Table 2.

Summary of the available literature on the use of Near Infrared Spectroscopy (NIRS) in predicting/diagnosing NEC in neonates.

Author and reference	Population studied	Study design	Primary outcomes/findings
Cortez et al. (15) (2011)	-n = 21 -preterm infants less than 30 weeks gestation enrolled (2 excluded from results)	-abdominal NIRS (left paraumbilical region) placed from within 48 h of birth for the first 14 days	-daily mean sRSO2 values decreased over first 9 days (p < 0.0001) followed by increase from day 10 to 14 (p = 0.0061) -sRSO2 was lower and FTOE higher in infants with feeding intolerance compared to those without (p = 0.0043) -higher sRSO2 and variability was associated with a healthy gut (n = 17) -neonates with NEC had low splanchnic rSO2s and decreased variability (n = 2) -very small study size—only two babies within their study cohort developed NEC
Fortune et al. (29) (2001)	-n = 40 -newborn infants -10 with acute abdomens (4 NEC) -29 controls -1 hypoxic ischaemic injury	-prospective, observational cohort study -cerebral and splanchnic (infraumbilical) regional TOI measured using NIRS -calculated CSOR -measured prior to surgery/admission and then daily until discharge	-neonates with abdominal pathology had lower CSOR (p < 0.001) -CSOR detected the presence of intra-abdominal pathology with a sensitivity of 90% (56–100) and specificity of 96% (82–100) -if CSOR <0.75 intestinal ischaemia was identified with a PPV of 0.75 (0.43–0.95) and excluded with a NPV of 0.96 (0.81–1.0)
Gay et al. (30) (2011)	-n = 29 premature piglets -3 developed NEC -11 died prematurely -15 served as controls	-serial abdominal (1 cm lateral to the umbilicus) NIRS recordings were taken of premature piglets who had received parenteral nutrition followed by enteral feeding -piglets monitored for developing NEC	-abdominal NIRS within 12 h of birth was significantly lower (p = 0.02) in infants who subsequently developed NEC compared with controls -for all time points measured, abdominal NIRS were significantly lower in the NEC group compared with controls (21% vs. 55%, p = 0.01). --the authors drew a sensible conclusion that these lower regional oxygenation readings with abdominal NIRS in piglets with NEC represented intestinal ischemia-reperfusion injury—a well-known theory for the pathogenesis of NEC -also demonstrated that in healthy piglets, when oxygen levels decreased during apnoeas, there was a decrease in the abdominal NIRS oxygenation (r = 0.96) which increased again once the apnoea resolved, demonstrating a clinical correlation with the gut NIRS readings
Howarth et al. (31) (2020)	-n = 48 -preterm infants <30w gestation -median BW 884 (range 460–1,600) grams, median GA 26 + 3 (23 + 0–29 + 6) weeks	-Cerebral oximetry measurements were performed using a NIRS monitor weekly for 60 min allowing measurement of cTOI from first week of life to 36 weeks post conceptional age	-276 NIRS measurements were completed, and 7 infants developed NEC -infants who developed NEC had significantly lower cTOI than those that did not (p = 0.011), even when adjusted for confounders including GA, BW, PDA, enteral feeds, gender, ethnicity, and Haemoglobin
Kalteren et al. (32) (2022)	-<32 w gestational age -n = 29 infants who received 58 RBCT -median GA 27.3 w	-prospective observational cohort study from March 2019 until December 2020 -measured urinary biomarkers for oxidative stress (8-isoprostane) and intestinal cell injury (I-FABP) shortly before and after RBCT -rsSO2 and rsSO2 variability were assessed simultaneously using INVOS 510°c oximeter placed in the infra umbilical region	-6 out of 29 developed NEC after RBCT -Urinary 8-isoprostane and I-FABP increased nearly 2 fold following RBCT (median 282–606 pg/ml and 4,732–6,968 pg/ml, p < 0.01) -this increase was more pronounced in infants who developed NEC -Changes in I-FABP correlated with changes in 8-isoprostane (rho = 0.623, p < 0.01) -Lower rsSO2 variability, but not higher mean rsSO2 was associated with higher 8-isoprostane and I-FABP levels after RBCT -RBCT are associated with signs of associated with concomitant signs of oxidative stress and intestinal injury, parallel with lower variability in splanchnic oxygenation -authors postulated that this may represent the early pathogenetic process of transfusion-associated NEC
