Necrotizing enterocolitis (NEC) is the most common cause of gastrointestinal emergency in premature infants, with a mortality rate of 20% to 30%.1 Disease pathology is thought to be multifactorial; however, it appears to involve the interplay of inflammation, bacterial translocation, and ischemia, which ultimately results in bowel necrosis.2 Determination of NEC onset is problematic due to the broad spectrum of clinical symptoms, from subtle signs such as feeding intolerance and mild abdominal distention, to complete cardiovascular collapse and shock.3 Enteral feeding, specific to type administered (human milk vs cow's milk) combined with underdeveloped intestinal function, and abnormal colonization of the gut microbiome, may increase the likelihood for NEC when immature intestinal mucosal surfaces become compromised, allowing for bacterial invasion.4 Studies have further suggested that inflammatory responses in the premature infant may be either inadequate or overexaggerated causing a mediator imbalance and altered intestinal blood flow.5 Over the past decade, a number of studies suggest that packed red blood cell (PRBC) transfusions administered to the significantly anemic preterm neonate may temporally contribute to a NEC subtype,6–9 although this phenomenon remains highly controversial. Patel and colleagues10 showed that, while the act of transfusion itself may not lead to NEC, the degree and severity of anemia do correlate with what has become known as transfusion-related NEC (TR-NEC) in the literature.
Autoregulatory processes in the premature infant's cerebral, renal, and mesenteric tissue beds are not complete. Loss of autoregulation may impair oxygen delivery in the anemic patient, increasing the risk for ischemic injury.11 Near-infrared spectroscopy (NIRS) measures a regional oxygenation saturation (rSO2) in a tissue bed, reflecting the difference between oxygenated and deoxygenated hemoglobin.12 Using NIRS technology, Fortune and colleagues13 suggest cerebral to splanchnic oxygenation ratios (CSORs) of less than 0.75 indicate infants who are at an increased risk for NEC development, as values below 0.75 may indicate suboptimal oxygen delivery and/or consumption within the splanchnic tissue bed. CSOR values (calculated as splanchnic/cerebral rSO2 values) were based on intact cerebral autoregulation (CAR) with decline in splanchnic (or mesenteric) readings. However, studies suggest that CAR, defined as maintenance of adequate blood flow during states of arterial pressure fluctuation, is lost in sick preterm infants during the first week of life,14 , 15 and may also become impaired with hypercapnia15 , 16 or NEC development.17
We present an extremely low birth-weight (ELBW) infant who was enrolled in a large prospective, observational study to evaluate tissue oxygen extraction trends measured by NIRS technology in cerebral, renal, and mesenteric tissue beds before, during, and 48 hours following PRBC transfusion. During the study period, the infant developed Bell's stage IIIA NEC18 complicated by pneumoperitoneum.
PRENATAL AND BIRTH HISTORY
Baby M was a 705-g, 24 weeks' gestation Caucasian female infant born to a 25-year-old gravida 3, para 2 mother. Prenatal maternal screens for rubella, hepatitis B, human immunodeficiency virus, and syphilis rapid plasma reagent were negative. Group beta streptococcus status was not obtained due to preterm delivery. She was born via emergency Cesarean delivery under maternal general anesthesia due to preterm labor and transverse lie. Membranes were ruptured at delivery, and amniotic fluid was clear and of normal volume. Following birth, Baby M was assigned Apgar scores of 3 at 1 minute, 5 at 5 minutes, and 4 at 10 minutes. In the delivery room, she was intubated during resuscitation for minimal respiratory effort, and one dose of exogenous surfactant was administered via endotracheal tube. Because she was born at a level II facility, she was prepared for transport to a tertiary care facility. Prior to transport, she received 2 normal saline boluses of 10 mL/kg and 1 dose of sodium bicarbonate. In addition, a dopamine infusion (12 μg/kg/min) was started for low mean arterial blood pressure as measured via an umbilical artery catheter. Her laboratory reports prior to transport included hematocrit 44% (hemoglobin 14.7 g/dL), platelets 232,000, white blood cells 3500, segmented neutrophils 26%, and bands 1%. She remained on conventional ventilation for transport with FiO2 0.80, ventilatory rate 40, positive end-expiratory pressure (PEEP) +5 cm H2O, and positive inspiratory pressure (PIP) 15 cm H2O, with arterial blood gas pH 7.07, pCO2 50, pO2 170, and base deficit 16.9. Ampicillin and gentamicin were administered after obtaining blood cultures and complete blood count with differential. Ground transport lasted approximately 1 hour and 40 minutes. Upon arrival to the receiving level III neonatal intensive care unit (NICU), Baby M was continued on conventional ventilation, and phototherapy was initiated for extreme bruising.
