Gastrointestinal perforation without intraperitoneal air-fluid level in neonatal pneumoperitoneum.

Twenty-one cases of neonatal pneumoperitoneum seen at the C.S. Mott Children's Hospital between 1970 and 1974 were reviewed. Of 14 cases with autopsy or surgically proven gastrointestinal perforation, six did not have an intraperitoneal air-fluid level on the initial cross-table lateral roentgenogram while eight did. Four cases of pneumoperitoneum secondary to mechanically assisted ventilation and air leak phenomenon had no air-fluid level. A postmortem study in one premature infant cadaver suggests that the air-fluid level may not be reliably demonstrated when the amount of intraperitoneal fluid is small. These data support the contention that an intraperitoneal air-fluid level does not uniformly occur in perforated viscus and that its absence, even in the presence of mechanically assisted ventilation, does not exclude the necessity for surgical intervention. In some cases, water-soluble contrast examination of the gastrointestinal tract appears to be a more reliable diagnostic adjunct to document a perforated viscus, when a definite intraperitoneal air-fluid level is not present.