Le Bouhellec et al. (33) (2021)	-n = 45 -mean GA of 31 weeks -mean BW 1,486 g	-assessed the ability of NIRS to distinguish those neonates with NEC soon after symptom onset -prospectively collected NIRS measurements of abdominal (infra-umbilically on the central abdominal wall) and cerebral regional tissue oxygen saturation (r-SO2), with values masked by an opaque cover. -Two physicians, blinded to the NIRS data, determined whether the gastrointestinal symptoms were related to NEC 10 days after symptom onset.	-Gastrointestinal symptoms were related to NEC in 23 patients and associated with other causes in 22 -Analysis of the 48 h of monitoring revealed comparable abdominal r-SO2 and splanchnic-cerebral oxygenation ratio (SCOR) in patients with and without NEC (r-SO2: 47.3 [20.4] vs. 50.4 [17.8], p = 0.59, SCOR: 0.64 [0.26] vs. 0.69 [0.24], p = 0.51). -Results were unchanged after NIRS analysis in 6-hour periods, and restriction of the analysis to severe NEC (i.e., grade 2 and 3, 57% of the NEC cases). -in this small study, NIRS monitoring was unable to individualize NEC in premature infants with acute gastrointestinal symptoms.
Marin et al. (34) (2013)	-n = 8 -preterm infants receiving RBCT -TR-NEC infants were 24-29w GA and BW 705-1,080 g -non-NEC infants were 27.6-30w GA and BW 980-1210g	-infants divided into those with NEC post transfusion (TR-NEC, n = 4) and those without (non-NEC, n = 4) -measured cerebral and mesenteric lower abdomen) oxygenation patterns before, during and 48 h after RBCT using NIRS -alculated mean baseline rSO2 change and CSOR	-TR-NEC group received larger mean volumes of total blood than non-NEC infants -TR-NEC group showed wider fluctuation above and below baseline in oxygen saturations than the non-NEC group
Patel et al. (35) (2014)	-n = 100 -preterm infants <32w GA and BW <1,500 g enrolled -8 with incomplete data excluded -divided into groups: infants with NEC (n = 14) and normal preterm infants without NEC (n = 78)	-2 year prospective cohort study -abdominal (right lower abdomen) NIRS measurements -taken 5 min every day for the first week and then the same day once weekly for the next 4 weeks -compared between those with and without NEC	-mean abdominal rSO2 in healthy preterm infants during the first week of life was significantly higher than those who later developed NEC (77.3% ± 14.4% vs. 70.7% ± 19.1%, p = 0.002) -infants who developed NEC had a greater variation in abdominal rSO2 during feeding for first 2 weeks of life -authors suggested that a rSO2 of ≤56% increases the likelihood of later developing NEC (86% sensitivity, 64% specificity, 96% NPV and 30% PPV) -abdominal rSO2 of ≤56% was independently associated with a significantly increased risk of NEC (OR 14.1; p = 0.01) -infants with PDA had significantly lower rSO2 than those without (p = 0.023)
Schat et al. (36) (2016)	-n = 33 -preterm infants -median GA 28w -median BW 1,235 g	-prospective observational cohort study -13 infants no NEC -20 NEC (10 uncomplicated, 10 complicated—Bells stage 3B or death) -mean 8 hly cerebral, liver (right costal arch)and infraumbilical regional oxygenation in those infants with no NEC and those with complicated and uncomplicated NEC in the first 48 h after symptoms developed	-no difference between those with NEC and no NEC regional oxygenation levels in the first 24 h after symptom onset -no significant difference in the first 24 h after symptom onset in regional oxygenation between infants with no NEC and definite NEC -significantly lower cerebral, liver and infraumbilical levels in those with complicated NEC in the first 24 h after onset of symptoms compared with those infants with uncomplicated NEC -cerebral regional oxygenation ≤71% in first 8 h after symptom onset predicted complicated NEC with sensitivity 100% and specificity 80% -liver oxygenation ≤59% in first 8 h after symptom onset predicted complicated NEC with sensitivity and specificity of both 100%