HOSPITALIZATION COURSE PRIOR TO STUDY ENROLLMENT
Prior to and throughout the study, Baby M remained on conventional ventilation. Dopamine was weaned off by 48 hours after birth. Cranial ultrasound at 72 hours was negative for intraventricular hemorrhage. Caffeine citrate was begun on day of life 6 in an attempt to wean from ventilator support. Trophic feedings of maternal breast milk were started on day of life 8. Ampicillin and gentamicin were given through day of life 7, although initial central line blood cultures from birth were negative. A peripheral blood culture was obtained on day of life 10 due to significant desaturations and metabolic acidosis. Within 24 hours, Serratia marcescens bacteremia was confirmed and intravenous meropenem every 12 hours was initiated. At time of enrollment (day of life 11), there were no further signs of sepsis, and she was receiving human milk enteral feeds without evidence of intolerance. At this time, her hematocrit was 28% (hemoglobin 9.5 g/dL), and the decision was made by the attending neonatologist to administer a PRBC transfusion of 15 mL/kg to infuse over 4 hours, prompting study enrollment. Prior to initiation of transfusion, the infant was tolerating 5 mL of nonfortified maternal breast milk over 1 hour via orogastric tube every 3 hours (56 mL/kg/d). Enteral feedings were held for transfusion. Arterial blood gas 8 hours prior to first transfusion showed compensated metabolic acidosis and diminished oxygenation (pH, 7.4; pCO2, 29; pO2, 44; bicarbonate, 17; and base deficit, −6.5). At this time conventional ventilation settings were assist-control with backup rate of 40 breaths per minute, FiO2 0.35, PIP 21 cm H2O, and PEEP of 6 cm H2O. At time of enrollment, her vital signs were stable, with heart rate 162 beats per minute, respiratory rate 62 per minute, mean arterial blood pressure 31 mmHg, and oxygen saturation 95%. No episodes of bradycardia or desaturations occurred in the past 24 hours prior to enrollment.
DATA COLLECTION PROCEDURES
Parental consent for the large prospective study and blood administration were obtained, and Baby M was placed on NIRS to analyze cerebral, mesenteric, and renal tissue oxygenation expressed as regional oxygen saturation (rSO2) values. We measured rSO2 values using a Food and Drug Administration–approved NIRS somatic oximeter (INVOS 5100C; Medtronic, Boulder, Colorado), which reflects the amount of oxygen delivered minus oxygen consumed at the tissue level.19 This specific device records rSO2 values, ranging from 15% to 95%, which reflects total oxygen bound to hemoglobin. NIRS probes were placed on the infant's infraumbilical region for mesenteric monitoring, center of forehead for cerebral monitoring, and left flank for renal monitoring. Data were electronically recorded every 30 seconds in real time during the study period. The decision to continue or hold feedings during PRBC transfusion was made by the attending neonatologist, independent of this study. All routine nursing and respiratory care was continued without interruption during data collection. Medical and nursing personnel were blinded to NIRS monitors, and NIRS machine alarms were silenced during the data collection period.
Baby M weighed 710 g on day 1 of study, and over the monitoring period, she received 2 full-volume (15 mL/kg) PRBC transfusions each administered over 4 hours, and each transfusion was separated by 67 hours. The decision to administer 2 full-volume transfusions was made by the attending neonatologist independent of this study. Baseline rSO2 means were calculated for the 22-minute period immediately prior to the first full-volume transfusion: mesenteric rSO2 mean = 63%, cerebral rSO2 mean = 63%, and renal mean = 26% (see Figure 1). For purposes of data description, means were averaged from raw rSO2 data in 30-minute epochs. Mesenteric oxygenation slowly declined, and within 2 hours into receiving the first transfusion, values dropped to 16% from 63%, with frequent signal dropout. During this same period, cerebral rSO2 values remained steady, ranging 59% to 63%, and renal oxygenation steadily rose from mid-20% to mid-30%. CSOR values prior to transfusion ranged from 0.79 to 0.95. For the 30-minute period immediately following transfusion completion, mesenteric 30-minute epoch mean = 15%, cerebral = 64%, and renal = 39%, with CSOR = 0.23. During the decline in mesenteric rSO2 values, SpO2 values ranged 90% to 98%, heart rate 162 to 166 beats per minute, and mean arterial blood pressure 33 to 45 mm Hg. She had no abdominal distention or discoloration, no emesis, and no signs of hemodynamic instability, and remained on conventional ventilation with FiO2 0.36, rate 40 per minute, PIP 20 cm H2O, PEEP + 6 cm H2O.
