Journal of Pediatric Surgery 50 (2015) 60–63
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Enteral nutrition in neonatal and pediatric extracorporeal life support: A survey of current practice Thomas J. Desmarais a,b, Yan Yan c, Martin S. Keller a, Adam M. Vogel a,⁎ a b c
Division of Pediatric Surgery, Washington University School of Medicine in Saint Louis, Saint Louis, MI, United States Geisel School of Medicine at Dartmouth, Hanover, NH, United States Department of Surgery, Washington University School of Medicine in Saint Louis, Saint Louis, MI, United States
a r t i c l e
i n f o
Article history: Received 29 September 2014 Accepted 6 October 2014 Key words: Extracorporeal life support ECLS ECMO enteral nutrition critical care
a b s t r a c t Purpose: The purpose of this study was to characterize enteral (EN) nutrition practices in neonatal and pediatric patients receiving extracorporeal life support (ECLS). Methods: A Web-based survey was administered to program directors and coordinators of Extracorporeal Life Support Organization centers providing neonatal and pediatric ECLS. The survey assessed patient and clinical factors relating to the administration of EN. Results: A total of 122 responses (122/521, 23.4%) from 96 institutions (96/187; 51.3%) were received. One hundred fifteen provided neonatal or pediatric ECLS, and 84.2% reported utilizing EN during ECLS. 55% and 71% of respondents provide EN ‘often’ or ‘always’ for venoarterial and venovenous ECLS, respectively. EN was reported as given ‘often’ or ‘always’ by 24% with increased vasopressor support, 53% with “stable” vasopressor support, and 60% with weaning of vasopressor support. Favorable diagnosis for providing EN includes respiratory distress syndrome, pneumonia, asthma, trauma/post-operative, pulmonary hemorrhage, and infectious cardiomyopathy. Vasopressor requirement and underlying diagnosis were the primary or secondary determinant of whether to provide EN 81% and 72% of the time. 38% reported an established protocol for providing EN. Conclusion: EN support is common but not uniform among neonatal and pediatric patients receiving ECLS. ECLS mode, vasopressor status, and underlying diagnosis play an important role in the decision to provide EN. © 2015 Elsevier Inc. All rights reserved.
1. Background Extracorporeal Life Support (ECLS) utilizes a series of established and evolving technologies to deliver life saving treatment to critically ill patients with reversible cardiac and pulmonary failure [1]. Malnutrition in critically ill patients is extremely common and is associated with increased mortality and morbidity including impaired immune function, impaired ventilator drive, prolonged ventilator dependence, and increased infections [2]. Consequently, adequate nutrition is essential to minimize physiologic complications of critical illness and to promote patient recovery. Enteral nutrition (EN) is the preferred method of caloric, protein, and micronutrient delivery and has been shown to reduce sepsis-associated morbidity and cost, improve intestinal immunologic function, and improve nitrogen balance in critically ill patients [3–5]. Alternatively, parenteral nutrition (PN) has been used to deliver daily caloric, protein, and micronutrient requirements [6], but is associated with several complications including intestinal villus hypoplasia, reduction of
intestinal absorptive function, increased bacterial translocation, centralline infections, hyperglycemia, and cholestasis [7]. Delivery of EN may be avoided because of concerns regarding inadequate intestinal perfusion and non-occlusive mesenteric ischemia with the development of necrotizing enterocolitis, intestinal ischemia, perforation, or gastrointestinal hemorrhage. Despite these concerns, hypoperfusion related intestinal complications while on ECLS remain an unproven risk. Several small studies have documented the feasibility and safety of EN in ECLS patients in the neonatal, pediatric, and adult populations [7–13]. Additionally, current guidelines for providing nutritional support of neonates simultaneously supported with ECLS recommend initiating enteral feeds when patients are clinically “stable” [14]. The optimum route for delivery of nutrition in neonates and children receiving ECLS is not well established and practice patterns have not been described. The goal of this study is to characterize current practice patterns of the administration and delivery of EN at centers treating neonatal and pediatric ECLS patients. 2. Methods
⁎ Corresponding author at: Division of Pediatric Surgery, Washington University School of Medicine in St. Louis, One Children's Place, Suite 5S40, Saint Louis, MO 63110, United States. Tel.: +1 314 454 6022; fax: +1 314 454 2442. E-mail address: [email protected] (A.M. Vogel). http://dx.doi.org/10.1016/j.jpedsurg.2014.10.030 0022-3468/© 2015 Elsevier Inc. All rights reserved.
The Washington University in St. Louis School of Medicine Institutional Review Board (#201302094) approved this study. The Extracorporeal Life Support Organization (ELSO) is an international consortium of health care professions and scientists who are dedicated to the
T.J. Desmarais et al. / Journal of Pediatric Surgery 50 (2015) 60–63
development and evaluation of novel therapies for support of failing organ systems [15]. Contact information for ECLS program directors and coordinators was obtained from participating centers identified on the ELSO Web site. The survey was administered through the REDCap electronic data capture tool hosted at Washington University in St Louis [16]. A link to the survey was sent electronically on June 4, 2013, with reminders sent on June 17, 2013, and again on July 8, 2013. The thirteen question Web-based survey was designed to assess nutritional implementation and delivery practices in neonatal and pediatric patients on ECLS. The survey (see Appendix A) sought to determine factors that might influence EN implementation such as ECLS mode, patient diagnosis, vasopressor support, and pharmacologic paralysis as well as administration preferences (gastric vs. post-pyloric and institutional unit-based “feeding” protocols). The data were compiled anonymously and analyzed as a composite. Responses are reported as ranges for continuous data and percentages for categorical data. Write-in answers are reported as direct quotes or grouped by theme.
