Transportation of patients on extracorporeal membrane oxygenation: a tertiary medical center experience and systematic review of the literature

ECMO remains a high-cost therapy with lifesaving potential in a select group of critically ill patients [3].Given the level of expertise needed for daily care of these patients, it is preferable that ECMO candidates be transferred to a specialized referral center. Because of the severe respiratory failure, patient transfer without ECMO is usually deemed to be too risky. Conversely, there are no data regarding the safety of inter-hospital ECMO transportations. Herein, we report a case series of patients transported under ECMO support to a referral hospital in Brazil with a survival rate of 57% and no major complications or deaths during transportation. Additionally, our systematic review of the literature showed a pooled survival rate for adult and pediatric patients of nearly two-thirds—with just 2 deaths reported in this cohort of 1481 patients—and without any other major adverse events resulting from the transportation itself.

Our data are compatible with the overall mortality reported in the latest publication by the Extracorporeal Life Support Organization (ELSO), in which the expected survival rate for adult venovenous extracorporeal support was 58% [56]. Similarly, using the available published data, we found a survival rate of 62% for adult patients transported while on ECMO. For the pediatric population, our pooled analysis retrieved a survival rate of 68%, in comparison with the 57% reported in ELSO guidelines [56]. Therefore, in this case series and in our overall analysis, we found no increase in mortality for ECMO support despite the need for patient transfer to a referral institution.

Concerns regarding the safety of transport of critically ill patients in need of extracorporeal support are an important question to be solved considering the recent global increase in ECMO support [5, 56, 57]. In the Cesar trial, patients who were randomized for the ECMO group were transferred to the referral center only after transport was considered safe by the ECMO Team, therefore delaying the initiation of support. As reported in the text, patients were not transported while on ECMO and, despite precautions, two deaths were reported during patient transfer [3]. Similarly, in a previous publication of 158 infants accepted for ECMO initiation, Boedy et al. reported 18 (39.1%) deaths associated with transport. Five infants died waiting for ECMO initiation and 13 died either during transport without ECMO assistance or, after arriving moribund, before ECMO could be started. Considering all these deaths occurred before ECMO initiation, the authors concluded that there may be a hidden mortality associated with ECMO transportation that is generally excluded when we look exclusively at ECMO-supported patients [58]. Therefore, a strategy of rapid ECMO initiation and patient transport while on ECMO support may be safer than the use of conventional mechanical ventilation during transfer to the referral center.

However, it is important to highlight that the presence of complications is common, and nearly a third of the analyzed studies reported at least one complication during transport. Sudden fall in tidal volume was the most common complication reported. Power failure, circuit rupture and other more severe complications were also reported, but no deaths or any adverse outcomes related to these complications during transport were described. It is interesting to note that the majority of complications (74%) were reported in one single study [43] (Additional file 5: Table 2s), suggesting that most reports on ECMO transportation did not focus on looking for adverse events. Certainly, the definition of complications during patient transportation varied between studies, making it difficult to understand the real size of this problem. It is very likely that the incidence of adverse events is much higher than described in this manuscript.

Our study has several limitations: (1) The absence of any scoring system in most of the studies makes it difficult to correlate expected mortality with final results. However, as previously described, overall mortality rate was compatible with the expected mortality previously published in the ELSO guidelines for ECMO-supported patients. (2) A publication bias may have affected our results. As observed in our analysis, low-quality studies were associated with a high mortality rate and, possibly, even higher mortality rates may be found in unpublished data. (3) No randomized clinical trial has directly evaluated the safety of transporting ECMO patients. Our data were extracted mainly from case series, and the results are limited by the inherent flaws of such studies. (4) The studies included in this manuscript span several years of ECMO transportation worldwide. Therefore, clinical and technical development, which may have influenced the presence of complications and death in ECMO patients throughout the years, are not addressed in this manuscript.