Gastrointestinal tract perforation: evaluation of MDCT according to perforation site and elapsed time.

OBJECTIVES

To evaluate multidetector computed tomography (MDCT) for the prediction of perforation site according to each gastrointestinal (GI) tract site and elapsed time.

METHODS

One hundred and sixty-eight patients who underwent MDCT before laparotomy for GI tract perforation were enrolled and allocated to an early or late lapse group based on an elapsed time of 7 h. Two reviewers independently evaluated the perforation site and assessed the following CT findings: free air location, mottled extraluminal air bubbles, focal bowel wall discontinuity, segmental bowel wall thickening, perivisceral fat stranding and localised fluid collection.

RESULTS

The overall diagnostic accuracy was 91.07 % and 91.67 % for reviewers 1 and 2, respectively, with excellent agreement (kappa 0.86). Accuracies (98.97 % and 97.94 %) and agreements (kappa 0.894) for stomach and duodenum perforation were higher than for other perforation sites. Strong predictors of perforation at each site were: focal bowel wall discontinuity for stomach, duodenal bulb and left colon, mottled extraluminal air bubbles for retroperitoneal duodenum and right colon, and segmental bowel wall thickening for small bowel. The diagnostic accuracy was not different between the early- and late-lapse groups.

CONCLUSIONS

MDCT can accurately predict upper GI tract perforation with high reliability. Elapsed time did not affect the accuracy of perforation site prediction.

KEY POINTS: Perforation of the stomach and duodenum can be accurately predicted with MDCT. Knowledge of CT findings predicting perforation site can improve diagnostic accuracy. Elapsed time does not significantly affect accuracy in predicting perforation sites.