61
Twenty-five respondents reported that there was at least one factor considered more important at their institution than the four factors discussed above. In the free response space provided, the most commonly listed factors included bowel functionality [14], severity of illness [2], lactic acidosis [2], presence of central cannulation [2], and cardiac arrest [2]. Other reasons included: “presence of feeding tube prior to heparinization” and “unit preference.” Forty-four of the respondents declared that there is a preferred anatomic site for the delivery of enteral feeds during ECLS. Of those reporting a preferred site for delivery of enteral feeds, 18 (41%) reported ‘gastric’ as the preferred site, while 26 (59%) reported ‘post pyloric’ as the preferred site. Thirty-six respondents, 38% of those answering the question, reported that their institution had an established protocol for managing enteral nutrition on ECLS. Fourteen respondents reported a pediatric only protocol, three a neonatal only protocol, and nineteen reported having both pediatric and neonatal protocols. 4. Discussion
3. Results Surveys were sent to 521 individuals from 187 institutions. One hundred twenty-two responses (23.4%) were received representing 96 institutions (51.3%). Of the 122 responses, 115 individuals reported providing neonatal or pediatric ECLS at the 90 centers they represented. Of these centers, 68.8% are located in the United States, 5.6% in the United Kingdom, and the remaining centers come from fifteen other countries. One hundred fourteen individuals continued the survey, and 96 (84.2%) reported providing enteral nutrition (EN), while 18 (15.8%) reported not providing enteral nutrition to their patients on ECLS. As summarized in Table 1, respondents were asked to rate how often their center provides EN to children on ECLS based on ECLS mode, patient diagnosis, level of vasopressor support, and pharmacologic paralysis. Respondents were also asked to rank which of these four categories would be considered the most important when deciding whether or not to provide EN (Table 2).
The results of our survey indicate that the vast majority of ECLS program directors and coordinators at ELSO centers provide EN to neonatal and pediatric patients receiving ECLS. Significant variability appears to exist with regard to the patient and cincal parameters that dictate the decision to initiate EN. ECLS patients are often the most critically ill and may gain significant benefit from EN [14]. However, there have been no rigorous, prospective studies investigating the role of EN on outcome in ECLS patients. The relationship of ECLS and the initiation of EN on gastrointestinal physiology has been explored in patients receiving ECLS. An analysis of a cohort of 16 neonatal VA ECLS patients showed an overall increase in intestinal permeability [10]. The initiation of enteral nutrition to seven patients did not result in any additional changes in intestinal permeability. Therefore, although intestinal integrity in ECLS patients may be compromised, it does not appear to deteriorate with EN. The results of intestinal hormone response (gastrin, cholecystokinin, and peptide-YY) to enteral nutrition in neonates supported on VA ECLS have also been evaluated in twelve patients and compared to twelve
Table 1 Utilization of enteral nutrition based on ECLS mode, diagnosis, vasopressor status, and paralysis. Variable
N (%)
Occasionally (%)
Some (%)
Often (%)
Always (%)
VA ECLS VV ECLS Neonatal diagnoses Meconium aspiration PPHN Respiratory distress syndrome Sepsis Congenital diaphragmatic hernia Congenital cardiac disease Pediatric diagnosis Pneumonia Asthma ARDS Trauma/postoperative Pulmonary hemorrhage Infectious cardiomyopathy Congenital cardiac disease Cardiac arrest Rhabdomyolysis Bone marrow transplant Hematologic transplant Vasopressor agent status Increasing support “Stable” support Weaning support Pharmacologic Paralysis
96 (100) 91 (94.8)
Never (%) 4 (4.1) 2 (2.2)
24 (25) 13 (14.3)
15 (15.6) 11 (12)
37 (38.5) 37 (40.7)
16 (16.7) 28 (30.8)
75 (78.1) 75 (78.1) 80 (83.3) 83 (86.5) 77 (80.2) 85 (88.5)
17 (22.7) 13 (17.3) 12 (24) 17 (20.5) 36 (46.8) 13 (15.3)
14 (18.7) 18 (24) 16 (20) 16 (19.3) 14 (18.2) 24 (28.2)
10 (13.3) 10 (13.3) 11 (13.8) 16 (19.3) 12 (15.6) 19 (22.3)
18 (24) 18 (24) 23 (28.8) 24 (29) 10 (13) 16 (18.8)
16 (21.3) 16 (21.3) 18 (22.5) 10 (12) 5 (6.5) 13 (15.3)
87 (90.6) 73 (76.0) 86 (89.6) 75 (78.1) 81 (84.4) 83 (86.5) 86 (89.6) 84 (87.5) 65 (67.7) 60 (62.5) 53 (55.2)
2 (2.3) 2 (2.7) 2 (2.3) 4 (5.3) 6 (7.4) 5 (6) 9 (10.5) 17 (20.2) 13 (20) 20 (33.3) 17 (32)
15 (17.2) 12 (16.4) 15 (17.4) 11 (14.7) 14 (17.3) 16 (19.3) 20 (23.3) 17 (20.2) 13 (20) 5 (8.3) 5 (9.4)
10 (11.5) 8 (11) 13 (15.1) 19 (25.3) 15 (18.5) 12 (14.5) 19 (22.1) 17 (20.2) 11 (17) 13 (21.7) 11 (20.8)
31 (35.6) 26 (35.6) 28 (32.6) 24 (32) 21 (26) 32 (38.6) 23 (26.7) 21 (25) 13 (20) 12 (20) 10 (18.9)
29 (33.3) 25 (34.2) 28 (32.6) 17 (22.7) 25 (30.9) 18 (21.7) 15 (17.4) 12 (14.3) 15 (23) 10 (16.7) 10 (18.9)
93 (96.9) 92 (95.8) 87 (90.6) 92 (95.8)
39 (42) 8 (8.7) 5 (5.7) 18 (19.6)
17 (18.3) 25 (27.2) 18 (20.7) 19 (20.7)
15 (16.1) 10 (10.9) 12 (14) 13 (14.1)
18 (19.4) 32 (34.8) 32 (36.8) 28 (30.4)
4 (4.3) 17 (18.5) 20 (23) 14 (15.2)
Reports the responses of survey participants who were asked to rate how often their institution provided enteral nutrition for patients on extracorporeal life support. ECLS: extracorporeal life support; VA: venoarterial, VV: venovenous, ECLS: extracorporeal life support, PPHN: persistent pulmonary hypertension of the newborn.
62
T.J. Desmarais et al. / Journal of Pediatric Surgery 50 (2015) 60–63
Table 2 The importance of mode, vasopressors, paralysis, and diagnosis when considering enteral nutrition for patients on extracorporeal life support. Variable
Respondents, N (%)
First (%)
Second (%)
Third (%)
Fourth (%)
ECLS mode Vasopressor requirements Pharmacologic paralysis Underlying diagnosis
89 (92.7) 89 (92.7) 89 (92.7) 92 (95.8)
7 (7.9) 38 (42.7) 9 (10.1) 39 (42.4)
8 (9) 34 (38.2) 18 (20.2) 27 (29.3)
30 (33.7) 13 (14.6) 28 (31.5) 16 (17.4)
44 (19.4) 4 (4.5) 34 (38.2) 10 (10.9)
“First” represents most important. ECLS: extracorporeal life support.
