Abstract
Background
Necrotizing enterocolitis (NEC) is the most common life-threatening gastrointestinal emergency in preterm and term neonates, with the majority of cases affecting neonates classified as very low birth weight (VLBW, bw <1500 g). Scores for neonatal acute physiology-perinatal extension-II (SNAPPE-II) and metabolic derangement acuity score (MDAS) have been developed and utilized to assess neonatal morbidity and mortality including the subset of VLBW neonates. Serial SNAPPE-II and MDAS scores have been reported in neonates with necrotizing enterocolitis to assist in surgical management, yielding mixed results.
Objective
To determine the relationship between clinical and/or laboratory deterioration using SNAPPE-II and MDAS scores measured at the time of NEC diagnosis and surgical management of NEC.
Methods
We retrospectively evaluated preterm neonates ≥23 weeks gestational age who developed pneumatosis intestinalis on radiographic imaging coupled with clinical signs of NEC. SNAPPE-II and MDAS scores were calculated within twelve hours of birth and within twelve hours of initial finding of pneumatosis intestinalis. Baseline characteristics and clinical variables between those who did and did not require surgical intervention were compared. Logistic regression and receiver - operator characteristics (ROC) curve analyses were also performed, and areas under the curve (AUC) computed, to assess the performance of SNAPPE-II and MDAS scoring systems to differentiate neonates with NEC in the two groups.
Results
Sixty-four neonates were evaluated in our study of which 20 required surgical management of NEC. While the baseline SNAPPE-II and MDAS scores did not differ between the surgical management and medical management only groups, when rescored within 12 h of NEC diagnosis, the surgical management group had significantly higher SNAPPE -II (38 (18.5–69) vs. 19 (10–34.5), p = .04) and MDAS (2.5 (1–3) vs. 1 (0–2), p = .0004) scores. The AUCs for MDAS 0.77 (95% CI 0.65-0.89 and 0.71 (95% CI 0.57–0.85) for SNAPPE-II, indicating an acceptable level of diagnostic ability of both scoring systems to differentiate between those who did and did not need surgical management.
Conclusion
SNAPPE II and MDAS scores performed within 12 h of NEC diagnosis may be useful in predicting which preterm VLBW neonates will require surgical intervention.
Introduction
Necrotizing enterocolitis (NEC) is a multifactorial, devasting intestinal process that causes significant morbidity and mortality in the neonatal intensive care unit (NICU) worldwide [Citation1]. NEC is the most common life-threating gastrointestinal emergency in preterm and term neonates. The incidence of NEC is approximately 0.5%, [Citation2] but in neonates less than 1500 g, the incidence increases to 3%−12% [Citation1]. Approximately 20% to 40% of neonates will develop NEC severe enough to require surgical intervention [Citation2]. The mortality of all neonates with NEC is between 15% to 30%, with the higher rate occurring in those with lower birth weight and gestational age [Citation2] and the highest rate in those neonates requiring surgical intervention [Citation3].
When there is concern for a neonate having developed NEC, the initial work up consists of laboratory studies and anteroposterior and lateral abdominal radiographs. [Citation4] It is important to assign disease severity to assist in monitoring, evaluating and treating the patient; furthermore, assigning severity to the disease can allow for specific counseling to the parents. The original definition was based on the staging criteria set forth by Dr. Martin Bell in 1978 [Citation5] that later expanded to include additional stages becoming the Modified Bell staging criteria [Citation6]. According to the Bell Staging Criteria, suspected NEC is identified as Stage 1, which includes neonates with mild symptoms and radiographic findings that demonstrate normal bowel, intestinal dilation or mild ileus [Citation5]. Stage 2 NEC includes neonates with definite disease, with abdominal distension and bloody stools and radiographic findings of pneumatosis intestinalis and/or portal vein gas [Citation5]. Stage 3 NEC is classified as neonates demonstrating clinical deterioration with possible bowel necrosis and in severe cases, pneumoperitoneum, which will require surgical intervention [Citation5]. There are multiple limitations to the Bell staging criteria [Citation7,Citation8], but nonetheless, it is the most accepted and widely used for management of NEC [Citation9]. Additional definitions of NEC have been established since the original descriptions, including newer scoring criteria through the Vermont Oxford Network definition, the Centers for Disease Control and Prevention definition, the Gestational Age-Specific Case Definition of NEC (UK), and the International Neonatal Consortium NEC Workgroup definition, underlying complexity of this medical diagnosis [Citation10].
