Identification of clinical parameters to increase the diagnostic yield of the non-emergent upper gastrointestinal series in pediatric outpatients

Background: Outpatient, non-emergent upper gastrointestinal (GI) series are frequently requested in children with no surgical history who have nonspecific symptoms such as abdominal pain, failure to thrive and vomiting. The positive yield of an upper GI series in these patients, and, thus, its utility, has not been studied.

Objectives: We evaluated the incidence of positive upper GI findings in children without a history of GI pathology or abdominal surgery in order to identify clinical indications associated with a greater diagnostic yield.

Materials and methods: Findings of upper GI series performed between October 2015 and October 2017 in three institutions in children younger than 18 years of age were retrospectively reviewed. The upper GI series protocol for each institution was also reviewed. Children with a medical or surgical GI history, children with insufficient history in the chart and those with an incomplete upper GI series were excluded from the study. Exam indications, patient demographics and clinical history were obtained from the electronic medical records.

Results: Of 1,267 children who underwent outpatient upper GI series, 720 (median age: 2 years) had no GI history and were included in the study. The most common indications were non-bilious vomiting (62%), reflux symptoms (28%) and abdominal pain (20%). Upper GI series were normal in 605/720 cases (84%), including 25/26 children with reported bilious emesis. Of the 115 positive studies, 78 (68%) showed only gastroesophageal reflux (GER) (median age: 11 months). Of the remaining 37 studies, 19 demonstrated esophageal findings. One case of malrotation without midgut volvulus was identified in a patient who presented with dysphagia and reflux symptoms. Using a multinomial logistic regression model and adjusting for other variables, reflux symptoms and younger patient age were independent predictors of GER on upper GI series (relative risk ratios of 2.2 and 0.9, respectively). Dysphagia and/or foreign body sensation and older patient age were independent predictors of the presence of esophageal findings (relative risk ratios of 3.3 and 1.1, respectively).

Conclusion: The yield of routine upper GI series in children with nonspecific symptoms, such as abdominal pain and vomiting, and no surgical history is low. Diagnostic yield was improved in older children and in those complaining of dysphagia and/or foreign body sensation. Routine upper GI series should be avoided in clinically well children with symptoms only of uncomplicated GER and no significant GI history. In children with a history of dysphagia and/or foreign body sensation, an esophagram/barium swallow can suffice.