ECLS patients receiving PN and eight patients not receiving ECLS support [17]. This analysis showed that hormone levels were significantly higher in patients receiving enteral nutrition and shown to be equivalent to age matched controls not being supported with ECLS. These studies demonstrate that utilizing EN on ECLS patients does not appear to adversely impact intestinal physiology and should not be used as a reason to withhold EN. Previous studies have demonstrated, feasability, safety, and potential benefits of enteral nutrition in ECLS patients. A comprehensive retrospective review of 29 consecutive pediatric patients supported with ECLS compared those who received EN with those who received PN [7]. The groups were similar in baseline characteristics including ECLS mode – although VV ECLS was more common in both groups. There were no complications associated with the utilization of enteral nutrition. The authors found no difference between the groups in the time needed to achieve caloric goals, a cost savings of $170 per day in the EN group, and a non-statistically significant survival benefit with EN. In a study comparing 16 neonates receiving EN to 35 neonates receiving PN, no differences were found in the rate of septic or other complications [13]. The groups were matched for diagnosis, all patients received VA ECLS, and all but two patients had received vasoactive medications. A five-year, single-center retrospective review of neonates receiving VA ECLS provides additional insight into factors impacting the decision to initiate EN [8]. Thirty-five of 112 neonates with congenital diaphragmatic hernia were not considered for EN. Of 77 patients considered for EN, 10 did not receive EN for hemodynamic instability, a short ECLS run, history of gastric retention, or clinician’s preference. The remaining 67 patients had EN initiated according to a standard feeding protocol. Of note, 87% of patients had a gastric feeding tube while the remainder had a trans-pyloric feeding tube. Most patients received vasopressors and there was no relationship between vasopressor use and discontinuation of EN. No major complications including bilious vomiting, blood-stained stools, or abdominal distention were identified. Additionally, approximately 80% of these patients were receiving EN while concomitantly receiving vasoactive medications after they reached a “stable state.” The authors reported no detectable clinical impact on gut function and no correlation with gastric residuals in these patients. These studies suggest that the routine use of enteral feeding in patients receiving VA ECLS is feasible, well tolerated, and not deleterious. Although not a specific focus of this survey, multiple studies of the use of EN in adult ECLS may add additional perspective to the neonatal and pediatric populations. A single-center retrospective review of 27 patients receiving VV ECLS for acute respiratory failure found that 96% received EN using a unit-based feeding protocol with the liberal use of prokinetic agents (95% by 48 hours) [11]. This cohort had 80% “tolerance” by two days and no serious adverse events (pulmonary aspiration, nosocomial pneumonia, intestinal ischemia, gastroinestinal bleeding, or other gastrointestinal complications). A second single-center retrospective series of 48 patients (35 VA ECLS and 13 VV ECLS) from 2005 to 2007, found that overall, 94% received EN with 69% of patients receving EN as their sole nutritional source [9]. In this group, 71% received prokinetic agents and average nutritional adequacy while on ECLS was 55%. Again, no adverse events were identified. A prospective observational study of 7 adult cardiothoracic VA ECLS patients (all were receiving multiple vasporessor agents) found nutritional adequacy
of 70% by 7 days without major adverse event. Finally, a retrospective, single-center review of 31 VA and 55 VV ECLS patients from 2007 to 2012 found an overall nutritional adequacy of 80% using a standard feeding protocol including prokinetic agents. Of note, patients who received continuous renal replacement therapy in conjunction with ECLS had a later initiation of feeds, longer time to first bowel movement, and more feeding intolerance. The use of paralysis and sedation did not appear to influence feeding tolerance. These studies demonstrate that a focus on early intiation of EN using standardized feeding protocols to address advancing feeds as well as feeding intolerance is safe and well tolerated. Not surprisingly, our survey results demostrate that a patient's underlying diagnosis as well as vasopressor support are important factors in a physician's decision to implement EN. The literature does not support the concept that the presence of EN, vasoactive medications, and ECLS predisposes to an inadequate splanchnic circulation and subsequent gastrointestinal complication such as necrotizing enterocolitis. Simarly, there does not appear to be a relationship between ECLS mode and the presence of pharmacologic paralysis with negative outcomes. Although underlying diagnosis seemed to have the greatest importance in the decision to start EN, there was a wide variablility among individual diagnoses. The most obvious response was the strong disinterest in initiating EN in neonates with congenital diaphragmatic hernia. However, the presence of intrathoracic gastrointestinal conents and the potential for tension physiology in this diagnosis represent a physiologically plausible risk and a legitimate reason for withholding EN. Other diagnoses, particularly in the absense of “clinical stability,” lack a biologically plausible a priori reason for witholding EN. However, since most respondents “always” or “often” provide EN to patients receiving ECLS, the survey results are consistent with current guidelines of providing nutritional support for neonates supported with ECLS that recommend initiating enteral feeds in “stable” patients [14]. There are several limitations to the survey and its interpretation. The overall response to the survey was low and therefore a large number of institutions' practice patterns and biases may not be identified. The high percentage of respondents reporting providing enteral nutrition to patients receiving ECLS may reflect such a responder bias. The survey was sent to program directors and coordinators of ELSO centers and represents their individual practice patterns and their perception of their institutions' practice; their responses may not accurately reflect additional ECLS provider practices. We hypothesize that respondents reporting an institutional feeding protocol for patients receiving ECLS (38% of centers) may suggest a less-than-enthusiastic institutional trend towards initiating EN. Since this was a voluntary survey, there was no opportunity to verify actual nutritional delivery practices. Finally, the nature of the survey did not make it feasible to collect additional important clinical data such as: nutritional adequacy; timing of initiation of EN and rate of increase to goal; barriers to continuous delivery of EN, and important outcome data such as gastrointestinal complications, infections, sepsis, fluid balance, duration of mechanical ventilation, and mortality. This study represents a first step to evaluate center-specific practice and bias toward providing EN to critically ill neonates and children receiving ECLS. A more detailed, multicenter retrospective analysis is needed to describe the relationship of important clinical and nutritional