Several neonatal risk scoring systems have been developed and utilized to assess a neonate’s morbidity and mortality during hospitalization [Citation11,Citation12]. Two of those, the score for neonatal acute physiology-perinatal extension-II (SNAPPE-II) and the metabolic derangement acuity score (MDAS) have been used in prior studies to determine if the scoring can help guide surgical management in preterm neonates with NEC [Citation13–17]. In general, these single center studies had heterogeneous study designs yielding conflicting results, suggesting the need for additional data sets and external validation. The goal of our study was to determine if using SNAPPE-II and MDAS scores can facilitate early prediction of patients who are at high risk of requiring surgical management.
Materials and Methods
Study design
This was a retrospective cohort study of VLBW (BW <1500 gm) neonates ≥ 23 weeks gestational age admitted to the Jack D. Weiler NICU of the Albert Einstein College of Medicine/Children’s Hospital at Montefiore from 2010 to 2019 who developed NEC. We included inborn neonates with at least one clinical sign of NEC and radiographic evidence of pneumatosis intestinalis confirmed by either two attending radiologists or by its presence on two images at different time points. Excluded were neonates with major congenital anomalies, those diagnosed with spontaneous intestinal perforation as well as those with incidental pneumatosis intestinalis finding, where abdominal evaluation was not a primary focus (e.g. radiographic studies ordered for central line placement, chest x ray with partial imaging of the abdomen). The study was approved by the Albert Einstein College of Medicine Institutional Review Board.
For each study neonate, we collected the following data from medical records: gestational age, birth weight, small for gestational age (SGA) status (<10th percentile weight for gestational age), large for gestational age (LGA) status (>90th percentile weight for gestational age), sex, mode of delivery, chorioamnionitis, maternal pregnancy induced hypertension/preeclampsia, singleton versus multiple gestation status, Apgar scores, antenatal and postnatal steroids administration, use of nonsteroidal anti-inflammatory drugs (NSAIDs), and the most recent patent ductus arteriosus (PDA) status prior to the NEC recorded. Additionally, we recorded weight, chronological and corrected gestational age at the time of NEC diagnosis, any packed red blood cell transfusion administration within 48 h prior to the diagnosis of NEC, type of enteral feeding prior to the diagnosis of NEC, and time from the diagnosis of NEC to the initial surgical evaluation. For the subset of neonates who underwent surgery, we recorded the type of surgical intervention (peritoneal drainage or laparotomy) and in those with laparotomy, whether the NEC diagnosis was confirmed with histopathology.
SNAPPE-II and MDAS definitions
Baseline SNAPPE-II and MDAS scores were obtained on NICU admission (score components were collected within 12 h of birth) and again at the time of NEC diagnosis (12 h before and 12 h after the radiographic findings of pneumatosis intestinalis). Clinical and laboratory components necessary for the SNAPPE-II scoring system are depicted in [Citation15] and for MDAS in [Citation17]; the worst physiologic parameter recorded was used in scores calculation.
Table 1. Variables and their associated points used to calculate SNAPPE-II scores.
Table 2. Variables and their associated criteria used to calculate MDAS scores.
Statistical analyses
Baseline clinical and laboratory data in patients with or without pneumoperitoneum and/or required surgery were summarized using standard descriptive statistics. The two groups were compared using the two-sample T-test or Wilcoxon rank sum test for continuous variables, and chi-square or Fisher’s exact test for categorical variables. The association of SNAPPE-II and MDAS scores, at the time of birth and diagnosis of NEC with pneumoperitoneum/surgery status was assessed using similar methods. Logistic regression models were fit to adjust for potential confounders and estimate odds ratios. In addition, receiver-operator characteristic (ROC) curves were generated and the area under the curves were computed to assess the ability of SNAPPE-II and MDAS to discriminate between patients who developed pneumoperitoneum and/or needed surgery from those who did not. Sensitivity and specificity at the “optimal threshold,” defined as the cut-point closest to the (0,1) point on the ROC curve were also estimated.
