Share this article on:

Prolonged Extracorporeal Membrane Oxygenator Support Among Neonates with Acute Respiratory Failure: A Review of the Extracorporeal Life Support Organization Registry

Prodhan, Parthak*; Stroud, Michael; El-Hassan, Nahed; Peeples, Sarah; Rycus, Peter§; Brogan, Thomas V.; Tang, Xinyu

doi: 10.1097/MAT.0000000000000006
Pediatric Circulatory Support

The objective of this study was to identify types of neonatal diseases associated with prolonged (≥21 days) extracorporeal membrane oxygenation (ECMO), characteristics of survivors and nonsurvivors among those requiring prolonged ECMO, and factors associated with mortality. Data were obtained from the Extracorporeal Life Support Organization registry over the period from January 1, 1998, through December 31, 2011, for all neonates (age <31 days), with respiratory failure as the indication for ECMO. The primary outcome was survival to hospital discharge. Survivors and nonsurvivors were compared for 1) patient demographics, 2) primary diagnosis, 3) pre-ECMO clinical course and therapies, and 4) ECMO course and associated complications. The most common diagnosis associated with prolonged ECMO support in neonates is congenital diaphragmatic hernia (CDH; 69%). Infants with meconium aspiration syndrome had the highest survival rate (71%) compared with other diagnoses analyzed (26.3%; p < 0.001). Nonsurvivors were more likely to experience complications on ECMO, and multivariate analysis showed that the need for inotropes while on ECMO support (odds ratio, 2.2 [95% confidence interval, 1.3–3.7]; p = 0.003) was independently associated with mortality. Neonates requiring prolonged ECMO support have a 24% survival to discharge. Many of these cases involve CDH. Complications are common with prolonged ECMO, but only receipt of inotropes was shown to be independently associated with mortality. This report may help guide clinical decision making and family counseling for neonates requiring prolonged ECMO support.

Supplemental Digital Content is available in the text.

From the *Department of Pediatrics, Division of Cardiology, College of Medicine, University of Arkansas for Medical Sciences, Arkansas Children’s Hospital, Little Rock, Arkansas; Department of Pediatrics, Division of Critical Care, College of Medicine, University of Arkansas for Medical Sciences, Arkansas Children’s Hospital, Little Rock, Arkansas; Department of Pediatrics, Division of Neonatology, College of Medicine, University of Arkansas for Medical Sciences, Arkansas Children’s Hospital, Little Rock, Arkansas; §Extracorporeal Life Support Organization, Ann Arbor, Michigan; Department of Pediatrics, Division of Critical Care, Seattle Children’s Hospital, Seattle, Washington; and Division of Biostatistics, College of Medicine, University of Arkansas for Medical Sciences, Arkansas Children’s Hospital, Little Rock, Arkansas.

Submitted for consideration May 2013; accepted for publication in revised form August 2013.

Disclosure: The authors have no conflicts of interest to report.

Reprint Requests: Parthak Prodhan, MD, College of Medicine, University of Arkansas for Medical Sciences, Arkansas Children’s Hospital, Little Rock, AR 72205. Email: prodhanparthak@uams.edu.

Extracorporeal membrane oxygenation (ECMO), a form of cardiopulmonary bypass adapted for use outside the operating room, has been used successfully to support children with acute cardiorespiratory failure, which is refractory to medical management.1–5 A subset of patients on ECMO support may require prolonged duration of ECMO support because of nonimprovement in the underlying disease process. The landscape and outcome of neonates with respiratory failure supported with prolonged ECMO (>21 days) in a large study cohort are largely unknown. The current information on neonates requiring prolonged ECMO support is limited to reports from single centers.6,7 However, these studies are limited by small sample size6,7 and are not focused specifically on the prolonged ECMO cohorts or on the spectrum of specific diagnostic groups.6,7

The objective of this study was to review data from the Extracorporeal Life Support Organization (ELSO)5 registry to provide information on the clinical characteristics, survival outcomes, and review factors associated with mortality for those requiring prolonged ECMO.

Back to Top | Article Outline

Materials and Methods

Patient Population

This retrospective study includes all neonates (age <31 days at the time of ECMO deployment) with respiratory failure as reported to the ELSO registry from January 1, 1998, to December 31, 2011. Among them, those receiving ECMO support for ≥21 days were included. A cutoff of 21 days was considered for the study based on prior reports on prolonged ECMO.8 Only data from the initial ECMO deployment were included for subjects who received ECMO support more than once. Exclusion criteria included age ≥31 days at the time of ECMO deployment, those with a cardiac diagnosis, and those without any listed diagnosis. No informed consent was obtained because the data set was limited to nonidentifiable elements. This study was approved by the Institutional Review Board (IRB) at our facility and by the ELSO registry.

