Outcome Comparison in Children Undergoing Extracorporeal Life Support Initiated at a Local Hospital by a Mobile Cardiorespiratory Assistance Unit or at a Referral Center

Purpose: To compare characteristics and outcome in children undergoing extracorporeal life support initiated in an extracorporeal life support center or at the patient's bedside in a local hospital, by means of a mobile cardiorespiratory assistance unit. Methods: A retrospective study in a single PICU during 6 years. Extracorporeal life support was started either in our center (control group) or in the local hospital (mobile cardiorespiratory assistance unit group). The data collected were demographics, markers of patient's preextracorporeal life support condition, and outcome. Results: One hundred twenty-six children underwent extracorporeal life support, 105 in the control group and 21 in the mobile cardiorespiratory assistance unit group. There was no difference between groups in terms of age, weight, or Pediatric Risk of Mortality II score. There was a significant difference in organ failure etiology between groups, with more respiratory cases in the mobile cardiorespiratory assistance unit group (76.2%) and more cardiac surgery cases in the control group (60%; p < 0.001). The duration of extracorporeal life support was longer in the mobile cardiorespiratory assistance unit group than in the control group (10 [1-36] vs 5 [0-33] d; p = 0.003). PICU length of stay and mortality (60% vs 47.6%; p = 0.294) were not significantly different between the two groups. To allow comparison of a more homogenous population, a subgroup analysis was performed including only respiratory failure patients from the two groups (R-control group [n = 22] and R-mobile cardiorespiratory assistance unit group [n = 16]). PICU length of stay was 17 (3-64) days in the R-control group and 23 (1-45) days in the R-mobile cardiorespiratory assistance unit group (p = 0.564), and PICU mortality rate was 54.5% in the R-control group and 43.8% in the R-mobile cardiorespiratory assistance unit group (p = 0.511). There was no difference between the R-groups for age, weight, Pediatric Risk of Mortality II score, and markers of kidney or liver dysfunction, and lactate blood levels. Conclusion: Extracorporeal life support can be safely initiated at children's bedside in the local hospital and then transported to the specialized referral center. Our results support the validity of an interregional organization of mobile cardiorespiratory assistance unit teams.