Does simulation improve clinical performance in management of postpartum hemorrhage?

Background Although simulation is now widely used to improve teamwork and communication, data demonstrating improvement in clinical outcomes are limited. Objective This study aimed to examine the clinical performance and outcomes associated with postpartum hemorrhage because of uterine atony following the implementation of a multidisciplinary simulation program. Study Design This was a prospective observational study of response to postpartum hemorrhage because of uterine atony in an academic medical center before (epoch 1: July 2017–June 2018) and after (epoch 2: July 2019–June 2020) implementing a multidisciplinary simulation program. A total of 22 postpartum hemorrhage simulations were performed from July 2018 to June 2019 involving more than 300 nursing, obstetrical, and anesthesia providers. The simulation program focused on managing postpartum hemorrhage events and improving teamwork and communication of the multidisciplinary teams. To evaluate the clinical effectiveness of the simulation program, the primary outcome was response to postpartum hemorrhage defined as the time from the administration of uterotonic medications to transfusion of the first unit of blood in the first 12 hours following delivery, comparing epoch 2 to epoch 1 following the implementation of a simulation program. Statistical analysis included the use of the Pearson chi-square test, Wilcoxon rank-sum test, Hodges-Lehmann statistic for differences, and bootstrap methods with a P value of Results Between July 1, 2017, and June 30, 2018, there were 12,305 patients who delivered, of which 495 patients (4%) required transfusion. Between July 1, 2019, and June 30, 2020, there were 12,414 patients who delivered, of which 480 patients (4%) required transfusion. When isolating cases of postpartum hemorrhage because of uterine atony in both transfused groups, there were 157 women in the presimulation group (epoch 1) and 165 women in the postsimulation group (epoch 2), respectively. There was no difference in age, race, parity, or perinatal outcomes between the 2 epochs. Women in epoch 2 began receiving blood products significantly earlier in the first 12 hours following delivery compared with women in epoch 1 (51 [range, 28–125] minutes vs 102 [range, 32–320] minutes; P=.005). In addition, there was a significantly decreased variation in the time from the administration of uterotonic medications to transfusion of blood in epoch 2 (P=.035). Furthermore, women in epoch 2 had significantly lower estimated blood loss than women in epoch 1 (1250 [range, 1000–1750] mL vs 1500 [range, 1000–2000] mL; P=.032). Conclusion The implementation of a multidisciplinary simulation program at a large academic center focusing on the management of postpartum hemorrhage was associated with an improved clinical response. Specifically, there were significantly faster times from the administration of uterotonic medications to transfusion of blood, decreased variance in the time from the administration of uterotonic medications to transfusion of blood, and lower estimated blood loss following the implementation of a simulation program. Because delay in treatment is a major cause of preventable maternal death in obstetrical hemorrhage, the results in our study provided clinical evidence that a simulation program may improve patient outcomes in such emergencies.