Anales de Pediatría (English Edition) (Oct 2018)
Correlation between pneumoperitoneum and surgical findings and morbidity and mortality in newborns with necrotising enterocolitis
Abstract
Introduction: Surgical intervention in necrotising enterocolitis (NEC) is correct when there is intestinal gangrene. This is evident when gangrene produces perforation and pneumoperitoneum, with this being the only universally accepted radiological indication for the surgical intervention of NEC. Objective: To perform an analysis on patients with surgically managed NEC, including determining how the decision to intervene is reached, the outcomes, and if patients with perforation had a pneumoperitoneum. Methods: Retrospective review of neonates with surgical NEC over a period of 10 years (2006–2015). An analysis was made of pre-surgical X-ray findings, which were compared with surgical ones, in addition to the morbidity and mortality, depending on the presence (N+) or absence (N−) of pneumoperitoneum. An evaluation was also made of the interobserver concordance with a paediatric radiologist blinded to the clinical reason using the kappa agreement index. Results: A total of 53 neonates were included in the study. Surgical treatment was indicated after observing pneumoperitoneum in 36%. In the remaining neonates, the surgical decision was made after noting a clinical and metabolic deterioration with classical X-ray findings. Intestinal perforation was observed in 39% of the N− neonates.There were no statistical differences between either group on analysing the excised intestinal length, days of intubation, starting of enteral nutrition, and the mortality rate. Comparisons in terms of duration of symptoms and total hospital stay were statistically significant (7 vs. 2 days, p = 0.008; 127 vs. 79 days, p = 0.003, respectively), with both being more favourable in the N+ group. These differences remained when the groups were adjusted by birthweight. Conclusions: Surgical indication has to be done on an ensemble of clinical and radiological evidence, as 39% of the neonates in the N− groups were perforated.In our study, the presence of a pneumoperitoneum did not correlate with a worse prognosis. Resumen: Introducción: La intervención quirúrgica en las enterocolitis necrosantes (EN) es precisa cuando existe gangrena intestinal, hecho evidente cuando produce perforación y neumoperitoneo, siendo este la única indicación radiológica aceptada universalmente para la intervención quirúrgica. Objetivo: Analizar a los pacientes intervenidos de EN, saber por qué se les intervino, cómo evolucionan y si los pacientes perforados presentan neumoperitoneo. Método: Estudio retrospectivo de una cohorte de recién nacidos con EN intervenidos durante un periodo de 10 años (2006-2015). Se analizan los hallazgos radiológicos preoperatorios y se correlacionan con los quirúrgicos y con la morbimortalidad, dependiendo de la presencia de neumoperitoneo (N+) o no (N–). Se evaluó la concordancia interobservador con radiólogo pediátrico enmascarado a la clínica mediante el índice de acuerdo kappa. Resultados: Se analizó a 53 pacientes. El 36% se intervino tras la visualización de neumoperitoneo; en el resto, la indicación fue deterioro clínico y metabólico, junto con hallazgos radiológicos asociados. En el 39% del grupo N– se objetivó perforación.No se encontraron diferencias significativas en ambos grupos con respecto a longitud intestinal resecada, días de intubación, día de inicio de nutrición enteral y mortalidad. La comparación entre duración de síntomas y estancia hospitalaria total en ambos grupos (N–/N+) fue significativa (7 vs. 2 días, p = 0,008; 127 vs. 79 días, p = 0,003 respectivamente), siendo más favorable en el grupo N+. Estas diferencias se mantuvieron al ajustar por peso. Conclusiones: La indicación quirúrgica ha de basarse en un conjunto de datos clínicos y radiológicos, ya que el 39% de los pacientes sin neumoperitoneo presentaron perforación.En nuestro estudio la presencia de neumoperitoneo no se correlaciona con peor pronóstico.