T.J. Desmarais et al. / Journal of Pediatric Surgery 50 (2015) 60–63
parameters to clinically meaningful outcomes. Such a study could provide information on best practices and inform multi-center, collaborative, prospective trials comparing different nutritional strategies. This is essential for improving morbidity and mortality in this patient population. Acknowledgments Thomas J. Desmarais was supported by a Doris Duke Charitable Research Foundation Grant to Washington University in Saint Louis. The authors wish to acknowledge the Extracorporeal Life Support Organization for their support as well as Mrs. Susan Phillips for her critical review of the manuscript. Appendix A. Nutritional Support in Neonatal and Pediatric Extracorporeal Life Support Survey 1. Institution represented by participant 2. Does your center ever administer enteral nutrition to children on ECLS? [yes/no] (if yes, continue to question 3; if no, go to question 13) 3. Please rate how often your center uses enteral nutrition in the following ECLS modes: a. Veno-arterial (VA) ECLS [never, occasionally, some, often, always] b. Veno-venous (VV) ECLS [never, occasionally, some, often, always] 4. Please rate how often your center utilizes enteral nutrition for the following diagnoses: For each diagnosis- [never, occasionally, some, often, always] a. Neonatal diagnoses i. Meconium aspiration ii. Persistent pulmonary hypertension iii. Respiratory distress syndrome iv. Sepsis v. Congenital diaphragmatic hernia vi. Congenital cardiac disease b. Pediatric diagnoses i. Pneumonia ii. Asthma iii. Acute respiratory distress syndrome iv. Trauma/postoperative v. Pulmonary hemorrhage vi. Infectious cardiomyopathy vii. Congenital cardiac viii. Cardiac arrest ix. Rhabdomyolysis x. Bone marrow transplant xi. Hematologic stem cell transplant 5. Please rate how often your center uses enteral nutrition in the following situations: a. Patient on vasopressor agent(s) i. Requiring increased support [never, occasionally, some, often, always] ii. “Stable” support [never, occasionally, some, often, always] iii. Weaning support [never, occasionally, some, often, always] b. Pharmacologically paralyzed [never, occasionally, some, often, always] 6. Please rank the following variables on their importance when considering initiation of enteral nutrition? [most important to least important] a. ECLS mode b. Vasopressor requirement c. Pharmacologic paralysis d. Underlying diagnosis 7. At your center, is there a factor considered more important than those in question 6 when deciding whether or not to initiated
63
enteral nutrition in a patient? [yes/no] (if yes go to question 8, if no continue to question 9) 8. Please use the space provided to let us know what that factor is. [blank space] 9. At your center, is there a preferred anatomic site for the delivery of enteral feeds? [yes/no] (if yes go to 10; if no go to 11) 10. What is the preferred anatomic site for delivering enteral feeds at your center? a. Gastric b. Post pyloric c. Other (free text response available) 11. Does your center have an established protocol for managing enteral nutrition for pediatric or neonatal patients on ECLS? a. Pediatric only b. Neonatal only c. Both d. Neither 12. Please use the space provided to report any further details that you feel would lead to a better understanding of how and when you use enteral nutrition. [blank space] 13. Would you be interested in participating in a prospective trial comparing enteral versus parenteral feeding in the child supported with ECLS? [yes/no]
References [1] Zwischenberger JB, Bartlett RH, Extracorporeal Life Support Organization. ECMO: extracorporeal cardiopulmonary support in critical care. Ann Arbor, MI: Extracorporeal Life Support Organization; 1995 717. [2] Heyland DK, Dhaliwal R, Drover JW, et al. Canadian clinical practice guidelines for nutrition support in mechanically ventilated, critically ill adult patients. JPEN J Parenter Enteral Nutr 2003;27(5):355–73. [3] de Lucas C, Moreno M, Lopez-Herce J, et al. Transpyloric enteral nutrition reduces the complication rate and cost in the critically ill child. J Pediatr Gastroenterol Nutr 2000;30(2):175–80. [4] Dellinger RP, Levy MM, Carlet JM, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008. Crit Care Med 2008;36(1):296–327. [5] Kreymann KG, Berger MM, Deutz NE, et al. ESPEN guidelines on enteral nutrition: intensive care. Clin Nutr 2006;25(2):210–23. [6] Singer P, Berger MM, Van den Berghe G, et al. ESPEN guidelines on parenteral nutrition: intensive care. Clin Nutr 2009;28(4):387–400. [7] Pettignano R, Heard M, Davis R, et al. Total enteral nutrition versus total parenteral nutrition during pediatric extracorporeal membrane oxygenation. Crit Care Med 1998;26(2):358–63. [8] Hanekamp MN, Spoel M, Sharman-Koendjbiharie I, et al. Routine enteral nutrition in neonates on extracorporeal membrane oxygenation. Pediatr Crit Care Med 2005;6 (3):275–9. [9] Lukas G, Davies AR, Hilton AK, et al. Nutritional support in adult patients receiving extracorporeal membrane oxygenation. Crit Care Resusc 2010;12(4):230–4. [10] Piena M, Albers MJ, Van Haard PM, et al. Introduction of enteral feeding in neonates on extracorporeal membrane oxygenation after evaluation of intestinal permeability changes. J Pediatr Surg 1998;33(1):30–4. [11] Scott LK, Boudreaux K, Thaljeh F, et al. Early enteral feedings in adults receiving venovenous extracorporeal membrane oxygenation. JPEN J Parenter Enteral Nutr 2004;28(5):295–300. [12] Umezawa Makikado LD, Flordelis Lasierra JL, Perez-Vela JL, et al. Early enteral nutrition in adults receiving venoarterial extracorporeal membrane oxygenation: an observational case series. JPEN J Parenter Enteral Nutr 2013;37(2):281–4. [13] Wertheim HF, Albers MJ, Piena-Spoel M, et al. The incidence of septic complications in newborns on extracorporeal membrane oxygenation is not affected by feeding route. J Pediatr Surg 2001;36(10):1485–9. [14] Jaksic T, Hull MA, Modi BP, et al. A.S.P.E.N. clinical guidelines: nutrition support of neonates supported with extracorporeal membrane oxygenation. JPEN J Parenter Enteral Nutr 2010;34(3):247–53. [15] Extracorporeal Life Support Organization. Available from: http://www.elsonet.org/. [16] Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42(2):377–81. [17] Hanekamp MN, Spoel M, Sharman-Koendjbiharie M, et al. Gut hormone profiles in critically ill neonates on extracorporeal membrane oxygenation. J Pediatr Gastroenterol Nutr 2005;40(2):175–9.
