Diagnostic accuracy of the IDEA protocol for non invasive diagnosis of rectosigmoid DE - a prospective cohort study.

Abstract

To test the accuracy of TVS applying the IDEA approach for suspected rectosigmoid DE and to determine the frequency of other pelvic diseases mimicking DE in patients undergoing surgery. Prospective single center observational study including consecutive women undergoing TVS for clinically suspected rectosigmoid DE followed by conservative or surgical therapy. TVS findings were compared with those obtained by laparoscopy and confirmed histologically. Of the 671 included patients, 128 women opted for medical therapy, and 6 patients decided for surgery but did not give consent to participate in the study. 537 women were enrolled in the final analysis. 279 (52 %) exhibited surgically confirmed rectosigmoid DE. The sensitivity and specificity, positive and negative predictive value (PPV, NPV), positive and negative likelihood ratio (LR+/-) and accuracy of TVS for diagnosing DE in the rectosigmoid were 93.5 %, 94.6 %, 94.9 %, 93.1 %, 17.24, 0.07, 94.04 %. 12 women who were clinically suspected for DE and mimicked sonographic signs fulfilling the IDEA criteria did exhibit other pathologies. Diagnoses were as follows: vaginal Gartner duct cyst (3/291;1.0 %), anorectal abscess (3/291; 1.0 %), rectal cancer (2/291;0.7 %), hydrosalpinx (2/291;0.7 %), metastatic endometrial cancer (1/291;0.35 %) and Crohn's disease (1/291;0.35 %). TVS for diagnosing colorectal DE applying the IDEA criteria is highly accurate for presurgical diagnosis. However, additional pelvic pathologies are encountered in 4-5 % of women attending for suspected rectosigmoid DE. These need to be taken into account when investigating patients for suspected DE.