428 Esophageal Food Bolus Impactions: Predictors of Persistent Impaction at Endoscopy and Strategies for Disimpaction

Abstract

INTRODUCTION: Esophageal food bolus impaction is a true gastrointestinal emergency. While most cases resolve spontaneously, up to 20% require emergent upper endoscopy (EGD). No prior studies have been done in the United States to explore how patient symptoms may predict persistent obstruction at the time of endoscopy. Our aim was to determine symptoms that reliably predict complete impaction in efforts to quickly recognize those who need emergent EGD at time of presentation. METHODS: This is a retrospective cohort study of 83 patients who presented to the emergency room with suspected food bolus impaction between 1/2013 and 5/2018 at a tertiary care academic medical center. Electronic health records were reviewed to determine presenting symptoms and initial treatment course. Exclusion criteria were age < 18, history of esophageal malignancy, or presence of an esophageal stent. Descriptive statistics, univariate and multivariate analysis were performed. RESULTS: During the study period 52 of 83 (63%) patients underwent an EGD for suspected food bolus impaction. Patients who underwent EGD were more likely to have received glucagon (P = 0.004) and to have presented with sialorrhea (P = 0.001), as compared to patients who did not undergo EGD. 42 of 52 (81%) patients who underwent EGD had an overt food bolus impaction identified. Predictors of overt impaction at EGD were glucagon administration (P = 0.02), underlying esophageal pathology (P = 0.01), and sialorrhea on presentation (P = 0.01). 30 of 42 (71%) patients with overt food impaction were found to have complete obstruction. These patients were more likely to have received glucagon (P = 0.04), have an earlier EGD (P = 0.04), and have symptoms of odynophagia (P = 0.003), sialorrhea (P = 0.001), and regurgitation (P = 0.006), as compared to patients with partial obstruction. On multivariate analysis, sialorrhea (P = 0.001) and odynophagia (P = 0.023) were symptoms predictive of EGD, and only sialorrhea (P = 0.01) was predictive of persistent food impaction at EGD. CONCLUSION: Sialorrhea and glucagon administration were independently associated with the need for EGD during admission. Sialorrhea was the only symptom predictive of persistent impaction found on EGD. Sialorrhea along with odynophagia were most predictive of complete obstruction as compared to partial. These symptoms on presentation may help triage patients in need of more emergent EGD.