Results
Sixty-four neonates with NEC were eligible for the study enrollment and data analyses: 20 developed pneumoperitoneum and/or required surgery (7 had pneumoperitoneum and surgery, 13 had surgery without radiographic demonstration of pneumoperitoneum), while 44 responded to medical treatment. Laparotomy was performed in 16 neonates, 3 had peritoneal drainage and 1 neonate had a peritoneal drain placed but later required laparotomy. Baseline demographic and clinical characteristics did not differ between those who did require surgery and who did not () except for the more males undergoing surgical management (p = 0.05). The difference in postnatal steroids use was caused by the higher use of stress doses of hydrocortisone in infants with surgical management of NEC (78.5%) compared to medically managed neonates (21.4%).
Table 3. Maternal and neonatal clinical characteristics.
Neonates in need of surgical management had significantly higher SNAPPE-II (38 (18.5–69) vs. 19 (10–34.5), p = .04) and MDAS (2.5 (1–3) vs. 1 (0–2), p = .0004) scores at the time of the initial necrotizing enterocolitis diagnosis, while no such difference in scores existed at birth (). The SNAPPE-II and MDAS scores remained significantly associated with need for surgical management after adjusting for male sex in logistic regression analysis; the estimated adjusted odds for surgery corresponding to a unit increase in SNAPPE-II score was 1.03 (95% CI 1.01–1.06, p = .02). The corresponding adjusted odds ratio for MDAS was 2.28 (95% CI 1.34–3.89, p = .03). In ROC analyses, the AUC for MDAS was 0.77 (95% CI 0.65–0.89), with sensitivity = 70%, specificity = 66% at the optimal threshold of 2.00. The AUC for SNAPPE-II was 0.71 (95% CI 0.57–0.85), with sensitivity = 70%, specificity = 60%, at the optimal threshold of 29.04 (). Since the presence of pneumoperitoneum is considered a definite indication for surgical management, we have constructed the ROC curves for both surgical and medical management groups without the 7 neonates presenting with pneumoperitoneum, to further assess the prediction of SNAPPE-II and MDAS scores; AUC values yielded similar results ().
Figure 1. ROC curves for MDAS and SNAPPE-II at NEC diagnosis on surgical vs medical NEC.
Figure 2. ROC curves for MDAS and SNAPPE-II at NEC diagnosis on surgical (without the neonates with pneumoperitoneum) vs medical NEC.
Table 4. SNAPPE-II and MDAS scores at birth and at time of NEC.
The diagnosis of NEC was confirmed operatively and via pathology in 15 neonates out of 17 who underwent laparotomy. Fifteen neonates died: 14 in the surgical management group (with NEC being the cause of death in 13 cases) and 1 in the medical management group. The median time from the diagnosis to surgical consultation/bedside evaluation was 6 (2.5–10) hours for the surgical intervention group vs 9 (3–18) hours for the medical management only group (p = .18); 13 out of 44 neonates in the medical management group did not receive surgical consultation, likely reflecting a milder clinical course.
Discussion
The decision to pursue surgical management options for NEC beyond the finding of pneumoperitoneum, the widely accepted indication for surgery, is a difficult one, and often reflects concerns over deteriorating clinical status despite maximal medical therapy [Citation18]. Our study, using SNAPPE-II and MDAS scoring systems, was conducted with the intention to aid clinicians in predicting surgical management in the era when NEC biomarkers are not sufficiently accurate to diagnose and predict the severity of NEC [Citation10,Citation18–20]. Our study showed significantly higher SNAPPE-II and MDAS scores in neonates who underwent surgery for NEC; both scoring systems composed from the data gathered at the time of the NEC diagnosis were predictive of surgical management within acceptable discrimination range of the model, with MDAS performing better than SNAPPE-II system.
