Evaluating when to transport a child for extracorporeal membrane oxygenation

Randomized clinical trials in perinatal lung disease have demonstrated the potentially life-saving benefits of extracorporeal membrane oxygenation (ECMO) support (1), but effective ECMO requires expert multidisciplinary care provided in specialty centers. Although relatively few critically ill children progress to require ECMO (2, 3), they must often be transferred to a regional center to receive or be evaluated for support (4). This process carries demonstrable risks, including misjudging the extent of critical illness in patients off-site, transporting them conventionally when they may deteriorate en route, or at times cannulating them simply in order to facilitate safe transfer.

In this issue of Pediatric Critical Care Medicine, Annicq et al. characterized the outcomes of children who were referred to a regional center for evaluation of ECMO support (5). The study focused on children referred for ECMO who were transported without the support of ECMO. In this population the study aimed to identify clinical factors associated with eventual receipt of ECMO. ECMO support was associated with worse oxygenation despite mechanical ventilation, higher levels of vasoactive medication support and a diagnosis of congenital diaphragmatic hernia. The authors highlighted that of 203 children transferred to a regional ECMO center, 60% were supported without ECMO, 9% died before ECMO support could be provided, and 32% received ECMO.

A substantial proportion of patients transported for ECMO evaluation are managed without ECMO (6). Since inter-hospital transport of critically ill children has inherent risks (7), and may add financial and social strain to their families, is it then wrong to transfer patients earlier in their course given that some will not require ECMO?

Delays in transfer may result in continued deterioration that results in death (5), and the efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR) trial demonstrated a survival benefit of referral to an ECMO center for consideration of treatment with ECMO (8). Thirty percent of patients in that trial did not receive ECMO but still had improved survival, arguably because they received care in an experienced referral center with a full armamentarium of therapeutic tools (9).

Although it was not the focus of the study by Annicq et al (5), the role of initiating ECMO prior to transport also requires comment. A key decision point in accepting a patient for ECMO evaluation is deciding if the patient would be more safely transported on ECMO. Initiating ECMO support at an external referral hospital requires the swift activation of a transport team that can ensure the safe cannulation of the child in a potentially unfamiliar environment then safely maintaining the child on ECMO throughout transport. The mobile ECMO team must include a skilled practitioner to place the cannulas and ECMO specialists to support the child and maintain the circuit. Despite the complexity, several teams have repeatedly reported their ability to safely initiate and transport patients on ECMO (10, 11). A mobile ECMO team may reduce the deaths that occurred after the child was accepted by the transporting team but before the child can be transported and initiated on ECMO support at the receiving hospital.

Nevertheless, some patients may die not because of inability to initiate ECMO at the outside hospital but due to delays in mobilizing a team and accepting the patient for transfer. The authors (5) compared time from referral to acceptance among children who died prior to transport and those who were successfully transported. The median time from initial referral to acceptance was approximately 30 minutes longer among the 9 children who were too unstable to transport and died, compared to the patients who were successfully transported. Based on this information, the authors suggested that efforts should be made to shorten time from referral to mobilization of transport teams. We agree that the time from referral to acceptance is a potentially promising quality metric because it is a measure of process that is independent of the patient and within the control of the ECMO referral center (12). It is also potentially modifiable and may be associated with better outcomes (5).

In summary, ECMO support likely improves survival in severe perinatal lung disease (1); care at an ECMO center may reduce the risk of mortality even if the patient does not receive ECMO (8, 9); and the association between increasing severity of illness and an increase in the risk of death prior to transport, all suggest that earlier transfer to an ECMO center may be justified (5, 6). At times, the safest way to transport a patient is first to place that patient on ECMO prior to transfer under the care of a mobile ECMO team. Furthermore, we agree that measuring the processes of care that are modifiable and which may potentially improve patient outcomes are useful quality metrics which we should strive to identify and improve upon.