general medicine nurses. INTERVENTIONS: Fifteen PGY-1 residents were randomized to complete all 16 weeks of their general medicine inpatient time on 1 medical nursing floor (intervention group with 43 nurses). Eighteen PGY-1 residents completed 16 weeks on 4 different general medical floors as per usual care (control group with 48 nurses). MAIN OUTCOMES AND MEASURES: The primary outcome was an assessment of team performance in physician-nurse simulation scenarios completed at 6 and 12 months. Interprofessional communication was assessed via a time-motion study of both work rounds and individual resident clinical work. Psychological safety and teamwork culture were assessed via surveys of both residents and nurses at multiple time points. RESULTS: Of the intervention and control PGY-1 residents, 8 of 15 (54%) and 8 of 18 (44%) were women, respectively. Of the nurses in the intervention and control groups with information available, 37 of 40 (93%) and 34 of 38 (90%) were women, respectively, and more than 70% had less than 10 years of clinical experience. There was no difference in overall team performance during the first simulation. At the 12-month simulation, the intervention teams received a higher mean overall score in leadership and management (mean [SD], 2.47 [0.53] vs 2.17 [0.39]; P = .045, Cohen d = 0.65) and on individually rated items were more likely to work as 1 unit (100% vs 62%; P = .003), negotiate with the patient (61% vs 10%; P = .001), support other team members (61% vs 24%; P = .02), and communicate as a team (56% vs 19%; P = .02). The intervention teams were more successful in achieving the correct simulation case outcome of negotiating a specific insulin dose with the patient (67% vs 14%; P = .001). Time-motion analysis noted intervention teams were more likely to have a nurse present on work rounds (47% vs 28%; P = .03). At 6 months, nurses in the intervention group were more likely to report their relationship with PGY-1 residents to be excellent to outstanding (74% vs 40%; P = .003), feel that the input of all clinical practitioners was valued (95% vs 53%; P < .001), and say that feedback between practitioners was delivered in a way to promote positive interactions (90% vs 60%; P = .003). These differences diminished at the 12-month survey. CONCLUSIONS AND RELEVANCE: In this randomized clinical trial, increased familiarity between nurses and residents promoted more rapid improvement of nursing perception of team relationships and, over time, led to higher team performance on complex cognitive tasks in medical simulations. Medical centers should consider team familiarity as a potential metric to improve physician-nursing teamwork and patient care. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT05213117.