Back to Top | Article Outline

Extracorporeal Life Support Organization Registry

The ELSO registry5 was founded in 1982 and collects data from member institutions on extracorporeal support in children and adults. Data are collected and sent from the contributing centers with a standardized data sheet containing patient demographics, diagnosis and procedure information, ECMO technique, complications, and patient outcomes. Currently, 200 centers voluntarily contribute data to the registry. Each individual member institution approves data reporting through its local IRB. Analyses and reports on the ELSO database including this report are approved by the Registry Committee of ELSO and analyses on deidentified data by the University of Michigan IRB.

Back to Top | Article Outline

Data Variables

Data relating to patient demographics, primary diagnosis, pre-ECMO course variables, pre-ECMO therapies, ECMO course data, and associated complications were analyzed. The end-point of the study was survival to hospital discharge. Primary International Classification of Diseases, Ninth Revision codes were reviewed by two authors independently. Any disagreement was resolved after review by a third author.

Back to Top | Article Outline

Statistical Analysis

Descriptive statistics were expressed as medians (25th percentile, 75th percentile) for continuous variables and percentages and frequencies for categorical variables. Mann–Whitney U tests were used for comparing distributions of continuous variables between nonsurvivors and survivors, whereas χ2 tests were used for comparing proportions of categorical variables between nonsurvivors and survivors. Kaplan-Meier curves for time to death based on admissions with different primary diagnoses were constructed, and the log-rank test was performed to determine overall difference in survival among diagnoses. Patients who were discharged from the hospital were considered to be censored. The multivariable logistic regression model with Firth penalized likelihood approach was fitted for mortality as a function of the presence of hemorrhage, central nervous system (CNS) infarction, renal dysfunction, arrhythmias, need for hemofiltration, inotropes, duration of ECMO, and total number of complications so as to assess the effects of different complications on mortality simultaneously. p Values <0.05 were considered to indicate statistical significance. All the data were analyzed using the R package version 2.15.0 (R Core Team, Vienna, Austria).

Back to Top | Article Outline

Results

Patient Characteristics

Four hundred thirty-nine patients met inclusion criteria. Among them, 334 (76%) died and 105 (24%) survived to hospital discharge. Figure 1 shows the temporal distribution of neonates requiring prolonged ECMO from 1998 to 2011. There were no significant differences in mortality by era (1998–2001 vs. 2002–2005 vs. 2006–2011; p = 0.36). Figure 2 depicts the number of patients, duration of ECMO support, and a corresponding survival percentage for different days on ECMO support. No survivors were noted among those on ECMO support for >43 days. There were no significant differences in survival by duration of ECMO before 43 days.

Figure 1

Figure 1

Figure 2

Figure 2

Table 1 shows clinical characteristics of the group and compares clinical characteristics of survivors and nonsurvivors. Nonsurvivors weighed less, were more frequently placed on ECMO in the first 7 days of life, and required shorter median times from intubation to ECMO support. Among neonates requiring prolonged ECMO support, those with congenital diaphragmatic hernia (CDH) represented the predominant primary diagnosis. The survival of CDH group (22%) was lower than the non-CDH group (28%). Among the CDH subgroup (data not shown), survivors differed significantly from nonsurvivors only by weight (3.1 kg [2.9–3.4] vs. 3 kg [2.7–3.3]; p = 0.04) (median [interquartile range (IQR)]; p value) among the compared clinical characteristics.

Table 1

Table 1

Table 2 shows a comparison between survivors and nonsurvivors for worst gas exchange variables and therapies provided before ECMO deployment. There were no significant differences between survivors and nonsurvivors. Among the subgroup of neonates with CDH (data not shown), survivors differed significantly from nonsurvivors for the use of pre-ECMO high frequency ventilation (57% vs. 79%; p < 0.001), bicarbonate therapy (50% vs. 32%; p = 0.01), and time off ECMO to extubation (549 hr [403–952] vs. 30 hr [0–312]; p < 0.001) (median [IQR]; p value).

Table 2

Table 2

Figure 3 shows survival curves for four major diagnoses (bacterial pneumonia, CDH, viral pneumonia, and meconium aspiration syndrome [MAS]). Median time to death with 95% confidence intervals (CI) was 35.5 days (26.8–41.8 days) for bacterial pneumonia, 36.8 days (34.2–40.3 days) for CDH, and 31.5 days (29.7 days, upper bound could not be estimated) for viral pneumonia. Meconium aspiration syndrome had a significantly better survival compared with the other three diagnoses (p < 0.001).

