OBJECTIVES: Increasing attention to patient safety in training hospitals may come at the expense of trainee autonomy and professional growth. This study sought to examine changes in medical trainees' self-reported behaviour after the institution-wide implementation of a rapid response system. METHODS: We conducted a two-point cross-sectional survey of medical trainees in 2006, during the implementation of a rapid response system, and in 2010, in a single academic medical centre. A novel instrument was used to measure trainee likelihood of calling for supervisory assistance, perception of autonomy, and comfort in managing decompensating patients. Non-parametric tests to assess for change were used and year of training was evaluated as an effect modifier. RESULTS: Response rates were 38% in 2006 and 70% in 2010. After 5 years of the full implementation of the rapid response system, residents were significantly more likely to report calling their attending physicians for assistance (rising from 40% to 65% of relevant situations; p < 0.0001). Year of training was a significant effect modifier. Interns felt significantly more comfortable in managing acutely ill patients; juniors and seniors felt significantly less concerned about their autonomy at 5 years after the implementation of the rapid response system. These changes were mirrored in the actual use of the rapid response system, which increased by 41% during the 5-year period after adjustment for patient volume (p < 0.0001). CONCLUSIONS: A primary team-focused implementation of a rapid response system was associated with durable changes in resident physicians' reported behaviour, including increased comfort with involving more experienced physicians and managing unstable patients.
OBJECTIVES: Increasing attention to patient safety in training hospitals may come at the expense of trainee autonomy and professional growth. This study sought to examine changes in medical trainees' self-reported behaviour after the institution-wide implementation of a rapid response system. METHODS: We conducted a two-point cross-sectional survey of medical trainees in 2006, during the implementation of a rapid response system, and in 2010, in a single academic medical centre. A novel instrument was used to measure trainee likelihood of calling for supervisory assistance, perception of autonomy, and comfort in managing decompensating patients. Non-parametric tests to assess for change were used and year of training was evaluated as an effect modifier. RESULTS: Response rates were 38% in 2006 and 70% in 2010. After 5 years of the full implementation of the rapid response system, residents were significantly more likely to report calling their attending physicians for assistance (rising from 40% to 65% of relevant situations; p < 0.0001). Year of training was a significant effect modifier. Interns felt significantly more comfortable in managing acutely ill patients; juniors and seniors felt significantly less concerned about their autonomy at 5 years after the implementation of the rapid response system. These changes were mirrored in the actual use of the rapid response system, which increased by 41% during the 5-year period after adjustment for patient volume (p < 0.0001). CONCLUSIONS: A primary team-focused implementation of a rapid response system was associated with durable changes in resident physicians' reported behaviour, including increased comfort with involving more experienced physicians and managing unstable patients.