Schat et al. (37) (2019)	-n = 30 -preterm infants <32 w GA -median GA 27.1 w -median BW 903 g	-case control study -10 infants with NEC and 20 control infants matched for GA/BW/presence of PDA -cerebral and intestinal (infraumbilical) regional oxygenation measured using NIRS 2 h daily for first 5 days and then weekly until 5 weeks of life or until NEC developed	-cerebral oxygenation was significantly decreased in those that later went on to develop NEC -cerebral regional oxygenation <70% within the first 48 h of life developed NEC significantly more often than those with cerebral regional oxygenation ≥70% [OR 9 (95% CI 1.33-61.14)] -no difference in intestinal regional oxygenation measurements in those with NEC and without NEC in the first week of life
Sood et al. (27) (2014)	-n = 57 infants -median gestational age of 27 weeks -received 147 RBCTs.	-monitored cerebral and sRSO2 (infraumbilical) in preterm infants receiving RBCTs -defined three time points (pre RBCT—12 h prior), during RBCT and post RBCT—24 h after RBCT) -also defined 3 groups (1 = no NEC within 7 days of RBCT [n = 120], 2 = NEC within 7 days prior [n = 19] and 3 = NEC within 7 days after RBCT [n = 8]).	-in group 1 and 2 rSO2 increased over RBCT periods but in group 3 rSO2 decreased over RBCT periods -RBCT, followed by a diagnosis of NEC, were characterised by lower heart rates pre-, during and post-RBCT, decline in sRSO2 and increase in cFTOE post-RBCT compared to RBCTs not associated with diagnosis of NEC. -Infants received a RBCT who then developed NEC were characterised by a higher variability in sRSO2, post RBCT reduction in sRSO2 and lower CSOR values post RBCT compared to pre RBCT -authors postulated that sRSO2 response to RBCT may potentially be a biomarker to identify infants more likely to develop TR-NEC after a RBCT
Stapleton et al. (38) (2007)	-n = 1	-case report -one infant with background of congenital heart disease who developed NEC -abdominal (midline below the umbilicus and above the pubic symphysis) and cerebral NIRS measurements performed 48 h after diagnosis of NEC was made	-initial abdominal NIRS readings showed low mesenteric rSO2 when compared with cerebral rSO2 (p < 0.0001) -after conservative medical treatment for NEC (NBM and IV antibiotics) mesenteric rSO2 improved compared with initial value
Zabaneh et al. (39) (2011)	-n = 2	-case report -12-day-old growth restricted infant with NEC whose twin did not develop NEC -abdominal NIRS infra umbilical) measured 48 h after NEC diagnosis made and measured at irregular intervals -measurements compared with asymptomatic twin	-mesenteric rSO2 were reduced in the twin with NEC. -mesenteric rSO2 returned to similar level as asymptomatic twin after bowel resection.

BW, birthweight); CGA, corrected gestational age; ELBW, extremely low birthweight; EPO, Erythropoietin; GA, gestational age; Hb, Haemoglobin; I-FABP, Intestinal fatty acid binding protein; IFN gamma, interferon gamma; IL-1 β. interleukin 1 beta; IL-6, interleukin 6; IL-8 interleukin 8; IL-10, interleukin 10; IL-17, interleukin 17; L-FABP, liver fatty acid binding protein; NEC, Necrotising Enterocolitis; NIRS, Near Infrared Spectroscopy; PCA, post conceptual age; RBCT, red blood cell transfusion; TANEC, transfusion associated NEC; TNF-α, Tumour necrosis factor alpha;TR-NEC, Transfusion related NEC; VLBW, very low birth weight.