Pulmonary hypertension (PH) in children may be idiopathic or due to a variety of underlying etiologies, including cardiovascular, respiratory, and systemic conditions (1). Although significant advancements in medical management have occurred over the past several decades, children with PH have higher average comorbidity scores and in-hospital all-cause mortality compared with the general inpatient pediatric population (2,3). Similarly, the disease burden and mortality of children with PH are significantly higher than for the general PICU population (4). As shown by our group, compared with the general PICU population, those with PH have a higher mortality rate (6.8% vs 2.3%), longer length of stay (4.4 vs 1.7 d), and are more likely to require invasive interventions including extracorporeal membrane oxygenation (ECMO) (4).
Knowledge about outcomes and hospital course of children with PH receiving ECMO is scarce.
Extracorporeal life support for patients with PH may be indicated in cases of life-threatening cardiorespiratory failure unresponsive to maximal medical therapies. Although the etiologies of PH in children are varied, the final common pathway in PH crises is typically right ventricular failure (1,5). Mechanical circulatory support has been shown to improve survival in adults with PH, especially when combined with lung transplantation (6–10). In children, a multicenter study utilizing the Kids’ Inpatient Database found a mortality rate of 39% for children with PH requiring ECMO compared with 8% in PH patients without ECMO (11). Small studies have found ECMO to be associated with good outcome in children with pulmonary vascular disease following congenital heart surgery (12–14). In contrast, one retrospective study showed poor outcome for children with PH who required ECMO prior to lung transplantation (15).
It is critical to better understand the growing and evolving populations of children with PH receiving mechanical circulatory support, so strategies and therapies to improve outcomes can be identified and further investigated. Given that pediatric PH is rare, a multicenter database like the Extracorporeal Life Support Organization (ELSO) registry is a valuable tool in addressing this important area of research.
This retrospective multicenter cohort study aims to describe the population of children with PH who require ECMO, identify factors associated with greater mortality, and develop a risk adjustment model predicting mortality for pediatric PH patients supported with ECMO.
MATERIALS AND METHODS
This retrospective multicenter cohort study used data from the ELSO Registry, a comprehensive database that collects ECMO data from 310 centers worldwide (16). Standardized data collection instruments are used at each center and entered through a secure web-based platform. Registry data include demographic information, pre-ECMO assessment (including laboratory, diagnostic, and clinical factors), and ECMO run characteristics (including mode, indication, complications, and outcome). Given that individual member centers receive approval by local institutional review boards to report de-identified data, additional human subjects’ research approval was not required.
The ELSO database was queried to identify patients between 28 days and 18 years old with a diagnosis of PH who received ECMO between January 1, 2007, and November 30, 2018. PH diagnoses were identified using International Classification of Diseases, 9th revision (ICD9: 416.0, 416.8) and ICD, 10th revision (ICD10: I27.0, I27.2, 127.21, 127.22, 127.24) codes. In an effort to exclude newborns with exclusive diagnoses of persistent PH of the newborn, congenital diaphragmatic hernia, and meconium aspiration syndrome, patients under 28 days old at the time of ECMO run were excluded. To reduce misclassification, an investigator (E.M.) reviewed all cases associated with diagnosis codes for both primary PH and congenital heart disease (CHD); these cases were reclassified as secondary PH if this was determined to be the more appropriate diagnosis.
The primary study outcome was in-hospital mortality. Predictor variables included patient factors (age, gender, and PH type), co-diagnoses, interventions, and therapies in the 24 hours pre-ECMO, laboratory values within 6 hours pre-ECMO, and ECMO run characteristics, complications, and discontinuation reasons. Co-diagnoses were chosen based on clinical relevance to PH and included heart failure, CHD, cardiogenic shock, myocarditis/cardiomyopathy, arrhythmia, cardiac tamponade, conduction abnormality, respiratory failure, pneumonia, bronchiolitis, pertussis, influenza, hemoptysis, and sepsis. Pre-ECMO interventions included mechanical circulatory support (cardiopulmonary bypass [CPB], ventricular assist device [VAD]), ventilatory support (conventional, high-flow oscillatory), pharmacologic therapies, and pre-ECMO cardiac arrest. ECMO run characteristics included mode, support type, duration, pump type, and bridge to transplant. Discontinuation reasons were defined by the ELSO Registry as “Expected Recovery,” “Died or Poor Prognosis,” and “ECMO Complication.” ECMO complications were queried individually and in organ-related groups (e.g., “any cardiovascular”). The majority of variables were dichotomized. Specific rationales for handling of each variable are provided (footnotes in Supplemental Table 1, Supplemental Digital Content 1, http://links.lww.com/PCC/B95).