Figure 3

Figure 3

Back to Top | Article Outline

Complications on ECMO

Table 3 depicts complications occurring while on ECMO. Nonsurvivors were more likely to have hemorrhage, CNS infarction, renal dysfunction, cardiac arrhythmias, and required hemofiltration and inotropic support. Overall, nonsurvivors had more aggregate complications compared with survivors (p ≤ 0.001). Multivariate analysis for complications showed that receiving inotropes while on ECMO support was independently associated with death before hospital discharge (odds ratio, 2.2; 95% CI, 1.3–3.7; p = 0.003). Among the subgroup of neonates with CDH (data not shown), survivors differed significantly from nonsurvivors for increased frequency of cardiac arrhythmias (1% vs. 12%; p = 0.01), receipt of inotropes 34% vs. 56%; p = 0.001), and total number of ECMO complications (4 [2.0–6.3] vs. 5.0 [3.0–8.0]; p = 0.03) on univariable comparison. We found no impact in utilization of inotropes while on ECMO support on mortality among those on venovenous (VV) and venoarterial ECMO for any of the diagnostic categories (see Figure, Supplemental Data Content 1, http://links.lww.com/ASAIO/A44).

Table 3

Table 3

Back to Top | Article Outline

Discussion

In this retrospective review of the ELSO registry data set among neonates who required prolonged ECMO support (>21 days): 1) CDH accounted for almost two-third of all cases, 2) overall survival to hospital discharge was only 24%, 3) neonates with MAS had a significantly better survival to hospital discharge compared with the other diagnostic groups (CDH, bacterial and viral pneumonia), and 4) the need for inotropes while on ECMO support was independently associated with death before hospital discharge.

Our study revealed that neonates with CDH account for 69% of patients requiring prolonged ECMO support for respiratory failure. In comparison, review of the neonatal ELSO registry data for the same era shows that CDH is the most common indication for ECMO in the United States and accounts for 29% of all neonatal ECMO patients.1 The survival among infants with CDH requiring prolonged ECMO (22%) is significantly lower compared with the overall report of ~50% survival from the ELSO registry and other single-center experience.5–11 Furthermore, our analysis did not show significant change in survival over the study period. Previously, in a single-center study, Tiruvoipati et al.6 showed that the mortality of CDH patients receiving ECMO support for >2 weeks was 82% (n = 2/11 survived). It can be speculated that the prolonged duration of ECMO and poor outcomes among neonates with CDH requiring prolonged ECMO may be related to the presence of severe pulmonary hypoplasia and irreversible pulmonary hypertension for which unfortunately, the ELSO registry database does not carry any information. Other factors previously identified to impact survival in this population, including gestational age, 5 minute Apgar score, PaO2, and PaCO2,6,7,9,10 were characteristics that did not affect survival in our analysis. Thus, these variables appear less significant once an infant with CDH requires ECMO for >21 days.

Although the number of patients with MAS requiring prolonged ECMO was small (n = 27), the survival to hospital discharge was significantly better (71%) compared with the other study groups (26.3%). However, the survival among these neonates with MAS is lower than overall survival for MAS (94%) reported for the ELSO registry. Radhakrishnan et al.12 speculate that more patients with MAS receive VV ECMO than their no-MAS counterparts, leading to less morbidity and thus less mortality in those infants.5 Surprisingly, there were no cases of isolated persistent pulmonary hypertension noted in this study. All patients reported to have pulmonary hypertension was in the setting of other underlying disease (MAS, CDH with/without lung agenesis). Interestingly, neonates requiring prolonged ECMO also showed a much lower survival to discharge than those requiring prolonged ECMO support among older children (38% vs. 24%). These differences may be related to the functional immaturity of neonates compared with older children and with a different case mix with CDH being the predominant diagnosis among the neonatal population.

Similar to previous results from prolonged ECMO in older children,8,13 we found that complications commonly occurred during ECMO support, with a median of three complications in survivors and nonsurvivors. In our study, inotropic support while on ECMO was independently associated with mortality, suggesting that ongoing cardiopulmonary dysfunction despite ECMO support may impact outcomes. This is similar to the report by Brogan et al. 8 among older children with respiratory failure on prolonged ECMO support. However, in contrast to their report, metabolic acidosis was not independently associated with mortality in this study. Other reports on prolonged ECMO in children have shown that ECMO complications did not influence either outcome or duration of ECMO.13

This study is subject to several limitations as data are abstracted in a retrospective manner into the ELSO registry. As a retrospective study of registry data, it is subject to considerable bias, both in reporting and in selection. Overall, 13% (57/439) were categorized as “other” as they carried nonspecific primary diagnosis of acute respiratory failure and respiratory distress syndrome, and the registry data set does not provide additional information to allow more precise classification of this group of patients. Furthermore, there is lack of standardization in the application of ECMO, which lends to large variability between and within centers to accurately exact indications and threshold for offering ECMO support. Also, ELSO does not release data on individual centers the data set is limited so as to investigate institutional variations. Finally, data on quality of survival and residual respiratory function are lacking, which is important especially in those requiring prolonged ECMO and survived to hospital discharge. Furthermore, among nonneonates discharged from the hospital, 1 year post-ECMO survival of 14% is discouraging but represented 75% of patients who survived to discharge.13 The current database is limited in providing long-term follow-up data on the survivors.