Maternal breast milk enteral feedings (56 mL/kg/d) via orogastric tube were reinitiated 4 hours following completion of the first transfusion. Mesenteric epoch means during the subsequent 24 hours ranged from 15% to 38%, with no abdominal distention or feeding intolerance. Mesenteric signal dropout began at 25 hours following transfusion completion, and sensor probes were replaced. At this time, she developed abdominal distention and a small amount of green-colored gastric content was aspirated. Feedings were discontinued. Mesenteric epoch means over the following 48 hours remained 15% to 32%.
A second full-volume 15-mL/kg PRBC transfusion was started 67 hours following completion of the first transfusion. At this time, she remained on conventional ventilation with the following settings: FiO2 0.28, rate 37 per minute, PIP 28 cm H2O, and PEEP + 6 cm H2O. Nothing by mouth (NPO) status continued during her second transfusion. Mesenteric rSO2 readings were intermittent and ranged from 15% to 30%, cerebral epoch means ranged 29% to 67%, and CSOR values ranged 0.21 to 0.50. Hemodynamic status remained stable, with SpO2 values 94% to 97%, mean arterial blood pressure 40 to 77 mm Hg, and heart rate 147. Arterial blood gases prior to the second blood transfusion showed compensated metabolic acidosis and diminished oxygenation (pH, 7.39; pCO2, 31; pO2, 43; bicarbonate, 18; and base deficit, −6.1). Over the next 36 hours, mesenteric readings continued to be very intermittent, ranging between 15% and 86%, cerebral epoch means ranged 16% to 90%, and renal epoch means ranged 29% to 51%. Approximately 38 hours following completion of the second transfusion, an abdominal radiograph was obtained for persistent abdominal distention, and pneumoperitoneum was identified. Peritoneal drains were placed, and mesenteric epoch means immediately rose to 77%, with cerebral epoch means 72% and renal epoch means 69% (Figure 2). CSOR value after drain placement dramatically increased (0.88-1.0).
HOSPITALIZATION COURSE FOLLOWING STUDY
Following peritoneal drain placement on day of life 17, fluconazole was added to antimicrobial therapy due to thrombocytopenia (platelet count = 34,000). An exploratory laparotomy was performed on day of life 27 for worsening abdominal distention. Intestinal adhesions to the liver were found, 12 cm of necrotic small bowel was removed, and an ileostomy with Hartman's pouch was created. Baby M remained nothing by mouth, with intermittent gastric suction continued until day of life 41. Trophic maternal breast milk enteral feedings were reinitiated on day of life 42. She continued to progress to full feedings over the next 2 weeks, and was eventually placed on human milk-fortified (human milk fortifier, 22 cal/oz) feedings via orogastric tube. However, she developed mild complications with dumping syndrome and slow weight gain. Therefore, her feedings were changed to a 50/50 mixture of 20 cal/oz of human milk and 20 cal/oz of Pregestimil. No further problems were encountered due to intestinal issues, and she was transferred back to birth hospital with an ileostomy and tolerating full orogastric feedings on day of life 66.
This case report describes events of an 11-day-old ELBW female infant with Serratia marcescens bacteremia who developed modified Bell's stage IIIA NEC.17 Previous studies have identified PRBC transfusions in anemic infants as an independent risk factor for the development of NEC.8 , 9 Although the pathogenesis is most likely multifactorial, most studies suggest transfusion-related NEC (TR-NEC) temporally occurs within 48 hours of receiving a PRBC transfusion.7 The infant presented in this case, Baby M, received 2 (15 mL/kg) PRBC transfusions over a period of 67 hours. During the monitoring period, she exhibited persistently diminished mesenteric rSO2 values that began during the first transfusion and continued for 6 days until intestinal perforation was diagnosed 38 hours following completion of the second full-volume PRBC transfusion. The most striking finding was that, during the period of initial decreased mesenteric oxygenation demonstrated by NIRS, she did not exhibit physiologic signs of hemodynamic instability and did not develop feeding intolerance until 25 hours following completion of the first transfusion.
