Test characteristics of the 3-view abdominal radiograph series in the diagnosis of intussusception

Objectives: To determine the test characteristics of the 3-view abdominal radiograph series to exclude intussusception in children presenting to the pediatric emergency department. Methods: We performed a single center retrospective review of children for whom supine, prorie, and lateral decubitus abdominal radiographs were performed as part of our standard diagnostic evaluation for intussusception. A pediatric radiologist masked to the patient's clinical data and outcome reviewed all radiographs. The criterion evaluated was whether air was visualized in the ascending colon on each of the 3 radiograph views. We determined the presence or absence of intussusception hierarchically by definitive radiological study report, surgical operative note, or clinical outcome. Results: We analyzed 179 patients, of whom 27 (15.1%) were diagnosed with intussusception. The test characteristics of the 3 view radiograph series in the diagnosis of intussusception when all 3 views had air in the ascending colon were sensitivity of 100% (95% confidence interval [CI] 87-100), specificity of 18.4% (95% CI, 12.3-24.6), likelihood ratio for a negative test of 0 (95% CI, 0.01-1.53), and negative predictive value (NPV) of 100% (lower bound 95% CI, 98%). When at least 2 views had air in the ascending colon, the test characteristics were sensitivity of 96.3% (95% CI, 89.2-100), specificity of 41.4% (95% CI, 33.6-49.3), likelihood ratio for a negative test of 0.09 (0.01-0.62), and NPV of 9.84 (95% CI, 95.2-100). Conclusions: Using specific criteria, the presence of air in the ascending colon on 2 or 3 abdominal radiograph views has the potential to substantially decrease the likelihood of or exclude intussusception.