Extracorporeal membrane oxygenation use in the first 24 hours following pediatric heart transplantation: Incidence, risk factors, and outcomes.

Godown J, Bearl DW, Thurm C, Hall M, Feingold B, Soslow JH, Mettler BA, Smith AH, Profita EL, Singh TP, Dodd DA.

Pediatr Transplant. 2019 Apr 11:e13414. doi: 10.1111/petr.13414. [Epub ahead of print]

PMID: 30973190

Similar articles

Select item 30973305


Take Home Points:


  • Primary graft dysfunction (requiring ECMO) after pediatric heart transplant is not uncommon and can lead to significant morbidity or mortality.
  • This retrospective study was designed using large administrative databases to find the incidence of, risk factors for, and outcomes of primary graft dysfunction and ECMO utilization after heart transplant. The results indicate that primary dysfunction is not rare (7.9% of transplants) and that outcomes are strongly associated with the duration of ECMO.

 Commentary from Dr. Timothy Pirolli (Dallas), section editor of Congenital Heart Surgery Journal Watch:  Primary graft dysfunction after heart transplant is often described as a “black box” of unknown etiology that requires extracorporeal membrane oxygenation (ECMO) support to allow the heart to recover, which may not occur. It constitutes 40% of mortality after transplant and often a single factor is not identified as the precipitating cause. The authors of this study sought to examine the incidence of, factors for and outcomes from the use of ECMO within 24 hours after heart transplant in children. They sought to examine two large administrative databases to utilize a unique data linkage to examine their questions.


The SPTR database includes information from every organ transplant since 1987. The PHIS database is an administrative/billing database that captures clinical and resource utilization data, including ICD-9 and -10 codes, from over 50 children’s hospitals. The researchers queried all heart transplants in children under 18 years old between 2002 and 2016 from the linked database and found 2820 patients. Primary graft dysfunction was then examined by using the PHIS to obtain billing codes for ECMO initiation within 24 hours of heart transplant. This was also the method to delineate the duration of ECMO therapy for each of these patients. Baseline demographics and data regarding the heart transplant were obtained. Analyses of this data using multivariate logistic regression models were used to assess risk factors for primary graft dysfunction. The authors also looked at each major transplant group (cardiomyopathy vs. congenital heart disease) to evaluate if there are separate risk factors for each. Outcomes evaluated included ECMO decannulation, re-transplantation and death.


The first key finding from this report is that 7.9% (n=224) of heart transplant recipients required ECMO within the first 24 hours after transplant. The median time on ECMO was 2 days with a maximum duration of 50 days. An evaluation of the multivariable logistic regression of patient demographics and primary graft dysfunction is summarized in table 2. It is not surprising that the authors found that younger patients and patients already on ECMO at time of transplant were independently associated with primary graft dysfunction. The risk factors for ECMO are also evaluated when the transplant recipients are separated into cardiomyopathy vs. congenital heart disease subgroups (Table 3.) Interestingly, only 14.3% (n=32) of the patients placed on ECMO died while on ECMO as seen in Figure 1. Of the 191 patients who were decannulated, 167 (87.4%) survived until discharge. The study also highlights that the survival to discharge decreases as the duration of the ECMO course increases, from 89% survival for 1-3 day ECMO runs to 18.8% survival for more than 10 days. This data is also presented graphically in Figure 2. The risks factors for in-hospital mortality for patients requiring post-transplant ECMO are presented in Table 5. Unsurprisingly, long ischemic times, length of ECMO and post-transplant dialysis were associated with in-hospital mortality. The need for post-transplant ECMO was associated with inferior in-house patient survival compared to those who did not need ECMO. However, for those who survived to discharge after ECMO, there was no long-term difference in survival when compared to those who did not require ECMO (Figures 3 and 4).


This unique method of assessing ECMO utilization in the first 24 hours after transplant by using large administrative databases gives a more thorough understanding of which patients are at risk for the need for post-transplant ECMO and which patients who do need ECMO will survive to discharge. The cynic in me points to the fact that hospitals are not going to forget to bill for ECMO and thus this data should be more complete than other voluntary collaborative databases. Still this data comes from just 28 hospitals and there is obviously much variability among transplant programs’ practices. The findings here will allow transplant programs to help guide their patients’ families into the short- and long-term outcomes from a post-transplant ECMO run. The results seem to be more favorable than expected. It is clear that this study does not allow for a better understanding of what the indication for ECMO was for these patients. And there is obviously the risks of errors and missing information in this data, but the summary of findings should give both practitioners and families hope that the need for post-transplant ECMO support is not a death sentence.

Tables and Figures: