The abdominal radiographic series in children with suspected bowel obstruction--should the second view be abandoned?

Abstract

To determine whether the addition of a second (horizontal beam) view to the supine abdominal radiograph improves diagnostic accuracy in the diagnosis of paediatric bowel obstruction. The abdominal radiographs of 72 paediatric patients were retrospectively evaluated for the presence of bowel obstruction by 22 independent observers (radiologists, surgeons and emergency physicians) of varying experience. Evaluation involved initial assessment of the supine radiograph alone, followed by reassessment with the addition of an upright or decubitus radiograph. When observers evaluated supine radiographs alone, their sensitivity was 66%, specificity 72% and accuracy 70%. With the addition of the second view sensitivity was 78%, specificity 75% and accuracy 76%. The mean change in accuracy was 6% with no statistically significant difference between radiologists, surgeons or emergency staff. The value of the routine second radiograph in the accuracy of diagnosis of bowel obstruction is limited, and its use should be restricted to selected cases in which diagnostic uncertainty persists.