The SNAPPE- II neonatal illness severity and mortality risk scoring system has been validated for all birth weight neonates with excellent discrimination utilizing large data sets collected within the first 12 h of life from neonates in Canada, Northern California (Kaiser Permanente) and New England [Citation21]. SNAPPE-II scores were then applied to a smaller group of VLBW neonates in Italy, confirming excellent discrimination result for this age group [Citation22], and lastly revalidated in a large population including all birth weight categories within the Vermont Oxford Network [Citation23]. Bonnard et al. [Citation15] used the SNAPPE-II scoring module retrospectively in VLBW with perforated NEC outside of its original intention and timing: the group determined the SNAPPE-II scores at multiple time points and concluded that the scoring system is a good predictor of mortality after peritoneal drain placement. Lastly, SNAPPE-II scores obtained both at the time of NEC diagnosis and surgical assessment were both predictive of surgical treatment of NEC, supporting our findings [Citation13].
The role of abnormal laboratory values in the pathogenesis of NEC including the subset of neonates who eventually needed surgery has been recognized and studied over the decades, with early reports stressing the importance of thrombocytopenia, neutropenia/bandemia, metabolic acidosis, hyponatremia and coagulopathy [Citation24–27]. Subsequent studies further evaluated the role of thrombocytopenia [Citation28,Citation29] and elevated C-reactive protein (CRP) in NEC [Citation30], recognizing both specificity and sensitivity limitations for these traditional laboratory tests. The search for novel, optimal biomarker(s) related to intestinal immaturity and inflammation is ongoing [Citation19]; however, testing for IL-8, claudin-3, intestinal fatty acid-binding protein, fecal calprotectin is not currently available for clinical use at this time. Mitochondrial deoxyribonucleic acid (mtDNA) emerged recently as a potential sensitive marker for intestinal injury in murine model of NEC. Mt DNA was proportionally increased in the ileum and blood to the degree of intestinal injury secondary to NEC [Citation31] and decreased in oxidative hepatic injury associated with NEC [Citation32].
The MDAS system, originally described in 2006 and combining readily available laboratory parameters with the presence of hypotension, has been found to be predictive of mortality related to peritoneal drain versus laparotomy in the initial approach in preterm neonates with intestinal perforation [Citation17]. The same group of investigators then reported the utility of serial MDAS recordings in the timing and decision making for operative intervention in neonates with NEC without radiographic findings of free air. In these retrospective studies, the MDAS scoring system application was associated with higher likelihood to achieve good outcomes, defined as full enteric feeding by discharge [Citation14,Citation15]. Furthermore, MDAS scores at the time of surgical assessment yielded “moderate” results in predicting surgery for neonates with the diagnosis of NEC [Citation13].
Our study results are best compared to the trial by Ibanez et al. [Citation13], who tested the utility of SNAPPE-II and MDAS scoring systems in prediction of surgery for neonates with a NEC diagnosis. Contrary to Ibanez et al. study, we used different time points for scoring calculations and enrolled exclusively VLBW neonates who had lower mean birth weights and gestational ages. While we computed the baseline scores within 12 h of birth and then again within 12 h of NEC diagnosis, Ibanez et al. study had scores computed at the moment of diagnosis and then again at the time of clinical deterioration/initial surgical assessment with the median time of 36 h for the surgical group (medical treatment group, not evaluated by surgeon was arbitrarily assigned the same elapsed time of 36 h for scoring calculations to allow group comparisons). Additionally, they included sizable number (37.5%) of neonates with Bell Stage I in their medical treatment group. In our unit, surgical assessment is strongly recommended for neonates with NEC Stage II and III; all surgical cases were evaluated at median time 6 h within the diagnosis and 31 out of 44 neonates in medical treatment group received surgical evaluation at median time 9 h, making 12 h time point for scoring calculations a reasonable choice. In agreement with Ibanez data, we reported similar AUC discrimination values for SNAPPE-II: 0.71 (95% CI 0.57–0.85) when all surgical cases are included and 0.68 (95% CI 0.50–0.87) when surgical cases without pneumoperitoneum were analyzed, mirroring Ibanez’s data at baseline (0.69, CI 95%, 0.57–0.80) and at 36 h 0.67, CI 95%, 0.55–0.80). Differing from Ibanez et al. we demonstrated very good prediction of MDAS scores with AUCs in acceptable discrimination range 0.7–0.8 irrespective of whether neonates presenting initially with pneumoperitoneum were included (0.77, 95% CI 0.65–0.89) or excluded (0.77, 95% CI 0.64–0.90) and overall performing better than MDAS in Ibanez study (0.59 with 95% CI 0.47–0.71 for baseline and 0.64 with 95% CI 0.52–0.77 at the 36 h time point).
