Accuracy and reliability of preoperative CT in perforated diverticulitis: Post hoc secondary analysis from the SCANDIV trial.

Abstract

Background and aims: Undetected cancers and fecal contamination in perforated diverticulitis may lead to failure of laparoscopic lavage. This study aimed to evaluate the diagnostic accuracy of preoperative computed tomography (CT) examinations and the interobserver agreement for detecting colonic malignancy and fecal contamination in patients with perforated diverticulitis.

Methods: This study included 198 patients with suspected perforated diverticulitis from the SCANDIV trial, randomized to laparoscopic lavage or primary resection. CT examinations of all patients were independently reviewed by two specialists in abdominal radiology, and features of fecal contamination and colonic malignancy were registered and compared to the surgical reports.

Results: At surgery, fecal contamination was described in 33 patients and was diagnosed on CT in 7 of those by both radiologists (sensitivity = 21%, specificity = 96%, positive predictive value (PPV) = 50%, and negative predictive value (NPV) = 86%) and in 25 patients by at least one of the radiologists (sensitivity = 76%, specificity = 63%, PPV = 29%, and NPV = 93%). Interobserver agreement between radiologists was 64% for fecal contamination (kappa = 0.18, SE = 0.05; p < 0.001). In 7 of 12 patients who turned out to have colon cancer, this was suspected at CT by at least one radiologist (sensitivity = 58%, specificity = 91%, PPV = 29%, NPV = 97%). Interobserver agreement between radiologists for suspected colonic malignancy was 92% (kappa = 0.46; SE = 0.12; p < 0.001).

Conclusion: In this selected cohort of patients with suspected perforated diverticulitis, re-evaluation of CT scans proved unreliable for detecting malignancies and fecal contamination. This suggests that preoperative CT cannot replace follow-up colonoscopy or colonography following laparoscopic lavage.