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Enteral nutrition in neonatal and pediatric extracorporeal life support: A survey of current practice Thomas J. Desmarais a,b, Yan Yan c, Martin S. Keller a, Adam M. Vogel a,⁎ a b c
Division of Pediatric Surgery, Washington University School of Medicine in Saint Louis, Saint Louis, MI, United States Geisel School of Medicine at Dartmouth, Hanover, NH, United States Department of Surgery, Washington University School of Medicine in Saint Louis, Saint Louis, MI, United States
a r t i c l e
i n f o
Article history: Received 29 September 2014 Accepted 6 October 2014 Key words: Extracorporeal life support ECLS ECMO enteral nutrition critical care
a b s t r a c t Purpose: The purpose of this study was to characterize enteral (EN) nutrition practices in neonatal and pediatric patients receiving extracorporeal life support (ECLS). Methods: A Web-based survey was administered to program directors and coordinators of Extracorporeal Life Support Organization centers providing neonatal and pediatric ECLS. The survey assessed patient and clinical factors relating to the administration of EN. Results: A total of 122 responses (122/521, 23.4%) from 96 institutions (96/187; 51.3%) were received. One hundred fifteen provided neonatal or pediatric ECLS, and 84.2% reported utilizing EN during ECLS. 55% and 71% of respondents provide EN ‘often’ or ‘always’ for venoarterial and venovenous ECLS, respectively. EN was reported as given ‘often’ or ‘always’ by 24% with increased vasopressor support, 53% with “stable” vasopressor support, and 60% with weaning of vasopressor support. Favorable diagnosis for providing EN includes respiratory distress syndrome, pneumonia, asthma, trauma/post-operative, pulmonary hemorrhage, and infectious cardiomyopathy. Vasopressor requirement and underlying diagnosis were the primary or secondary determinant of whether to provide EN 81% and 72% of the time. 38% reported an established protocol for providing EN. Conclusion: EN support is common but not uniform among neonatal and pediatric patients receiving ECLS. ECLS mode, vasopressor status, and underlying diagnosis play an important role in the decision to provide EN. © 2015 Elsevier Inc. All rights reserved.
1. Background Extracorporeal Life Support (ECLS) utilizes a series of established and evolving technologies to deliver life saving treatment to critically ill patients with reversible cardiac and pulmonary failure [1]. Malnutrition in critically ill patients is extremely common and is associated with increased mortality and morbidity including impaired immune function, impaired ventilator drive, prolonged ventilator dependence, and increased infections [2]. Consequently, adequate nutrition is essential to minimize physiologic complications of critical illness and to promote patient recovery. Enteral nutrition (EN) is the preferred method of caloric, protein, and micronutrient delivery and has been shown to reduce sepsis-associated morbidity and cost, improve intestinal immunologic function, and improve nitrogen balance in critically ill patients [3–5]. Alternatively, parenteral nutrition (PN) has been used to deliver daily caloric, protein, and micronutrient requirements [6], but is associated with several complications including intestinal villus hypoplasia, reduction of
intestinal absorptive function, increased bacterial translocation, centralline infections, hyperglycemia, and cholestasis [7]. Delivery of EN may be avoided because of concerns regarding inadequate intestinal perfusion and non-occlusive mesenteric ischemia with the development of necrotizing enterocolitis, intestinal ischemia, perforation, or gastrointestinal hemorrhage. Despite these concerns, hypoperfusion related intestinal complications while on ECLS remain an unproven risk. Several small studies have documented the feasibility and safety of EN in ECLS patients in the neonatal, pediatric, and adult populations [7–13]. Additionally, current guidelines for providing nutritional support of neonates simultaneously supported with ECLS recommend initiating enteral feeds when patients are clinically “stable” [14]. The optimum route for delivery of nutrition in neonates and children receiving ECLS is not well established and practice patterns have not been described. The goal of this study is to characterize current practice patterns of the administration and delivery of EN at centers treating neonatal and pediatric ECLS patients. 2. Methods
⁎ Corresponding author at: Division of Pediatric Surgery, Washington University School of Medicine in St. Louis, One Children's Place, Suite 5S40, Saint Louis, MO 63110, United States. Tel.: +1 314 454 6022; fax: +1 314 454 2442. E-mail address: [email protected] (A.M. Vogel). http://dx.doi.org/10.1016/j.jpedsurg.2014.10.030 0022-3468/© 2015 Elsevier Inc. All rights reserved.
The Washington University in St. Louis School of Medicine Institutional Review Board (#201302094) approved this study. The Extracorporeal Life Support Organization (ELSO) is an international consortium of health care professions and scientists who are dedicated to the
T.J. Desmarais et al. / Journal of Pediatric Surgery 50 (2015) 60–63
development and evaluation of novel therapies for support of failing organ systems [15]. Contact information for ECLS program directors and coordinators was obtained from participating centers identified on the ELSO Web site. The survey was administered through the REDCap electronic data capture tool hosted at Washington University in St Louis [16]. A link to the survey was sent electronically on June 4, 2013, with reminders sent on June 17, 2013, and again on July 8, 2013. The thirteen question Web-based survey was designed to assess nutritional implementation and delivery practices in neonatal and pediatric patients on ECLS. The survey (see Appendix A) sought to determine factors that might influence EN implementation such as ECLS mode, patient diagnosis, vasopressor support, and pharmacologic paralysis as well as administration preferences (gastric vs. post-pyloric and institutional unit-based “feeding” protocols). The data were compiled anonymously and analyzed as a composite. Responses are reported as ranges for continuous data and percentages for categorical data. Write-in answers are reported as direct quotes or grouped by theme.