The variables needed for SNAPPE-II and MDAS scores can easily be extracted from the electronic health record and computed almost instantly. Following these scores over time may prevent omission of important clinical or laboratory data and may increase the objectivity of clinical assessments of a neonate with NEC. Similar to the imperfect scoring systems used when managing neonatal pain, sedation, neonatal abstinence syndrome, early onset of sepsis, neonatal transition, it remains unclear whether following prospective SNAPPE-II and MDAS scores would improve the management of neonate with NEC.
The obvious limitation of our single center study is its retrospective design of prospectively collected clinical research data. Furthermore, while we think the 12 h interval used for collection of variables constituting SNAPPE-II and MDAS scores is a previously reported and reasonable one, it’s also a purely arbitrarily chosen time point by the authors.
In conclusion, the two scoring systems, SNAPPE-II and MDAS, may serve as an ancillary tool when the decision for surgical intervention is in question and the same time the mortality and the morbidity of surgical NEC neonates have not been reduced significantly.
Disclosure statement
No potential conflict of interest was reported by the author(s).
References
- Gregory KE, Deforge CE, Natale KM, et al. Necrotizing enterocolitis in the premature infant: neonatal nursing assessment, disease pathogenesis, and clinical presentation. Adv Neonatal Care. 2011;11(3):155–166.
- Dominguez KM, Moss RL. Necrotizing enterocolitis. Clin Perinatol. 2012;39(2):387–401.
- Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med. 2011;364(3):255–264.
- Rich BS, Dolgin SE. Necrotizing enterocolitis. Pediatr Rev. 2017;38(12):552–559.
- Bell M, Ternberg J, Feigin R, et al. Neonatal necrotizing enterocolitis. Therapeutic decisions based upon clinical staging. Ann Surg. 1978;187(1):1–7.
- Kliegman RM, Walsh MC. Neonatal necrotizing enterocolitis: pathogenesis, classification, and spectrum of illness. Curr Probl Pediatr. 1987;17(4):213–288.
- Gordon PV, Swanson JR, Attridge JT, et al. Emerging trends in acquired neonatal intestinal disease: is it time to abandon Bell’s criteria? J Perinatol. 2007;27(11):661–671.
- Gephart SM, Gordon PV, Penn AH, et al. Changing the paradigm of defining, detecting, and diagnosing NEC: perspectives on Bell’s stages and biomarkers for NEC. Semin Pediatr Surg. 2018;27(1):3–10.
- Juhl SM, Hansen ML, Gormsen M, et al. Staging of necrotizing enterocolitis by Bell’s criteria is supported by a statistical pattern analysis of clinical and radiological variables. Acta Paediatr. 2019;108(5):842–848.
- Patel RM, Ferguson J, McElroy SJ, et al. Defining necrotizing enterocolitis: current difficulties and future opportunities. Pediatr Res. 2020;88(Suppl 1):10–15.
- Garg B, Sharma D, Farahbakhsh N. Assessment of sickness severity of illness in neonates: review of various neonatal illness scoring systems. J Matern Fetal Neonatal Med. 2018;31(10):1373–1380.
- Richardson DK, Tarnow-Mordi WO, Escobar GJ. Neonatal risk scoring systems. Can they predict mortality and morbidity? Clin Perinatol. 1998;25(3):591–611.
