In a retrospective analysis 1, the patient records of 98 children under 15 years of age who had ingested a corrosive substance were analysed. Acids (usually vinegar) caused more severe oesophageal injuries than alkalies. None of the children had ingested lye. Prolonged dysphagia and drooling were accurate (sensitivity 100%, specificity 90%) predictors of a scar-forming corrosive injury of the oesophagus. Routine oesophagoscopy is therefore not indicated.
In another retrospective analysis 2, the patient records of 115 children who had ingested a corrosive substance were analysed between 1976-1994. All complications were associated with the ingestion of a strong alkali (sensitivity 100%). The ingestion of lye resulted in complications in 36.8% of the children whereas washing machine detergents resulted in a complication in only 2.7% of the cases (p > 0.01). Endoscopy was no more sensitive in detecting oesophageal injury than "a symptom or sign". The authors conclude that the risk of complications is low and endoscopy is not indicated in asymptomatic children. Endoscopy is recommended in symptomatic children who have ingested lye or ammonia, and in children with severe symptoms who have ingested dishwasher detergents or acids, and in children with respiratory difficulties. Endoscopy should not be performed before 6 hours have elapsed from the ingestion of the substance in order to avoid underestimation of the severity of the injury.
In a third retrospective analysis 3, the records of 162 children who presented with accidental caustic substance ingestion were analyzed. An endoscopy was performed in all patients within 12 to 24 hours of the substance being ingested. Mild esophageal lesions were identified in 143 of 162 patients (88.3%), and severe (third degree) esophageal lesions in 19 patients (11.7%). The risk of severe esophageal lesions without signs and/or symptoms was very low (OR 0.13, 95% CI 0.02 to 0.62).The presence of 3 or more symptoms was seen as an important predictor of severe esophageal lesions (OR 11.97, 95% CI 3.49 to 42.04). Multivariate analysis showed that the presence of symptoms is the most significant predictor of severe esophageal lesions (OR 2.3, 95% CI 1.57 to 3.38). The authors conclude that the incidence of patients with third-degree lesions without any early symptoms and/or signs is very low, and an endoscopy could be avoided. The risk of severe damage increases proportionally with the number of signs and symptoms, and an endoscopy is always mandatory in symptomatic patients.
Comment: The quality of evidence is upgraded by strong association.
Primary/Secondary Keywords