61
Twenty-five respondents reported that there was at least one factor considered more important at their institution than the four factors discussed above. In the free response space provided, the most commonly listed factors included bowel functionality [14], severity of illness [2], lactic acidosis [2], presence of central cannulation [2], and cardiac arrest [2]. Other reasons included: “presence of feeding tube prior to heparinization” and “unit preference.” Forty-four of the respondents declared that there is a preferred anatomic site for the delivery of enteral feeds during ECLS. Of those reporting a preferred site for delivery of enteral feeds, 18 (41%) reported ‘gastric’ as the preferred site, while 26 (59%) reported ‘post pyloric’ as the preferred site. Thirty-six respondents, 38% of those answering the question, reported that their institution had an established protocol for managing enteral nutrition on ECLS. Fourteen respondents reported a pediatric only protocol, three a neonatal only protocol, and nineteen reported having both pediatric and neonatal protocols. 4. Discussion
3. Results Surveys were sent to 521 individuals from 187 institutions. One hundred twenty-two responses (23.4%) were received representing 96 institutions (51.3%). Of the 122 responses, 115 individuals reported providing neonatal or pediatric ECLS at the 90 centers they represented. Of these centers, 68.8% are located in the United States, 5.6% in the United Kingdom, and the remaining centers come from fifteen other countries. One hundred fourteen individuals continued the survey, and 96 (84.2%) reported providing enteral nutrition (EN), while 18 (15.8%) reported not providing enteral nutrition to their patients on ECLS. As summarized in Table 1, respondents were asked to rate how often their center provides EN to children on ECLS based on ECLS mode, patient diagnosis, level of vasopressor support, and pharmacologic paralysis. Respondents were also asked to rank which of these four categories would be considered the most important when deciding whether or not to provide EN (Table 2).
The results of our survey indicate that the vast majority of ECLS program directors and coordinators at ELSO centers provide EN to neonatal and pediatric patients receiving ECLS. Significant variability appears to exist with regard to the patient and cincal parameters that dictate the decision to initiate EN. ECLS patients are often the most critically ill and may gain significant benefit from EN [14]. However, there have been no rigorous, prospective studies investigating the role of EN on outcome in ECLS patients. The relationship of ECLS and the initiation of EN on gastrointestinal physiology has been explored in patients receiving ECLS. An analysis of a cohort of 16 neonatal VA ECLS patients showed an overall increase in intestinal permeability [10]. The initiation of enteral nutrition to seven patients did not result in any additional changes in intestinal permeability. Therefore, although intestinal integrity in ECLS patients may be compromised, it does not appear to deteriorate with EN. The results of intestinal hormone response (gastrin, cholecystokinin, and peptide-YY) to enteral nutrition in neonates supported on VA ECLS have also been evaluated in twelve patients and compared to twelve
Table 1 Utilization of enteral nutrition based on ECLS mode, diagnosis, vasopressor status, and paralysis. Variable
N (%)
Occasionally (%)
Some (%)
Often (%)
Always (%)
VA ECLS VV ECLS Neonatal diagnoses Meconium aspiration PPHN Respiratory distress syndrome Sepsis Congenital diaphragmatic hernia Congenital cardiac disease Pediatric diagnosis Pneumonia Asthma ARDS Trauma/postoperative Pulmonary hemorrhage Infectious cardiomyopathy Congenital cardiac disease Cardiac arrest Rhabdomyolysis Bone marrow transplant Hematologic transplant Vasopressor agent status Increasing support “Stable” support Weaning support Pharmacologic Paralysis
96 (100) 91 (94.8)
Never (%) 4 (4.1) 2 (2.2)
24 (25) 13 (14.3)
15 (15.6) 11 (12)
37 (38.5) 37 (40.7)
16 (16.7) 28 (30.8)
75 (78.1) 75 (78.1) 80 (83.3) 83 (86.5) 77 (80.2) 85 (88.5)
17 (22.7) 13 (17.3) 12 (24) 17 (20.5) 36 (46.8) 13 (15.3)
14 (18.7) 18 (24) 16 (20) 16 (19.3) 14 (18.2) 24 (28.2)
10 (13.3) 10 (13.3) 11 (13.8) 16 (19.3) 12 (15.6) 19 (22.3)
18 (24) 18 (24) 23 (28.8) 24 (29) 10 (13) 16 (18.8)
16 (21.3) 16 (21.3) 18 (22.5) 10 (12) 5 (6.5) 13 (15.3)
87 (90.6) 73 (76.0) 86 (89.6) 75 (78.1) 81 (84.4) 83 (86.5) 86 (89.6) 84 (87.5) 65 (67.7) 60 (62.5) 53 (55.2)
2 (2.3) 2 (2.7) 2 (2.3) 4 (5.3) 6 (7.4) 5 (6) 9 (10.5) 17 (20.2) 13 (20) 20 (33.3) 17 (32)
15 (17.2) 12 (16.4) 15 (17.4) 11 (14.7) 14 (17.3) 16 (19.3) 20 (23.3) 17 (20.2) 13 (20) 5 (8.3) 5 (9.4)
10 (11.5) 8 (11) 13 (15.1) 19 (25.3) 15 (18.5) 12 (14.5) 19 (22.1) 17 (20.2) 11 (17) 13 (21.7) 11 (20.8)
31 (35.6) 26 (35.6) 28 (32.6) 24 (32) 21 (26) 32 (38.6) 23 (26.7) 21 (25) 13 (20) 12 (20) 10 (18.9)
29 (33.3) 25 (34.2) 28 (32.6) 17 (22.7) 25 (30.9) 18 (21.7) 15 (17.4) 12 (14.3) 15 (23) 10 (16.7) 10 (18.9)
93 (96.9) 92 (95.8) 87 (90.6) 92 (95.8)
39 (42) 8 (8.7) 5 (5.7) 18 (19.6)
17 (18.3) 25 (27.2) 18 (20.7) 19 (20.7)
15 (16.1) 10 (10.9) 12 (14) 13 (14.1)
18 (19.4) 32 (34.8) 32 (36.8) 28 (30.4)
4 (4.3) 17 (18.5) 20 (23) 14 (15.2)
Reports the responses of survey participants who were asked to rate how often their institution provided enteral nutrition for patients on extracorporeal life support. ECLS: extracorporeal life support; VA: venoarterial, VV: venovenous, ECLS: extracorporeal life support, PPHN: persistent pulmonary hypertension of the newborn.
62
T.J. Desmarais et al. / Journal of Pediatric Surgery 50 (2015) 60–63
Table 2 The importance of mode, vasopressors, paralysis, and diagnosis when considering enteral nutrition for patients on extracorporeal life support. Variable
Respondents, N (%)
First (%)
Second (%)
Third (%)
Fourth (%)
ECLS mode Vasopressor requirements Pharmacologic paralysis Underlying diagnosis
89 (92.7) 89 (92.7) 89 (92.7) 92 (95.8)
7 (7.9) 38 (42.7) 9 (10.1) 39 (42.4)
8 (9) 34 (38.2) 18 (20.2) 27 (29.3)
30 (33.7) 13 (14.6) 28 (31.5) 16 (17.4)
44 (19.4) 4 (4.5) 34 (38.2) 10 (10.9)
“First” represents most important. ECLS: extracorporeal life support.