Acceptable reference rSO2 ranges for stable premature infants are reported as 66% to 83% in cerebral beds, 64% to 87% in renal beds, and 32% to 66% in mesenteric beds.20 However, mesenteric readings are highly variable compared with cerebral and renal rSO2 trends. Mintzer and colleagues21 found that, during calm states, mesenteric values were associated with large variance (16% from baseline), while renal had intermediate (6% from baseline) and cerebral had minimal variance (3% from baseline). To minimize variance, these investigators suggest that averaging time data into longer epochs reduced variability. They further suggest that optimal baseline data should be gathered in ≤15-minute epochs to produce the least amount of variability while still being able to detect shorter true physiologic declines in tissue oxygentation.21 Patel and colleagues22 suggest infants who develop NEC have increased rSO2 variance, but lower overall measurements compared with those without disease, concluding intestinal dysfunction and ischemia are key components in NEC development.
Zabaneh and colleagues23 described rSO2 differences in premature twins with and without bowel ischemia, finding persistently lower rSO2 mesenteric readings and decreased CSOR values in the twin with surgical NEC compared with higher readings in the twin without disease. Cortez and colleagues24 reported similar findings in NEC rSO2 trends among twins. In their study, twin ELBW infants exhibited “high” rSO2 baselines, and then developed significant decline to 15% (±3) with frequent signal dropout of 90% or more. Investigators conclude that this specific trend may be due to increased intra-abdominal free air; yet further state that the combination of signal dropout and persistently low rSO2 readings was more likely due to severely impaired mesenteric tissue oxygenation.24 The NIRS findings from Baby M demonstrated similar NIRS findings that may support these conclusions, as necrotic bowel requiring intestinal resection was identified intraoperatively.
Decreased rSO2 values may be related to diminished oxygen delivery, alteration in hemoglobin oxygen affinity, or gross increase in oxygen consumption4; however, other factors must be considered if variability loss or diminished rSO2 mesenteric readings are encountered (see Table 1). Proper probe placement and accurate skin adhesion must be ensured. Collection of peritoneal or intraluminal gas, fluid, or air may lengthen infrared path length and distort rSO2 readings, as depth of infrared light penetrance is approximately 1.5 cm for most NIRS sensors.12 Presence of stool, especially meconium, can act as a chromophore and reduce the accuracy of rSO2 readings.25 Finally, intestinal structure and function may affect rSO2 trends. Peristalsis, changes in bowel wall thickness, or vascular and tissue blood flow changes associated with enteral feedings may also contribute to rSO2 fluctuations and variability. Oxygen delivery improves in the postprandial state in healthy preterm infants, exhibited by a rise in rSO2 values approximately 30 to 60 minutes after bolus feeds.26 Therefore, prudent evaluation into underlying pathology or associated events is essential when assessing changes in mesenteric oxygenation trends. The significant rSO2 signal dropout that occurred in Baby M may have been due to air accumulation as peritoneal perforation developed, or worsening oxygen delivery and ischemia, or both. To ensure accurate readings were being obtained, we replaced and repositioned the mesenteric sensor probe until readings were observed. We believe that if this technology data had been available to the care team, it may have prompted additional investigation earlier in the course.
Several studies suggest loss of CAR may be identified using NIRS noninvasive bedside monitoring.27 Collectively, studies have found greater stability and less variability in cerebral oxygenation compared with somatic tissue beds.20 , 21 Impaired CAR may predispose cerebral tissues to hypo- or hyperperfusion relative to blood pressure changes,28 increasing the risk for ischemia and/or hemorrhage. Neurodevelopmental delay is a major morbidity associated with NEC, and may be related to cerebral pressure passivity associated with sepsis, inflammation, and hypotension.14 , 17 As a measurement of oxygen extraction in cerebral tissue beds, NIRS offers a practical approach to identifying premature infants who may be at risk for ischemic injury and loss of CAR.27 Cerebral oxygenation stability was demonstrated from Baby M NIRS data throughout and following the first PRBC transfusion; however, cerebral rSO2 30-minute epoch means demonstrated wide variability during and following the second PRBC transfusion during the same time she developed feeding intolerance and gastric residuals. Wide variability (16%-90%) in cerebral epoch means continued until resolution of the pneumoperitoneum after peritoneal drain placement. At this time, cerebral rSO2 readings stabilized and variability minimized. Cerebral oxygenation mean varied as much as 74% from baseline, suggesting that cerebral pressure passivity may have occurred during pneumoperitoneum development; however, intervention and stabilization led to restoration to baseline rSO2 values. Because routine physiologic monitoring did not reveal gross hemodynamic instability, periods of cerebral hypoperfusion were not readily appreciated without the additional information provided by NIRS tracings.
