Interprofessional Extracorporeal Membrane Oxygenation Cardiopulmonary Resuscitation Simulations Aimed at Decreasing Actual Cannulation Times: A Longitudinal Study.

BACKGROUND: Since 2013, the cardiac intensive care unit (CICU) at Children's National has conducted annual extracorporeal membrane oxygenation cardiopulmonary resuscitation (ECPR) simulations that focus on team dynamics, room setup, and high-quality CPR. In 2019 and 2020, the simulations were expanded to include the surgical and extracorporeal membrane oxygenation (ECMO) teams in an effort to better understand and improve this process. METHODS: During a 4-week period in 2019, 7 peripheral ECPR simulations were conducted, and through a 3-week period in 2020, 7 central ECPR simulations were conducted. Participants in each session included: 8 to 10 CICU nurses, 1 CICU attending, 1 to 2 ICU or cardiology fellows, 1 cardiovascular surgery fellow or attending, and 1 ECMO specialist. For each session, the scenario continued until the simulated patient was on full cardiopulmonary bypass. An ECMO trainer was used for peripheral simulations and a 3-dimensionally-printed heart was used for central cannulations. An ECMO checklist was used to objectively determine when the patient and room were fully prepared for surgical intervention, and simulated cannulation times were recorded for both groups. A retrospective chart review was conducted to compare actual cannulation times before and after the intervention period, and video was used to review the events and assist in dividing them into medical versus surgical phases. Control charts were used to trend the total ECPR times before and after the intervention period, and mean and P values were calculated for both ECPR times and for all other categorical data. RESULTS: Mean peripheral ECPR times decreased significantly from 71.7 to 45.1 minutes ( P = 0.036) after the intervention period, and this was reflected by a centerline shift. Although we could not describe a similar decrease in central ECPR times because there were only 6 postintervention events, the times for each of these events were shorter than the historical mean of 37.8 minutes. There was a trend in improved survival, which did not meet significance both among patients undergoing peripheral ECPR (15.4% ± 10% to 43.8% ± 12.4%, P = 0.10) and central ECPR (36.4% ± 8.4% to 50% ± 25%, P = 0.60). The percentage of time dedicated to the medical phases of the actual versus simulated procedures was very consistent among both peripheral (33.0% vs. 31.9%) and central (39.6% vs. 39.8%) cannulations. CONCLUSIONS: We observed a significant decrease in peripheral cannulation times at our institution after conducting interprofessional ECPR simulations taken to the establishment of full cardiopulmonary bypass. The use of an ECMO trainer and a 3-dimensionally-printed heart allowed for both the medical and surgical phases of the procedure to be studied in detail, providing opportunities to streamline and improve this complex process. Larger multisite studies will be needed in the future to assess the effect of efforts like these on patient survival.

Pump in Parallel-Mechanical Assistance of Partial Cavopulmonary Circulation Using a Conventional Ventricular Assist Device.

Mechanical assistance of systemic single ventricle is effective in pulling blood through a cavopulmonary circuit. In patients with superior cavopulmonary connection, this strategy can lead to arterial desaturation secondary to increased inferior caval flow. We hypothesized that overall augmentation in cardiac output with mechanical assistance compensates for the drop in oxygen saturation thereby maintaining tissue oxygen delivery (DO2). Bidirectional Glenn (BDG) was established in seven swine (25 kg) after a common atrium had been established by balloon septostomy. Mechanical circulatory assistance of the single ventricle was achieved using an axial flow pump with ventricular inflow and aortic outflow. Cardiac output, mean pulmonary artery pressure (PAP), common atrial pressure (left atrial pressure [LAP]), arterial oxygen saturation (SaO2), partial pressure of arterial oxygen (PaO2), and DO2 were compared between assisted and nonassisted circulation. Significant augmentation of cardiac output was achieved with mechanical assistance in BDG circulation (BDG: median [interquartile range {IQR}], 0.8 [0.9-1.15] L/min versus assisted BDG: median [IQR], 1.5 [1.15-1.7] L/min; p = 0.05). Although oxygen saturations and PaO2 trended to be lower with assistance (SaO2; BDG: median [IQR], 43% [32-57%]; assisted BDG: median [IQR], 32% [24-35%]; p = 0.07) (PaO2; BDG: median [IQR], 24 [20-30] mm Hg; assisted BDG: median [IQR], 20 [17-21] mm Hg; p = 0.08), DO2 was unchanged with mechanical assistance (BDG: median [IQR], 94 [35-99] ml/min; assisted BDG: median [IQR], 79 [63-85] ml/min; p = 0.81). No significant change in the LAP or PAP was observed. In the setting of superior cavopulmonary connection/single ventricle, the systemic ventricular assistance with a ventricular assist device (VAD) leads to increase in cardiac output. Arterial oxygen saturations however may be lower with mechanical assistance, without any change in DO2.