ECLS patients receiving PN and eight patients not receiving ECLS support [17]. This analysis showed that hormone levels were significantly higher in patients receiving enteral nutrition and shown to be equivalent to age matched controls not being supported with ECLS. These studies demonstrate that utilizing EN on ECLS patients does not appear to adversely impact intestinal physiology and should not be used as a reason to withhold EN. Previous studies have demonstrated, feasability, safety, and potential benefits of enteral nutrition in ECLS patients. A comprehensive retrospective review of 29 consecutive pediatric patients supported with ECLS compared those who received EN with those who received PN [7]. The groups were similar in baseline characteristics including ECLS mode – although VV ECLS was more common in both groups. There were no complications associated with the utilization of enteral nutrition. The authors found no difference between the groups in the time needed to achieve caloric goals, a cost savings of $170 per day in the EN group, and a non-statistically significant survival benefit with EN. In a study comparing 16 neonates receiving EN to 35 neonates receiving PN, no differences were found in the rate of septic or other complications [13]. The groups were matched for diagnosis, all patients received VA ECLS, and all but two patients had received vasoactive medications. A five-year, single-center retrospective review of neonates receiving VA ECLS provides additional insight into factors impacting the decision to initiate EN [8]. Thirty-five of 112 neonates with congenital diaphragmatic hernia were not considered for EN. Of 77 patients considered for EN, 10 did not receive EN for hemodynamic instability, a short ECLS run, history of gastric retention, or clinician’s preference. The remaining 67 patients had EN initiated according to a standard feeding protocol. Of note, 87% of patients had a gastric feeding tube while the remainder had a trans-pyloric feeding tube. Most patients received vasopressors and there was no relationship between vasopressor use and discontinuation of EN. No major complications including bilious vomiting, blood-stained stools, or abdominal distention were identified. Additionally, approximately 80% of these patients were receiving EN while concomitantly receiving vasoactive medications after they reached a “stable state.” The authors reported no detectable clinical impact on gut function and no correlation with gastric residuals in these patients. These studies suggest that the routine use of enteral feeding in patients receiving VA ECLS is feasible, well tolerated, and not deleterious. Although not a specific focus of this survey, multiple studies of the use of EN in adult ECLS may add additional perspective to the neonatal and pediatric populations. A single-center retrospective review of 27 patients receiving VV ECLS for acute respiratory failure found that 96% received EN using a unit-based feeding protocol with the liberal use of prokinetic agents (95% by 48 hours) [11]. This cohort had 80% “tolerance” by two days and no serious adverse events (pulmonary aspiration, nosocomial pneumonia, intestinal ischemia, gastroinestinal bleeding, or other gastrointestinal complications). A second single-center retrospective series of 48 patients (35 VA ECLS and 13 VV ECLS) from 2005 to 2007, found that overall, 94% received EN with 69% of patients receving EN as their sole nutritional source [9]. In this group, 71% received prokinetic agents and average nutritional adequacy while on ECLS was 55%. Again, no adverse events were identified. A prospective observational study of 7 adult cardiothoracic VA ECLS patients (all were receiving multiple vasporessor agents) found nutritional adequacy
of 70% by 7 days without major adverse event. Finally, a retrospective, single-center review of 31 VA and 55 VV ECLS patients from 2007 to 2012 found an overall nutritional adequacy of 80% using a standard feeding protocol including prokinetic agents. Of note, patients who received continuous renal replacement therapy in conjunction with ECLS had a later initiation of feeds, longer time to first bowel movement, and more feeding intolerance. The use of paralysis and sedation did not appear to influence feeding tolerance. These studies demonstrate that a focus on early intiation of EN using standardized feeding protocols to address advancing feeds as well as feeding intolerance is safe and well tolerated. Not surprisingly, our survey results demostrate that a patient's underlying diagnosis as well as vasopressor support are important factors in a physician's decision to implement EN. The literature does not support the concept that the presence of EN, vasoactive medications, and ECLS predisposes to an inadequate splanchnic circulation and subsequent gastrointestinal complication such as necrotizing enterocolitis. Simarly, there does not appear to be a relationship between ECLS mode and the presence of pharmacologic paralysis with negative outcomes. Although underlying diagnosis seemed to have the greatest importance in the decision to start EN, there was a wide variablility among individual diagnoses. The most obvious response was the strong disinterest in initiating EN in neonates with congenital diaphragmatic hernia. However, the presence of intrathoracic gastrointestinal conents and the potential for tension physiology in this diagnosis represent a physiologically plausible risk and a legitimate reason for withholding EN. Other diagnoses, particularly in the absense of “clinical stability,” lack a biologically plausible a priori reason for witholding EN. However, since most respondents “always” or “often” provide EN to patients receiving ECLS, the survey results are consistent with current guidelines of providing nutritional support for neonates supported with ECLS that recommend initiating enteral feeds in “stable” patients [14]. There are several limitations to the survey and its interpretation. The overall response to the survey was low and therefore a large number of institutions' practice patterns and biases may not be identified. The high percentage of respondents reporting providing enteral nutrition to patients receiving ECLS may reflect such a responder bias. The survey was sent to program directors and coordinators of ELSO centers and represents their individual practice patterns and their perception of their institutions' practice; their responses may not accurately reflect additional ECLS provider practices. We hypothesize that respondents reporting an institutional feeding protocol for patients receiving ECLS (38% of centers) may suggest a less-than-enthusiastic institutional trend towards initiating EN. Since this was a voluntary survey, there was no opportunity to verify actual nutritional delivery practices. Finally, the nature of the survey did not make it feasible to collect additional important clinical data such as: nutritional adequacy; timing of initiation of EN and rate of increase to goal; barriers to continuous delivery of EN, and important outcome data such as gastrointestinal complications, infections, sepsis, fluid balance, duration of mechanical ventilation, and mortality. This study represents a first step to evaluate center-specific practice and bias toward providing EN to critically ill neonates and children receiving ECLS. A more detailed, multicenter retrospective analysis is needed to describe the relationship of important clinical and nutritional