- Ibanez V, Couselo M, Marijuan V, et al. Could clinical scores guide the surgical treatment of necrotizing enterocolitis? Pediatr Surg Int. 2012;28(3):271–276.
- Tepas JJ, Sharma R, Leaphart CL, et al. Timing of surgical intervention in necrotizing enterocolitis can be determined by trajectory of metabolic derangement. J Pediatr Surg. 2010;45(2):310–313.
- Bonnard A, Zamakhshary M, Ein S, et al. The use of the score for neonatal acute physiology-perinatal extension (SNAPPE II) in perforated necrotizing enterocolitis: could it guide therapy in newborns less than 1500 g? J Pediatr Surg. 2008;43(6):1170–1174.
- Tepas JJ, 3rd, Leaphart CL, Plumley D, et al. Trajectory of metabolic derangement in infants with necrotizing enterocolitis should drive timing and technique of surgical intervention. J Am Coll Surg. 2010;210(5):847–854.
- Tepas JJ, 3rd, Sharma R, Hudak ML, et al. Coming full circle: an evidence-based definition of the timing and type of surgical management of very-low-birth weight (<1000 g) with signs of acute intestinal perforation. J Pediatr Surg. 2006;41(2):418–422.
- Frost BL, Modi BP, Jaksic T, et al. New medical and surgical insights into neonatal necrotizing enterocolitis: a review. JAMA Pediatr. 2017;171(1):83–88.
- Neu J. Necrotizing enterocolitis: the future. Neonatology. 2020;117(2):240–244.
- Markel TA, Engelstad H, Poindexter BB. Predicting disease severity of necrotizing enterocolitis: how to identify infants for future novel therapies. J Clin Neonatol. 2014;3(1):1–9.
- Richardson DK, Corcoran JD, Escobar GJ, et al. SNAP-II and SNAPPE-II: simplified newborn illness severity and mortality risk scores. J Pediatr. 2001;138(1):92–100.
- Gagliardi L, Cavazza A, Brunelli A, et al. Assessing mortality risk in very low birthweight infants: a comparison of CRIB, CRIB-II, and SNAPPE-II. Arch Dis Child Fetal Neonatal Ed. 2004;89(5):F419–F422.
- Zupancic JA, Richardson DK, Horbar JD, et al. Revalidation of the score for neonatal acute physiology in the Vermont oxford network. Pediatrics. 2007;119(1):e156–e163.
- O'Neill JA, Stahlman MT, Meng HC. Necrotizing enterocolitis in the newborn: operative indications. Ann Surg. 1975;182(3):274–279.
- Buras R, Guzzetta P, Avery G, et al. Acidosis and hepatic portal venous gas: indications for surgery in necrotizing enterocolitis. Pediatrics. 1986;78(2):273–277.
- O’Neill JA, Jr, Holcomb GW. Jr. Surgical experience with neonatal necrotizing enterocolitis (NNE). Ann Surg. 1979;189(5):612–619.
- Gupta SK, Burke G, Herson VC. Necrotizing enterocolitis: laboratory indicators of surgical disease. J Pediatr Surg. 1994;29(11):1472–1475.
- Ververidis M, Kiely EM, Spitz L, et al. The clinical significance of thrombocytopenia in neonates with necrotizing enterocolitis. J Pediatr Surg. 2001;36(5):799–803.
- Kenton AB, O'Donovan D, Cass DL, et al. Severe thrombocytopenia predicts outcome in neonates with necrotizing enterocolitis. J Perinatol. 2005;25(1):14–20.
- Pourcyrous M, Korones SB, Yang W, et al. C-reactive protein in the diagnosis, management, and prognosis of neonatal necrotizing enterocolitis. Pediatrics. 2005;116(5):1064–1069.
- Bindi E, Li B, Zhou H, et al. Mitochondrial DNA: a biomarker of disease severity in necrotizing enterocolitis. Eur J Pediatr Surg. 2020;30(1):85–89.
- Bindi E, Alganabi M, Biouss G, et al. Hepatic oxidative injury: role of mitochondrial dysfunction in necrotizing enterocolitis. Pediatr Surg Int. 2021;37(3):325–332.