327 - Risk factors for early death in neonates diagnosed with necrotizing enterocolitis

Poster Number: 327
Publication Number: 327.133

William H. Stafford, University of Tennessee Health Science Center College of Medicine, Memphis, TN, United States; Jennifer Davidson, UTHSC, Memphis, TN, United States; Ajay J. Talati, UTHSC, Memphis, TN, United States; Selamawit Tesema, University of Tennessee Health Sciences Center, Memphis, TN, United States; Walker Laird, University of Tennessee Health Science Center College of Medicine, Memphis, TN, United States

Background: Necrotizing enterocolitis (NEC) causes morbidity and mortality in preterm infants. Incidence of NEC is estimated to be 4-7% of VLBW births. NEC requiring surgical intervention or resulting in death is most concerning in the neonatal intensive care unit (NICU). Pathogenesis is multifactorial with no model to predict NEC severity. Several patient factors, exposures, and biomarkers may help in diagnosis and risk stratification, but none are used systematically.

Objective: Identify the clinical and laboratory findings of NEC resulting in mortality at a level IV NICU.

Design/Methods: Retrospective review was completed for infants diagnosed with NEC (any stage) at a single institution level IV NICU from January 2017 to December 2020. Clinical, laboratory, and radiologic data was collected from prior to NEC diagnosis, time of diagnosis, and treatment period following diagnosis. Subjects were divided: death within 48hr of diagnosis or arrival to our unit, survivors past 48hr of diagnosis, and death at any time point. Statistical analysis was performed to determine significant difference between groups with p value set at .05.  

Results: 134 cases were identified; 11 (8.2%) died within 48hr, 104 (77.6%) survived past 48hr, and 30 (22.4%) died prior to discharge from NICU. Demographic and pre-NEC characteristic data, seen in table 1, were similar between the groups. Laboratory data at time of diagnosis can be seen in table 2. Serum creatinine and degree of acidosis on blood gas were significantly different between early deaths and survivors. Other lab values were similar between the groups. Clinical factors such as need for vasopressor support, documented respiratory distress, free air on imaging, higher NEC stage, and surgical intervention were all statistically more likely in early death group. There were no significant differences between the early death group and all death group in all categories of data.  

Conclusion(s): In our cohort, several characteristics were more likely to be seen in infants with severe NEC. Infants who had worse signs of organ dysfunction (respiratory distress, worse acidosis, elevated Cr, need for pressors) were more likely to die within 48 hours. No significant demographic differences (i.e gestational age, birth weight, maternal drug use) were seen between our groups. Interestingly, type of feeds, amount of feeds, caloric content of feeds, and duration of TPN had no impact on mortality. This information may be helpful when counseling families about prognosis of patients who have NEC with evidence of organ dysfunction.