T.J. Desmarais et al. / Journal of Pediatric Surgery 50 (2015) 60–63
parameters to clinically meaningful outcomes. Such a study could provide information on best practices and inform multi-center, collaborative, prospective trials comparing different nutritional strategies. This is essential for improving morbidity and mortality in this patient population. Acknowledgments Thomas J. Desmarais was supported by a Doris Duke Charitable Research Foundation Grant to Washington University in Saint Louis. The authors wish to acknowledge the Extracorporeal Life Support Organization for their support as well as Mrs. Susan Phillips for her critical review of the manuscript. Appendix A. Nutritional Support in Neonatal and Pediatric Extracorporeal Life Support Survey 1. Institution represented by participant 2. Does your center ever administer enteral nutrition to children on ECLS? [yes/no] (if yes, continue to question 3; if no, go to question 13) 3. Please rate how often your center uses enteral nutrition in the following ECLS modes: a. Veno-arterial (VA) ECLS [never, occasionally, some, often, always] b. Veno-venous (VV) ECLS [never, occasionally, some, often, always] 4. Please rate how often your center utilizes enteral nutrition for the following diagnoses: For each diagnosis- [never, occasionally, some, often, always] a. Neonatal diagnoses i. Meconium aspiration ii. Persistent pulmonary hypertension iii. Respiratory distress syndrome iv. Sepsis v. Congenital diaphragmatic hernia vi. Congenital cardiac disease b. Pediatric diagnoses i. Pneumonia ii. Asthma iii. Acute respiratory distress syndrome iv. Trauma/postoperative v. Pulmonary hemorrhage vi. Infectious cardiomyopathy vii. Congenital cardiac viii. Cardiac arrest ix. Rhabdomyolysis x. Bone marrow transplant xi. Hematologic stem cell transplant 5. Please rate how often your center uses enteral nutrition in the following situations: a. Patient on vasopressor agent(s) i. Requiring increased support [never, occasionally, some, often, always] ii. “Stable” support [never, occasionally, some, often, always] iii. Weaning support [never, occasionally, some, often, always] b. Pharmacologically paralyzed [never, occasionally, some, often, always] 6. Please rank the following variables on their importance when considering initiation of enteral nutrition? [most important to least important] a. ECLS mode b. Vasopressor requirement c. Pharmacologic paralysis d. Underlying diagnosis 7. At your center, is there a factor considered more important than those in question 6 when deciding whether or not to initiated
63
enteral nutrition in a patient? [yes/no] (if yes go to question 8, if no continue to question 9) 8. Please use the space provided to let us know what that factor is. [blank space] 9. At your center, is there a preferred anatomic site for the delivery of enteral feeds? [yes/no] (if yes go to 10; if no go to 11) 10. What is the preferred anatomic site for delivering enteral feeds at your center? a. Gastric b. Post pyloric c. Other (free text response available) 11. Does your center have an established protocol for managing enteral nutrition for pediatric or neonatal patients on ECLS? a. Pediatric only b. Neonatal only c. Both d. Neither 12. Please use the space provided to report any further details that you feel would lead to a better understanding of how and when you use enteral nutrition. [blank space] 13. Would you be interested in participating in a prospective trial comparing enteral versus parenteral feeding in the child supported with ECLS? [yes/no]
References [1] Zwischenberger JB, Bartlett RH, Extracorporeal Life Support Organization. ECMO: extracorporeal cardiopulmonary support in critical care. Ann Arbor, MI: Extracorporeal Life Support Organization; 1995 717. [2] Heyland DK, Dhaliwal R, Drover JW, et al. Canadian clinical practice guidelines for nutrition support in mechanically ventilated, critically ill adult patients. JPEN J Parenter Enteral Nutr 2003;27(5):355–73. [3] de Lucas C, Moreno M, Lopez-Herce J, et al. Transpyloric enteral nutrition reduces the complication rate and cost in the critically ill child. J Pediatr Gastroenterol Nutr 2000;30(2):175–80. [4] Dellinger RP, Levy MM, Carlet JM, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008. Crit Care Med 2008;36(1):296–327. [5] Kreymann KG, Berger MM, Deutz NE, et al. ESPEN guidelines on enteral nutrition: intensive care. Clin Nutr 2006;25(2):210–23. [6] Singer P, Berger MM, Van den Berghe G, et al. ESPEN guidelines on parenteral nutrition: intensive care. Clin Nutr 2009;28(4):387–400. [7] Pettignano R, Heard M, Davis R, et al. Total enteral nutrition versus total parenteral nutrition during pediatric extracorporeal membrane oxygenation. Crit Care Med 1998;26(2):358–63. [8] Hanekamp MN, Spoel M, Sharman-Koendjbiharie I, et al. Routine enteral nutrition in neonates on extracorporeal membrane oxygenation. Pediatr Crit Care Med 2005;6 (3):275–9. [9] Lukas G, Davies AR, Hilton AK, et al. Nutritional support in adult patients receiving extracorporeal membrane oxygenation. Crit Care Resusc 2010;12(4):230–4. [10] Piena M, Albers MJ, Van Haard PM, et al. Introduction of enteral feeding in neonates on extracorporeal membrane oxygenation after evaluation of intestinal permeability changes. J Pediatr Surg 1998;33(1):30–4. [11] Scott LK, Boudreaux K, Thaljeh F, et al. Early enteral feedings in adults receiving venovenous extracorporeal membrane oxygenation. JPEN J Parenter Enteral Nutr 2004;28(5):295–300. [12] Umezawa Makikado LD, Flordelis Lasierra JL, Perez-Vela JL, et al. Early enteral nutrition in adults receiving venoarterial extracorporeal membrane oxygenation: an observational case series. JPEN J Parenter Enteral Nutr 2013;37(2):281–4. [13] Wertheim HF, Albers MJ, Piena-Spoel M, et al. The incidence of septic complications in newborns on extracorporeal membrane oxygenation is not affected by feeding route. J Pediatr Surg 2001;36(10):1485–9. [14] Jaksic T, Hull MA, Modi BP, et al. A.S.P.E.N. clinical guidelines: nutrition support of neonates supported with extracorporeal membrane oxygenation. JPEN J Parenter Enteral Nutr 2010;34(3):247–53. [15] Extracorporeal Life Support Organization. Available from: http://www.elsonet.org/. [16] Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42(2):377–81. [17] Hanekamp MN, Spoel M, Sharman-Koendjbiharie M, et al. Gut hormone profiles in critically ill neonates on extracorporeal membrane oxygenation. J Pediatr Gastroenterol Nutr 2005;40(2):175–9.