Data analysis was performed using StataSE (version 15; StataCorp, College Station, TX). Chi-square tests were used to compare distributions of categorical variables. Continuous normally distributed variables were presented as mean and sd and compared by t tests. Non-normally distributed continuous data were presented as median and interquartile range (IQR) and compared with Wilcoxon rank-sum tests. Univariate logistic regression analyses were used to analyze the effect of predictors on hospital mortality. Results were presented in odds ratios (ORs) and 95% CIs. The p values of less than 0.05 were considered statistically significant. To generate adjusted ORs to assess independent effects, each variable was adjusted only for those variables found to have a p value of less than 0.05 in univariate analyses.
Multivariable prediction models for different sets of variables (pre-ECMO, ECMO related, all combined) were then developed as follows: 1) variables with p values less than 0.05 in univariate analyses were entered in a multivariable logistic model and 2) stepwise backward selection was performed by dropping variables with p values greater than 0.05 one at a time until only variables with a p value less than 0.05 remained. To assess discrimination of the models, receiver operating characteristic curves were generated and compared using areas under the curve (AUC).
We identified 605 patients between 28 days and 18 years old with a diagnosis of PH, representing 634 ECMO runs (23 patients had two runs, three patients had three runs). Over the 10-year study period, a median of 62 ECMO runs for children with PH was identified per year (range, 31–90; Fig. 1). The mortality rate ranged from 41.8% to 58.7% (median, 51.6%). Cases per year did not vary significantly (p = 0.53). Mortality in patients with PH requiring ECMO was 51.3% compared with 44.8% (p = 0.001) for patients requiring ECMO without a diagnosis of PH.
Demographic and Pre-ECMO Patient Characteristics
Demographic and clinical characteristics of the cohort are presented in Table 1. Median age was 11 months (IQR, 4 mo to 5 yr), and 51% were male. Race/ethnicity was most frequently white (53%) followed by black (15%) and Hispanic (14.0%). Secondary PH cases represented the majority of the cohort (60.9%). The most frequently reported co-diagnoses included CHD (46.8%), respiratory failure (27.9%), heart failure (18.9%), and pneumonia (16.3%).
Characteristics prior to ECMO cannulation are presented in Table 2. The majority of patients were receiving conventional mechanical ventilation (70.3%); median intubation duration was 31 hours (IQR, 9–150). Nearly a third of cases (30.1%) had a cardiac arrest at some point prior to cannulation; 12.8% received CPB and 2.5% were supported with a VAD within the preceding 24 hours. Nitric oxide was used prior to ECMO in 58.5% of cases. The most frequently used vasoactive infusions included epinephrine (51.3%), milrinone (34.7%), and dopamine (25.0%). Thirty-two percent of patients were receiving at least three vasoactive infusions in the day prior to cannulation. At the time closest to cannulation (within 6 hr), median oxygen saturation was 82% (IQR, 60–93) and median pH was 7.24 (IQR, 7.12–7.34).
ECMO Support and Complications
ECMO run characteristics and complications are summarized in Table 3. ECMO support type included pulmonary (43.1%), cardiac (40.2%), and extracorporeal cardiopulmonary resuscitation (ECPR) (16.7%). Median ECMO run duration was 160 hours (IQR, 87–279); the majority of cannulations were venoarterial (80.4%). For the small proportion who received venovenous ECMO (13.6%), almost half had a co-diagnosis of pneumonia (36/86, 42%). The most frequently reported individual complications were inotropic support (48.7%), mechanical/circuit complications (39.6%), hemorrhage (38.5%), and renal replacement therapy (RRT) (37.5%). Longer ECMO duration was significantly associated with any cardiac complication (p = 0.024), pulmonary hemorrhage (p = 0.002), RRT (p < 0.001), mechanical/circuit complications (p < 0.001), and infections (p < 0.001). No significant difference was noted in ECMO duration with neurologic complications or metabolic acidosis.