Back to Top | Article Outline

Conclusions

Neonates receiving prolonged ECMO support (>21 days) for respiratory failure have a much poorer survival to discharge (24%) compared with overall results reported by the ELSO registry (75%). Furthermore, the need for prolonged ECMO appears to be related to the underlying diagnosis as CDH accounts for almost two-third of cases. The survival among this group was almost half of those reported for overall survival in the ELSO registry. Complications on ECMO are common, but only the receipt of inotropes while on ECMO support (suggestive of poor cardiorespiratory support) increased the odds of mortality.

Back to Top | Article Outline

References

1. Brown KL, Sriram S, Ridout D, et al. Extracorporeal membrane oxygenation and term neonatal respiratory failure deaths in the United Kingdom compared with the United States: 1999 to 2005. Pediatr Crit Care Med. 2010;11:60–65
2. Karimova A, Brown K, Ridout D, et al. Neonatal extracorporeal membrane oxygenation: Practice patterns and predictors of outcome in the UK. Arch Dis Child Fetal Neonatal Ed. 2009;94:F129–F132
3. Roy BJ, Rycus P, Conrad SA, Clark RH. The changing demographics of neonatal extracorporeal membrane oxygenation patients reported to the Extracorporeal Life Support Organization (ELSO) Registry. Pediatrics. 2000;106:1334–1338
4. Hintz SR, Suttner DM, Sheehan AM, Rhine WD, Van Meurs KP. Decreased use of neonatal extracorporeal membrane oxygenation (ECMO): how new treatment modalities have affected ECMO utilization. Pediatrics. 2000;106:1339–1343
5. ELSO. Registry Report of the Extracorporeal Life Support Organization. 2012 Ann Arbor, MI International Summary
6. Tiruvoipati R, Vinogradova Y, Faulkner G, Sosnowski AW, Firmin RK, Peek GJ. Predictors of outcome in patients with congenital diaphragmatic hernia requiring extracorporeal membrane oxygenation. J Pediatr Surg. 2007;42:1345–1350
7. Seetharamaiah R, Younger JG, Bartlett RH, Hirschl RBCongenital Diaphragmatic Hernia Study Group. . Factors associated with survival in infants with congenital diaphragmatic hernia requiring extracorporeal membrane oxygenation: A report from the Congenital Diaphragmatic Hernia Study Group. J Pediatr Surg. 2009;44:1315–1321
8. Brogan TV, Zabrocki L, Thiagarajan RR, Rycus PT, Bratton SL. Prolonged extracorporeal membrane oxygenation for children with respiratory failure. Pediatr Crit Care Med. 2012;13:e249–e254
9. Ssemakula N, Stewart DL, Goldsmith LJ, Cook LN, Bond SJ. Survival of patients with congenital diaphragmatic hernia during the ECMO era: An 11-year experience. J Pediatr Surg. 1997;32:1683–1689
10. Hoffman SB, Massaro AN, Gingalewski C, Short BL. Predictors of survival in congenital diaphragmatic hernia patients requiring extracorporeal membrane oxygenation: CNMC 15-year experience. J Perinatol. 2010;30:546–552
11. Guner YS, Khemani RG, Qureshi FG, et al. Outcome analysis of neonates with congenital diaphragmatic hernia treated with venovenous vs venoarterial extracorporeal membrane oxygenation. J Pediatr Surg. 2009;44:1691–1701
12. Radhakrishnan RS, Lally PA, Lally KP, Cox CS Jr. ECMO for meconium aspiration syndrome: Support for relaxed entry criteria. ASAIO J. 2007;53:489–491
13. Green TP, Moler FW, Goodman DM. Probability of survival after prolonged extracorporeal membrane oxygenation in pediatric patients with acute respiratory failure. Extracorporeal Life Support Organization. Crit Care Med. 1995;23:1132–1139
14. Gupta P, McDonald R, Chipman CW, et al. 20-year experience of prolonged extracorporeal membrane oxygenation in critically ill children with cardiac or pulmonary failure. Ann Thorac Surg. 2012;93:1584–1590
Keywords:

prolonged; extracorporeal membrane oxygenation; acute respiratory failure; neonates; outcomes

Supplemental Digital Content

Back to Top | Article Outline
Copyright © 2014 by the American Society for Artificial Internal Organs