Immediate Complications of Upper Endoscopy in Children

Abstract

Background: Esophagogastroduodenoscopy (EGD) is an essential diagnostic and therapeutic procedure for the evaluation of gastrointestinal disorders in children. Although generally considered a safe procedure, EGD has the potential for various complications. However, the available estimates of the incidence and type of these complications are inconsistent. Methods: We conducted a cross-sectional retrospective study to determine the frequency and determinants of immediate complications in all children (0-18 years) who underwent EGD at 13 pediatric facilities between November 1999 and December 2003 recorded in Pediatric Endoscopy Database System - Clinical Outcomes Research Initiative (PEDS-CORI). The database was analyzed for the procedure indication, ASA (American Society of Anesthesiologists) class, gender, age, anesthesia type, type of complication (recorded during or shortly after the procedure), and unplanned interventions. We used Chi-square analysis to compare complication rates across levels of categorical variables, and Mantel-Haenzel test for ordered variables. Results: We analyzed 10,236 procedures performed in 9234 patients. Immediate complications were reported in 239 procedures (2.3%; 95% CI: 2.0% to 2.6%). The most common complications were hypoxia (157, 1.5%) and bleeding (28, 0.3%), followed by respiratory distress (9, 0.09%), wheezing (9, 0.09%), arrhythmia (7, 0.07%), and nausea/vomiting (6, 0.06%). No perforations or deaths were reported. Complication rates were significantly higher in the youngest age group (p < 0.0001), highest ASA class (p < 0.02), and in the presence of a fellow (p < 0.01). General anesthesia was associated with significantly lower complication rates as compared to IV sedation (1.2% vs. 3.7%; p < 0.0001). Of those procedures with a complication reported, 171 (71.5%) also reported an unplanned intervention. The most common interventions among this group were: oxygen administration during the procedure (47.7%), early termination of procedure (10.5%), and sedation reversal (3.4%). Conclusions: This is the largest multi-center study of complication rates of pediatric EGD ever performed. The overall immediate complication rate of pediatric EGD (2.3%) is higher than that reported in adult patients (0.13%). All complications were nonfatal, and most were hypoxia related and reversible. Informed consent should reflect the frequency and type of these complications. Young age, higher ASA class, and IV sedation are risk factors for developing complications.