Univariate and Multivariable Models of Factors Associated With Mortality
Univariate and multivariable predictors of mortality in patients with PH are presented in Supplemental Table 1 (Supplemental Digital Content 1, http://links.lww.com/PCC/B95). Among those with PH, mortality has a bimodal distribution according to age. The youngest (1–6 mo) and oldest (> 5 yr) ages at admission were associated with the greatest mortality rates at 54.2% (OR, 1.6; 95% CI, 1.1–2.3) and 62.3% (OR, 2.2; 95% CI, 1.5–3.4), respectively (reference group, 2–5 yr). Notably, the reference age group had the highest proportion of children with CHD (46% compared with 41% in the 1- to 6-mo group and 12.5% in > 5-yr group; p < 0.001). No significant differences in mortality were identified by gender or PH type. A co-diagnosis of pneumonia was associated with significantly lower mortality (OR, 0.5; 95% CI, 0.3–0.8). A history of pre-ECMO arrest, pre-ECMO blood gas results with pH less than 7.12, Paco2 greater than 75, Pao2 less than 35, and arterial oxygen saturation (Sao2) less than 60% were also significant univariate predictors of mortality but did not remain significant in the adjusted analyses.
Similarly, although ECMO duration of greater than 280 hours was associated with increased mortality in univariate analysis, this lost significance in the adjusted analysis. Patients who received ECPR had a mortality rate of 72.6% (adjusted OR, 3.8; 95% CI, 2.2–6.3). Cardiac-related ECMO complications associated with mortality in univariate analysis included cardiopulmonary resuscitation (CPR), inotropic support, myocardial stun, and tamponade. In the adjusted analyses, only myocardial stun retained significance (adjusted OR, 6.5; 95% CI, 1.4–29.7). Hypertension requiring vasodilator therapy emerged as a significant negative predictor when adjusting for other significant univariate predictors (OR, 0.6; 95% CI, 0.4–0.9). Hemorrhage (including pulmonary, intracranial, and other), seizures, RRT, and pH less than 7.2 were all significant predictors of mortality in both univariate and adjusted analyses. All variables with p values of less than 0.05 in the adjusted analyses are presented in Figure 2.
Univariate predictors were entered into three different models: 1) pre-ECMO factors (age, pneumonia, pre-ECMO arrest, pH, Paco2, Pao2, Sao2); 2) ECMO-related factors (duration, ECPR, any cardiac complication, any neurologic complication, pulmonary hemorrhage, any hematologic complication, RRT, mechanical complication, pH); and 3) both sets of variables combined. The following significant factors were identified in the final models: 1) pre-ECMO (age, absence of pneumonia, and pH < 7.12; AUC, 0.62); 2) ECMO related (ECPR, any neurologic complication, pulmonary hemorrhage, RRT, and metabolic acidosis; AUC, 0.72); and 3) all variables combined (AUC, 0.75) (p < 0.001) (Fig. 3).
Mortality in Children Following Successful ECMO Discontinuation
ECMO was discontinued most often because of expected recovery (68.7%; n = 434); however, 127 (29.3%) of these patients did not survive to discharge. This cohort of patients who died despite expected recovery was examined in more detail (Supplemental Table 2, Supplemental Digital Content 2, http://links.lww.com/PCC/B96). A bimodal distribution was noted with respect to ECMO duration, with the majority of survivors falling into either the less than 24-hour or greater than 280-hour group. Patients in the survivor group were significantly less likely to fall in the 6-month to 5-year age group as well as to have had cardiovascular, neurologic, and hematologic ECMO complications, pulmonary hemorrhage, and RRT on ECMO. No significant differences were noted in cardiac arrest, intubation duration, acidosis, or vasoactive infusion requirements pre-ECMO. Of the 29 patients who were reportedly on ECMO as a bridge to transplant, 14 (48.3%) survived to discharge. Eleven patients (38%) received transplants (six lungs, five hearts); three of these patients died prior to discharge (one lungs, two hearts).
