Factors associated with bowel resection among infants with intussusception in the United States.

Background: Intussusception is the most common cause of infant bowel obstruction. Because delays in diagnosis can lead to severe outcomes, differentiating milder cases from those with potentially serious outcomes is important.

Objective: The objective of this study was to identify factors associated with bowel resection among intussusception cases using data from a large nationwide study, which investigated the association between intussusception and Rotashield.

Methods: We examined characteristics of 376 intussusception cases not associated with Rotashield use. Cases were confirmed by a radiologic procedure, surgery, or autopsy. Clinical characteristics of infants with and without bowel resection were compared.

Results: During the week before hospitalization, 93% of the 376 infants with intussusception had vomiting, 72% reported bloody stool, 63% had hemoccult positive stool, 51% had diarrhea, 43% reported fever, and 14% had documented fever. Surgery was performed on 209 cases (56%). Of these 209 cases, 33% (67/209) required bowel resection. Documented fever on admission significantly increased the risk of bowel resection (odds ratio, adjusted for race and sex, 2.7; 95% confidence interval, 1.2-6.0). Among infants with intussusception, the presence of a reported symptom for at least 2 days before hospital admission was also an independent predictor of bowel resection (adjusted odds ratio, 2.7; 95% confidence interval, 1.5-4.8).

Conclusions: Bowel resection appears to be more likely among intussusception patients with documented fever and symptoms for at least 2 days. However, because resection also occurred among those without fever or prolonged symptoms, severe disease must also be considered in absence of these symptoms.