374 Practice Patterns in Patients Admitted for Denovo Acute Food Impaction: A Multi-Center Retrospective Study

Abstract

INTRODUCTION: Food impactions are a medical emergency for gastroenterologists. Despite their common occurrence, recent data shows high variability in self-reported practice patterns of how gastroenterologists approach food impactions. This may lead to delayed diagnosis and care for patients, particularly in eosinophilic esophagitis (EoE). There is limited data available assessing real life practice habits and recommendations based on endoscopic findings. METHODS: This is a multi-centered, single-system retrospective chart review of all patients admitted with de novo food impaction between Jan 2013 and Sept 2018. Data collected includes demographics, clinical, endoscopic, and discharge recommendations. Patients with diagnosis of achalasia, EoE, malignancy, or stricture were excluded. Our primary aim was to characterize GI practice habits and discharge planning. A planned secondary analysis comparing practice habits in patients with and without presumed EoE was done. RESULTS: 199 patients were admitted with food impaction during the study period, 157 meeting inclusion criteria. 36 had endoscopic features of EoE, with 33 having EoE as the presumptive diagnosis. 109 had a food bolus present on endoscopy. To remove this bolus, the ‘push through’ method was used in 29.9% of cases, roth nets in 26.1%, and graspers in 22.9%. On sub-group analysis, there was no statistical difference regarding endoscopic technique between those with and without presumed EoE. Esophageal biopsy was obtained in 40.1% of cases, but in 75.8% of patients with presumed EoE vs. 32.0% without (P ≤ 0.001) . At discharge, 79% of patients were started on a proton pump inhibitor (PPI), 43.9% were instructed to have a follow up endoscopy, with providers more likely to both start a PPI (P = 0.01) and refer for follow up (P = 0.02) if EoE was suspected. CONCLUSION: We conducted a multi-center, single-system study on the real life practice of gastroenterologists evaluating patients with de novo food impaction. EoE was suspected in 33 patients with a confirmed diagnosis in 63.6%. Push technique continued to be used as frequently in patients suspected of having EoE. Only 78.8% of patients underwent esophageal biopsy and 93% were started on PPI, though both statistically more frequent in patients with suspected EoE. There remains room for quality improvement in the treatment of de novo food impactions in patients suspected of having EoE and those who are not. A standardized protocol for food impactions would facilitate prompt diagnosis and treatment.