ECMO in Primary Versus Secondary PH
Pre-ECMO and ECMO-related characteristics comparing patients with primary and secondary PH are shown in Supplemental Table 3 (Supplemental Digital Content 3, http://links.lww.com/PCC/B97). No differences were noted in mortality rates. Compared with secondary PH, those with diagnoses of primary PH were significantly older, less likely to have co-diagnoses of heart failure, CHD, and arrhythmias, and less likely to be supported with CPB prior to ECMO. Although no difference was noted in intubation or ECMO duration, those with primary PH were more likely to be supported with high-flow oscillatory ventilation and had significantly higher mean airway pressures. There was no significant difference in pre-ECMO usage of sildenafil or nitric oxide, but prostacyclin therapy was used significantly more often in patients with primary PH. Patients with primary PH were more likely to have a pulmonary etiology, but no differences were noted in ECMO mode. Bridge to transplant was significantly more common in the primary PH cohort.
This multicenter study is the first to describe the epidemiology, clinical characteristics, and prognostic factors in children with PH requiring ECMO. The majority of ECMO runs used a venoarterial cannulation strategy, and 16% of patients were cannulated in the setting of ECPR. Mortality in patients with PH was significantly higher than for those without PH. Several factors associated with worse prognosis were identified, including age, co-diagnosis of pneumonia, pre-ECMO and ECMO-related acidosis, ECPR, pulmonary hemorrhage, RRT, and neurologic complications.
Prior to ECMO cannulation, the majority of patients in this cohort received intensive medical management including mechanical ventilation, nitric oxide, and vasoactive medications. Median duration of intubation prior to ECMO was relatively short (31 hr), which may reflect the limited tolerance of children with right ventricular dysfunction to positive-pressure ventilation (17,18). Blood gas analyses prior to ECMO most frequently demonstrated acidemia, hypercarbia, and hypoxia. These factors are likely associated with increased pulmonary vascular resistance (PVR), which may contribute to hemodynamic compromise necessitating ECMO (5). Although not significant in univariate analyses, vasodilator therapy pre-ECMO did emerge as a significant negative predictor of mortality in the adjusted analysis.
Cardiac arrest was documented prior to ECMO in 30% of patients. Notably, only 16.7% of patients were reported in the ECPR group, suggesting that just under 15% of patients had a cardiac arrest unrelated to immediate ECMO cannulation. In a prior study of children with PH admitted to the PICU, only 4.5% of all PH patients received CPR during admission (4), suggesting that children with PH who go on to require ECMO are an even sicker subgroup of PICU patients. Consistent with this, a study of ECMO prior to thoracic organ transplantation noted that 40% of children with PH on ECMO as a bridge to lung transplantation had a pre-ECMO cardiac arrest (19).
Although the vast majority of patients were cannulated onto venoarterial ECMO, ECMO support type was divided evenly between cardiac and pulmonary, likely reflecting the intimate relationship between right heart failure and lung disease associated with increased PVR in PH (5). The small subset of patients who were cannulated with venovenous ECMO (13.6%) likely represent a group of patients with less severe PH secondary to lung disease or those with intercurrent respiratory infections, given that 42% had a co-diagnosis of pneumonia. RRT was required in 38% of ECMO runs, likely reflective of multisystem organ failure. Median ECMO duration was nearly 1 week, and two thirds of patients were decannulated in the setting of expected recovery. Nearly 5% of patients (n = 29) were placed on ECMO as a bridge to transplant. Less than half (n = 11) ultimately received a transplant. Overall, 15 of the 29 patients died (51.7%), including three following transplantation. This mortality rate is similar to a report of pediatric ECMO as a bridge to heart transplant in which 52% (n = 16) of patients died (19). However, in that study, only one of the 14 patients on ECMO as a bridge to lung transplant survived (19). Comparison with our cohort is limited because type of organ transplant is only known if the transplant procedure code is present.
Overall, we demonstrate that children with PH who require ECMO have significantly greater odds of mortality compared with those without PH (51.3% vs 44.8%; p = 0.001). Given the high risk for cardiopulmonary collapse in children with PH who have intercurrent illness or progressive disease (1,5,20), this is not surprising. Furthermore, children with PH may be more difficult to wean from ECMO, especially when invasive hemodynamic data, such as pulmonary artery pressure monitoring, are not available to guide decision-making. On the other hand, the absolute difference in mortality rates of 6.5% is somewhat reassuring and suggests that, like all children who require ECMO, those with PH still have a nearly 50% chance of survival. Given this, our findings should not discourage ECMO cannulation in children with PH but rather provide insight into how underlying disease severity may contribute to outcome. Furthermore, the frequency of PH patients requiring ECMO and the corresponding annual mortality rates do not vary significantly over the 10-year study period. This suggests that a small proportion of children with PH still require invasive mechanical circulatory support, and once this is required, mortality rate remains high despite ongoing advancements in PH-directed therapies over time.
