Diagnostic accuracy of physical examination, transvaginal sonography, rectal endoscopic sonography, and magnetic resonance imaging to diagnose deep infiltrating endometriosis

Abstract

To compare the value of physical examination, transvaginal sonography (TVS), rectal endoscopic sonography (RES), and magnetic resonance imaging (MRI) for the assessment of different locations of deep infiltrating endometriosis (DIE). Retrospective longitudinal study. Tertiary university gynecology unit. Ninety-two consecutive patients with clinical evidence of pelvic endometriosis. Physical examination, TVS, RES, and MRI, performed preoperatively. Descriptive statistics, calculation of likelihood ratios (LR(+) and LR(-)) of physical examination, TVS, RES, and MRI for DIE in specific locations confirmed by surgery/histology. The sensitivity and LR(+) and LR(-) values of physical examination, TVS, RES, and MRI were, respectively, 73.5%, 3.3, and 0.34, 78.3%, 2.34, and 0.32, 48.2%, 0.86, and 1.16, and 84.4%, 7.59, and 0.18 for uterosacral ligament endometriosis; 50%, 3.88, and 0.57, 46.7%, 9.64, and 0.56, 6.7%, -, and 0.93, and 80%, 5.51, and 0.23 for vaginal endometriosis; and 46%, 1.67, and 0.75, 93.6%, -, and 0.06, 88.9%, 12.89, and 0.12, and 87.3%, 12.66, and 0.14 for intestinal endometriosis. The MRI performs similarly to TVS and RES for the diagnosis of intestinal endometriosis but has higher sensitivity and likelihood ratios for uterosacral ligament and vaginal endometriosis.