The Natural History of Management and Complications of Esophageal Food Impaction

Abstract

Introduction: The use of endoscopic pull removal over push technique in treatment of food impaction has been recommended, although recent data suggest safety may be equivalent. We sought to evaluate the incidence of food impaction, change in technique, and safety of food bolus removal in a fixed population over 45 years. Methods: For this population based study, all recorded episodes of esophageal food impaction in Olmsted County, MN from 1967-2012 were identified. Food impaction was defined as evidence of esophageal obstruction necessitating medical assessment in proximity to eating. Charts were reviewed for impaction therapy, etiology and other characteristics. Results: 652 impactions were identified in 480 patients (56.6 (18-99) yo, 60.6% male). The incidence of impaction increased from 1.9/yr (1970s) to peak of 26.5/yr (2000s). 439 (67.3%) were cleared via EGD with 176 (27.0%) by removal and 258 (39.6%) by push. Clearance via meds (90 (13.8%) glucagon, 19 (2.9%) nitroglycerine) and spontaneously (92 (14.1%)) was less frequent. Treatment changed over time (by decade, 1980s-2010s) with push (14.1%-41.3%) and med use (14.1%-20.6%) increasing, and spontaneous (26.6%-6.3%) and removal (43.8%-20.6%) decreasing. Common underlying conditions included obstructing lesion on EGD, GERD, EoE, radiation, esophageal surgery, achalasia or neurologic disorders (Table 1). Complication rate increased from 0/64 (0%) in the 1980s to 6/63 (9.5%) in the 2010s. 31 (4.8%) tears and 2 (0.3%) perforations occurred, all with endoscopy (P<0.001). Complication rate was similar in push (22, 8.3%) and removal (11, 6.2%) techniques (P=0.5), with all per-forations occurring with push (P=0.5). 1 perforation occurred in EoE (2.4%) and 1 with obstructive lesion (0.4%). Obstruction on EGD was not associated with increased complications (6.9% vs 8.7%, P=0.6) when compared to those without but EoE did have more complications than those without (13.4% vs 4.1%, P=0.004). Association with EoE (1.6%-19.0%) and GERD (15.6%-41.2%) increased over time but association with obstructive lesions decreased (77.8%-60.0%). Other etiologies were relatively stable.Table: Table. Population demographicsFigure: Food impaction treatment and complications over study period.Table: Table. Impaction characteristicsConclusion: 1. Push technique does not appear to have increased risk of complication compared to pull removal. 2. Increasing incidence of food impaction was found in association with EoE and GERD over the study period. 3. An increasing complication rate was seen and may be related to EoE, with higher rate of complications in this population.