CT predictors for differentiating benign and clinically worrisome pneumatosis intestinalis in children beyond the neonatal period.

PURPOSE

To determine whether computed tomographic (CT) findings can help differentiate between benign and clinically worrisome causes of pneumatosis intestinalis (PI) in children.

MATERIALS AND METHODS

This retrospective study was approved by the institutional review board, and requirement for informed consent was waived. Data were stored in a secured and HIPAA-compliant fashion. CT reports from an 8-year period (July 2000-July 2008) were reviewed to determine all cases with a diagnosis of PI. In these cases, demographic information, clinical presentation, underlying medical condition, and CT findings were reviewed. The cases were grouped into one of two final diagnostic groups: clinically worrisome versus benign PI (diagnosis of exclusion, resolution documented at serial imaging without therapeutic intervention). In each case, the CT findings reviewed included distribution (small bowel, large bowel), extent (mild, moderate, extensive), and morphologic characteristics (linear, cystic, both) of the PI and associated findings such as soft-tissue bowel wall thickening, periintestinal soft-tissue stranding, free air, free fluid, portal venous gas, small-bowel obstruction, and bowel dilatation. Associations between CT findings and benign or clinically worrisome PI were assessed with logistic regression models.

RESULTS

There were 44 cases identified. Final diagnostic categories for PI included benign (n = 15) and associated underlying bowel disease (n = 29; definitive in 26 and suspected but not defined in three). The following findings were significant (expressed as percentage of clinically worrisome PI vs percentage of benign): soft-tissue bowel wall thickening (51.2% vs 13.3%, P = .0167), free peritoneal fluid (82.8% vs 33.3%, P = .002), extent of PI (extensive 17.2% vs 69%, P < .001), and periintestinal soft-tissue stranding (55.2% vs 20.0%, P = .0228). Distribution, free peritoneal air, and characteristic morphology (linear vs cystic) were not associated with clinically worrisome PI (all P > .05).

CONCLUSION

The cystic or linear pattern of pneumatosis in children is not a useful CT sign to differentiate benign from clinically worrisome PI. CT findings that include soft-tissue thickening of the bowel wall, free fluid, periintestinal soft-tissue stranding, and the extent of PI can be useful in differentiating these entities.