To report our institutional experience of venoarterial extracorporeal membrane oxygenation (ECMO) in children with septic shock and circulatory collapse.
Retrospective case series.
Intensive care unit of a tertiary pediatric referral center.
Forty-five children with refractory septic shock who received venoarterial ECMO for hemodynamic support.
We measured mean arterial pressure and inotropes before cannulation, ventilator settings, oxygenation, site and cause of infection, time on ECMO, complications of ECMO relating to the circuit or anticoagulation, survival to hospital discharge, and functional outcome assessment. Between July 1988 and October 2006, 441 children at our institution received extracorporeal life support for a variety of indications. Forty-five (10%) with septic shock received venoarterial ECMO specifically for hemodynamic support. Eighteen (40%) of these had suffered cardiac arrest and were receiving chest compressions immediately before cannulation. The median time spent on ECMO was 84 hrs (range, 32–135). There were mechanical problems with the ECMO circuit requiring intervention in 17 (38%) patients, such as oxygenator or pump head failure, clots in the circuit, or cannulae malposition. This caused no long-term harm in any but one of the patients, who died during a circuit change. Eleven patients (24%) had clinically apparent episodes of bleeding that required surgical intervention or blood transfusion. Twenty-one (47%) patients survived to hospital discharge. Atrioaortic cannulation through a sternotomy incision was associated with an improvement in survival to hospital discharge (73% of those with central cannulation survived vs. 44% without, p = .05). No survivors had severe disability at long-term follow-up.
Extracorporeal membrane oxygenation can be safely used to resuscitate and support children with sepsis and refractory shock. Sepsis and multiorgan failure should not be considered a contraindication to ECMO. This study adds support to existing guidelines.
From the Intensive Care Unit (GM, WB, DB, AT) and Clinical Epidemiology and Biostatistics Unit (SD), The Royal Children's Hospital, Melbourne; and the Department of Paediatrics (WB), The University of Melbourne, Australia.
*See also p. 498.
Supported, in part, by a National Health and Medical Research Council postgraduate medical scholarship (No. 359273) (GM).
The authors have not disclosed any potential conflicts of interest.