Background
Foreign body (FB) ingestions are common in children presenting to the emergency department. Health care providers are quickly challenged to determine which children need urgent endoscopy for diagnostic or therapeutic reasons. We performed a retrospective study to determine if esophageal injury caused by FB ingestion correlated to presenting signs or symptoms, location of impaction, duration of impaction, or denomination of coin (as this was the most commonly ingested FB).
Methods
A retrospective chart review of children between birth and 17 years of age who presented for esophagogastroduodenoscopy for removal of upper gastrointestinal FB was performed. Demographic data collected from all children included age, sex, and race. For children with FB ingestion, the type of FB, location of the FB, underlying gastrointestinal pathology, duration of impaction, and endoscopic findings were recorded. Descriptive analysis of the data was performed using means, medians, SD, and percentages; χ2 test was used to test the association between categorical variables.
Results
Over a 10-year period of review, a total of 3279 esophagogastroduodenoscopies were performed; 248 (7.8%) were done for FB removal. The mean age for children having endoscopy for FB removal was 3.9 (SD, 3.2) years (median, 3.1 years); there was a slight male predominance (male/female ratio = 1.6:1). The vast majority (81%) of retained FBs was coins. Most of the FBs were located in the upper esophagus (68%). Success rate for retrieval was greater for esophageal FBs (99%) than for more distally located FBs (70%; P < 0.001). Mucosal ulceration, seen in 59 children (30%), was related to a complaint of substernal pain but not vomiting, respiratory distress, or drooling. The finding of esophageal ulceration was not related to location of coin impaction or denomination of ingested coin but was related to duration of impaction and the unexpected finding of FB during chest radiograph. Underlying pathology was found more commonly in children with meat bolus impaction (100%) than in children with other FB ingestions (3.6%; P < 0.001).
Conclusions
Ingestion of FBs by children remains a significant problem faced by emergency department personnel. In our study, a complaint of substernal chest pain in children with an esophageal FB predicted esophageal ulceration. Also, esophageal FBs unexpectedly found on chest radiograph or known to be present greater than 72 hours were more likely to have esophageal ulceration. These clinical and historic clues can help direct appropriate prompt referral for endoscopic removal.
