Foreign body ingestion in children: Definition of a nomogram to predict surgical or endoscopic intervention
Cecilia Mantegazza, Simona Ferraro, Davide Biganzoli, Francesca Destro, Paolo Quitadamo, Sara Isoldi, Filippo Torroni, Monica Malamisura, Sara Renzo, Lorenzo Fioretti, Paolo Gandullia, Giacomo Tantari, M. Maino, Giorgio Fava, Matteo Bramuzzo, Silvia Zingarella, Maria Teresa Illiceto, Lorenzo Norsa, M. Pellegrino, Luca Maria Antoniello, Paolo Orizio, Alessio Nanni, Fabio Cisarò, Enrico Felici, Claudio Romano, Salvatore Oliva, Elia Biganzoli, Giuseppe Marano, Cristina Bucci, Mariano Caldore, Manuel Murciano, Giulia Chiarazzo, Martina Ichino, Francesco Macchini, Alessandra Marinari, Giovanni Di Nardo, Antonio Marseglia, Marco Deganello Saccomani, Debora Sala, Elia Balestra, Silvia Iuliano
Abstract
BACKGROUND AND AIMS: Foreign body ingestion (FBI) in children requires early identification to prevent adverse outcomes and may necessitate endoscopic or surgical intervention. This study aims to develop a nomogram that identifies children who require urgent surgical or endoscopic intervention by using the patient's medical history and clinical parameters collected at admission. METHODS: This study is a retrospective review (01/2015-12/2020) of a multicenter case series of children admitted for FBI. Data from 5864 records from 24 hospitals in Italy were analyzed. Logistic regression models were used to establish the probability of requiring surgical or endoscopic intervention based on patient history and clinical characteristics. The nomogram representing the results from the multivariable model was reported to examine the propensity for surgery/endoscopy. RESULTS: The study identified a significant association between intervention and various factors, including type of foreign body (blunt: reference category, disk battery (odds ratio OR:4.89), food bolus (OR:1.88), magnets (OR:2.61), sharp-pointed (OR:1.65), unknown (OR:1.02)), pre-existing diseases or conditions (OR 3.42), drooling (OR 10.91), dysphagia (OR 5.58), vomiting (OR 3.30), retrosternal pain (OR 5.59), abdominal pain (OR 1.58), hematemesis (OR 2.82), food refusal/poor feeding (OR 2.99), and unexplained crying (OR 2.01). The multivariable regression model showed good calibration and discrimination ability, with an area under the ROC curve of 0.77. CONCLUSIONS: This study developed the first nomogram to predict the probability of the need for surgical or endoscopic intervention in children with FBI, based on the information collected at admission. The nomogram will aid clinicians in identifying children who require early intervention to prevent adverse outcomes.