Although the literature is limited, a few studies provide data for comparison. Utilizing the Kids’ Inpatient Database, Nasr et al (11) identified a mortality rate of 39% in children with pulmonary arterial hypertension (PAH) requiring ECMO. Compared with children with PAH who did not require ECMO, those requiring ECMO were younger and had higher odds of acute kidney injury, sepsis, neurologic, and thrombotic complications compared with controls (11). However, this study excluded those with CHD and did not compare the mortality rate to children without PH receiving ECMO. Therefore, it is possible that this population was not as sick as those reported into the ELSO database. In a small study of pediatric lung transplant recipients over an 18-year period, 10% of patients required perioperative ECMO; severe PH represented 31% of these ECMO cases with a mortality rate of 50% (15), consistent with our study’s findings.
We demonstrated a bimodal distribution in mortality by age with the youngest (1–6 mo) and oldest (> 5 yr) group of patients having the greatest mortality, consistent with a prior study of PICU admissions with PH (4). This may reflect the fact that the lowest-risk age group had the highest proportion of children with CHD, which may be associated with greater surgical or medical treatment options. Secondary PH was more common than primary PH, which is also consistent with the PH population in the PICU (4) as well as with several large registry studies (21,22). Compared with the general PICU, in which patients with primary PH have greater odds of mortality (4), there was no significant difference between groups in this cohort. Perhaps this reflects the common endpoint of right ventricular failure in both primary and secondary PH patients who go on to require ECMO, a marker for greater overall illness severity in this subpopulation of pediatric PH.
A co-diagnosis of pneumonia was associated with a significantly greater odds of survival in both univariate and multivariable analyses. This likely reflects that pneumonia, unlike progression of underlying disease, is often treatable with antibiotic therapy and supportive care, including ECMO. No characteristics related to pre-ECMO support were ultimately predictors of mortality in adjusted analyses. The fact that intubation duration did not predict mortality suggests that timing of ECMO cannulation is not remarkable in this cohort, in contrast to the pediatric respiratory failure population (23,24), which may reflect differences in pathophysiology. Pre-ECMO blood gas results, which were associated with mortality in univariate analyses, may have lost significance in adjusted analyses due to collinearity. However, pre-ECMO acidosis (pH < 7.12) emerged as a significant predictor in the multivariable pre-ECMO prediction model.
Although inotropic support prior to ECMO was not associated with mortality, requirement for vasoactive infusions on ECMO was significantly related to mortality in univariate analysis. ECMO run factors significantly associated with greater odds of mortality in the multivariable prediction model included ECPR, pulmonary hemorrhage, neurologic complications, RRT, and metabolic acidosis. In our cohort, survival was 27% in patients who received ECPR. This is slightly lower overall when compared with the general PICU literature, with survival rates ranging from 33% to 42% (25–30). Perhaps ECPR is associated with lower survival in children with PH because elevated right ventricular pressure makes adequate coronary perfusion during CPR particularly challenging (1,5). Hypertension requiring vasodilator support on ECMO was the only ECMO complication associated with significantly greater odds of survival.
Notably, ECMO duration was not associated with mortality in the adjusted model. This suggests that the complications experienced while on ECMO are more significant than the duration of ECMO support itself. Consistent with this, almost a third of patients who were decannulated in the setting of expected recovery did not survive. Further examination of this subgroup revealed that these children were more likely to have suffered complications on ECMO compared with those who survived in the setting of expected recovery. This reinforces that number and severity of complications on ECMO should be considered when assessing prognosis and determining optimal timing of decannulation. Most notably, pulmonary hemorrhage, metabolic acidosis, neurologic complications, and RRT should raise concern for worse prognosis. In an effort to reduce ECMO duration and associated complications, longer-term mechanical support alternatives, such as lung assist devices, should be studied further as a potential transition from ECMO while awaiting transplant or recovery (31).