Renal oxygenation rSO2 30-minute epoch means gradually rose following blood transfusions and throughout the study period. However, the beginning renal baseline was substantially lower than reference ranges, and may have been related to anemia, as improvement in tissue oxygenation was demonstrated following blood administration. Because this improvement was not mirrored in the intestinal tissue bed, perfusion-reperfusion injury may have occurred. Previous studies suggested that, after a transfusion, a decreased mesenteric blood flow state occurred postprandial compared with normal controls. This physiologic reduction in oxygen delivery could increase the risk for intestinal ischemia.29 However, it is also possible that anemia-related or infectious ischemic mechanisms were present prior to blood administration.
This case report demonstrates the importance of recognizing, understanding, and appropriately interpreting diminished rSO2 trends when implementing NIRS technology into routine NICU practice. Baby M experienced persistent and prolonged reduction in mesenteric rSO2 values that ultimately resulted in stage IIIA NEC with pneumoperitoneum. In this particular case, routine physiologic monitoring, including pulse oximetry, did not provide indication of organ hypoperfusion or impending NEC development. Perhaps the use of NIRS in conjunction with predictive scoring criteria, such as GutCheckNEC,30 would increase the ability to identify infants in early stages of NEC development. However, further validation studies are required before these practices can be adopted into routine clinical practice. The ability of NIRS to capture possible tissue ischemia suggests this noninvasive technology provides valuable data that can help guide clinical decision-making for our vulnerable preterm population.
1. Patel RM, Kandefer S, Walsh MC, et al Causes and timing of death in extremely premature infants from 2000 through 2011. New Engl J Med. 2015;372(4):331–340.
2. Petrosyan M, Guner YS, Williams M, Grishin A, Ford HR. Current concepts regarding the pathogenesis of necrotizing enterocolitis
. Pediatr Surg Int. 2009;25(4):309–318.
3. Bradshaw WT. Necrotizing enterocolitis
: etiology, presentation, management, and outcomes. J Perinat Neonatal Nurs. 2009;23(1):87–94.
4. Marin T, Strickland O. Transfusion-related necrotizing enterocolitis
: a conceptual framework. Adv Neonatal Care. 2013;13(2):166–174.
5. Denning TL, Bhatia AM, Kane AF, Patel RM, Denning PL. Pathogenesis of NEC
: role of the innate and adaptive immune response. Semin Perinatol. 2017;41(1):15–28.
6. Marin T, Moore J, Kosmetatos N, et al Red blood cell transfusion–related necrotizing enterocolitis
in very-low-birthweight infants: a near-infrared spectroscopy
investigation. Transfusion. 2013;53(11):2650–2658.
7. Mohamed A, Shah PS. Transfusion associated necrotizing enterocolitis
: a meta-analysis of observational data. Pediatrics. 2012;129(3):529–540.
8. Stritzke AI, Smyth J, Synnes A, Lee SK, Shah PS. Transfusion-associated necrotising enterocolitis in neonates. Arch Dis Child Fetal Neonatal Ed. 2013;98(1):F10–F14.
9. Wan-Huen P, Bateman D, Shapiro DM, Parravicini E. Packed red blood cell transfusion is an independent risk factor for necrotizing enterocolitis
in premature infants. J Perinatol. 2013;33(10):786–790.
10. Patel RM, Knezevic A, Shenvi N, et al Association of red blood cell transfusion, anemia and necrotizing enterocolitis
in very low-birth-weight infants. JAMA. 2016;315(9):889–897.
