OBJECTIVE: Pneumoperitoneum diagnosed by plain radiography is often a sign of gastrointestinal perforation and is unusual (0.17%) in patients on continuous ambulatory peritoneal dialysis (CAPD). These patients are prone to bacterial peritonitis, which can have overlapping clinical findings with perforated viscus. Because CT is often used to screen symptomatic CAPD patients, the reliability of pneumoperitoneum as a diagnostic sign is examined. MATERIALS AND METHODS: Records of 433 CAPD patients were examined; of these, 56 patients had had CT while on CAPD (rupture, 6; nonrupture, 50). Plain radiography and CT were examined for free air and fluid. Patients were classified according to the amount of free air detected. RESULTS: Patients with ruptured viscus had pneumoperitoneum in 5 of 6 cases by CT and 4 of 6 by plain radiography. Excluding free air, ruptured viscus could only be diagnosed in 1 of the 6 patients by CT. Patients without rupture had pneumoperitoneum in 15 of 50 cases by CT and 2 of 46 by plain radiography. No instance of pneumoperitoneum was discovered on plain radiography and not CT. CONCLUSION: Pneumoperitoneum was found by CT in 30% of nonrupture CAPD patients, therefore, CT appeared to be more sensitive than plain radiography for its detection. The presence, quantity, and distribution of free air are not helpful in separating perforations from nonperforations. The lack of pneumoperitoneum on CT was found to be a useful but not absolute diagnostic sign to exclude gastrointestinal rupture.
OBJECTIVE: Pneumoperitoneum diagnosed by plain radiography is often a sign of gastrointestinal perforation and is unusual (0.17%) in patients on continuous ambulatory peritoneal dialysis (CAPD). These patients are prone to bacterial peritonitis, which can have overlapping clinical findings with perforated viscus. Because CT is often used to screen symptomatic CAPD patients, the reliability of pneumoperitoneum as a diagnostic sign is examined. MATERIALS AND METHODS: Records of 433 CAPD patients were examined; of these, 56 patients had had CT while on CAPD (rupture, 6; nonrupture, 50). Plain radiography and CT were examined for free air and fluid. Patients were classified according to the amount of free air detected. RESULTS:Patients with ruptured viscus had pneumoperitoneum in 5 of 6 cases by CT and 4 of 6 by plain radiography. Excluding free air, ruptured viscus could only be diagnosed in 1 of the 6 patients by CT. Patients without rupture had pneumoperitoneum in 15 of 50 cases by CT and 2 of 46 by plain radiography. No instance of pneumoperitoneum was discovered on plain radiography and not CT. CONCLUSION: Pneumoperitoneum was found by CT in 30% of nonrupture CAPD patients, therefore, CT appeared to be more sensitive than plain radiography for its detection. The presence, quantity, and distribution of free air are not helpful in separating perforations from nonperforations. The lack of pneumoperitoneum on CT was found to be a useful but not absolute diagnostic sign to exclude gastrointestinal rupture.