Comparison of 2-view abdominal radiographs with ultrasound in children with suspected intussusception

Abstract

Background: The clinical presentation of intussusception is variable; therefore, screening with either abdominal radiography (AXR) or abdominal ultrasound (US) is often used, although the optimal method is not known.

Purpose: This study aimed to compare the utility of AXR with that of the US in children with suspected intussusception.

Methods: Retrospective cohort of children age 3 months to 3 years presenting to a pediatric emergency department (ED) between 2007 and 2009. Inclusion criteria were as follows: (1) presentation to the ED for suspected intussusception and (2) both 2-view AXR and US performed during the ED visit. An AXR was deemed negative for intussusception if air was visualized in the ascending colon on 2 views and transverse colon on the supine view; US results were obtained from the radiologist report at the time of presentation. Criterion standard measures for intussusception were contrast enema, operative report, or follow-up.

Results: A total of 286 children were included, with mean (SD) age 16.1 (9.1) months; 62.2% were male, and 43.7% were African American. Intussusception was present in 61 subjects (21.3%). Abdominal radiography had sensitivity of 62.3% (95% confidence interval [CI], 50.1%-74.5%) and specificity of 86.7% (95% CI, 82.2%-91.1%), whereas US had a sensitivity of 98.4% (95% CI, 95.2%-100.0%) and specificity of 96.4% (95% CI, 94.0%-98.9%). Ultrasound had a greater negative predictive value (99.5%; 95% CI, 98.6%-100.4%) compared with AXR (89.4%; 95% CI, 85.4%-93.5%). Abdominal radiography had a greater false-positive rate (13.3% vs 3.6%) and greater false-negative rate (37.8% vs 1.6%), compared with US.

Conclusions: Ultrasound is superior to AXR as a screening method for establishing and excluding the diagnosis of intussusception. The poor test characteristics of 2-view AXR suggest that it should not be used as a primary screening method in cases of suspected intussusception.