This study has several important limitations. Given the retrospective design, the study relied on accurate database entry at each site based on chart review. A standardized registry input process and ongoing data quality monitoring likely mitigate this to some degree. Additionally, the ELSO database does not contain echocardiography or catheterization data to confirm PH diagnoses or stratify patients based on severity or etiology. Furthermore, the lack of granularity in the database makes it challenging to differentiate those with PH secondary to coexisting temporary disease processes, such as pneumonia or acute respiratory distress syndrome, from those with primary PH. Finally, this study excluded patients under 28 days old in an effort to eliminate persistent PH of the newborn and other neonatal conditions associated with PH. Therefore, study results are not generalizable to the neonatal population. Nonetheless, this study provides the largest multicenter assessment of pediatric PH patients requiring ECMO over a 10-year period.
This is the first study to describe ECMO in pediatric PH. Children with PH who require ECMO support have a significantly greater odds of mortality compared with those without PH. Risk factors for mortality include age, absence of pneumonia, pre-ECMO acidosis, ECPR, pulmonary hemorrhage, neurologic complications, and acidosis while on ECMO. Identification of those PH patients requiring ECMO who are at even higher risk for mortality may inform clinical decision-making, including timing of ECMO decannulation, and improve our ability to prognosticate for patients and families.
1. Abman SH, Hansmann G, Archer SL, et al.; American Heart Association Council on Cardiopulmonary, Critical Care, Perioperative and Resuscitation; Council on Clinical Cardiology; Council on Cardiovascular Disease in the Young; Council on Cardiovascular Radiology and Intervention; Council on Cardiovascular Surgery and Anesthesia; and the American Thoracic Society: Pediatric pulmonary hypertension
: Guidelines from the American Heart Association and American Thoracic Society. Circulation 2015; 132:2037–2099
2. Maxwell BG, Nies MK, Ajuba-Iwuji CC, et al. Trends in hospitalization for pediatric pulmonary hypertension
. Pediatrics 2015; 136:241–250
3. Frank DB, Crystal MA, Morales DL, et al. Trends in pediatric pulmonary hypertension
-related hospitalizations in the United States from 2000-2009. Pulm Circ 2015; 5:339–348
4. Balkin EM, Steurer MA, Delagnes EA, et al. Multicenter mortality
and morbidity associated with pulmonary hypertension
in the pediatric intensive care unit. Pulm Circ 2018 Jan-Mar;8(12045893217745785. Epub 2017 Dec 18.
5. Oishi P, Fineman JR. Pulmonary hypertension
. Pediatr Crit Care Med 2016; 17:S140–S145
6. Liu KS, Tsai FC, Huang YK, et al. Extracorporeal life support
: A simple and effective weapon for postcardiotomy right ventricular failure. Artif Organs 2009; 33:504–508
7. Tudorache I, Sommer W, Kühn C, et al. Lung transplantation for severe pulmonary hypertension
–awake extracorporeal membrane oxygenation
for postoperative left ventricular remodelling. Transplantation 2015; 99:451–458
8. Moser B, Jaksch P, Taghavi S, et al. Lung transplantation for idiopathic pulmonary arterial hypertension on intraoperative and postoperatively prolonged extracorporeal membrane oxygenation
provides optimally controlled reperfusion and excellent outcome. Eur J Cardiothorac Surg 2018; 53:178–185
9. de Perrot M, Granton JT, McRae K, et al. Impact of extracorporeal life support
on outcome in patients with idiopathic pulmonary arterial hypertension awaiting lung transplantation. J Heart Lung Transplant 2011; 30:997–1002
10. Fuehner T, Kuehn C, Hadem J, et al. Extracorporeal membrane oxygenation
in awake patients as bridge to lung transplantation. Am J Respir Crit Care Med 2012; 185:763–768
11. Nasr VG, Faraoni D, DiNardo JA, et al. Adverse outcomes in neonates and children with pulmonary artery hypertension supported with ECMO. ASAIO J 2016; 62:728–731
