Urinary excretion of low-osmolar contrast media during small-bowel follow-through studies in infants is not definitive evidence of bowel perforation.

Abstract

Presence of contrast agent in the urinary system in infants after small-bowel follow-through study with low-osmolar contrast media has been described as a sign of bowel perforation. To evaluate how often the presence of contrast agent in the bladder after small-bowel follow-through is a reliable sign of bowel perforation or necrosis. From the radiology information system, we retrieved imaging reports of infants evaluated with small-bowel follow-through and findings of contrast agent in the bladder. We retrieved demographic and clinical information from the medical records. Presence of bladder contrast medium was considered true-positive evidence of bowel perforation or necrosis if confirmed by pneumoperitoneum, extraluminal contrast agent, surgery or pathology within 3days of the small-bowel follow-through. False-positives for bowel perforation or necrosis were based on surgical findings or clinical follow-up. Of the 207 infants who had small-bowel follow-through, 18 infants (12 boys; mean age 50days, range 14days to 8.5months) had contrast medium in the bladder after the small-bowel follow-through. Fifteen of the 18 (83.3%) had a history of prematurity and 11 had prior abdominal surgery. Four of the 18 (22.2%) had bowel perforation or necrosis at surgery or pathology performed more than 3days after the small-bowel follow-through and were considered indeterminate and excluded. Eight of the remaining 14 infants (57.1%) had bowel perforation or necrosis based on surgical evidence of perforation or pathology confirmation of necrosis (n=6), pneumoperitoneum (n=1) or contrast agent leakage from enterocutaneous fistula (n=1). Six of the 14 (42.9%) were false-positives, without evidence of bowel perforation or necrosis based on clinical follow-up (n=4) or surgery (n=2). Demonstration of urinary contrast agent post small-bowel follow-through with low-osmolar contrast medium in newborns/infants with complex medical problems is not a definitive indication of bowel perforation or necrosis. More than one-third of our patients with contrast medium in the bladder did not have bowel perforation or necrosis.