97 - Diagnostic accuracy of the newly proposed clinical definition of Necrotizing Enterocolitis (NEC): a retrospective chart review study

Poster Number: 97
Publication Number: 97.237

Mira C. Malavia, University of Missouri-Kansas City School of Medicine, Kansas City, MO, United States; Venkatesh Sampath, Childrens Mercy Kansas City, Kansas City, KS, United States; Hung-Wen Yeh, Children's Mercy Hospitals and Clinics, Kansas City, MO, United States

Background:

Necrotizing Enterocolitis (NEC) is a life-threatening disease in premature infants associated with significant mortality and long-term morbidity. The modified Bell’s criteria have been the primary classification system for NEC. The lack of specificity of Bell’s criteria for NEC diagnosis and its inability to discern NEC from other clinical conditions such as spontaneous ileal perforation (SIP), feeding intolerance and septic ileus has prompted newer diagnostic criteria. Caplan et al. criteria emphasize the timing of disease onset, clinical variables, and radiographic findings to improve precision of diagnosis, prognostication, and research.

Objective:

1. To determine agreement between Bell’s criteria and Caplan criteria for NEC diagnosis


2. To evaluate whether the day of life 10 cut-off for SIP vs. NEC proposed by Caplan et al. accurately reflects surgical diagnosis of NEC and SIP

Design/Methods: A retrospective chart review of preterm infants born between 2010-2021 at Children’s Mercy Hospital was performed. Each infant was reassigned a diagnosis with modified Bell’s and Caplan criteria based on the bedside diagnosis, documented clinical signs, radiography, and surgical findings. Agreement between documented clinical diagnosis, modified Bell’s and Caplan criteria were compared using the Kappa statistic κ.

Results: 204 patients were identified with 222 diagnoses of NEC (160) or SIP (59) based on Bell’s criteria. Compared to infants who developed NEC, infants with SIP had a lower gestational age, lower 5-minute APGAR score, and earlier onset of intestinal perforation (Table 1). Applying the Caplan criteria, 14 infants (8.8%) with perforation originally regarded as NEC would have been re-classified as SIP based on the perforation occurring before day 10 of life (Table 2 ). Among infants with medical NEC (n=111), applying the Caplan criteria reclassified 22 infants (13.8%) with feeding intolerance and 10 infants (6.2%) with septic ileus. The agreement in diagnosis by the Caplan criteria and the initial diagnosis for NEC and SIP (n = 181 diagnoses) was statistically significant (κ = 0.82 , p < 0.001). 106 infants with surgery-based diagnoses agreed with the modified Bell’s and surgical criteria (κ = 0.95, p < 0.001).

Conclusion(s): This study demonstrates that applying Caplan’s criteria for SIP vs. NEC would have reclassified true surgical NEC as SIP based on day of life criteria in 9% of preterm infants. Caplan criteria did discriminate medical NEC vs. septic ileus vs. feeding intolerance better than bedside diagnosis.