11. De Smet D, Vanderhaegen J, Naulaers G, Van Huffel S. New measurements for assessment of impaired cerebral autoregulation using near-infrared spectroscopy
. Adv Exp Med Biol. 2009;645:273–278.
12. Marin T, Moore J. Understanding near-infrared spectroscopy
. Adv Neonatal Care. 2011;11(6):382–388.
13. Fortune P, Wagstaff M, Petros A. Cerebro-splanchnic oxygenation ratio (CSOR) using near infrared spectroscopy may be able to predict splanchnic ischaemia in neonates. Intensive Care Med. 2001;27:1401–1407.
14. Soul JS, Hammer PE, Tsuji M, et al Fluctuating pressure-passivity is common in the cerebral circulation of sick premature infants. Pediatr Res. 2007;61(4):467–473.
15. O'Leary H, Gregas MC, Limperopoulos C, et al Elevated cerebral pressure passivity is associated with prematurity-related intracranial hemorrhage. Pediatrics. 2009;124(1):302–309.
16. Kaiser JR, Gauss CH, Williams DK. The effects of hypercapnia on cerebral autoregulation in ventilated very low birth weight infants. Pediatr Res. 2005;58(5):931–935.
17. Schat TE, van der Laan ME, Schurink M, et al Assessing cerebrovascular autoregulation in infants with necrotizing enterocolitis
using near-infrared spectroscopy
. Pediatr Res. 2016;79(1-1):76–80.
18. Walsh MC, Kliegman RM. Necrotizing enterocolitis
: treatment based on staging criteria. Pediatr Clin North Am. 1986;33(1):179–201.
19. Dullenkopf A, Frey B, Baenziger O, Gerber A, Weiss M. Measurement of cerebral oxygenation state in anaesthetized children using the INVOS 5100 cerebral oximeter. Pediatr Anesth. 2003;13(5):384–391.
20. McNeill S, Gatenby JC, McElroy S, Engelhardt B. Normal cerebral, renal and abdominal regional oxygen saturations using near-infrared spectroscopy
in preterm infants. J Perinatol. 2010;31(1):51–57.
21. Mintzer JP, Parvez B, Chelala M, Alpan G, LaGamma EF. Quiescent variability of cerebral, renal, and splanchnic regional tissue oxygenation
in very low birth weight neonates. J Neonatal Perinatal Med. 2014;7(3):199–206.
22. Patel AK, Lazar DA, Burrin DG, et al Abdominal near-infrared spectroscopy
measurements are lower in preterm infants at risk for necrotizing enterocolitis
. Pediatr Crit Care Med. 2014;15(8):735–741.
23. Zabaneh R, Cleary J, Lieber C. Mesenteric oxygen saturations in preterm twins with and without necrotizing enterocolitis
. Pediatr Crit Care Med. 2011;12(6):e404–e406.
24. Cortez J, Gupta M, Amaram A, Pizzino J, Sawhney M, Sood BG. Noninvasive evaluation of splanchnic tissue oxygenation
using near-infrared spectroscopy
in preterm neonates. J Matern Fetal Neonatal Med. 2011;24(4):574–582.
25. Said M, Niforatos N, Rais-Bahrami K. Validation of near infrared spectroscopy to measure abdominal somatic tissue oxygenation
in neonates. J Neonatal Perinatal Med. 2013;6(1):23–30.
26. Dani C, Pratesi S, Barp J, et al Near-infrared spectroscopy
measurements of splanchnic tissue oxygenation
during continuous versus intermittent feeding method in preterm infants. J Pediatr Gastroenterol Nutr. 2013;56(6):652–656.
27. Wong FY, Leung TS, Austin T, et al Impaired autoregulation in preterm infants identified by using spatially resolved spectroscopy. Pediatrics. 2008;121(3):e604–e611.
28. Verhagen EA, Hummel LA, Bos AF, Kooi EMW. Near-infrared spectroscopy
to detect absence of cerebrovascular autoregulation in preterm infants. Clin Neurophysiol. 2014;125(1):47–52.
29. Nowicki PT. Ischemia
and necrotizing enterocolitis
: Where, when, and how. Semin Pediatr Surg. 2005;14(3):152–158.
30. Gephart SM, Wetzel C, Krisman B. Prevention and early recognition of necrotizing enterocolitis
, a tale of two tools: eNEC and GutCheckNEC
. Adv Neonatal Care. 2014;14(3):201–210.