Predictors of spontaneous passage of ingested foreign bodies in adults: twelve years of experience.

Abstract

Foreign body ingestion in adults is commonly encountered in clinical practice. The therapeutic approach of whether to follow-up or extract is often controversial. We aimed to explore predictors for spontaneous passage of ingested foreign bodies by focusing on foreign body type, length, and location of impaction. We performed a 12-year retrospective single-center study. Logistic regression analysis was done to identify predictors of spontaneous passage. Overall, 365 patients with foreign body ingestion were included. The rate of spontaneous passage was 53.7% in general, while the spontaneous passage rate was 47.9% in food impaction, 44.3% in sharp objects, 88.7% in blunt objects and only 22.2% in long blunt objects (> 6cm). On regression analysis, esophageal location was associated with a higher impaction rate and lower spontaneous passage vs. stomach and small and large intestine (OR 0.15, 95% CI 0.07-0.31, OR 0.18, 95% CI 0.09-0.37 and OR 0.02, 95% CI 0.003-0.14), respectively. Performing Receiver operating characteristics (ROC) analysis found that the maximal length above which the foreign body will fail to pass spontaneously was 3.5cm in the stomach and 3cm in the small intestine, with area under the curve (AUC) of 0.8509 in stomach and 0.8073 in small intestine. Endoscopic removal was needed for all esophageal foreign bodies, and all foreign bodies more than 3.5cm above the duodenum. Spontaneous passage of ingested foreign body in a selected cohort of patients depends on foreign body type, location, and length.