12. Dhillon R, Pearson GA, Firmin RK, et al. Extracorporeal membrane oxygenation
and the treatment of critical pulmonary hypertension
in congenital heart disease. Eur J Cardiothorac Surg 1995; 9:553–556
13. Kolovos NS, Bratton SL, Moler FW, et al. Outcome of pediatric patients treated with extracorporeal life support
after cardiac surgery. Ann Thorac Surg 2003; 76:1435–1441
14. Alsoufi B, Al-Radi OO, Gruenwald C, et al. Extra-corporeal life support following cardiac surgery in children: Analysis of risk factors and survival in a single institution. Eur J Cardiothorac Surg 2009; 35:1004–1011
15. Puri V, Epstein D, Raithel SC, et al. Extracorporeal membrane oxygenation
in pediatric lung transplantation. J Thorac Cardiovasc Surg 2010; 140:427–432
16. Thiagarajan RR, Barbaro RP, Rycus PT, et al.; ELSO Member Centers: Extracorporeal Life Support
Organization Registry International Report 2016. ASAIO J 2017; 63:60–67
17. Bronicki RA, Penny DJ, Anas NG, et al. Cardiopulmonary interactions. Pediatr Crit Care Med 2016; 17:S182–S193
18. Cheifetz IM. Cardiorespiratory interactions: The relationship between mechanical ventilation and hemodynamics. Respir Care 2014; 59:1937–1945
19. Kirshbom PM, Bridges ND, Myung RJ, et al. Use of extracorporeal membrane oxygenation
in pediatric thoracic organ transplantation. J Thorac Cardiovasc Surg 2002; 123:130–136
20. Ivy DD, Abman SH, Barst RJ, et al. Pediatric pulmonary hypertension
. J Am Coll Cardiol 2013; 62:D117–D126
21. Haworth SG, Hislop AA. Treatment and survival in children with pulmonary arterial hypertension: The UK pulmonary hypertension
service for children 2001-2006. Heart 2009; 95:312–317
22. Van Loon RLE, Roofthooft MTR, Hillege HL, et al. Pediatric pulmonary hypertension
in the Netherlands: Epidemiology and characterization during the period 1991 to 2005. Circulation2011;124:1755–1764
23. Domico MB, Ridout DA, Bronicki R, et al. The impact of mechanical ventilation time before initiation of extracorporeal life support
on survival in pediatric respiratory failure: A review of the extracorporeal life support
registry. Pediatr Crit Care Med 2012; 13:16–21
24. Zabrocki LA, Brogan TV, Statler KD, et al. Extracorporeal membrane oxygenation
for pediatric respiratory failure: Survival and predictors of mortality
. Crit Care Med 2011; 39:364–370
25. Alsoufi B, Al-Radi OO, Nazer RI, et al. Survival outcomes after rescue extracorporeal cardiopulmonary resuscitation in pediatric patients with refractory cardiac arrest. J Thorac Cardiovasc Surg 2007; 134:952–959.e2
26. Morris MC, Wernovsky G, Nadkarni VM. Survival outcomes after extracorporeal cardiopulmonary resuscitation instituted during active chest compressions following refractory in-hospital pediatric cardiac arrest. Pediatr Crit Care Med 2004; 5:440–446
27. Raymond TT, Cunnyngham CB, Thompson MT, et al.; American Heart Association National Registry of CPR Investigators: Outcomes among neonates, infants, and children after extracorporeal cardiopulmonary resuscitation for refractory inhospital pediatric cardiac arrest: A report from the national registry of cardiopulmonary resuscitation. Pediatr Crit Care Med 2010; 11:362–371
28. Thiagarajan RR, Laussen PC, Rycus PT, et al. Extracorporeal membrane oxygenation
to aid cardiopulmonary resuscitation in infants and children. Circulation 2007; 116:1693–1700
29. Lasa JJ, Rogers RS, Localio R, et al. Extracorporeal Cardiopulmonary Resuscitation (E-CPR) during pediatric in-hospital cardiopulmonary arrest is associated with improved survival to discharge: A report from the American Heart Association’s get with the guidelines-resuscitation (GWTG-R) registry. Circulation 2016; 133:165–176
30. Wolf MJ, Kanter KR, Kirshbom PM, et al. Extracorporeal cardiopulmonary resuscitation for pediatric cardiac patients. Ann Thorac Surg 2012; 94:874–879
31. Strueber M, Hoeper MM, Fischer S, et al. Bridge to thoracic organ transplantation in patients with pulmonary arterial hypertension using a pumpless lung assist device. Am J Transplant 2009; 9:853–857