The diagnostic accuracy of an abdominal radiograph with signs and symptoms of intussusception

Abstract Objective The objective of this study was to compare the diagnostic accuracy of an abdominal ultrasound to that of a highly suggestive abdominal radiograph combined with signs and symptoms of intussusception. Design This was a retrospective cross-sectional study of children 3 years or younger with signs and symptoms of intussusceptions who presented to a pediatric emergency department (ED). Univariate analysis, multivariate analysis, and diagnostic accuracy of clinical characteristics and radiographic findings were derived. Results A highly suggestive abdominal radiograph (14.80; 5.85-37.45), right upper quadrant mass (8.90; 1.14-69.47), vomiting (2.54; 1.36-4.76), and abdominal pain (2.45; 1.36-4.40) were found to be significantly associated with intussusception by univariate analysis. Vomiting (2.80; 1.34-5.85), abdominal pain (2.75; 1.33-5.69), and bloody stools (2.70; 1.07-6.81) were independently associated with intussusceptions by multivariate analysis. Bloody stools were time dependent. Bloody stools occurred in those patients with intussusception at a median time of 24 hours, from the time the patient started with signs and symptoms to the time of presentation to the ED, vs those without bloody stools presenting at a median time of 11 hours. The combination of a highly suggestive abdominal radiograph, abdominal pain, lethargy, and vomiting was highly specific (95%) for intussusception, comparable to that of an ultrasound (93%). In patients with this combination, all were found to have intussusception by enema or surgery. Conclusions Ultrasound is not needed before an enema for the diagnosis of intussusception for those with a highly suggestive abdominal radiograph, abdominal pain, lethargy, and vomiting.